research on clinical gait analysis

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Front Med Technol

Present and future of gait assessment in clinical practice: Towards the application of novel trends and technologies

Abdul aziz hulleck.

1 Mechanical Engineering Department, Khalifa University, Abu Dhabi, United Arab Emirates

Dhanya Menoth Mohan

2 School of Mechanical and Aerospace Engineering, Monash University, Clayton Campus, Melbourne, Australia

Nada Abdallah

3 Weill Cornell Medicine, New York City, NY, United States

Marwan El Rich

Kinda khalaf.

4 Biomedical Engineering Department, Khalifa University, Abu Dhabi, United Arab Emirates

5 Health Engineering Innovation Center, Khalifa University, Abu Dhabi, United Arab Emirates

Despite being available for more than three decades, quantitative gait analysis remains largely associated with research institutions and not well leveraged in clinical settings. This is mostly due to the high cost/cumbersome equipment and complex protocols and data management/analysis associated with traditional gait labs, as well as the diverse training/experience and preference of clinical teams. Observational gait and qualitative scales continue to be predominantly used in clinics despite evidence of less efficacy of quantifying gait.

Research objective

This study provides a scoping review of the status of clinical gait assessment, including shedding light on common gait pathologies, clinical parameters, indices, and scales. We also highlight novel state-of-the-art gait characterization and analysis approaches and the integration of commercially available wearable tools and technology and AI-driven computational platforms.

A comprehensive literature search was conducted within PubMed, Web of Science, Medline, and ScienceDirect for all articles published until December 2021 using a set of keywords, including normal and pathological gait, gait parameters, gait assessment, gait analysis, wearable systems, inertial measurement units, accelerometer, gyroscope, magnetometer, insole sensors, electromyography sensors. Original articles that met the selection criteria were included.

Results and significance

Clinical gait analysis remains highly observational and is hence subjective and largely influenced by the observer's background and experience. Quantitative Instrumented gait analysis (IGA) has the capability of providing clinicians with accurate and reliable gait data for diagnosis and monitoring but is limited in clinical applicability mainly due to logistics. Rapidly emerging smart wearable technology, multi-modality, and sensor fusion approaches, as well as AI-driven computational platforms are increasingly commanding greater attention in gait assessment. These tools promise a paradigm shift in the quantification of gait in the clinic and beyond. On the other hand, standardization of clinical protocols and ensuring their feasibility to map the complex features of human gait and represent them meaningfully remain critical challenges.

1. Introduction

Changes in the signature of gait, or the unique sequential walking pattern in humans, reveal key information about the status and progression of numerous underlying health challenges, from neurological and musculoskeletal conditions to cardiovascular and metabolic disease, and to ageing-associated ambulatory dysfunction and trauma. Accurate reliable identification of gait patterns and characteristics in clinical settings, as well as monitoring and evaluating them over time, enable effective tailored treatment, inform predictive outcome assessment, and an allow for an overall better practice of precision medicine.

In clinical gait assessment, both a person's “ability” to walk and “how” the individual walks are highly relevant. The walking ability of a person is typically based on two main aspects: how far can an individual walk and what is his/her tolerance level ( 1 ). For example, for post stroke gait assessment, 3-, 6-, or 10 min walk tests are used, in addition to Functional Ambulation Category (FAC), Short Physical Performance Battery (SPPB), and/or Motor Assessment Scale (MAS). Other clinical subjective assessment scales include the Unified Parkinson Disease Rating Scale (UPDRS) the Scale for the Rating and Assessment of Ataxia (SARA)], the Alzheimer's Disease Assessment Scale (ADAS), the Expanded Disability Status Scale (EDSS) the High-level MobilitARTIy Assessment Tool (HiMAT), and the Dynamic Gait Index ( 2 ). The quality of gait or “how” the person walks, on the other hand, highly depends on the quantification of gait patterns and accurate identification of specific gait characteristics. Despite evidence of the advantages of quantitative instrumented gait analysis (IGA) in clinical practice and recommendations by the National Institute for Health and Clinical Excellence (NICE) ( 3 ) identifying IGA is the preferrable choice for “gait-improving surgery”, it remains not well leveraged in clinical settings due to the high cost/cumbersome equipment and complex protocols/data analysis associated with traditional gait labs, as well as diverse training, experience and preference of clinical teams ( 3 – 5 ). Moreover, the use of IGA by allied health professionals, such as physiotherapists, occupational therapists and orthotists, and training also remain non standardized and limited ( 5 – 7 ).

Observational gait analysis continues to be popular among clinicians due to its inherent simplicity, availability, and low cost ( 8 ). On the other hand, the validity, reliability, specificity, and responsiveness ( 9 , 10 ) of these qualitative methods are controversial and increasingly being questioned ( 6 ). Furthermore, there is evidence to suggest that subjective clinical assessment scales may not be sensitive to disease severity and specific characteristics and may limit understanding of underlying disease mechanisms, hence adversely impacting optimal treatment ( 11 ). Examples of such scales include Multiple Sclerosis (MS), where subjective measures, such as the Expanded Disability Status Scale (EDSS), the Multiple Sclerosis Severity Scale (MSSS), Multiple Sclerosis Walking Scale (MSWS), and Multiple Sclerosis Functional Composite (MSFC), continue to be widely used in clinical practice. These scales have been criticized for lack of sensitivity ( 12 ), high interrater variability ( 13 ), as well as being prone to practice effects and variability ( 14 , 15 ). Similarly, clinical assessment of Parkinson's disease (PD) using the Unified Parkinson's Disease Rating Scale (UPDRS) is subjective and largely dependent on the expertise and experience of the clinicians, as well as the severity of the disease ( 16 ). In Stroke patients, assessment tests such as Functional Ambulation Category (FAC), Short Physical Performance Battery (SPPB), and/or Motor Assessment Scale (MAS) are typically employed, along with qualitative observational/visual gait analysis (using naked eye or video images). Nevertheless, the validity, reliability, specificity, and responsiveness of these qualitative methods are questioned ( 9 ), and although they may be useful for the rudimentary evaluation of some gait parameters, they are not adequate for analyzing the multifaceted aspects of gait variability and complexity ( 17 ).

Instrumented gait analysis (IGA), which can provide accurate and precise quantitative measurement of gait patterns and characteristics, has long been the gold standard for gait assessment in research practice ( 18 ). IGA generally refers to the use of instrumentation to capture and analyze a variety of human gait parameters (spatiotemporal, kinematic, and kinetic measures). Traditional IGA systems include motion capture systems, and force plates, instrumented walkways, and treadmills, while more recent systems comprise of miniaturized wearable sensing system, computational platforms and modalities ( 18 ). Literature on the clinical applicability and efficacy of IGA indicates that IGA-based quantitative assessment can improve the diagnosis, outcome prediction, and rehabilitation of various gait impairments as compared to conventional observational scales and techniques for gait dysfunction in a wide spectrum of diseases including MS, PD, Stroke, and Cerebral Palsy ( 9 – 13 ). A recent review on the clinical efficacy of IGA confirms that there is strong evidence that 3-D gait analysis, or 3DGA; has the potential to alter and reinforce treatment decisions; increases confidence in treatment planning and agreement among clinicians; can better identify diagnostic groups and expected treatment outcomes; and overall can improve patient outcomes if recommendations are followed ( 19 ).

Emerging at an unprecedented rate, wearable sensing systems and associated computational modalities are rapidly transforming the quality and accessibility of healthcare, spanning multiple applications from neurology and orthopedics to cardiovascular, metabolic, and mental health. Magnetic (e.g., magnetometers), inertial measurement (e.g., accelerometers and gyroscopes), and force sensors (e.g., insole foot pressure) nowadays offer unprecedented data capture opportunities that can overcome limitations of non-wearable devices due to their low-cost, less setup-time and complexity, lightweight, and portability, making them ideal for out-of-lab and continuous monitoring in the clinic and beyond ( 20 ). Magneto-inertial measurement units (MIMUs), in conjunction with force pressure sensors, have the capability of capturing spatiotemporal, kinematic, and kinetic gait data ( 2 ) rendering the concept of a mobile gait lab a reality. Such labs can inherently overcome the limitations of IGA traditional labs, providing less costly and cumbersome tools with potential for gait assessment in natural environments (clinics, homes, sports arenas, etc.), user friendly interfaces, and the opportunity to provide continuous real-time feedback to clinicians and patients, as well as tele rehabilitation capabilities. In addition, wearable systems allow for easy synchronization with other physiological measurement systems, including EMG, ECG, and EEG, towards the acquisition of invaluable multimodal continuous physiological data in various settings.

This scoping review aims to provide a summary of the current state of clinical gait assessment, including shedding light on gait pathologies and clinical indices and scales, as well as a roadmap for the development of future gait mobile labs- highlighting the clinical validity and reliability of the latest devices and data interpretation algorithms. The word novel in the title of this review reflects recent emergence/implementation of the technologies reviewed and/or recent commercialization. This includes wearable technologies, as well as AI-driven computational platforms. The remainder of the review is structured as follows: Section 2 describes the adopted methodology, including the approach, search strategy and selection criteria. Section 3 details clinical gait pathologies, relevant parameters, as well as current clinical gait assessment tools, scales, and indices, while Section 4 presents gait assessment technologies applicable to clinical settings, including state-of-the-art imaging techniques and wearable technologies, algorithms, and novel AI-driven computational platforms. Section 5 deliberates on the concept of a mobile gait lab for clinical applications. Section 6 highlights the limitations, while Section 7 presents the conclusive remarks and future work.

This review is aimed at summarizing various clinical gait pathologies and associated parameters, applicable gait analysis techniques and gait indices, and the latest trends in wearables systems and algorithms. To address this broader research objective, the authors adopted a scoping review approach rather than a systematic review approach. As reported in ( 21 ), scoping reviews are ideal for addressing a broader scope with a more expansive inclusion criterion.

2.1. Search criteria

A keyword search was performed in PubMed, Web of Science, Medline, and ScienceDirect databases, using a combination of search terms from the following groups: 1. (normal gait OR pathological gait OR gait parameters OR gait indices), 2. (gait assessment OR gait analysis), 3. (wearable systems OR wearable algorithms), 4. (inertial measurement units OR accelerometer OR gyroscope OR magnetometer OR insole sensors OR electromyography sensors). No limit for the year of publication was set, however, the search was last updated in December 2021. Only articles written in the English language were considered in this review. In addition, the reference list of the included articles was checked to identify additional relevant publications meeting the inclusion criteria. The literature search and data extraction were carried out independently by two authors (AAH, DMM) and any inconsistencies and disagreements discrepancies were resolved through following discussions with the other authors (NA, MER, KK).

This scoping review included original published works and review articles which met the following inclusion/exclusion criteria: (i) studies addressing various gait disorders and associated gait parameters, (ii) studies focusing on instrumented gait analysis techniques and gait indices, (iii) studies evaluating the use, validity, and reliability of wearable-based gait measurement devices/systems for measuring gait events, and evaluating and assessing gait dysfunction, (iv) studies concerning the applicability of sensor fusion techniques and algorithms applicable for wearable-based systems with application to gait analysis. The title and/or abstract of the studies were initially screened for suitability. The full-text articles meeting the inclusion criteria were obtained for data extraction and synthesis. A flowchart explaining the same is shown in Figure 1 .

Publication selection process.

3. Clinical gait pathologies and parameters

3.1. normal gait cycle and parameters.

Normal gait can be defined as a series of rhythmic, systematic, and coordinated movements of the limbs and trunk that results in the forward advancement of the body's center of mass ( 22 ). A result of intricate dynamic interactions between the central nervous system and feedback mechanisms ( 23 ), walking is characterized by individual gait cycles and functional phases ( Figure 2 ). A gait cycle consists of two main phases, stance, and swing, which are further divided into five and three functional phases, respectively. The stance phase corresponds to the duration between heel strike and toe-off of the same foot, constituting approximately 60% of the gait cycle. The swing phase begins with toe-off and ends with heel contact of the same foot and occupies 40% of the cycle. As each functional phase contributes to successfully accomplishing the goal of walking, healthy gait involves cyclic and complementary movements of the limbs under control. It is characterized by stance stability; toe clearance during the swing; pre-positioning at swing; sufficient step length; as well as mechanical and metabolic efficiency ( 24 ). Table 1 provides gait parameter ranges based on studies on healthy adults. Determining an appropriate normal range for many of the features is highly challenging as individuals exhibit a wide range of gait patterns across different age groups and gender ( 17 ).

Normal gait cycle (adapted from 1 ).

Gait parameters for healthy individuals ( 1 ).

3.2. Gait parameters associated with pathology

Gait disorders are typically associated with deficits in the brain, spinal cord, peripheral nerves, muscles, joints, or bones. Some medical conditions leading to pathological gait include but not limited to muscular dystrophy, myelodysplasia, cerebral palsy, arthritis, osteoarthritis, head injury, lower limb amputation, multiple sclerosis, rheumatoid, spinal cord injury, parkinsonism, and stroke ( 25 ).

In neuromuscular conditions, the loss of central control affects the motion. In general, patients walk slower than healthy individuals and with compromised spatiotemporal, kinematic, and kinetic parameters. In older adults, a walking speed decline of 0.7% per year is observed, along with significant changes in cadence and step length. The aging population also exhibits lower knee extension at heel-strike and knee flexion during the swing phase ( 23 , 26 ). The following subsections describe some of the most common gait disorders and associated pathological parameters. The associated impacted parameters are summarized in Table 2 .

Effect of pathology on gait disorders.

3.2.1. Neurological gait disorders in elderly people

Gait ailments associated with aging lead to reduction in the quality of life and increased morbidity and mortality. Elderly patients exhibit complex gait disorders, and their dual task ability deteriorates due to a decline in their central resources ( 23 , 26 ).

Specific gait dysfunction noted in the elderly population are summarized as follows:

3.2.1.1. Hypokinetic-rigid gait disorders

Shuffling with a reduced step height and stride length characterizes hypokinetic gait disorder ( 27 ). Reduced arm swing with slow turning movements is also present in isolation. Festination, when patients use rapid small steps to maintain the feet beneath the forward moving trunk, is also observed. Ataxic elements include broad stance width and an increased variability in timing and amplitude of steps ( 27 ). Gait associated with underlying diseases, such as Parkinson's disease, cerebrovascular disease, and ventricular widening, is classified within hypokinetic-rigid gait disorders ( 27 , 28 ).

3.2.1.2. Cautious and careless gait

Defined as gait during which people move slowly with a wider base, and shorter stride, with minimal trunk movement, while the knees and elbows are bent. Whereas careless gait is when patients appear overly confident and walk insensitively fast. Careless gait is due to confusion and delirium associated with old age ( 27 ).

3.2.1.3. Dyskinetic gait or involuntary movements

Patients with post-anoxic encephalopathy exhibit bouncing gait and stance. This is also observed in patients with Parkinson's disease-causing excessive trunk movements contributing to falls. Several dystonic patients are reported to walk on their toes ( 27 , 28 ).

3.21.1.4. Psychogenic gait disorders

Gait dysfunction is common in elderly people due to adverse effects of drugs leading to extrapyramidal side-effects, sedation, orthostatic hypotension, behavioral abnormalities, or ataxia ( 27 , 28 ).

3.2.1.5. Fluctuating or episodic gait disorders

Elderly people often exhibit fluctuating or episodic gait disorder after exercise due to fatigue, and it might be an indication of underlying vascular or neurogenic limping. Freezing gait is part of hypokinetic-rigid syndrome ( 27 , 28 ).

3.2.2. Gait disorders in Parkinson's disease

PD is a neurological disorder which leads to cognition, where gait impairment deteriorates with disease progression, increasing reliance on cognition to control gait. Due to cognitive impairment with PD, the ability to compensate for gait disorders diminishes, leading to further gait impairment. PD is characterized by deficit in amplitude and gait speed, along with increased gait variability ( 29 ).

3.2.3. Gait in diabetic peripheral neuropathy

Neuropathy of motor, sensory, and autonomic components of the nervous system are one of the many complications of Type II Diabetes (T2D). An intact central and peripheral nervous system are essential to initiate and control healthy gait, along with sufficient muscle strength, bone, and joint movements in complete range for normal locomotion. Patients diagnosed with T2D take extra steps when walking in straight paths and during turns, along with an overall reduction in walking speed, step length, cadence, and fewer acceleration patterns as compared to age-matched healthy controls. Joint range of motion is also altered in T2D, where patients with diabetic peripheral neuropathy exhibit a reduced range of motion at the ankle joint in dorsi and plantar flexion and a reduced flexion and extension range of motion at the knee joint in both, as compared to non-diabetic people ( 24 ).

3.2.4. Post stroke gait

Hemiplegia after stroke contributes to significant reduction in gait performance. In stroke survivors, function of the cerebral cortex is usually impaired, whilst that of spinal cord is preserved. Dysfunction is typically demonstrated by a marked asymmetrical deficit. Decreased walking speed and cadence, in addition to longer gait cycle and double limb support as compared to healthy individuals. For hemiplegic stroke survivors, a reduced peak extension of the hip joint in late stance, varying peak lateral pelvis displacement, knee flexion and decreased plantarflexion of ankle at toe off are reported. The GRF (Ground Reaction Force) pattern is characterized as asymmetric, along with decreased amplitude of joint moments, at the lower limb joints on the paretic side ( 30 ).

3.2.5. Total hip arthroplasty (THA)

Large deficits in gait speed ( 31 ), stride length ( 32 , 33 ), sagittal hip range of motion ( 32 , 33 ), hip abduction moment-coronal plane ( 31 ), and negligible changes in transverse plane hip range of motion ( 31 ), deficiency in single limb support time ( 31 ), are reported in patients post THA as compared to healthy controls. Peak hip extension is typically reduced, whereas peak hip flexion remains similar as compared to controls. In addition, peak hip abduction moment is reduced along with peak hip external rotation moment ( 34 ).

3.3. Clinical gait assessment measures and indices

The use of observational gait analysis and subjective rating sales continues to be widespread in clinical settings, both as a diagnostic tool and as a prognostic measure, as previously mentioned. Although these techniques can be useful for the initial rudimentary evaluation of some gait parameters, the validity, reliability, specificity, and responsiveness of these qualitative methods are highly questionable. Researchers have therefore proposed various pathology-specific gait indices and summary measures ( 35 ) based on commercially available technologies with accepted levels of accuracy Table 3 . summarizes the current clinical gait summary measures, discrete and continuous gait indices, and non-linear approaches reported in literature, along with advantages and disadvantages.

Clinical gait measures and indices ( 123 , 124 ).

4. Gait assessment technologies applicable to clinical settings

In the past couple of decades, remarkable technological advancement has been witnessed in the field of gait assessment and analysis, particularly in gait assessment technology. Instrumented walkways, both portable and non-portable, became a good alternative to complicated, bulky and non-portable traditional gait labs. These systems (for example the Walkway and StrideWay from Tekscan Inc., Boston, United States) are now widely used in research and to a limited extent in clinical practice. They typically include low-profile floor walkway systems equipped with grids of embedded sensors below the surface, which record foot-strike patterns as a function of time and space as an individual walks across the platform, and dedicated software which computes the various spatiotemporal gait measures Although these instrumented mats involve less setup time and are generally simple to operate as compared to traditional IGA labs, they are expensive, restrictive to specific operational environment to over-ground trials ( 36 , 37 ).

Marker-based optical motion capture (Mocap) is another rapidly emerging technology effective for obtaining 3D kinematic movement data. Passive Mocap systems [e.g., Vicon (Vicon Motion Systems Ltd, Oxford, United Kingdom) and ELITE optoelectronic system (BTS S. p .A., Milano, Italy)], include retro-reflective markers (that reflect the light emitted by high-resolution infrared cameras) attached to specific anatomic landmarks. The location of the marker is identified by decoding the camera images. Here, the markers must be calibrated for identification before the recording session commences. Active Mocap systems (e.g., Optotrak motion capture system; Northern Digital Inc., Waterloo, Canada), on the other hand, use light-emitting diode (LED) markers (reflect their own light powered by a battery), which are automatically identified ( 38 , 39 ). In the context of clinical relevance, although such systems yield extremely accurate reliable data, operational factors including infrastructure, non-portability, high cost, additional time required for initial set-up and calibration, operational complexity, and restrictions to indoor setup impose hurdles to their functional deployment in clinics and rehabilitation centers ( 84 ). Recently, more portable cost-effective alternatives, such as Microsoft Kinect (based on a depth sensor-based markerless motion capture solution) became the application of choice ( 40 ).

Optoelectronic systems (e.g., Optogait®, Microgate, Italy) have also been used to capture spatiotemporal gait parameters. These mainly consist of a transmitting and a receiving bar containing an infrared light. Interruptions of the communication between the emitter and receiver are detected by the system to calculate the various gait parameters ( 41 ).

An evolution in the measurement of gait kinetic parameters can also be witnessed in the last two decades. These parameters include ground reaction forces, and intersegmental joint reaction forces, moments, and powers. Instrumented walkways offer dynamic plantar pressure mapping but are expensive and do not provide joint kinetic data. Force plates are also used in various gait analysis studies ( 38 , 39 , 42 ). These are able to provide intersegmental joint reaction forces by using the ground reaction forces measured along with inverse dynamics models (Winters book) Chen et al. ( 93 ) developed a novel remote sensing technology called “Electrostatic Field Sensing (EFS)” for measuring human gait including stepping, walking, and running, and further extended the work to post-stroke gait. This technology is credited with several advantages, such as being non-contact, affordable, and allows long-time monitoring ( 43 ). Shoe insole systems represent another category of gait quantification tools and techniques. These systems are designed to allow for the recording of both dynamic plantar pressure and spatiotemporal data. F-scan (Tekscan Inc., Boston, United States) is an ultra-thin in-shoe pressure measurement system utilizing Force-Sensitive Resistive films (FSR) technology ( 44 ).

The characteristics of different measurement systems applicable to clinical settings are summarized in Table 4 , and the pros and cons of these systems are listed in Table 5 .

Portable wearable gait assessment tools.

Pros and cons of different IGA systems.

Computational pipeline using computer vision techniques has been proposed as an ecological and precise method to quantify gait in children with neurodevelopmental disorders, along with the pose estimation software to obtain whole-body gait synchrony and balance ( 45 ). Speed, arm swing, postural control, and smoothness (or roughness) of movement features of gait for Parkinson's patients were extracted using videos processed by ordinal random forest classification model. Significant correlation between clinician labels and model estimates was reported, which provides gait impairment severity assessment in Parkinson's disease using single patient video, thereby reducing the need for sophisticated gait equipment ( 46 ). Computer vision-based gait assessment tools promise frequent gait monitoring using minimal resources ( 46 ). Deep learning to detect human subject in 2D images and then combining 3D sensing data to measure gait features has proven to be more robust than depth cameras in gait parameter acquisition ( 47 ).

4.1. Imaging techniques for gait assessment

As previously mentioned, marker-based optoelectronic systems are currently the most widely used systems in IGA among both research and clinical communities. On the other hand, one of the main sources of error inherent to these systems is the degree of movement of the skin, muscle, and other soft tissues, or the so- called soft tissue artifacts (STA), under the markers in relation to bony landmarks, hence violating the rigid body assumption underlying these methods ( 48 , 49 ). Moreover, STA varies by marker location in a unique and unpredictable manner, particularly during dynamic activities, which can make it unreliable for clinical applications ( 50 ).

Although not yet widespread in biomechanics, computer vision based markerless gait assessment methods offer a promising tool for gait assessment in research, as well clinical and sports biomechanics applications. By leveraging modern technologies, such as improved solvers, advanced image features and modern machine learning, markerless vision-based systems can reduce the required number of cameras, incorporating moving cameras, increasing the number of tracked individuals, and offering robust detection and fitting in diverse environments. On the other hand, issues such as accuracy and field-based feasibility remain to be addressed ( 51 ).

Three-dimensional imaging techniques have been successfully used to directly determine bone movements during walking as a gold reference standard to validate/improve current motion capture techniques ( 54 ). For example, researchers have resorted to quantifying STA by comparing with reference 3D kinematics of bone reconstructed from fluoroscopy-based tracking ( 53 ). Fluoroscopy has also emerged as a means for tracking position and orientation of underlying skeletal anatomy of the foot/ankle ( 54 ). Although single plane fluoroscopy yielded large errors when used to evaluate the accuracy of multi-segment foot models ( 49 ), dual fluoroscopy (DF) was found reliable and is considered as the current reference standard to compare joint angles ( 55 ). Combined with 2D/3D registration, video-fluoroscopy allows for accurate quantification of 3D joint motion free of STA ( 56 ). High-speed dual fluoroscopy (DF) has been reported to measure in-vivo bone motion of the foot and ankle with sub-millimeter and sub-degree errors ( 57 ). DF has also been used to evaluate multi-segment foot models and reported good agreement between DF and skin-marker data for the first metatarsal and sagittal plane measurements of the longitudinal arch ( 48 ).

Various researchers investigated the use of DF for clinical applications . In-vivo dual fluoroscopy was used to quantify the hip joint kinematics of patients with Femoroactabular impingement syndrome (FAIS) relative to asymptomatic, morphologically normal control participants during standing, level walking, incline walking and an unweighted functional activity. The kinematic position of the hip joint was obtained by registering projections of 3D computed Tomography models with DF images ( 58 ). Knee kinematic profiles were also obtained using 3D video-fluoroscopy and compared to actual and nominal flexion-extension, internal-external rotations, and antero-posterior translations profiles with optical mocap during stair climbing ( 59 ). Joint function for total talonavicular replacement after a complex articular fracture was evaluated using a full body gait analysis and 3D joint kinematics based on single-plane fluoroscopy ( 60 ). The 3D video fluoroscopic analysis was performed to assess joint motion of the replaced ankle ( 60 ). DF and CT imaging techniques were both employed to calculate in-vivo hip kinematics, along with model-based tracking, to compare the effect of different coordinate systems ( 61 ). Since marker-based systems are unable to accurately analyze talocrural or subtalar motion because the talus lacks palpable landmarks to place external markers ( 54 ), digitized video fluoroscopy was reportedly used to determine the sagittal plane motion of the medial longitudinal arch during dynamic gait ( 62 ). Characteristics of knee joint motion were also analyzed in 6DOF during treadmill walking using a dual fluoroscopy imaging system at different speeds ( 63 ).

DF uses anatomical landmarks visible on 3D CT reconstructions which substantially reduces errors due to STA ( 58 ). Computed tomography (CT) scans of participants are usually needed in DF to determine bone position from the DF images. Single plane fluoroscopy is restricted to 2D motion capture, while using a second FS allows for a full 3D analysis although a single gantry system has lower radiation than the biplane system with reported ionizing radiation levels of 10 µSv per trial ( 54 ). Stationary image intensifiers and static systems have a restricted field of view limiting their application to highly restricted movements ( 56 ). Moving fluoroscopes, consisting of a fluoroscopic unit mounted on a moving trolley which moves with the subject and is controlled by wire sensors to ensure that it remains in the field of view of the image intensifier ( 56 ), provide an enhanced field of view ideal for dynamic scenarios and moving joints.

Fluoroscopic systems designed for precise capture of bone movement and joint kinematics, unlike optical or inertial systems, are not yet commercially available, generally requiring in-house instrumentation and further performance evaluation. The evaluation would typically include determining the resolution of the hardware imaging chain, assessing how the hardware and software reduce or eliminate various distortions, and measuring static and dynamic accuracies and precisions based on precisely known motions and positions ( 64 ). Image quality is a major determinant of error in fluoroscopic applications ( 62 ). Pulse imaging of fluoroscopes, such as pulse width, limits image quality at a given frame rate. Increasing the pulse rate, which is function of pulse width, may add to radiation exposure, leading to an important tradeoff consideration between image quality and radiation exposure ( 63 ). Moving video-fluoroscopes reported lower gait velocity, step length, and cadence as compared to control conditions, indicating altered time distance parameters towards those of slow walking ( 56 ). So far, dynamic MRI used to define in- vivo talocrural and subtalar kinematics ( 65 ) does not allow data collection during normal gait.

Continued multidisciplinary collaborative efforts among biomechanists, imaging and computer vision experts, and clinicians are essential for fully leveraging these highly promising techniques in clinical applications.

4.2. Portable wearable systems for gait assessment

Wearable technology – the use of body-worn sensors to measure the characteristics of human locomotion, has recently emerged as an efficient, convenient, and most importantly, inexpensive option to quantitative gait analysis for both clinical and research-based applications ( Figure 3 ). In general, it uses individual sensor elements, such as accelerometers, gyroscopes, magneto resistive sensors, force/pressure sensors, goniometers, inclinometers, and electromyographic (EMG) sensors, or combined as an inertial measurement unit (IMU) ( 66 ). In comparison to conventional counterparts (e.g., walkway and camera based Mocap), wearable sensing enables continuous gait monitoring (> 2 h) outside the lab or clinic, allowing for replication of natural patterns of walking. Moreover, gait patterns over an ample distance could be measured as opposed to limited walking distance in a lab-based setting.

Wearable gait lab

Accelerometers are often used in gait analysis for assessing stability and risk of fall. In a study which used a single tri-axial accelerometer mounted on the sacrum to analyze the risk of fall among 80 participants, accelerometry-based techniques were found to be able to detect subjects with increased risk of fall by employing appropriate machine learning techniques ( 66 ). In ( 67 ), a 3D accelerometer attached to the lower back was used for stability assessment of older adults. The applicability of a single accelerometer, worn on the back was further examined in ( 68 ), which highlights promising results for implementation in routine clinical practices. Considerable work has also been carried out to assess the consistency of gait characteristics obtained from accelerometers, where discrepancies in sensors positioning yield to critical errors ( 69 ). Furthermore, in ( 70 ), the authors have provided a comprehensive review on the use of accelerometry-based gait analysis techniques and their application to clinical settings.

Gyroscopes are also increasingly employed for gait studies. These devices measure angular velocity and are often combined with accelerometers and other micro-electromechanical systems (MEMS) devices to enhance performance through sensor fusion techniques. They have found applications in step detection, gait event detection, segmental kinematics, and more. For instance, a single gyroscope placed in the instep of the foot was successfully used to detect gait events, including heel strike, foot flat, heel off, and toe-off ( 71 ). Another study involved two gyroscopes, mounted on the lower left and right side of the waist to calculate walking steps and step length ( 72 ).

Magnetometers measure the magnetic field direction and intensity at a specific point. In combination with other inertial sensors (accelerometers and gyroscopes), they form a so-called inertial measurement unit (IMU), which can produce a drift-free estimation of gait parameters ( 73 ). Sophisticated commercialized IMUs (Physiolog 5 IMU, Gait Up, Switzerland, MTw Awinda, Xsens Technologies B.V., Netherlands), as well as in-house developed systems, were equally used for gait studies ( 74 ). In the context of human motion analysis, IMUs are employed for several possible goals, for example, to estimate the joint angles ( 74 ), to detect the risk of fall in an elderly population, long term monitoring of activities and symptoms ( 75 ), measurement of gait events, spatiotemporal parameters ( 76 , 77 , 78 ), ground reaction forces and moments ( 79 ), and estimation of gait symmetry ( 80 ). Mariani et al. (2010) used IMUs to measure foot kinematics in a study involving both young and elderly and reported the suitability of the system to clinical practice ( 81 ). Parisi et al. developed a low-cost system with a single IMU attached to the lower trunk to examine the gait characteristics of both hemiparetic and normal control subjects through measurement of spatiotemporal parameters, which showed excellent correlation with the parameters obtained from a standard reference system ( 78 ).

Insole systems for gait measurement and analysis represent a major category, which is cost-effective, portable, and applicable for both indoor and outdoor settings. Over the years, various technologies were developed ( 82 ), tested, and commercialized. These include capacitive sensors (Pedar system, Novel GmbH, Germany) ( 83 ), force-sensing resistors (FSR) (F-Scan, Tekscan Inc., United States) ( 84 ), and piezoresistive sensors (FlexiForce system, Tekscan, United States and ParoTec system, Paromed, Germany) ( 82 ). Researchers have adopted different approaches about the design, fabrication, and applications of insole systems. Both prefabricated and in-house fabricated insole systems have been tested for healthy and pathological gait ( 85 , 86 ). Some studies have also integrated inertial measurement units (IMU) with shoe insoles to enhance their capabilities. Despite the fact that these shoe-based systems have successfully been used for various gait analysis applications, they suffer from some drawbacks, such as (i) distortion of the flexible contact surface due to repeated loading, which leads to changes in the sensor response, (ii) drift in the output due to prolonged load application that causes heat inside the shoe, and (iii) need for subject-specific calibration that may alter accuracy ( 87 ). Mancinelli et al. (2012) presented ActiveGait – a novel sensorized shoe system for real-time monitoring of gait deviations associated with Cerebral Palsy in children. They reported that the severity of gait deviations can be estimated with an accuracy greater than 80% using the features derived from the center of pressure trajectories gathered from the shoe system ( 88 ). In ( 87 ), the authors designed a novel flexible foot insole system using an optoelectronic sensing technology for monitoring plantar pressure deviations in real-time. The system consists of an array of 64 sensing elements and onboard electronics for signal processing and transmission. Experimental validation was conducted on healthy subjects while walking at self-selected slow and normal speed. A commercial force plate (AMTI, Watertown, United States) was used as a reference system for benchmarking. Jagos et al. (2017), on the other hand, developed the eSHOE, which consists of four FSR sensors, a three-axis accelerometer, and a three-axis gyroscope, and reported good agreement with the gait parameters obtained from the GAITRite mat ( 89 ). Various other studies have also examined the applicability of shoe-based systems for gait analysis ( 85 , 90 – 92 ).

Another class of sensors that found major applications in gait studies is electromyography (EMG) sensors. Surface EMG is a non-invasive technique used to measure muscle activity. In ( 93 ), Lee et al. proposed a method using EMG signals to obtain biometrics from gait for personal identification methods. Another study adopted EMG techniques to understand the co-contraction patterns of thigh muscle during free walking using surface EMG ( 94 ). These research efforts emphasize the importance of wearable sensors in the study of human gait. The wearable systems discussed in this section are summarized in Table 6 .

Instrumented gait analysis (IGA) systems and their features.

Although emerging new wearable technologies promise to enhance gait assessment and rehabilitation, there is limited research on the use of wearable technology to assess gait and mobility and its efficacy in clinical settings. According to a recently published review by Peters et. al. on the use of wearable technology to assess gait and mobility in stroke patients ( 95 ), most of the available studies are intervention studies conducted in laboratory settings that have used sensors to investigate change in cadence, step time variability, and gait speed. As wearable technologies continue to progress in affordability and accessibility, it is expected that such technologies would enable the gathering of movement-related data in “real-world” and various clinical settings. Importantly, these researchers indicated that so far only a limited number of studies examined reliability and validity of existing wearable devices, highlighting the need for more studies to examine psychometric and other properties when collecting gait and mobility information to determine which wearable technologies are most effective. Another recent review on the evaluation of the use of wearables in PD also indicates that novel technologies and wearables have the potential to enable early or differential diagnosis of PD, monitoring of motion state, prevention, or reduction of off-stage status, and assessing of movement complications. On the other hand, more research is required for the validation and the identification of more accurate markers of PD progression ( 96 ). Importantly, these authors warn that wearable devices may not be appropriate in cases of severe motor impairment, off-stage state, cognitive impairment, and for elderly patients and that further research is required for clinical validation.

4.3. Wearable-based gait computational algorithms

Besides sensor technology, sensor fusion algorithms play a critical role in predicting the accuracy/precision of these wearable-based systems. Most of the research has focused mainly on gait feature detection, daily physical activity monitoring, and gait data classification targeting disease diagnosis and user recognition. These algorithms are based on different data mining and AI technology, including machine learning, fuzzy computing, wavelet transforms, genetic algorithms, and data fusions. Alaqtash et al. ( 97 ) developed an intelligent fuzzy computational algorithm for characterizing gait in healthy, as well as impaired subjects. McCamley et al. established a method to calculate initial and final contact of gait using continuous wavelet transforms, employing waist-mounted inertial sensors ( 98 ). Another study cited the use of a single accelerometer mounted at the lower trunk and a corresponding algorithm to identify gait spatiotemporal parameters ( 68 ). A real-time gait event detection algorithm was proposed in ( 99 ) making use of adaptive decision rules. Further in ( 100 ), an original signal processing algorithm is developed to extract heel strike, toe strike, heel-off, and toe-off from an accelerometer positioned on the feet.

A novel gyroscope only (GO) algorithm was proposed in ( 101 ) to calculate knee angle through the integration of gyroscope-derived knee angular velocity. A zero-angle update algorithm was implemented to eliminate drift in the integral value. In addition, published work on noise-zero crossing (NZC) gait phase algorithm was also adapted. This method is applicable for continuous monitoring of gait data. Nukala et al. used support vector machines (SVM), KNN, binary decision trees (BDT), and backpropagation artificial neural network (BP-ANN) to classify the gait of patients from normal subjects, where features extracted from raw signals from gyroscopes and accelerometers were used as inputs. This study reported the highest overall classification accuracy of 100% with BP-ANN, 98% with SVM, 96% with KNN, and 94% with BDT ( 102 ).

Li et al. proposed DTW algorithm, sample entropy method, and empirical mode decomposition to calculate 3 main gait features of post-stroke subjects: symmetry, complexity character, and stepping stability. A k-nearest neighbor (KNN) classifier trained on the acquired features showed a promising result (area under the curve (AUC) of 0.94), which suggests the feasibility of such techniques to automatic gait analysis systems ( 43 ). Rastegari et al. employed a feature selection technique called maximum information gain minimum correlation (MIGMC) to extract gait data of subjects with Parkinson's Disease ( 103 ). The performance of several machine learning classifiers, including Support Vector Machines, Random Forest, AdaBoost, Bagging, and Naïve Bayes were also assessed to test the power of the feature set obtained.

The use of novel computational platforms, including Machine Learning, Support Vector Machine, and Neural Network approaches, are increasingly commanding greater attention in gait and rehabilitation research. Although their use in clinical settings are not yet well leveraged, these tools promise a paradigm shift in stroke gait quantification and rehabilitation, as they provide means for acquiring, storing and analyzing multifactorial complex gait data, while capturing its non-linear dynamic variability and offering the invaluable benefits of predictive analytics ( 1 ). A recent review article discussed the potential value of ML in gait analysis towards quantification and rehabilitation ( 104 ). The authors concluded that further evidence is required although preliminary data demonstrates that the control strategies for gait rehabilitation benefit from reinforcement learning and (deep) neural-networks due to their ability to capture participants' variability. This review paper demonstrated the success of ML techniques in detecting gait disorders, predicting rehabilitation length, and control of rehabilitation devices. Further work is needed for verification in clinical settings.

4.4. Data-driven gait rehabilitation in clinical settings

Quantitative gait assessment is invaluable towards disease-specific and patient specific rehabilitation/therapeutic interventions. Spatiotemporal, kinematic, and kinetic parameters obtained during instrumented gait assessment can help clinicians benchmark, devise strategies, and evaluate the effect of various rehabilitation interventions. Gait disorders not only affect these parameters, and patterns and time spent in the various gait phases, but can also highly impact gait symmetry, and regularity, depending on the disease and severity ( 105 ). Increasing evidence supports a data-driven physical rehabilitation approach to the treatment of functional gait disturbance ( 106 ). There are multiple examples in literature on the effective use of quantitative gait measures towards more effective data-driven rehabilitation. A recent review by Biase et al. ( 107 ) studied the most relevant technologies used to evaluate gait features and the associated algorithms that have shown promise to aid diagnosis and symptom monitoring towards rehabilitation in Parkinson's disease (PD) patients. They reported physical kinematic features of pitch, roll and yaw rotations of the foot during walking, based on which feature extraction and classification techniques, such as principal component analysis (PCA) and support vector machines (SVM) method were used to classify the PD patients. They also used gait features, including step duration, rise and fall gradients of the swing phase, as well as standard deviation of the minima as quantitative measures, for benchmarking and monitoring PD motor status during rehabilitation. Interestingly, this review sheds light on need to change the evaluated gait features as a function of disease progression. Another study was Pistacchi et al. ( 108 ) suggested spatiotemporal gait parameters, such as speed and step length, where reduced step length seems to be a specific feature of Parkinson's disease gait particularly in early disease stages. On the other hand, asymmetry, step shuffling, double-limb support and increased cadence are more common in mild to moderate stages, while advanced stages are more frequent freezing of gait (FOG) and motor blocks, reduced balance and postural control, motor fluctuations and dyskinesia ( 109 ). Researchers have also investigated the evaluation of ambulatory systems for gait analysis post hip replacement ( 110 ). They found gait characteristics such as stride length and velocity, as well as thigh and shank rotations different from healthy individuals and recommended their use to monitor post-surgical rehabilitation efficacy. Spatiotemporal gait parameters, such as step length, width and cadence have been used ( 111 ) to assess the effect of swing resistance and assistance rehabilitation on gait symmetry in hemiplegic patients. Investigators have also studied whether specific variables measured routinely at a rehabilitation center were predictors of gait performance of hemiparetic stroke patients ( 112 ). They found that motor control and balance were the best predictors of gait performance. A recent review article on assessment methods of post stroke gait suggests that multiple spatiotemporal, kinematic, and kinetic parameters can be useful in diagnosing post-stroke gait dysfunction and as quantitative measures to evaluate rehabilitation outcomes ( 1 ). Spatiotemporal characteristics of post-stroke gait include reduced step or stride length, increased step length on the hemiparetic side, wider base of support, greater toe-out angle, reduced walking speed and cadence. Stride time, stance period on both lower limb, and double support time are also increased, in addition to less time in stance and more time in swing phase for the paretic side, as well as asymmetries in spatial and temporal factors. Kinematic parameters associated with hemiplegic gait (reduced mean peak extension of the hip joint in late stance, alterations in the lateral displacement of the pelvis and flexion of the knee, and decreased plantarflexion of the ankle at toe-off, in addition to a significant decrease in peak hip and knee flexion during the swing phase, reduced knee extension prior to initial contact, as well as decreased ankle dorsiflexion during swing), and kinetic parameters (asymmetric patterns, as well as decreased amplitudes of the joint moments and joint powers at the hip, knee, and ankle joints on the paretic side) can be used as quantitative means to design and evaluate effective rehabilitation ( 113 – 115 ). IGA has also been successfully used to quantify and improve gait dysfunction associated with ageing and assess the risk of falling ( 116 ). Spatiotemporal gait parameters such as velocity, swing time, stride length, stride time- and double support time variability, as well as heel strike and toe off angles, and foot clearance, have been suggested as plausible indicative quantitative measures ( 116 ) to assess the risk of falling in elderly subjects. Inertial sensor-equipped shoes additionally provided heel strike and toe off angles, and foot clearance ( 116 ). The study ( 117 ) summarizes that multi-component exercise therapy which consisted of strength, ROM exercise, balance, flexibility and stretching exercises, circuit exercise training, and gait training was found to enhance gait function for individuals suffering with diabetic peripheral neuropathy compared to control groups using spatiotemporal gait parameters like velocity, cadence, step length, step time, double support time, stride length, stride time, ankle ROM. Gait assessment has potential to develop patient training paradigms for overcoming gait disorders ( 111 ).

5. Mobile gait lab for clinical applications and beyond

In recent decades, the healthcare field has witnessed a tremendous interest in the use of wearable sensing modalities and AI-driven data management/analysis techniques for patient diagnosis, monitoring, and rehabilitation. The portability, lightweight, ease of use, and high-power efficiency are some of the factors that promote applicability to a clinical platform.

There are few examples in literature demonstrating the potential success of using wearable-based systems for gait assessment in clinical settings. Prajapati et al. assessed the walking activity of inpatients with subacute stroke using commercial accelerometers attached above the ankle. They found that the walking bouts were shorter in duration and gait was more asymmetric ( 118 ). Studies have established test-retest reliability and accuracy of different sensor technologies; however, further validation trials are recommended prior to any clinical use. Hsu et al. assessed the test-retest reliability of an accelerometer-based system with infrared assist for measuring spatiotemporal parameters, including walking speed, step length, and cadence, as well as trunk control parameters, including gait symmetry, gait regularity, acceleration root mean square, and acceleration root mean square ratio of healthy subjects in hospital ( 119 ). This study showed excellent test-retest reliability of the parameters considered, and thus highlighting the reliability of an infrared assisted, trunk accelerometer-based device for clinical gait analysis. Another study investigated the concurrent validity and test-retest reliability of gait parameters (cadence, gait velocity, step time, step length, step time variability, and step time asymmetry) acquired from elderly subjects, using a tri-axial accelerometer attached to the center of body mass ( 120 ). In comparison to a reference GAITRite system, the acquired parameters showed good validity and reliability. Poitras et al. performed a systematic review of 42 studies assessing the reliability and validity of wearable sensors, specifically, IMUs, for quantifying the joint motion ( 121 ). Evidence suggests that IMU could be an alternative solution to an expensive motion capture system, as it shows good validity for lower-limb analysis involving fewer complex tasks. However, more work is needed to draw a better conclusion with regards to its reliability, as well as to standardize the protocol to get more accurate data in a clinical setting. Importantly, additional research efforts are needed to examine the responsiveness of wearables in free-living conditions in hospital settings.

6. Limitations

This review aimed to summarize available published work on the present and future of gait analysis in clinical settings. The focus was to highlight current systems, scales, and indices, as well as recent technology-driven gait characterization and analysis approaches and their applicability to clinical settings. Within this context, pathological gait associated with different disease, as well as ageing was briefly discussed. As such, this article may have not covered the complete spectrum of gait pathologies and associated parameters. A scoping (non-systematic) search methodology was selected to broaden the scope and integration of the three main aspects of focus (gait pathology, clinical assessment, recent tools, and technologies). In addition, we do not recommend any specific protocol over the other, as most of the papers incorporate different inclusion/exclusion criteria for subject selection, as well as different sampling sizes, which may render comparisons unrealistic.

7. Conclusive remarks and future work

This scoping review aimed to shed light on the status of gait assessment in clinical settings, as well as the state-of-the-art emerging tools and technologies and their potential clinical applicability. Clinical gait analysis continues to rely mainly on observational gait and quantitative scales and is hence subjective and suffers from variability and the lack of sensitivity influenced by the observer's background and experience. Based on the reviewed literature, quantitative IGA-based gait analysis, commonly used in research labs, has the capability of providing clinicians with accurate and reliable gait data for informed diagnosis and continuous monitoring. On the other hand, several factors, including high cost and infrastructure challenges; data variability, complexity, and multidimensionality; lack of sufficient knowledge and standardized training in clinical environments; and time constraints, continue to limit its wide-spread deployment. Rapidly emerging smart wearable technology and AI, including Machine Learning, Support Vector Machine, and Neural Network approaches, are increasingly playing a bigger role in gait assessment. Although their use in clinical settings is not yet well leveraged, these tools promise an unprecedented paradigm shift in the quantification of gait in the clinic and beyond, as they provide means for acquiring, storing, and analyzing multifactorial complex gait data, while capturing its non-linear dynamic variability and offering the invaluable benefits of predictive analytics.

Researchers are also paying increased attention to multisource and multi-modality sensor fusion approaches, which can further add value by integrating the output of multiple sensors to capture the complexity and variability of gait. Multimodality sensor fusion also allows for simultaneous monitoring of various physiological signals during locomotion, such as EMG, ECG, and EEG, where fusing these with various gait measures (spatiotemporal, kinematic, and kinetic) can shed light on underlying health conditions and disease etiology towards better informed outcome prediction and clinical decisions. As the volume of data from the variety of sensors, including electroencephalography, electro-oculography, electro-cardiography, and electromyography, motion capture and force sensors data, substantially increases, more AI-driven sophisticated data management and modeling are needed to quantify and interpret complex network AI/NN models. Models which include static and dynamic features, combined with sophisticated data reduction and individualized feature selection of the most relevant gait characteristics are needed to close the loop for this paradigm shift. Future work is warranted on a multidisciplinary level: to validate the clinical applicability and integration of the various sensing modalities, to ensure proper synchronization of the various systems for accurate continuous real-time monitoring, to develop and validate fast and reliable computational platforms, and to implement modular user-friendly interfaces easy to use in any environment.

In summary, instrumented gait analysis is a well-established tool for the quantitative assessment of gait dysfunction which could be effectively used for functional diagnosis, treatment/surgery/rehabilitation/planning, and progression monitoring for a wide spectrum of disease. The literature indicates that recent advancement in wearable technology and computationally advanced data analytics, including AI, can overcome the challenges of traditional gait labs, allowing for less costly, portable, and relatively simple gait testing protocols in clinical settings, as well as user-friendly data management, analysis, and interpretation computational platforms. On the other hand, the development of clinically driven standardized methodology and procedures is of paramount significance and remains largely unaddressed. These standardized practices should not only focus on quantitative gait diagnosis but should also incorporate sophisticated objective measures and 3-D dynamic gait profiles and markers for monitoring progress and outcome prediction and evaluation. Proper gait protocols should be devised and leveraged towards identifying gait characteristics that could be effectively used as early disease diagnostic markers. Importantly, training clinical teams at various levels, from doctors and surgeons to physiotherapists and other allied health professionals, on properly using these novel assessment and computational tools is equally important and warrants an equally rapid paradigm shift in training and practice in clinical settings towards patient-specific precise medicine.

This publication is based upon work supported by the HEIC at Khalifa University of Science and Technology.

Author contributions

AAH, MR and KK conceived the idea. AAH, DMM, KK and MR formulated the objective for this review. AAH designed the search strategy, conducted abstract screening and full text review, extracted the data, and drafted the manuscript. KK, NA, DMM, and AAH contributed to writing the manuscript. DMM and NA performed a part of the literature survey, including abstract screening, full text review, and data extraction. MR, and KK provided significant guidance on the content of the manuscript, overall supervision, and critical feedback. All authors contributed to the article and approved the submitted version.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Loading metrics

Open Access

Peer-reviewed

Research Article

Clinical gait analysis using video-based pose estimation: Multiple perspectives, clinical populations, and measuring change

Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Visualization, Writing – original draft, Writing – review & editing

Affiliations Center for Movement Studies, Kennedy Krieger Institute, Baltimore, Maryland, United States of America, Department of Physical Medicine and Rehabilitation, The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States of America

Roles Data curation, Writing – review & editing

Affiliations Center for Movement Studies, Kennedy Krieger Institute, Baltimore, Maryland, United States of America, Department of Neuroscience, The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States of America

Roles Resources, Writing – review & editing

Affiliation Department of Neurology, The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States of America

Roles Conceptualization, Funding acquisition, Methodology, Project administration, Supervision, Writing – original draft, Writing – review & editing

* E-mail: [email protected]

• Jan Stenum, 
• Melody M. Hsu, 
• Alexander Y. Pantelyat, 
• Ryan T. Roemmich

• Published: March 26, 2024
• https://doi.org/10.1371/journal.pdig.0000467
• Peer Review
• Reader Comments

Gait dysfunction is common in many clinical populations and often has a profound and deleterious impact on independence and quality of life. Gait analysis is a foundational component of rehabilitation because it is critical to identify and understand the specific deficits that should be targeted prior to the initiation of treatment. Unfortunately, current state-of-the-art approaches to gait analysis (e.g., marker-based motion capture systems, instrumented gait mats) are largely inaccessible due to prohibitive costs of time, money, and effort required to perform the assessments. Here, we demonstrate the ability to perform quantitative gait analyses in multiple clinical populations using only simple videos recorded using low-cost devices (tablets). We report four primary advances: 1) a novel, versatile workflow that leverages an open-source human pose estimation algorithm (OpenPose) to perform gait analyses using videos recorded from multiple different perspectives (e.g., frontal, sagittal), 2) validation of this workflow in three different populations of participants (adults without gait impairment, persons post-stroke, and persons with Parkinson’s disease) via comparison to ground-truth three-dimensional motion capture, 3) demonstration of the ability to capture clinically relevant, condition-specific gait parameters, and 4) tracking of within-participant changes in gait, as is required to measure progress in rehabilitation and recovery. Importantly, our workflow has been made freely available and does not require prior gait analysis expertise. The ability to perform quantitative gait analyses in nearly any setting using only low-cost devices and computer vision offers significant potential for dramatic improvement in the accessibility of clinical gait analysis across different patient populations.

Author summary

People that experience a stroke or are diagnosed with Parkinson’s disease often have mobility impairments such as slow walking speeds, shortened steps and abnormal movement of the legs during walking. It is a challenge for clinicians to measure and track the multitude of walking parameters that can indicate recovery or progression of disease in an objective and quantitative manner. We present a new workflow that allows a user to analyze the gait pattern of a person walking recorded with only a single video obtained with a smartphone or other digital recording device. We test our workflow is 3 groups of participants: persons with no gait impairment, persons post-stroke, and persons with Parkinson’s disease. We show that a user can perform these video-based gait analyses by recording videos with views from either the side or the front, which is important given the space restrictions in most clinical areas. Our workflow can produce accurate results as compared with a gold standard three-dimensional motion capture system. Furthermore, the workflow can track changes in gait, which is needed to measure changes in mobility over time that may occur because of recovery or progression of disease. This work offers potential for dramatic improvement in the accessibility of clinical gait analysis across different patient populations.

Citation: Stenum J, Hsu MM, Pantelyat AY, Roemmich RT (2024) Clinical gait analysis using video-based pose estimation: Multiple perspectives, clinical populations, and measuring change. PLOS Digit Health 3(3): e0000467. https://doi.org/10.1371/journal.pdig.0000467

Editor: Mengling Feng, National University Singapore Saw Swee Hock School of Public Health, SINGAPORE

Received: May 4, 2023; Accepted: February 12, 2024; Published: March 26, 2024

Copyright: © 2024 Stenum et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The dataset of unimpaired gait is available from http://bytom.pja.edu.pl/projekty/hm-gpjatk . The stroke and PD datasets contain videos with identifiable information and are therefore not available. Code for our workflow is available at https://github.com/janstenum/GaitAnalysis-PoseEstimation/tree/Multiple-Perspectives .

Funding: We acknowledge funding from the NIH (grant R21 HD110686 to RTR), RESTORE Center Pilot Project Award (to RTR via NIH grant P2CHD101913), the American Parkinson Disease Association (grant 964604 to RTR), and the Sheikh Khalifa Stroke Institute at Johns Hopkins Medicine to RTR. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Walking is the primary means of human locomotion. Many clinical conditions–including neurologic damage or disease (e.g., stroke, Parkinson’s disease (PD), cerebral palsy), orthopedic injury, and lower extremity amputation–have a debilitating effect on the ability to walk [ 1 – 3 ]. Quantitative gait analysis is the foundation for effective gait rehabilitation [ 4 ]: it is critical that we objectively measure and identify specific deficits in a patient’s gait and track changes. Unfortunately, there are significant limitations with the current state-of-the-art. Marker-based motion capture laboratories are considered the gold standard measurement technique, but they are prohibitively costly and available largely to select hospitals and research institutions. Other commercially available technologies (e.g., gait mats, wearable systems) only provide predefined parameters (e.g., spatiotemporal data or step counts), are relatively costly, and require specific hardware. There is a clear need for new technologies that can lessen these barriers and provide accessible and clinically useful gait analysis with minimal costs of time, money, and effort.

Recent developments in computer vision have enabled the exciting prospect of quantitative movement analysis using only digital videos recorded with low-cost devices such as smartphones or tablets [ 5 – 7 ]. These pose estimation technologies leverage computer vision to identify specific “keypoints” on the human body (e.g., knees, ankles) automatically from simple digital videos [ 8 , 9 ]. The number of applications of pose estimation for human health and performance has increased exponentially in recent years due to the potential for dramatic improvement in the accessibility of quantitative movement assessment [ 6 , 7 , 10 ]. We have previously used OpenPose [ 8 ]–a freely available pose estimation algorithm–to develop and test a comprehensive video-based gait analysis workflow, demonstrating the ability to measure a variety of spatiotemporal gait parameters and lower-limb joint kinematics from only short (<10 seconds) sagittal (side view) videos of individuals without gait impairment [ 11 ]. Others have also used a variety of approaches to combine pose estimation outputs and neural networks to estimate different aspects of mobility [ 5 , 12 – 16 ].

This foundational work in using pose estimation for video-based gait analysis has demonstrated significant potential of this emerging technology. There are now prime opportunities to build upon what has already been developed and progress toward direct clinical applications. In moving toward clinical application, we considered the needs for: 1) flexible approaches that can accommodate different perspectives based on the space constraints of the end user (e.g., a clinician may only have access to a long, narrow hallway or hospital corridor where a sagittal recording of the patient is not possible), 2) testing and validation directly in clinical populations with gait dysfunction, 3) measurement of clinically relevant gait parameters that are of particular relevance to specific populations, and 4) the ability to measure changes in gait that occur in response to a change in speed.

Here, we present a novel, versatile approach for performing clinical gait analysis using only simple digital videos. First, we developed and tested a novel workflow that performs a gait analysis using frontal plane recordings of a person walking either away from or toward the camera ( Fig 1 ). Our approach is based on tracking the size of the person as they appear in the video image (measured with keypoints from OpenPose) and using trigonometric relationships to estimate depth and, ultimately, spatial parameters such as step length and gait speed ( Fig 2 ; see expanded description in Methods ). Second, we test both our frontal and sagittal workflows directly in two clinical populations with gait impairments that result from neurologic damage or disease (persons post-stroke or with Parkinson’s disease).

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

We recorded three-dimensional (3D) motion capture and digital videos of gait trials performed by persons post-stroke and persons with Parkinson’s disease (A). We analyzed digital videos of the frontal (C Front ) and sagittal plane (C Sag ) with OpenPose to track anatomical keypoints (B). We developed workflows to perform a gait analysis, independently, for videos of the frontal and sagittal plane (C). See Methods section for detailed information about the frontal and sagittal plane post-processing workflows. Note that the ‘Calculate depth-change time-series’ step in the frontal workflow contains multiple sub-steps including tracking the pixel size of the torso and low-pass filtering (see S4 Fig for justification of tracking method and smoothing). We compared spatiotemporal gait parameters and joint kinematics from our workflows to parameters obtained with 3D motion capture (D).

https://doi.org/10.1371/journal.pdig.0000467.g001

A person of size (height) s stands at two distances from a frontal plane camera (C Front ; panel A): an initial reference depth ( d Ref ) and at a depth-change (Δ d i ). The size in pixels of the person at each depth are denoted by s Ref and s i . From trigonometric relationships we derive a relationship between pixel size and depth-change (B, see Methods for detailed explanation; f , focal length of camera; x IP , position of image plane of camera; x Cam , position of camera lens; x Ref , initial position of person; x i , position of person following depth-change). The predicted pixel sizes of a person standing at increasing depths closely tracks manually annotated pixel sizes, which shows that we can use pixel size to estimate depth-changes (C). Summary of our frontal plane workflow (D): OpenPose tracks anatomical keypoints, we find gait cycle events, calculate a time-series of pixel size, and calculate depth-change at which point step lengths and step times can be derived.

https://doi.org/10.1371/journal.pdig.0000467.g002

Development and testing of a novel approach for gait videos recorded in the frontal plane

We first validated our frontal plane approach during overground walking in a group of young participants without gait impairment (we have previously demonstrated the accuracy of obtaining gait parameters using sagittal plane videos in the same dataset of unimpaired participants [ 11 ]). We then compared spatiotemporal gait parameters (step time, step length and gait speed; averaged values for a single walking bout) simultaneously obtained with 3D motion capture and with frontal plane videos positioned to capture the person walking away from one camera and toward the other camera (data collection setup shown in Fig 3A ).

We recorded digital videos of the frontal plane where the person walked toward one camera and away from the other camera (A). We compared spatiotemporal gait parameters (B, step time; C, step length; D, gait speed) between the two digital videos and 3D motion capture (see S1 Table ).

https://doi.org/10.1371/journal.pdig.0000467.g003

Step time showed average differences (negative values denote greater values in video data; positive values denote greater values in motion capture data) and errors (absolute difference) up to one and two motion capture frames (motion capture recorded at 100 Hz; 0.01 and 0.02 s), respectively, between motion capture and frontal plane video ( Fig 3B and S1 Table ). The 95% limits of agreement between motion capture and frontal plane videos ranged from −0.03 to 0.05 s, suggesting that 95% of differences with motion capture fell within this interval. Step length showed average differences and errors up to about 0.02 and 0.03 m, respectively, between motion capture and frontal plane videos ( Fig 3C ). The 95% limits of agreement between motion capture and frontal plane videos ranged from −0.052 to 0.094 m. Gait speed showed average differences and error up to 0.04 and 0.06 m s −1 , respectively, with 95% limits of agreement ranging between −0.11 and 0.17 m s −1 ( Fig 3D ). Correlations for all spatiotemporal gait parameters between motion capture and frontal plane videos were strong (all r values between 0.872 and 0.981, all P <0.001; Fig 3B–3D ).

Testing of video-based gait analysis in persons with neurologic damage or disease

Next, we evaluated both our sagittal and frontal plane workflows in two patient populations with neurologic damage or disease (persons post-stroke and persons with PD). We compared spatiotemporal gait parameters (step time, step length, and gait speed), lower-limb sagittal plane joint kinematics, and condition-specific, clinically relevant parameters (stroke: step time asymmetry and step length asymmetry; PD: trunk inclination) simultaneously obtained with 3D motion capture and with sagittal and frontal plane videos (data collection setup shown in Fig 4A ). Note that frontal videos are limited to spatiotemporal gait parameters and that joint kinematics and trunk inclination can only be obtained from sagittal videos within our current workflows.

We recorded digital videos of the frontal and sagittal plane during gait trials (A). We compared spatiotemporal gait parameters (B, step time; C, step length; D, gait speed) and gait asymmetry (E, step time asymmetry; F, step length asymmetry) between the two digital videos and 3D motion capture. We also compared lower-limb joint kinematics at the hip, knee and ankle obtained with sagittal videos and motion capture for the paretic (G) and non-paretic (H) limbs (MAE, mean absolute error). Gait parameters are calculated as session-level averages of four gait trials at either preferred or fast speeds (see Table 1 ).

https://doi.org/10.1371/journal.pdig.0000467.g004

We present gait parameters as averaged values across four overground walking bouts each at 1) preferred and 2) fast speeds (see S2 Table for values of gait parameters). For preferred speed trials we instructed participants to walk at their preferred speed; for fast speed trials we instructed participants to walk at the fastest speed that they felt comfortable. Of the four trials at each speed, there were two trials of the participants walking away from the frontal camera (with the left side against the sagittal camera) and two trials walking toward the frontal camera (with the right side against the sagittal camera). We intend our workflows to have clinical applications and therefore present values as session-level values (i.e., the results that would be obtained as if the four walking trials were treated as a single clinical gait analysis); we report more detailed comparisons at the level of single trial averages and step-by-step comparisons in the supplement ( S3 and S4 Tables).

Testing in persons post-stroke

We then tested how well our workflows could measure gait parameters in persons post-stroke. Step time showed average differences and errors of zero and one motion capture frames (recorded at 100 Hz; 0 and 0.01 s), respectively, between motion capture and sagittal videos; and average differences and errors of two and five motion capture frames (0.02 and 0.05 s), respectively, between motion capture and frontal videos ( Fig 4B and Table 1 ). The 95% limits of agreement spanned a narrower interval (−0.04 to 0.04 s) for sagittal videos than frontal videos (−0.09 to 0.10 s). Correlations of step time between motion capture and videos were strong ( Fig 4B ; all r ≥0.980).

https://doi.org/10.1371/journal.pdig.0000467.t001

Step length showed average differences and errors of about 1 and 3 cm between motion capture and sagittal videos and average differences and errors of about −3 and 7 cm between motion capture and frontal videos ( Fig 4C and Table 1 ). The 95% limits of agreement spanned intervals of −0.058 to 0.079 m for sagittal videos and −0.154 to 0.087 m for frontal videos. Correlations of step length between motion capture and videos were strong ( Fig 4C ; r ≥0.922).

Gait speed showed average differences and errors of 0.02 and 0.04 m s −1 between motion capture and sagittal videos and average differences and errors of −0.07 and 0.10 m s −1 between motion capture and frontal videos ( Fig 4D and Table 1 ). The 95% limits of agreement spanned intervals of −0.11 to 0.14 m s −1 for sagittal videos and −0.20 to 0.06 m s −1 for frontal videos. Correlations of gait speed between motion capture and videos were strong ( Fig 4D ; r ≥0.981).

Step time asymmetry showed average differences and errors of 0.01 and 0.03 between motion capture and sagittal videos and average differences and errors of 0.02 and 0.07 between motion capture and frontal videos ( Fig 4E and Table 1 ). The 95% limits of agreement spanned intervals of −0.04 to 0.07 for sagittal videos and −0.10 to 0.14 for sagittal videos. Correlations of step time asymmetry between motion capture and videos were strong ( Fig 4E ; all r ≥0.865).

Step length asymmetry showed average differences and errors of −0.002 and 0.050 between motion capture and sagittal videos and average differences and errors of −0.042 and 0.106 between motion capture and frontal videos ( Fig 4F and Table 1 ). The 95% limits of agreement spanned intervals of −0.142 to 0.138 for sagittal videos and −0.291 to 0.208 for frontal videos. Correlations of step length asymmetry were strong between motion capture and sagittal videos ( Fig 4F ; r = 0.890) but weak between motion capture and frontal videos ( Fig 4F ; r = 0.230).

The average mean absolute errors of lower-limb sagittal plane joint kinematics of the paretic and non-paretic limbs were 3.3°, 4.0°, and 6.3° at the hip, knee, and ankle, respectively, between motion capture and sagittal videos ( Fig 4G and 4H ).

Testing in persons with Parkinson’s disease

We next evaluated the performance of the video-based gait analysis in persons with PD ( Fig 5A ). Step time showed average differences and errors of zero and one motion capture frames (0 and 0.01 s) between motion capture and sagittal videos and average differences and errors of one and three motion capture frames (0.01 and 0.03 s) between motion capture and frontal videos ( Fig 5B and Table 1 ). The 95% limits of agreement spanned intervals of −0.02 to 0.02 s for sagittal videos and −0.03 to 0.05 s for frontal videos. Correlations of step time between motion capture and videos were strong ( Fig 5B ; all r ≥0.961).

We recorded digital videos of the frontal and sagittal plane during gait trials (A). We compared spatiotemporal gait parameters (B, step time; C, step length; D, gait speed) between the two digital videos and 3D motion capture. We compared trunk inclination between sagittal plane videos and motion capture (E). We also compared lower-limb joint kinematics at the hip, knee and ankle obtained with sagittal videos and motion capture for the right (F) and non-paretic (G) limbs (MAE, mean absolute error). Gait parameters are calculated as session-level averages of four gait trials at either preferred or fast speeds (see Table 1 ).

https://doi.org/10.1371/journal.pdig.0000467.g005

Step length showed average differences and errors of about −1 and 2 cm between motion capture and sagittal videos and average differences and errors of −5 and 7 cm between motion capture and frontal videos ( Fig 5C and Table 1 ). The 95% limits of agreement spanned intervals of −0.044 to 0.023 m for sagittal videos and −0.150 to 0.048 m for frontal videos. Correlations of step length between motion capture and videos were strong ( Fig 5C ; all r ≥0.959).

Gait speed showed average differences and errors of −0.02 and 0.03 m s −1 between motion capture and sagittal videos and average differences and errors of −0.12 and 0.15 m s −1 between motion capture and frontal videos ( Fig 5D and Table 1 ). The 95% limits of agreement spanned intervals of −0.07 to 0.03 m s −1 for sagittal videos and −0.28 to 0.04 m s −1 for frontal videos. Correlations of gait speed between motion capture and videos were strong ( Fig 5D ; all r ≥0.982).

Trunk inclination showed average differences and errors of 0° and 1.5° between motion capture and sagittal videos ( Fig 5E and Table 1 ; trunk inclination can only be extracted from sagittal videos, not frontal videos).

The average mean absolute errors of left and right lower-limb sagittal plane joint kinematics were 2.7°, 3.5°, and 4.8° at the hip, knee, and ankle, respectively, between motion capture and sagittal videos ( Fig 5F and 5G ).

Measuring changes in gait that occur due to changes in gait speed

Next, to evaluate how accurately video analysis can track within-participant gait changes, we calculated the changes in spatiotemporal gait parameters that accompanied the increase in gait speed from preferred to fast speed gait trials in persons post-stroke and with PD ( Fig 6A ). The change in step time as a result of faster walking in persons post-stroke showed average differences and errors of zero and two motion capture frames (0 and 0.02 s) when compared between motion capture and sagittal videos and average differences and errors of zero and four motion capture frames (0 and 0.04 s) when compared between motion capture and frontal videos ( Fig 6B and Table 2 ). The 95% limits of agreement of change in step time of post-stroke walking spanned intervals of −0.03 to 0.03 s for sagittal videos and −0.08 to 0.07 s for frontal videos.

We recorded digital videos of the frontal and sagittal plane during gait trials at preferred and fast speeds (A). We compared spatiotemporal gait parameters (B, step time; C, step length; D, gait speed) between the two digital videos and 3D motion capture. Subscripts Δv of gait parameters denote changes in the gait parameter due to speed-increases from preferred to fast speed walking trials. We calculated gait parameters as the difference between the session-level averages of preferred and fast speed trials (see Table 2 ).

https://doi.org/10.1371/journal.pdig.0000467.g006

https://doi.org/10.1371/journal.pdig.0000467.t002

In persons with PD, the change in step time showed average differences and error of zero and two motion capture frames (0 and 0.02 s) between motion capture and sagittal videos and average differences and errors of zero and three motion capture frames (0 and 0.03 s) between motion capture and frontal videos ( Fig 6B and Table 2 ). The 95% limits of agreement of change in step time of PD walking spanned intervals of −0.02 to 0.02 s for sagittal videos and −0.05 to 0.04 s for frontal videos. Correlations of change in step time between motion capture and videos were strong ( Fig 6B ; all r ≥0.828).

The change in step length as a result of faster walking in persons post-stroke showed average differences and errors of about 0 and 2 cm between motion capture and sagittal videos and average differences and errors of about −1 and 5 cm between motion capture and frontal videos ( Fig 6C and Table 2 ). The 95% limits of agreement of change in step length of post-stroke walking spanned intervals of −0.031 to 0.037 m for sagittal videos and −0.088 to 0.075 m for frontal videos.

Change in step length in persons with PD showed average differences and errors of about 0 and 2 cm between motion capture and sagittal videos and average differences and errors of about −3 and 7 cm between motion capture and frontal videos ( Fig 6C and Table 2 ). The 95% limits of agreement of change in step length of PD walking spanned intervals of −0.022 to 0.028 m for sagittal videos and −0.122 to 0.070 m for frontal videos. Correlations of change in step length between motion capture and videos were strong ( Fig 6C ; all r ≥0.763).

The change in gait speed from preferred to fast speed gait trials in persons post-stroke showed average differences and errors of 0.01 and 0.04 m s −1 between motion capture and sagittal videos and average differences and errors of −0.02 and 0.06 m s −1 between motion capture and frontal videos ( Fig 6D and Table 2 ). The 95% limits of agreement of change in gait speed of post-stroke walking spanned intervals of −0.09 to 0.11 m s −1 for sagittal videos and −0.14 to 0.11 m s −1 for frontal videos.

Finally, in persons with PD, measured change in gait speed showed average differences and errors of 0 and 0.03 m s −1 between motion capture and sagittal videos and average differences and errors of −0.07 and 0.11 m s −1 between motion capture and frontal videos ( Fig 6D and Table 2 ). The 95% limits of agreement of change in gait speed of PD walking spanned intervals of −0.04 to 0.04 m s −1 for sagittal videos and −0.19 to 0.06 m s −1 for frontal videos. Correlations of change in gait speed between motion capture and videos were strong ( Fig 6D ; all r ≥0.949).

Factors that affect accuracy of the frontal video-based gait analysis workflow

We noted that step length errors were occasionally large when calculated from frontal videos (up to nearly 30% of the average step length). We have previously described factors such as the position of the person relative to the camera that influence step length errors when calculated from sagittal videos [ 11 ]. Similarly, we wanted to identify and understand factors that influence step length errors from videos recorded in the frontal plane.

First, we considered that greater depth of the person relative to the frontal plane camera may lead to less precise step length estimates ( S1 Fig ). We partitioned the analysis of step length errors into videos from the frontal plane where the person walks away from the camera or toward the camera because OpenPose may track keypoints differently when viewing the front of the person (when walking toward) or the back of the person (when walking away). We found that step length errors increased with greater depth from the camera, so that the person’s size appeared smaller in the image. Step length errors were more affected by depth when the person walked away from the camera compared to walking toward the camera: from average step length errors of about 7 cm nearest the camera (beginning of trial when the person walks away from the camera; end of the trial when the person walks toward the camera), average errors increased up to about 16 cm when the person walked away, with a more modest increase of up to 11 cm when the person walked toward the camera. This suggests that precision may decrease as the person appears smaller, likely due to less precise keypoint tracking by OpenPose.

We also considered whether a scaling effect influenced step length errors so that longer steps had greater errors. We found that step length errors were not influenced by the magnitude of step length ( S2 Fig ).

We noted time-lags in the gait cycle detection of the frontal videos relative to motion capture that could have influenced step length errors (this analysis could only be performed for the unimpaired participant dataset, in which motion capture and video recordings were synchronized). The timing of gait cycle detection differed depending on walking direction: when the person walked away from the camera, gait cycle timings were, on average, four motion capture frames (~0.04 s) before the timing detected from motion capture, and 15 motion capture frames (~0.15 s) after motion capture when the person walked toward the camera ( S3 Fig , panel A). Using gait event timings from motion capture to calculate step lengths from frontal videos, there was a statistical difference in step length errors when the person walked away from the camera ( P = 0.024), but not when the person walked toward the camera ( P = 0.501; S3 Fig , panel B). The average step length error decreased from about 2 to 1 cm in the unimpaired participant dataset when using gait event timing from the motion capture data in the videos where the person walked away from the camera.

Last, we considered that walking direction relative to the frontal plane camera may have influenced the accuracy of gait parameters. In the unimpaired participant dataset, in which two frontal plane cameras simultaneously captured the same walking trial from different vantage points (see Fig 3A ), we noted a minor overestimation of gait speed by an average of 0.04 m s −1 from the camera that the person walked away from compared to the camera that the person walked toward ( S1 Table ). We observed similar, albeit exaggerated, trends in the stroke and PD datasets. When comparing the average gait speed differences between motion capture and the frontal plane camera, gait speed was overestimated by 0.13 and 0.21 m s −1 for stroke and PD, respectively, when the person walked toward the frontal plane camera; the overestimation was only minor at 0.01 and 0.03 m s −1 for stroke and PD, respectively, when the person walked toward the camera ( S3 Table ). The overestimation of gait speed was accompanied by greater errors when comparing the frontal camera to motion capture: average errors were 0.14 and 0.23 m s −1 for stroke and PD, respectively, when the person walked away from the camera; errors were only 0.06 and 0.08 m s −1 when the person walked toward the camera ( S3 Table ).

The trends of overestimation and greater errors from frontal plane recordings where the person walked away from the camera were mirrored in the results of step length: there were greater overestimations and errors of step length when the person walked away from the camera (average overestimations of 0.056 and 0.082 m and errors of 0.084 and 0.092 m for stroke and PD, respectively) compared to when the person walked toward the camera ( S3 Table ; average overestimations of 0.013 and 0.021 m and errors of 0.062 and 0.055 m for stroke and PD, respectively). This suggests that spatial gait parameters obtained from a frontal plane camera are influenced by walking direction and that the greatest precision was obtained when the person walked toward the camera. Furthermore, this also suggests that the accuracy of gait parameters presented here, when calculated as session-level averages, can be improved if using only gait trials with the same walking direction.

In this study, we demonstrated a new approach for performing clinical gait analyses using simple videos recorded using low-cost devices and a workflow that leverages a freely available pose estimation algorithm (OpenPose) for video-based movement tracking. We showed that this novel approach can perform accurate gait analyses 1) from videos recorded from multiple perspectives (e.g., frontal or sagittal viewpoints), 2) across a diverse range of persons with and without gait impairment, 3) that capture clinically relevant and condition-specific aspects of gait, and 4) that measure within-participant changes in gait as a result of changes in walking speed. These findings demonstrate the versatility and accessibility of video-based gait analysis and have significant potential for clinical applications.

Interest in video-based, markerless gait analysis has accelerated rapidly. Previous studies have used various approaches to move quantitative clinical gait analysis outside of the laboratory or research center and directly into the home or clinic [ 5 , 6 , 13 – 15 , 17 ]. Here, we aimed to develop a single approach that addressed several outstanding needs, including the needs to accommodate multiple different types of environments/viewing perspectives, use of datasets in multiple clinical populations with gait impairment, measurement of both spatiotemporal gait parameters and lower extremity two-dimensional kinematics, and measurement of within-participant changes in gait. It is also notable that we achieved accurate results using multiple different video recording devices with different sampling rates. By comparing our results against gold standard motion capture measurements, we provide data about the accuracy of all findings with respect to the current state-of-the-art.

Our findings also enable us to progress toward development of a series of best practices for video-based clinical gait analysis. Unsurprisingly, we found that video-based gait analyses generated from videos recorded using a sagittal viewpoint generally led to stronger correlations with motion capture data and lower error when compared to videos recorded from frontal viewpoints. This was particularly evident in gait parameters that require especially high levels of precision (e.g., step length asymmetry in persons post-stroke). Similar to our previous work [ 11 ], we also found that video-based measurements of ankle kinematics were generally less accurate (relative to motion capture) than measurements of hip or knee kinematics in persons with or without gait impairment. Therefore, when using the current iteration of our workflow, a user is likely to obtain best results by recording a sagittal video (if possible) and targeting measurement of spatiotemporal gait parameters and more proximal lower limb kinematics. We emphasize that our single-camera, video-based approach is not intended to reach marker-based motion capture levels of accuracy that other multi-camera approaches may target [ 6 , 18 , 19 ] or that may be required by various scientific disciplines (e.g., biomechanics, human motor control), but rather offers clinicians and other end-users access to a reasonably accurate approach for clinical gait analysis that requires minimal time and only a single video recording device.

It is informative to consider the accuracy of our workflow relative to reported test-retest minimal detectable change or minimal clinical important difference values of the population of interest. For example, a meaningful change in gait speed is often reported as 0.10 m s −1 [ 20 ], but may vary from 0.05 up to 0.30 m s −1 depending on the population studied [ 21 – 32 ]. The average errors of our video-based measurements relative to motion capture generally fall within these margins, suggesting that gait speed is likely to be reliably measured in many populations (e.g., older adults, post-stroke, PD, following hip fracture, cerebral palsy, multiple sclerosis) using our workflow. Minimal detectable changes in gait kinematics may also be dependent on the population of interest, with estimates ranging from about 4° to 11° of lower-limb sagittal plane kinematics [ 26 , 28 , 33 – 36 ]. Average errors of sagittal plane hip and knee kinematics in our study were less than 4°, while errors at the ankle were up to 6.8°, suggesting that hip and knee kinematics from our workflow can be accurately tracked while continued improvement in measurement of ankle angles is needed.

There remain additional significant hurdles to widespread implementation of video-based clinical gait analyses. There is a crucial unmet need for improved ease of use, as the user currently must have access to specific computing hardware (i.e., pose estimation is most efficient when using a graphics processing unit (GPU)), download all relevant software, record the videos, and manually process each video through the workflow. This generates an output that is contained within the software. This process is not well-suited for users without some level of technical expertise; there is an important need for new technologies that can streamline these steps and remove much of the technical know-how and burden of manual processing. Furthermore, there is a need for validation in additional adult and pediatric clinical populations, as previous work has shown that existing pose estimation algorithms have difficulty with tracking patient populations with anatomical structures that likely differ significantly from the images used to train the algorithms [ 13 ]. Thirteen of the participants with stroke used a cane; we did not observe instances where OpenPose mistakenly identified the cane as a limb. Lastly, it is likely that accuracy will continue to improve in the future as both computer vision algorithms and methods for data post-processing continue to advance. In this study, we used a pre-trained network [ 8 ], while a different network that was trained to be specific to both gait and clinical condition may further improve accuracy (the challenges of existing pre-trained networks for human pose estimation in movement science have been well-documented [ 37 ]).

In this study, we developed and tested a novel approach for video-based clinical gait analysis. We showed that this approach accommodates multiple viewing perspectives, provides accurate and clinically relevant gait analyses (as compared to 3D motion capture) across multiple participant populations with and without gait impairment, and tracks within-participant changes in gait that are relevant to rehabilitation and recovery outcomes. All software needed to perform these analyses is freely available at https://github.com/janstenum/GaitAnalysis-PoseEstimation/tree/Multiple-Perspectives , where we also provide a series of detailed instructions to assist the user. There is an urgent need to begin to move these emerging technologies with potential for significant clinical applications toward more user-friendly solutions.

Materials and methods

Participants.

We recruited 44 individuals post-stroke (15 female, 29 male; age 61±11 years (mean±SD); body mass 90±23 kg; height 1.73±0.11 m) and 19 individuals with PD (6 female, 13 male; age 67±7 years; body mass 77±14 kg; height 1.71±0.09 m) to participate in the study; all participants were capable of walking independently with or without an assistive device. All participants gave written informed consent before enrolling in the study in accordance with the protocol approved by The Johns Hopkins School of Medicine Institutional Review Board (Protocol IRB00255175). Additionally, we used a publicly available dataset [ 38 ] of overground walking sequences from 32 unimpaired participants (10 women, 22 men) made available at http://bytom.pja.edu.pl/projekty/hm-gpjatk . The dataset included synchronized 3D motion capture files and digital video recordings of the walking sequences. The publicly available dataset does not contain identifiable participant information and faces have been blurred in the video recordings. Our analysis of the publicly available videos was deemed exempt by The Johns Hopkins University School of Medicine Institutional Review Board.

Protocol and data collection

Participants visited our laboratory for one day of testing. They first performed ten-meter walk tests at their preferred speed and the fastest speed at which they felt comfortable walking. Participants then performed eight overground walking trials (four trials at each preferred and fast speeds) across a walkway of 4.83 m.

We mounted two commercially available tablets (Samsung Galaxy Tab A7) on tripods positioned to capture frontal (C Front ) and sagittal (C Sag ) plane views of the overground walking trials (video recordings occurred at a 30-Hz sampling rate; see Fig 1 for overview). Of the eight total walking trials, the participant walked away from the frontal plane camera with the left side turned to the sagittal plane camera during four of the trials; during the other four trials, the participant walked toward the frontal plane camera with the right side turned to the sagittal plane camera. Tablet cameras obtained videos with 1920 × 1080 pixel resolution. The frontal-view tablet was positioned 1.52 m behind the start/end of the walkway and the sagittal-view tablet was positioned 3.89 m to the side of the midpoint of the walkway. The tablet positions were chosen to achieve the longest walkway in which the person remained visible to both frontal and sagittal tablets, given the space restrictions of the laboratory. The frontal- and sagittal-view tablets were rotated to capture portrait and landscape views, respectively. The height of the frontal-view camera was set so that the entire participant remained visible when they were nearest the camera (about 0.85 m). The height of the sagittal-view camera was about 1.18 m so that the participant appeared in the middle of the image as they travelled across the walkway.

We simultaneously recorded walking trials using ten cameras (Vicon Vero, Denver, CO, USA) as part of a marker-based, 3D motion capture system at 100 Hz. We placed reflective markers on the seventh cervical vertebrae (C7), tenth thoracic vertebrae, jugular notch, xiphoid process, and bilaterally over the second and fifth metatarsal heads, calcaneus, medial and lateral malleoli, shank, medial and lateral femoral epicondyles, thigh, greater trochanter, iliac crest, and anterior and posterior superior iliac spines (ASIS and PSIS, respectively).

In the previously published dataset of unimpaired adults without gait impairment, we used a subset of the data (sequences labelled s1 ) that consisted of a single walking bout of approximately 5 m that included gait initiation and termination. We excluded data for one participant because the data belonged to another subset with diagonal walking sequences. We used data from two digital cameras (Basler Pilot piA1900-32gc, Ahrensburg, Germany) that simultaneously recorded frontal plane views of the person walking away from one camera and toward the other camera (see Fig 3A for overview). The digital cameras obtained videos with 960 × 540 pixel resolution captured at 25 Hz. The average distance from the starting position of the participants to the cameras were 2.50 and 7.28 m for the camera that recorded the participant walking away and toward, respectively. Cameras were mounted on tripods and the height was about 1.3 m. Motion capture cameras (Vicon MX-T40, Denver, CO, USA) recorded 3D marker positions at 100 Hz. Markers were placed on the seventh cervical vertebrae, tenth thoracic vertebrae (T10), manubrium, sternum, right upper back and bilaterally on the front and back of the head, shoulder, upper arm, elbow, forearm, wrist (at radius and ulna), middle finger, ASIS, PSIS, thigh, knee, shank, ankle, heel, and toe.

Data processing and analysis

Motion capture data from the participants with stroke or PD were smoothed using a zero-lag 4 th order low-pass Butterworth filter with a cutoff frequency of 7 Hz. The motion capture data from the participants without gait impairment in the publicly available dataset had already been smoothed. We identified left and right heel-strikes and toe-offs as the positive and negative peaks, respectively, of the anterior-posterior left or right ankle markers relative to the torso [ 39 ].

All digital video data were processed in two steps: 1) using OpenPose to automatically detect and label two-dimensional coordinates of various anatomical keypoints, 2) post-processing in MATLAB using custom-written code. The OpenPose analysis was similar for all video data, whereas we divided the post-processing workflows into two separate pipelines for videos capturing frontal or sagittal plane views.

• We ran the OpenPose demo over sequences of the video recordings that contained each walking bout. We have previously used a cloud-based service to run OpenPose with remote access to GPUs. Here we used a local computer with a GPU (NVIDIA GeForce RTX 3080) so that videos containing identifiable participant information were not shared with any third-party services.
• Videos were analyzed in OpenPose using the BODY_25 keypoint model that tracks the following 25 keypoints: nose, neck, mid-hip and bilateral keypoints at the eyes, ears, shoulders, elbows, wrists, hips, knees, ankles, heels, halluces, and fifth toes.
• The output of the OpenPose analysis yielded: 1) JSON files for every video frame containing pixel coordinates of each keypoint detected in the frame, and 2) a new video file in which a stick figure that represents the detected keypoints is overlaid on the original video recording.
• We changed the pixel coordinate system so that the positive vertical was directed upward and that positive horizontal was directed toward the participant’s left side.
• We visually inspected and corrected errors in left-right identification of the limbs. In all, 362 (less than 1% of the 131,519 frames in total) frontal video frames were corrected.
• We gap-filled keypoint trajectories using linear interpolation for gaps spanning to up 0.12 s.
• We identified events of left and right gait cycles by local maxima and minima of the vertical distance between the left and right ankle keypoints. Gait events on the left limb were detected at positive peaks and gait events on the right limb were detected at negative peaks in trials where the participants walked away from the frontal plane camera; and vice versa in trials where the participants walked toward the camera. In order to unify the nomenclature of gait events across motion capture data and sagittal and frontal plane video data, we refer to the gait events of the frontal plane analysis as heel-strikes.

• We changed the pixel coordinate system so that positive vertical was direction upward and positive horizontal was the direction of travel.
• We visually inspected and corrected errors in left-right identification of the limbs. In all, 5,369 (about 3.5% of the 153,669 frames in total) of sagittal video frames were corrected.
• We gap-filled keypoint trajectories using linear interpolation for gaps spanning up to 0.12 s.
• We smoothed trajectories using a zero-lag 4 th order low-pass Butterworth filter with a cutoff frequency at 5 Hz.
• We calculated a scaling factor to dimensionalize pixel distance. The scaling factor was as a ratio of a known distance in the line of progression relative to the pixel distance. We used the distance between strips of tape on the walkway.
• We identified left and right heel-strikes and toe-offs as the positive and negative peaks, respectively, of the horizontal trajectories of the left or right ankle keypoints relative to the mid-hip keypoint.

We cross-referenced gait events that had independently been identified in motion capture data and sagittal or frontal plane video data to ensure that all gait parameters were obtained based on the same gait cycles.

We calculated the following spatiotemporal gait parameters:

• Step time: duration between consecutive bilateral heel-strikes.
• Step length (we used two methods to calculate step lengths): 1) as the horizontal distance between ankle markers or keypoints at instants of heel-strike and 2) as the distance travelled by the torso between consecutive bilateral heel-strikes. We used the distance travelled by the torso because the distances between the ankles at a heel-strike instant cannot be obtained from frontal plane videos. When comparing step lengths between motion capture and sagittal plane videos, we used the distance between the ankles; all step length comparisons with frontal plane data used the distance travelled by the torso. Step length methods were highly correlated ( r = 0.938) with an average difference of −0.069 m, suggesting that the distance travelled by the torso was about 7 cm longer than the distance between the ankles ( S5 Fig ).
• Gait speed: step length divided by step time.

In stroke and PD data, we calculated paretic/non-paretic or left/right step times and step lengths, respectively. Paretic/left step time is the duration from non-paretic/right heel-strike until paretic/left heel-strike; vice versa for non-paretic/right step times. Paretic/left step length, calculated as the distance between the ankles, is the distance at paretic/left heel-strike; vice versa for non-paretic/right step lengths. Paretic/left step length, calculated as the distance travelled by the torso, is the distance travelled from non-paretic/right heel-strike to paretic/left heel-strike; vice versa for non-paretic/right step lengths.

We calculated the changes in spatiotemporal gait parameters that accompany speed-changes (i.e., shorter step times, longer step lengths, and faster gait speeds) from the preferred and fast speed trials in the stroke and PD data. This allowed us to test how well gait changes can be tracked using video recordings.

There are several commonly observed, clinically relevant gait impairments in stroke (e.g., gait asymmetry [ 41 ]) and PD (e.g., stooped posture [ 42 ])–thus, for each population we calculated condition-specific gait parameters. We calculated step time asymmetry and step length asymmetry (difference between steps divided by sum of steps) in stroke gait and trunk inclination in PD gait. Trunk inclination was calculated as the angle relative to vertical between the mid-hip and neck keypoints at heel-strikes in the sagittal plane videos and the angle between the C7 and right PSIS markers at heel-strikes in the motion capture data. During initial comparisons we found an offset (mean±SD 12.0°±1.5°) between motion capture and sagittal plane video data; we subtracted a fixed offset of 12° from trunk inclination in the sagittal plane video data in order to create a better numeric comparison with the motion capture data. The offset is a consequence of the fact that video-based keypoints and markers track similar anatomical regions, but do not track the exact same anatomical locations [ 37 , 43 ].

We calculated sagittal plane lower limb joint kinematics at the hip, knee, and ankle using two-dimensional coordinates from the motion capture data and the sagittal plane video data. We used markers at the greater trochanter and lateral femoral epicondyles and keypoints at the hip and knee to calculate hip angles; markers at the greater trochanter, lateral femoral epicondyles and lateral malleoli and keypoints at the hip, knee, and ankle to calculate knee angles; markers at the lateral femoral epicondyles, lateral malleoli, and 5 th metatarsal and keypoints at the knee, ankle, and hallux to calculate ankle angles.

From our stroke and PD datasets, we compared gait parameters at three levels of comparisons: at the step level calculating parameters for individual steps, as averages across single gait trials, and at the session level calculated as averages across several gait trials. In total there were 2,684 individual gait cycles (1,790 for stroke, 709 for PD and 185 for unimpaired), 527 gait trials (352 for stroke, 144 for PD and 31 for unimpaired) and 124 session level averages (88 for stroke and 36 for PD). We present session level gait parameters for stroke and PD and trial level for unimpaired data in the main text of the manuscript; we show results at the trial and step level in the S3 and S4 Tables.

In the stroke and PD datasets, we compared gait parameters obtained during trials that were simultaneously recorded by motion capture, sagittal plane videos, and frontal plane videos (see Fig 1 for overview). Note that some parameters (joint kinematics and trunk inclination) can only be obtained with motion capture data and sagittal plane videos.

In the dataset with unimpaired participants, we compared spatiotemporal gait parameters obtained during trials that were simultaneously captured with motion capture data and with two frontal cameras positioned to capture the participant walking away from one camera and toward the other camera (see Fig 3A for overview).

Statistical analyses

We compared gait parameters obtained with motion capture and video by calculating differences, errors (absolute differences) and 95% limits of agreement (mean differences ± 1.96 × SD). We assessed correlations by calculating Pearson correlation coefficients.

Supporting information

S1 fig. step length errors and differences of frontal plane workflow relative to person’s distance to camera..

https://doi.org/10.1371/journal.pdig.0000467.s001

S2 Fig. Step length errors and differences of frontal plane workflow relative to magnitude of step length.

https://doi.org/10.1371/journal.pdig.0000467.s002

S3 Fig. Influence of gait event timings on step length errors when using frontal plane workflow.

https://doi.org/10.1371/journal.pdig.0000467.s003

S4 Fig. Evaluation of tracking methods and smoothing using frontal plane workflow.

https://doi.org/10.1371/journal.pdig.0000467.s004

S5 Fig. Comparison of two methods to calculate step length.

https://doi.org/10.1371/journal.pdig.0000467.s005

S1 Table. Comparison of spatiotemporal gait parameters of the unimpaired group.

https://doi.org/10.1371/journal.pdig.0000467.s006

S2 Table. Spatiotemporal gait parameters for stroke and PD groups.

https://doi.org/10.1371/journal.pdig.0000467.s007

S3 Table. Comparison of spatiotemporal gait parameters of stroke and PD groups calculated as trial averages.

https://doi.org/10.1371/journal.pdig.0000467.s008

S4 Table. Comparison of spatiotemporal gait parameters of stroke and PD groups calculated for individual steps.

https://doi.org/10.1371/journal.pdig.0000467.s009

• View Article
• Google Scholar
• PubMed/NCBI
• 4. Perry J. & Burnfield J. M. Gait Analysis: Normal and Pathological Function. 2nd ed. New Jersey: SLACK Inc. (2010).

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 15 February 2024

Validation of portable in-clinic video-based gait analysis for prosthesis users

• Anthony Cimorelli 1 ,
• Ankit Patel 2 , 3 ,
• Tasos Karakostas 1 , 4 &
• R. James Cotton   ORCID: orcid.org/0000-0001-5714-1400 1 , 4  

Scientific Reports volume  14 , Article number:  3840 ( 2024 ) Cite this article

667 Accesses

1 Citations

1 Altmetric

Metrics details

• Biomedical engineering
• Bone quality and biomechanics
• Outcomes research

Despite the common focus of gait in rehabilitation, there are few tools that allow quantitatively characterizing gait in the clinic. We recently described an algorithm, trained on a large dataset from our clinical gait analysis laboratory, which produces accurate cycle-by-cycle estimates of spatiotemporal gait parameters including step timing and walking velocity. Here, we demonstrate this system generalizes well to clinical care with a validation study on prosthetic users seen in therapy and outpatient clinics. Specifically, estimated walking velocity was similar to annotated 10-m walking velocities, and cadence and foot contact times closely mirrored our wearable sensor measurements. Additionally, we found that a 2D keypoint detector pretrained on largely able-bodied individuals struggles to localize prosthetic joints, particularly for those individuals with more proximal or bilateral amputations, but after training a prosthetic-specific joint detector video-based gait analysis also works on these individuals. Further work is required to validate the other outputs from our algorithm including sagittal plane joint angles and step length. Code for the gait transformer and the trained weights are available at https://github.com/peabody124/GaitTransformer .

Similar content being viewed by others

Walking naturally after spinal cord injury using a brain–spine interface

Henri Lorach, Andrea Galvez, … Grégoire Courtine

Semantic reconstruction of continuous language from non-invasive brain recordings

Jerry Tang, Amanda LeBel, … Alexander G. Huth

An asynchronous wireless network for capturing event-driven data from large populations of autonomous sensors

Jihun Lee, Ah-Hyoung Lee, … Arto Nurmikko

Introduction

Gait impairments are a common target for rehabilitation. The most widely used outcome measures are the 10-m walk test or 6-min walk test which measure walking speed and endurance 1 , but do not capture more detailed walking biomechanics. Alternatively, a motion analysis laboratory uses optical motion capture and force plates to obtain precise estimates of joint kinematics and kinetics and compute temporal and spatiotemporal gait parameters 2 . However, the cost, time, and equipment required for formal gait analysis preclude performing it frequently, during clinical encounters, or outside the laboratory. There is a substantial need for clinically-usable tools that fill the gap between formal gait assessment and performance-based outcome measures to enable routine, quantitative characterization of gait and its associated impairments. These tools would enable more sensitive outcome measures to follow patients’ progress with therapy as well as to better power research to improve interventions. It would also enable routine screening of gait parameters in the clinic. This has many potential applications, such as allowing early detection of gait parameters associated with a risk of falling 3 , which could then enable earlier interventions with fall prevention strategies.

Given the importance of analyzing gait and mobility, numerous approaches have been explored including wearable sensors and video analysis. Many different types of wearable sensors and algorithms have been described ranging from wrist devices that estimate step counts to placing numerous sensors over the body to estimate more complete kinematics 4 , 5 , 6 . An advantage of wearables is they can enable ubiquitous monitoring of activity throughout the day 7 . However, they can often require extensive and time-consuming calibration, require complicated and often proprietary algorithms to extract the relevant information, and have a wide range of reported accuracies 8 .

In recent years, human pose estimation using deep learning has seen an extremely rapid advance yielding numerous approaches that can estimate 3D joint locations and pose 9 , 10 , 11 . However, the performance measure for the majority of this computer vision research is the accuracy of estimating 3D joint locations and not biomechanically motivated kinematics or gait parameters. Using multiple cameras, joint locations can be triangulated in 3D to produce more accurate estimates and these systems have been validated on numerous aspects of gait 12 , 13 , 14 , 15 , 16 , 17 , 18 . In general, the need for multiple cameras makes these systems less portable and amenable for use in a clinic, although OpenCap has shown it is possible using only a computer and two smartphones with calibrated positions 19 . Approaches have been developed that can analyze cycle-by-cycle gait from monocular video 20 , 21 , but they have not been validated on data acquired in the clinic or on clinical populations. Alternatively, other approaches train a neural network to directly map a sequence of 2D keypoints to average gait parameters that have been tested on clinical populations but do not enable analyzing individual gait cycles 22 , 23 .

We developed an algorithm, the Gait Transformer, trained on a large clinical gait laboratory dataset of paired videos and motion capture data 24 . This Gait Transformer decomposes Human Pose Estimation (HPE) trajectories of walking into individual gait cycles to produce accurate estimates of gait event timing and walking velocity, when tested on that dataset . However, artificial intelligence (AI) algorithms can be sensitive to changes in the data distribution. In this case, there are numerous differences between videos of gait collected in the real world or clinic and those from the gait laboratory. Thus, validating how this algorithm generalizes to data collected in clinical settings—the primary goal—is critical to enabling its use. The goals of this work include: (1) describe our combined system including smartphone application, wearable sensors, Pose Pipeline 25 and Gait Transformer 24 for clinical gait analysis, (2) validate the performance of this system on data acquired in the clinic, (3) identify under which conditions it performs well and when it is less reliable.

We performed this validation on prosthesis users and selected this population for several reasons. From a technical perspective, this is a challenging test of this system as the limbs of prosthesis users often appear visually different than able-bodied individuals, and it was previously unknown if pretrained HPE algorithms will generalize to prosthesis users. In addition, some prosthesis users walk with significant gait deviations compared to able-bodied individuals 26 , which further challenges the Gait Transformer. From a clinical perspective, we expect that routine access to video-based gait analysis would enable better outcomes to monitor improvements in walking with therapy or with adjustments to prosthetic components, but this will ultimately need to be empirically validated through clinical trials. Previous studies have demonstrated that lower limb prosthesis users saw an improvement in various gait parameters including walking speed, distance walked, spatiotemporal measures and biomechanics with specialized therapeutic interventions 27 , 28 . However, these studies used performance-based outcome measures and those that analyzed biomechanics were limited to laboratory settings over level ground. The ability to perform routine, quantitative gait analysis could identify improvements in the quality of walking during therapy, such as increased symmetry and time spent in single stance on the prosthetic limb. This increased level of detail for analyzing gait may allow for more sensitive outcome measurements to demonstrate how an individual’s gait continues to improve when their walking velocity plateaus. This could enable physical therapists to demonstrate to insurance agencies that patients are still making progress and justify additional sessions. Or, more sensitive measures could provide data for prosthetists to show advanced components also improve quality of walking. As prosthetic components continue to advance, it is becoming increasingly difficult to get high-end prosthetic components covered by payers and there is a need to develop systems that can quantify prosthetic gait in real-world settings 29 . Due to the many limitations associated with current gait analysis systems, prosthetists typically rely on observational gait analysis when performing dynamic alignment of a prosthetic device 30 . Therefore, having access to routine quantitative gait analysis could assist prosthetists with dynamic alignment during clinical visits.

Data collection

Mobile acquisition and wearable sensors.

Video and sensor data is acquired on an Android smartphone using a custom app to synchronize recording from both modalities (Fig.  1 ). The mobile phone is mounted on a 3-axis gimbal (DJI Osmo Mobile 2) to improve the video stability when following subjects during ambulation. The video was obtained in portrait orientation at 1080 × 1920 resolution at 30 frames per second.

The wearable sensors are a custom design and have previously been described 31 , 32 . For the purpose of this study, the wearable sensor data was used solely for validation. The sensors stream data from the IMU to the smartphone over Bluetooth Low-Energy (BLE). The IMU is an ICM-20948 and outputs 3-axis accelerometer and gyroscope data at 562.5 Hz. Magnetometer data is available but is not used in this study and we typically do not stream it in order to optimize the BLE bandwidth. The sensors can also acquire two channels of EMG data, but this was not used in these experiments. Prior to experiments, the magnetometer is calibrated in the location where data will be collected by rotating around each of the axes and accounting for hard and soft iron distortions. IMU data is fused on the sensor using a complementary filter to compute the 3D orientation, and the estimated orientation is streamed over BLE. Compared to our prior works using silicone-encapsulated wearable sensors attached to the skin, in this work the sensors were placed in a 3D-printed case with Velcro \(\circledR\) on the outside for attachment to the body.

( A ) Android cell phone, two wearable sensors and gimbal. ( B ) One sensor placed on the shank of a definitive transtibial prosthesis with adhesive Velcro \(\circledR\) . ( C ) Two sensors placed on the shank/thigh of a diagnostic transfemoral. ( D ) Screenshot from gait analysis app, with video in the top panel and sensor data in the bottom panel.

Clinical population and data annotation

This study was approved by the Northwestern University Institutional Review Board. All methods were performed in accordance with the relevant guidelines and regulations. All participants gave written consent. Participants shown in images in this manuscript gave additional consent for the use of their images and videos in scientific publications and their faces were masked. Video of gait and other activities was obtained from a convenience sample of lower-limb prosthesis users seen in an outpatient prosthetics clinic or participating in outpatient physical therapy at Shirley Ryan AbilityLab. For each subject, we recorded their age, height, level, and (bi)laterality of amputation, the etiology of the amputation, the type of prosthetic components, and their Medicare Functional Classification Level (K-level). Walking data was collected in either prosthetic or therapy clinics. In the case of therapy, videos were obtained as participants performed their usual therapy.

Sensors were placed on both the shank and thigh of the prosthetic limb(s), with up to four sensors used in the case of individuals with bilateral amputations. They were attached either with double-sided Velcro \(\circledR\) attached to the prosthetic pylon, and socket (for transfemoral amputees) or with a Velcro \(\circledR\) strap around the thigh. They were placed laterally, with the IMU Z axis pointed laterally and the X axis pointed down. In this work, only the data from the shank IMU was used to detect the prosthetic limb swinging.

After data collection was complete, videos were annotated with the following categories by the author: the activity being performed (e.g., overground walking, treadmill walking, running, other therapeutic tasks), the view (frontal, sagittal, or a mixture), and the subjective accuracy of keypoint tracking of the prosthetic limb with a 3-point Likert scale ranging from 1 (poor) to 3 (good). Specifically 1 indicated the keypoints do not locate the prosthetic joints and tracking is frequently inaccurate, 2 indicated they locate the joint but are intermittently inaccurate, and 3 indicates the locate the joint well throughout the video. The individual performing the manual annotation was the main author and they were blinded to the keypoint confidence predictions prior to annotation. Whether the prostheses were visible or occluded by clothing was also annotated for each video. Time points when the participant entered and exited 10-m areas indicated by tape on the ground were also recorded to compute ground truth velocity.

Velocity annotation

Tape was placed on the ground at 10-m spacing in locations where subjects would typically walk such as the hallway in the prosthetic clinic and in multiple locations in the therapy gyms. The start and end times when subjects completed a straight overground walk between these markers were retrospectively annotated. This provided both ground truth measurements of their walking velocity and identified specific video segments where participants were walking, as the collected data contained a mixture of activities.

In this work, we focus our analysis on video segments where individuals are performing above-ground walking in a forward direction during the time window where they were walking between the two pieces of tape, which we refer to as timed walking segments. We focus on the 10-m annotation periods both to determine the accuracy of the system when computing gait speed and because these were segments where individuals were known to be walking and were in view of the camera.

Data processing

The input to the gait transformer is a series of 3D keypoint locations. To obtain these, the video was processed with PosePipe 25 , a human pose estimation pipeline based on DataJoint 33 that simplifies running cutting-edge HPE algorithms on video. The steps used in the pipeline include (1) a tracking algorithm 34 to compute bounding box tracks for all people in the scene followed by (2) manually annotating the bounding box for the subject of interest undergoing gait analysis. (3) Then we perform top-down 2D keypoints detection in each frame using the MMPose toolbox 35 , specifically using an HRNet 36 trained on the COCO dataset 37 using distribution aware relative keypoints 38 , (4) the 2D keypoint trajectories are then lifted to 3D joint locations 39 .

We also used DeepLabCut (DLC) 40 to train a custom 2D keypoint detector for a subset of prosthesis users with videos where keypoint detection was performing poorly. This is completed by manually annotating both intact and prosthetic hip, knee and ankle joints. We computed the 2D keypoints on those same videos with this model and replaced any prosthetic joints computed by MMPose with the estimates from DLC. This corrected set of keypoints was then passed to the lifting step and then to the gait transformer.

We flagged any frames as clipped whenever any of the keypoints of the leg came within 10 pixels of the edge of the screen.

Gait transformer

The sequence of 3D joint location was mapped onto the relative timing of four gait events (right and left foot contact and toe off) as well as the pelvis velocity using the Gait Transformer 24 . This is trained on a large dataset of walking videos with synchronous marker-based motion capture and force plate data from our clinical gait laboratory. For training, data was aligned in the sagittal direction using the medial orientation of the pelvis. It also outputs sagittal plane joint kinematics including foot position relative to the pelvis, foot velocity, and hip and knee angles, all of which we do not focus on in this work. We refer the reader to the 24 for details of the architecture and training of the gait transformer. Compared to that work, we retrained the gait transformer and excluded bilateral elbows and wrists as we found occasionally the use of assistive devices (i.e., canes or crutches) in a less common pattern would trigger false detection of steps.

The Gait Transformer was applied to the lifted 3D joint trajectories. Training samples from the gait laboratory are typically only a few gait cycles long and we found that it did not generalize to inference on much longer sequences. Videos acquired in the clinic were much longer than a few gait cycles, so we applied it on a sliding window of 90 frames (3 s), which covers at least one complete stride for the majority of subjects. For each position of the sliding window, we used the middle output other than the beginning and end where we used the corresponding half of the sequence to pad the output.

Sensor processing

In order to validate the accuracy of the gait transformer event timing of the prosthetic limb, we used gyroscope data from the wearable sensor on the prosthetic shank to detect the prosthetic-limb swing phase. Sensor data is timestamped to the smartphone time. Gyroscope data were sampled at a nominal sampling rate of 562.5 Hz. The Android system time of each Bluetooth packet is also stored and linear regression is used to calibrate the sensor timebase against the Android time, typically with an updated sampling rate of 1–2 Hz different than the nominal value.

We also noted that the video start timestamp showed some latency compared to the sensor timestamps. This has been resolved in more recent versions of our smartphone application with an API that acquires a more precise, per-frame timestamp. We used the hip and knee sagittal plane angles from the gait transformer to adjust for this timing error by finding the offset that minimized the mean squared error between the gyroscope on the shank and the change in that angle computed from the gait transformer outputs. This was typically around 170 ms.

We detected a prosthetic limb swing from the gyroscope on the prosthetic shank. With sensors placed on the lateral side of the shank, the z-axis is roughly aligned to be perpendicular to the sagittal plane. We applied an 8th-order low-pass Chebyshev filter to the gyroscope with a cutoff of 35 Hz. All negative values were zeroed out and a median filter of 360 ms was applied to deglitch a few strides where participants caught their toe and a brief reversal was seen in the gyroscope midtrace. These deglitched positive segments were identified as swings with the time the sign became negative identified as the start and end of swing periods.

Sensor versus video cadence

We compared the cadence estimated with the gait transformer over 10-m walking segments to that computed from the sensors. We computed the cadence from video by averaging \(\dot{\omega }\) over the timed walking segment and converting from strides in rad/s to steps/min (i.e.  \(c_v=\frac{120}{2\pi (t_e-t_s)}\int _{t_s}^{t_e} \dot{\omega }(t) \, \partial t\,\) ). We computed cadence from the sensors using the average stride time of the prosthetic limb side over the steps in the time window: \(c_s=120 / \frac{s_j-s_i}{j-i}\) , where \(i\) and \(j\) index the first and last sensor swing time, respective, that fall within the timed walking period.

Matching video and sensor foot contacts

We also compared the time when the end of swing from the gyroscopes to the foot contact time from the gait transformer. This analysis only reflects a bound, because detecting when the prosthetic shank stops rotating forward in swing (i.e., when the gyroscope swaps signs) approximates the end of swing time but is not the actual time the foot contacts the ground and will typically occur slightly before true foot contact (Fig.  2 ). For each walking segment, we computed the offset that produced the closest matches between the end of prosthetic leg swings detected by the sensors and the time of prosthetic foot contact estimated from the Gait Transformer and Kalman filter (typically about 100ms). After this we measured the foot contact detection accuracy as the residual timing error between the offset sensor times and the detected video times. We also measured the fraction of events that were detected with a window of 500 ms.

Example gyro and video timing information. Top trace is the z-component of the gyroscope mounted on the tibial shank. The time point where the gyroscope goes from positive to negative (where heel stops rotating forward relative to knee) is identified as a close proxy for foot contact. Bottom plot shows the sin component of the quadrature output for heel strike, with the positively directed zero crossing marker. There is a close correspondence between detected foot contacts from the gyro (blue) and the video (yellow). Vertical bars correspond to the annotated boundaries of the 10-m walk test.

Participant demographics

From 19 participants, we annotated 231 timed walking segments during level walking, with 79 in the frontal plane, 67 in the sagittal plane and the remainder either oblique or changing. When restricted to only the frontal plane, timed walking segments were obtained from a total of 16 participants (Table 1 ).

Usability of the system

Our system made it easy to collect gait data in clinical situations, including physical therapy and prosthetic appointments. Wearable sensors took less than 30 seconds to apply and remove and would require no setup time if not using sensors for validation studies. The sensors connect to the cell phone through our Android app, so acquiring data is as easy as pressing the video record button in our app. Currently, running the analysis pipeline requires technical skills, but we hope to fully automate this in the future. Data was easily and routinely obtained from subjects in both therapy and prosthetic appointments. This was particularly true for the frontal view, but obtaining clear sagittal views in hallways was often challenging due to space limitations. No subjects withdrew from the study. Example visualizations from the system are shown in Figs.  3 and 4 .

Example visualizations from the gait transformer. Keypoints on the ankle become colored red when in contact with the ground. Overlaid traces show the foot position, foot velocity, pelvis velocity, hip and knee angles, and the quadrature encoded gait timing.

Traces from 5 s of walking from the portable system. The left column shows the outputs from the Gait Transformer. Traces from the left leg are shown in blue and from the right leg are shown in red. The additional dashed traces in the foot-down plot correspond to the phase-shifted quadrature encoding. The right column shows the raw sensor data from the thigh and shank. For gyro and accel plots, the three colors correspond to the x, y, and z IMU axis. For the quaternion plots, there are four colors for the individual quaternion components. Blue and red dots indicate the foot down events on the left and right, respectively.

2D keypoint accuracy

Our 2D keypoint detection used an MMPose model, which is trained on the COCO dataset and primarily contains able-bodied individuals. We found that it frequently generalized to prosthesis users, but not always, and that quality of 2D keypoint tracking varied across subjects. In some cases, the detector would eschew localizing the prosthetic joint of the subject of interest and track the corresponding joint of a nearby therapist. While 2D keypoint detectors, including MMPose, output a per-joint confidence estimate, this has not been systematically tested for prosthesis users. For each video, we manually annotated our subjective estimate of tracking quality from 1 (poor) to 3 (good). For each timed walking segment, we also computed the average confidence that MMPose reported for the prosthetic ankle. As the 2D keypoint algorithms have primarily been trained on able-bodied individuals, we first wanted to analyze the accuracy of the keypoint confidence values on prosthetic limbs compared to manual annotation. Figure  5 a shows the histogram of ankle keypoint qualities, which demonstrates that the confidence estimates from the keypoint detector align with our manual annotation. In fact, the interquartile ranges of ankle confidence conditioned on each of our annotated qualities were nearly non-overlapping (IQR for keypoint quality 1: [0.44,0.60]; 2: [0.59,0.80], 3: [0.77,0.87]).

After validating that the ankle joint confidence is valid in prosthesis users, we then conditioned on the prosthetic level and whether clothing was covering the prosthesis or not, which revealed two trends (Fig.  5 b). First, ankle tracking was worse for bilateral prosthesis users compared to unilateral prosthesis users. Secondly, among unilateral prosthesis users, ankle tracking was worse for those with higher levels of amputation. Ankle tracking was particularly poor in the case of a prosthesis user participant with a bilateral transfemoral amputation and another with a hip disarticulation.

Accuracy of ankle keypoints for prosthesis users with off-the-shelf 2D keypoint detection algorithm. ( A ) shows the histogram of average ankle qualities for videos stratified by manual annotation of quality (with 3 being the best), showing that the estimated quality corresponded with our annotation. The ( B ) shows the average ankle quality stratified by prosthetic level and whether the prostheses were covered by clothing. TT transtibial, TF transfemoral, HD hip-disarticulation, BL bilateral.

In comparison to the MMPose model trained on COCO 37 , our DLC model that was specifically trained on our prosthetic user was able to perform much better, Fig.  6 . Note that we did not test the generalization of this model to new users, and all prosthesis users we analyzed with our DLC model had 20 frames manually annotated and were included in the training dataset.

Examples of poor keypoint detection using pretrained algorithm (top row) that were corrected by training with DLC (bottom row).

Viewpoint sensitivity and clipping

We also classified segments as clipped or not if any of the 2D keypoints of the leg hit the edge of view on more than 1% of the frames, because we found the transformer was sensitive to the errors this produced. This occurred in none of the 79 frontal views, in 16 of the 67 sagittal views, and in 7 of the 85 mixed or oblique views. This was due to the somewhat limited space in clinical settings, where in hallways it can be hard to track far enough to the side of a person of interest to frame them with room for error.

Velocity and cadence accuracy

Accuracy of gait transformer for velocity (top row) and cadence (bottom row) under different conditions (columns). Text insets note the correlation coefficient (r), mean absolute error (MAE), and number of walking segments (n). The first column is all videos in the frontal plane. The second column is those where the average ankle detection quality was greater than 0.7. The third column is the excluded segments reprocessed with DLC. The last are videos acquired in the sagittal plane.

We compared the velocity estimated from the video acquired in the frontal plane with the gait transformer to the velocity computed from the manually annotated times as participants walked over a 10-meter interval between the tape on the ground. We found the gait transformer velocity for videos acquired in the frontal plane was quite accurate compared to ground truth, with a correlation (r) of 0.95 and a mean absolute error (MAE) of 0.14 m/s (Fig.  7 a). We also found that the cadence from the gait transformer was a close fit to the sensor data with an MAE of 4.8 steps/min, with most of this error coming from outliers (Fig.  7 e).

We repeated this analysis, excluding segments where the average prosthetic ankle quality was less than 0.7 (Fig.  7 b,f). Notably, we found this removed almost all the error estimating cadence (r = 0.998, MAE = 1.0). The accuracy of the estimated velocity also improved slightly (MAE = 0.13 m/s). We then used our custom prosthetic keypoint detector trained in DLC to replace the prosthetic joints in the excluded segments (Fig.  7 c,g). We found this improved cadence detection for most sessions, but in particular, there were still several outliers for one individual with a slower cadence compared to other individuals.

We found the performance was quite poor when tested on video acquired in the sagittal plane (Fig.  7 d,h). This was relatively unsurprising since the gait transformer is trained on video acquired in the frontal plane. We attempted to improve the viewpoint invariance by augmenting the training process by randomly rotating the lifted 3D keypoints. We found that while it resulted in a tighter correlation with the outputs, both in the gait laboratory validation dataset and on the prosthetic data it resulted in a bias that underestimated velocity at faster speeds.

Step time accuracy

Over the frontal view timed walking segments, we detected and matched 889 prosthetic foot contact events and were unable to detect only 12 of these events. For each walking segment, we also computed the mean absolute error of the residuals. The average of this over all sessions was 72 ms. For sessions with good ankle tracking, we detected 490 foot contact events and missed only 1 event, with an average error over sessions of 45 ms.

Previously, we trained an algorithm for video-based gait analysis on a large clinical gait laboratory dataset of paired videos and motion capture data 24 . While the algorithm validated well on this dataset, there are differences between the training data and the intended application of the algorithm that could result in poor generalization. These include recording portrait videos while walking with the patient through a therapy clinic where other people are present. Our training dataset also had a predominance of children, who are most commonly analyzed in these laboratories. For any clinical application of AI, it is critical to evaluate the external validity (or out-of-domain generalization in machine learning parlance) for the intended use case.

In this study, we evaluated the performance of our gait transformer when tested on prosthesis users walking in therapy or outpatient clinic. This was a powerful stress test for our algorithm, as there are several properties of prosthesis users that might cause a failure to generalize including the visual appearance of prostheses and prosthetic gait patterns. This study also highlighted the power of an interpretable pipeline, with understandable features at multiple stages, such as 2D keypoint detection accuracy or failures to detect a single step. In development, this aspect of the pipeline was extremely important as it enables identifying and alleviating points of failure throughout the different stages of the pipeline to ensure we can trust the outputs from the transformer and make adjustments when necessary to improve accuracy.

Large datasets are a critical driving force of AI algorithms and have enabled impressive advances in HPE in recent years. However, public HPE datasets, like COCO 37 , primarily contain able-bodied individuals. Perhaps unsurprisingly, we found that a 2D keypoint detector trained on COCO did a poor job detecting the ankles of some prosthesis users. This was most pronounced with more proximal or bilateral amputations. This is most likely due the greater visual difference between these limbs and joints compared to the able-bodied individuals in the dataset, resulting in worse out-of-domain generalization of the algorithms. In these cases, tools like DLC 40 make it relatively easy to train a custom keypoint detector. However, using DLC is still labor intensive as it requires manual annotation of each video. Future work will look to improve 2D keypoint detection for prosthetic limbs using self-supervised-learning. Towards this goal, we recently developed a markerless motion capture system and validated it on prosthesis users 41 , 42 Importantly, this work also speaks to the need for more work on AI fairness for people with disabilities 43 , 44 .

When we removed the videos with poor ankle tracking quality and only included videos where the 2D keypoints were accurately localized, we found that our algorithm performed well on videos of prosthesis users for video acquired in the frontal plane. Specifically, we had a mean absolute error for the velocity of 0.13 m/s and for the cadence of 1 step/min. Whether this is sufficiently accurate ultimately depends on the clinical question. One study found the minimally clinically important difference for walking velocity of prosthesis users as 0.21 m/s 45 , which is greater than our algorithm. However, for older adults, it has been suggested a small meaningful change in walking speed is 0.05 m/s 46 . Our results also do not indicate whether analyzing longer segments of walking would reduce the error, or whether this arises from a bias for given individuals that would persist over longer recordings.

The requirement to record video in the frontal plane to obtain accurate results is a limitation of our current approach. However, given the difficulties obtaining good videos in sagittal planes while walking in hallways and therapy gyms with other people present, it is also the most convenient approach for its intended use setting. An argument for sagittal videos is that they should enable more accurate estimates of many important sagittal plane kinematics. Our system outputs many of these and performs well on the training data but testing the external validity of these outputs is important future work. Specific to prosthesis users, testing the external validity of joint kinematics will be an area of challenge as prosthetic components attempt to mimic anatomic motion but do not move in an exact manner as able-bodied joints. As the gait transformer was primarily trained on individuals with intact limbs, this may affect the ability to accurately estimate prosthetic joint kinematics.In this study we validated the accuracy of the gait transformer’s velocity, cadence and step timing measures as these were measures we could easily validate with wearable sensors in the clinic. However, our long-term goal is to develop a system that can accurately output several spatiotemporal and kinematic gait parameters to allow clinicians to further quantify the quality of gait. Further validation to determine the external validity of our portable gait analysis system to measure other spatiotemporal and kinematic gait parameters in prosthesis users is a high priority, which will leverage our recently developed markerless motion capture system 41 , 42 .

Several other groups have looked at the ability to estimate gait parameters from monocular video. Stenum and colleagues analyzed sagittal videos from a dataset 47 of 32 healthy individuals walking down a walkway a fixed distance from a mounted camera using OpenPose, and showed they could accurately estimate gait event timing, sagittal plane joint angles, step length and walking velocity 20 . A pipeline similar to ours using 2D keypoints followed by lifted 3D joints was used with height-informed skeletal refinement prior to extracting gait parameters of healthy individuals walked towards a stationary camera and produced accurate estimates of step timing, length, and walking velocity 21 . Similarly to our work, both of these computed features on individual gait cycles, but in comparison, they were not externally validated on clinical populations or on video acquired in the community or clinic. A neural network can also be trained to directly map 2D keypoint trajectories from the video onto average gait parameters, and such an approach has been used on gait laboratory data from children 22 and subsequently we showed a similar approach with stroke survivors 23 . However, this approach does not allow for examining the cycle-by-cycle variability of gait parameters. Finally, by triangulating keypoints detected from multiple cameras it is possible to estimate more accurate 3D keypoint locations and perform inverse kinematics to fit biomechanical models to them 16 , 17 , 41 , 42 . OpenCap is a particularly portable version of this that only requires two iPhones 19 .

There are several future directions that we anticipate will improve this system. One is better fusion with additional modalities of data, including the camera depth channel and inertial measurements from the wearable sensors. In this work, we have focused purely on monocular video as this is the most widely available modality. We are also enthusiastic about integrating physics-based modeling in the inference process 48 , 49 , 50 , 51 and ways to combine this with self-supervised learning 52 . It is also important to make this system easier to use in a higher-throughput manner. One need is to automate the annotation step to robustly identify the subject of interest. A potential solution is through the use of QR codes placed on the participant during data collection which would allow the computer vision to identify the person of interest without requiring manual annotation. In this work, we used the periods where 10-meter walk tests were annotated to select what to analyze, but this is only a small fraction of the data we acquired. Utilizing activity recognition to identify when the subject is walking, using the detected pose to identify viewpoint, and the Kalman error to determine when walking is being reliably tracked could help automate analysis of the remainder of the data. We recently developed a 3D lifting algorithm that produces calibrated confidence estimates of the joint locations, and integrating this could also help determine the trustworthiness of outputs 53 . Finally, there are many other clinically meaningful gait parameters available in our dataset that we could train the gait transformer to output and validate, including step width and center of mass.

Data availability

The raw video datasets generated and/or analysed during the current study are not publicly available due to videos from clinical settings with faces visible. The 2D and 3D keypoint trajectories and the annotated velocities and cadences will also be made available from the corresponding author on reasonable request. Code for the gait transformer and the trained weights are available at https://github.com/peabody124/GaitTransformer .

Moore, J. L. et al. A core set of outcome measures for adults with neurologic conditions undergoing rehabilitation: A clinical practice guideline. J. Neurol. Phys. Ther. 42 , 174–220. https://doi.org/10.1097/NPT.0000000000000229 (2018).

Article   PubMed   PubMed Central   Google Scholar  

Richard, J., Levine, D. & Whittle, M. Whittle’s Gait Analysis 5th edn. (Elsevier, 2012).

Hamacher, D., Singh, N. B., Van Dieën, J. H., Heller, M. O. & Taylor, W. R. Kinematic measures for assessing gait stability in elderly individuals: A systematic review. J. R. Soc. 8 , 1682–1698. https://doi.org/10.1098/rsif.2011.0416 (2011).

Article   CAS   Google Scholar  

Prasanth, H. et al. Wearable sensor-based real-time gait detection: A systematic review. Sensors 21 , 2727. https://doi.org/10.3390/s21082727 (2021).

Article   PubMed   PubMed Central   ADS   Google Scholar  

Weygers, I. et al. Inertial sensor-based lower limb joint kinematics: A methodological systematic review. Sensors 20 , 673. https://doi.org/10.3390/s20030673 (2020).

Picerno, P. 25 years of lower limb joint kinematics by using inertial and magnetic sensors: A review of methodological approaches. Gait Posture 51 , 239–246. https://doi.org/10.1016/j.gaitpost.2016.11.008 (2017).

Article   PubMed   Google Scholar  

Rast, F. M. & Labruyère, R. Systematic review on the application of wearable inertial sensors to quantify everyday life motor activity in people with mobility impairments. J. NeuroEng. Rehabil. 17 , 148. https://doi.org/10.1186/s12984-020-00779-y (2020).

Klöpfer-Krämer, I. et al. Gait analysis: Available platforms for outcome assessment. Injury 51 (Suppl 2), S90–S96 (2020).

Zheng, C. et al. Deep Learning-Based Human Pose Estimation: A Survey (2020).

Liu, W. & Mei, T. Recent advances of monocular 2D and 3D human pose estimation: A deep learning perspective. ACM Comput. Surv. 2022 , 3524497. https://doi.org/10.1145/3524497 (2022).

Article   Google Scholar  

Muhammad, Z.-U.-D., Huang, Z. & Khan, R. A review of 3D human body pose estimation and mesh recovery. Digit. Signal Process. 128 , 103628. https://doi.org/10.1016/j.dsp.2022.103628 (2022).

Nakano, N. et al. Evaluation of 3D markerless motion capture accuracy using OpenPose with multiple video cameras. Front. Sports Active Living 2 , 50. https://doi.org/10.3389/fspor.2020.00050 (2020).

Pagnon, D., Domalain, M. & Reveret, L. Pose2Sim: An end-to-end workflow for 3D markerless sports kinematics—Part 1: Robustness. Sensors 21 , 6530. https://doi.org/10.3390/s21196530 (2021).

Needham, L. et al. The development and evaluation of a fully automated markerless motion capture workflow. J. Biomech. 144 , 111338. https://doi.org/10.1016/j.jbiomech.2022.111338 (2022).

Kanko, R. M., Laende, E., Selbie, W. S. & Deluzio, K. J. Inter-session repeatability of markerless motion capture gait kinematics. J. Biomech. 121 , 110422. https://doi.org/10.1016/j.jbiomech.2021.110422 (2021).

Kanko, R. M. et al. Assessment of spatiotemporal gait parameters using a deep learning algorithm-based markerless motion capture system. J. Biomech. 122 , 110414. https://doi.org/10.1016/j.jbiomech.2021.110414 (2021).

Kanko, R., Laende, E., Davis, E., Selbie, W. S. & Deluzio, K. J. Concurrent assessment of gait kinematics using marker-based and markerless motion capture. J. Biomech. https://doi.org/10.1101/2020.12.10.420075 (2020).

McGuirk, T. E., Perry, E. S., Sihanath, W. B., Riazati, S. & Patten, C. Feasibility of Markerless motion capture for three-dimensional gait assessment in community settings. Front. Hum. Neurosci. 16 , 867485 (2022).

Uhlrich, S. D. et al. OpenCap: 3D human movement dynamics from smartphone videos. PLoS Comput. Biol. 19 , e1011462. https://doi.org/10.1101/2022.07.07.499061 (2022).

Stenum, J., Rossi, C. & Roemmich, R. T. Two-dimensional video-based analysis of human gait using pose estimation. PLOS Comput. Biol. 17 , e1008935. https://doi.org/10.1371/journal.pcbi.1008935 (2021).

Article   CAS   PubMed   PubMed Central   ADS   Google Scholar  

Azhand, A., Rabe, S., Müller, S., Sattler, I. & Heimann-Steinert, A. Algorithm based on one monocular video delivers highly valid and reliable gait parameters. Sci. Rep. 11 , 14065. https://doi.org/10.1038/s41598-021-93530-z (2021).

Kidziński, Ł et al. Deep neural networks enable quantitative movement analysis using single-camera videos. Nat. Commun. 11 , 1–10. https://doi.org/10.1038/s41467-020-17807-z (2020).

Lonini, L. et al. Video-based pose estimation for gait analysis in stroke survivors during clinical assessments: A proof-of-concept study. Digit. Biomark. 6 , 9–18. https://doi.org/10.1159/000520732 (2022).

Cotton, R. J., McClerklin, E., Cimorelli, A., Patel, A. & Karakostas, T. Transforming Gait: Video-Based Spatiotemporal Gait Analysis. In 2022 44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC) 115–120. https://doi.org/10.1109/EMBC48229.2022.9871036 (2022).

Cotton, R. J. PosePipe: Open-Source Human Pose Estimation Pipeline for Clinical Research. arXiv:2203.08792 [cs, q-bio] (2022).

Krajbich, J. I., Pinzur, M. S., Potter, B. K. & Stevens, P. M. Atlas of Amputations and Limb Deficiencies: Surgical, Prosthetic, and Rehabilitation Principles (American Academy of Orthopaedic Surgeons, 2016). arXiv:NkN2xwEACAAJ .

Wong, C. K. et al. Exercise programs to improve gait performance in people with lower limb amputation: A systematic review. Prosthet. Orthot. Int. 40 , 8–17 (2016).

Article   PubMed   ADS   Google Scholar  

Highsmith, M. J. et al. Gait training interventions for lower extremity amputees: A systematic literature review. Technol. Innov. 18 , 99–113 (2016).

Gard, S. A. Use of quantitative gait analysis for the evaluation of prosthetic walking performance. J. Prosthet. Orthot. https://doi.org/10.1097/00008526-200601001-00011 (2006).

Brinkmann, P. M. &  Stevens, J. T. Clinical Considerations of Observational Gait Analysis. In Atlas of Amputations and Limb Deficiencies Surgical, Prosthetic, and Rehabilitation Principles 81–95 (2016).

Cotton, R. J. & Rogers, J. Wearable Monitoring of Joint Angle and Muscle Activity. In 2019 IEEE 16th International Conference on Rehabilitation Robotics (ICORR) 258–263. https://doi.org/10.1109/ICORR.2019.8779538 (IEEE, 2019).

Cotton, R. J. Kinematic Tracking of rehabilitation patients with markerless pose estimation fused with wearable inertial sensors. IEEE 15th International Conference on Automatic Face & Gesture Recognition (2020).

Yatsenko, D. et al. DataJoint: Managing big scientific data using MATLAB or Python . https://doi.org/10.1101/031658 (2015).

Wojke, N., Bewley, A. & Paulus, D. Simple Online and Realtime Tracking with a Deep Association Metric . arXiv:1703.07402 [cs]. arXiv:1703.07402 (2017).

Contributors, M. OpenMMLab Pose Estimation Toolbox and Benchmark (2020).

Sun, K., Xiao, B., Liu, D. & Wang, J. Deep high-resolution representation learning for human pose estimation. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition 5686–5696. https://doi.org/10.1109/CVPR.2019.00584 (2019).

Lin, T.-Y. et al. Microsoft COCO: Common Objects in Context. In Computer Vision - ECCV 2014 Lecture Notes in Computer Science (eds Fleet, D. et al. ) 740–755 (Springer International Publishing, 2014).

Zhang, F., Zhu, X., Dai, H., Ye, M. & Zhu, C. Distribution-Aware Coordinate Representation for Human Pose Estimation. In 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 7091–7100. https://doi.org/10.1109/CVPR42600.2020.00712 (2020).

Liu, J., Rojas, J., Liang, Z., Li, Y. & Guan, Y. A Graph Attention Spatio-temporal Convolutional Network for 3D Human Pose Estimation in Video (2020). arXiv:2003.14179 .

Mathis, A. et al. DeepLabCut: Markerless pose estimation of user-defined body parts with deep learning. Nat. Neurosci. 2018 , 1. https://doi.org/10.1038/s41593-018-0209-y (2018).

Cotton, R. J. et al. Markerless Motion Capture and Biomechanical Analysis Pipeline. In IEEE International Consortium for Rehabilitation Robotics , (arXiv). arXiv:2303.10654 .

Cotton, R. J. et al. Improved Trajectory Reconstruction for Markerless Pose Estimation. In 45th Annual International Conference of the IEEE Engineering in Medicine and Biology Society . arXiv:2303.02413 .

Guo, A., Kamar, E., Vaughan, J. W., Wallach, H. & Morris, M. R. Toward Fairness in AI for People with Disabilities: A Research Roadmap . arXiv (2019).

Trewin, S. AI Fairness for People with Disabilities: Point of View . arXiv:1811.10670 [cs] (2018).

Carse, B., Scott, H., Davie-Smith, F., Brady, L. & Colvin, J. Minimal clinically important difference in walking velocity, gait profile score and two minute walk test for individuals with lower limb amputation. Gait Posture 88 , 221–224. https://doi.org/10.1016/j.gaitpost.2021.06.001 (2021).

Perera, S., Mody, S. H., Woodman, R. C. & Studenski, S. A. Meaningful change and responsiveness in common physical performance measures in older adults. J. Am. Geriatr. Soc. 54 , 743–749. https://doi.org/10.1111/j.1532-5415.2006.00701.x (2006).

Kwolek, B. et al. Calibrated and synchronized multi-view video and motion capture dataset for evaluation of gait recognition. Multimed. Tools Appl. 78 , 32437–32465. https://doi.org/10.1007/s11042-019-07945-y (2019).

Yuan, Y., Wei, S.-E., Simon, T., Kitani, K. & Saragih, J. SimPoE: Simulated character control for 3D human pose estimation. In 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 7155–7165, https://doi.org/10.1109/CVPR46437.2021.00708 (2021).

Xie, K. et al. Physics-based Human Motion Estimation and Synthesis from Videos . arXiv:2109.09913 [cs] (2021).

Shimada, S., Golyanik, V., Xu, W. & Theobalt, C. PhysCap: Physically Plausible Monocular 3D Motion Capture in Real Time . arXiv:2008.08880 [cs] (2020).

Shi, M. et al. MotioNet: 3D Human motion reconstruction from monocular video with skeleton consistency. ACM Trans. Graph. 40 , 1–15 (2020).

Gong, K. et al. PoseTriplet: Co-evolving 3D Human Pose Estimation, Imitation, and Hallucination under Self-supervision . arXiv:2203.15625 (2022).

Pierzchlewicz, P. A., Cotton, R. J., Bashiri, M. & Sinz, F. H. Multi-hypothesis 3D human pose estimation metrics favor miscalibrated distributions . arXiv:2210.11179 (2022).

Download references

Acknowledgements

This work was generously supported by the Research Accelerator Program of the Shirley Ryan AbilityLab. We would like to thank Emoonah McClerklin for help with data collection.

Author information

Authors and affiliations.

Shirley Ryan AbilityLab, Chicago, USA

Anthony Cimorelli, Tasos Karakostas & R. James Cotton

Department of Neuroscience, Baylor College of Medicine, Houston, USA

Ankit Patel

Department of Electrical & Computer Engineering, Rice University, Houston, USA

Department of Physical Medicine and Rehabilitation, Northwestern University, Evanston, USA

Tasos Karakostas & R. James Cotton

You can also search for this author in PubMed   Google Scholar

Contributions

R.J.C. and A.C. designed the study, collected the data and analyzed the data. T.K. and A.P. provided guidance on the development and validation on the gait transformer algorithm. T.K. provided the dataset for training the gait tranformer algorithm and R.J.C. trained it on this data. All authors were involved in the preparation of the manuscript.

Corresponding author

Correspondence to R. James Cotton .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Supplementary Information.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Cimorelli, A., Patel, A., Karakostas, T. et al. Validation of portable in-clinic video-based gait analysis for prosthesis users. Sci Rep 14 , 3840 (2024). https://doi.org/10.1038/s41598-024-53217-7

Download citation

Received : 17 January 2023

Accepted : 30 January 2024

Published : 15 February 2024

DOI : https://doi.org/10.1038/s41598-024-53217-7

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Clinical Gait Analysis and Musculoskeletal Modeling

• First Online: 24 December 2013

Cite this chapter

• Karelia Tecante 4 ,
• Frank Seehaus 4 ,
• Bastian Welke 4 ,
• Gavin Olender 4 ,
• Michael Schwarze 4 ,
• Sean Lynch 4 &
• Christoph Hurschler 4  

1727 Accesses

1 Citations

Gait analysis goal is to investigate the mechanics of the muscle, the relationships between muscles and bones and the motions of joints. However, for a deeper analysis of the internal force acting on the human body research has focused on multi-body modeling and simulation. The aim is to integrate the elements of the musculoskeletal system and the joint mechanics in order to better understand what has been learned through in vivo and in vitro experiments. This chapter presents a general overview of musculoskeletal modeling and simulation in the clinical gait area.

• Musculoskeletal modeling
• Segmentation

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

• Available as PDF
• Read on any device
• Instant download
• Own it forever
• Available as EPUB and PDF
• Compact, lightweight edition
• Dispatched in 3 to 5 business days
• Free shipping worldwide - see info
• Durable hardcover edition

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Shihab, A., & Moataz, E. (2011). Development and validation of a three-dimensional biomechanical model of the lower extremity. In V. Klika (Ed.), Theoretical biomechanics . ISBN: 978-953-307-851-9, InTech, DOI: 10.5772/24156 .

Erdemir, A., Scott M., Walter H., van den Bogert, A. J., et al. (2007). Model-based estimation of muscle forces exerted during movements. Clinical Biomechanics (Bristol, Avon) , 22 (2), 131–154.

Google Scholar  

Schwarze, M., Hurschler, C., Seehaus, F., Oehler, S., & Welke, B. (2013). Loads on the prosthesis-socket interface of above-knee amputees during normal gait: Validation of a multi-body simulation. Journal of Biomechanics , 46 (6), 1201–1206.

Welke, B., Schwarze, M., Hurschler, C., Calliess, T., & Seehaus, F. (2013). Multi-body simulation of various falling scenarios for determining resulting loads at the prosthesis interface of transfemoral amputees with osseointegrated fixation. Journal of Orthopaedic Research , 31 (7), 1123–1129.

Simon, S. R. (2004). Quantification of human motion: Gait analysis-benefits and limitations to its application to clinical problems. Journal of Biomechanics , 37 (12), 1869–1880.

Hausdorff, J. M., Merit, E. C., Renée F., Jeanne Y. W., & Ary L. G. (1998). Gait variability and basal ganglia disorders: Stride-to-stride variations of gait cycle timing in parkinson’s disease and Huntington’s disease. Movement Disorders , 13 (3), 428–437.

Hausdorff, J. M., Apinya, L., Merit, E. C., Amie, L. P., David, K., Ary, L. G., et al. (2000). Dynamic markers of altered gait rhythm in amyotrophic lateral sclerosis. Journal of Applied Physiology , 88 (6), 2045–2053.

Hausdorff, J. M., Schaafsma, J. D., Balash, Y., Bartels, A. L., Gurevich, T., & Giladi, N. (2003). Impaired regulation of stride variability in Parkinson’s disease subjects with freezing of gait. Experimental Brain Research , 149 (2), 187–194.

Bowen, J. D., & Gerry S. M. B. A. (2010). Gait Assessment. In The hip and pelvis in sports medicine and primary care (pp. 71–86). New York: Springer.

Klets, O., Riad, J. & Gutierrez-Farewik, E. M. (2010). Personalized musculoskeletal modeling of lower extremities based on magnetic resonance imaging data of 15 patients with hemiplegic cerebral palsy, IUTAM Symposium on human movement analysis and simulation, 2010 September 13th–15th, Belgium: Leuven.

Klets, O. (2011). Subject-specific musculoskeletal modeling of the lower extremities in persons with unilateral cerebral palsy. Licentiate dissertation. Stockholm: KTH Royal Institute of Technology.

Delp, S. L., & Loan, J. P. (1995). A graphics-based software system to develop and analyze models of musculoskeletal structures. Computers in Biology and Medicine , 25 (1), 21–34.

Arnold, E. M., Samuel, R. W., Richard L. L., & Scott L. D, (2010). A model of the lower limb for analysis of human movement. Annals of Biomedical Engineering , 38 (2), 269–279.

Delp, S. L., Loan, J. P., Melissa, G. H., Felix E. Z., Eric L. T., & Joseph M. R. (1990). An interactive graphics-based model of the lower extremity to study orthopaedic surgical procedures. IEEE Transactions on Biomedical Engineering , 37 (8), 757–767.

Lenaerts, G., Ward, B., Frederik, G., Michiel, M., Arthur, S., Van der Perre, G., et al. (2009). Subject-specific hip geometry and hip joint centre location affects calculated contact forces at the hip during gait. Journal of Biomechanics , 42 (9), 1246–1251.

Correa, T. A., Baker, R., & Pandy, M. G. (2011). Accuracy of generic musculoskeletal models in predicting the functional roles of muscles in human gait. Journal of Biomechanics , 44 (11), 2096–2105.

Article   Google Scholar  

Cuypers, R., Tang, Z., Luther, W., & Pauli, J. (2008). Efficient and accurate femur reconstruction using model-based segmentation and superquadric shapes. Proceedings of the Fourth IASTED International Conference , 619 (007), 99.

Brunner, G., Nambi, V., Yang, E., Kumar, A., Virani, S. S., Kougias, P., et al. (2011). Automatic quantification of muscle volumes in magnetic resonance imaging scans of the lower extremities. Magnetic Resonance Imaging , 29 (8), 1065–1075.

Schmid, J., Kim, J., & Magnenat-Thalmann, N. (2011). Extreme leg motion analysis of professional ballet dancers via MRI segmentation of multiple leg postures. International Journal of Computer Assisted Radiology and Surgery , 6 (1), 47–57.

Schmid, J., Sandholm, A., Chung, F., Thalmann, D., Delingette, H., & Magnenat-Thalmann, N. (2009). Musculoskeletal simulation model generation from MRI data sets and motion capture data. In Recent advances in the 3D physiological human (pp. 3–19). Berlin: Springer.

Teran, J., Sifakis, E., Blemker, S. S., Ng-Thow-Hing, V., Lau, C., & Fedkiw, R. (2005). Creating and simulating skeletal muscle from the visible human data set. IEEE Transactions on Visualization and Computer Graphics , 11 (3), 317–328.

Nikravesh, PE. (1998). Computer-aided analysis of mechanical systems, Prentice-Hall Inc, NJ: Englewood Cliff.

Seireg, A., & Arvikar, R. J. (1973). A mathematical model for evaluation of forces in lower extremeties of the musculo-skeletal system. Journal of Biomechanics , 6 (3), 313–326.

Raikova, R. (1992). A general approach for modelling and mathematical investigation of the human upper limb. Journal of Biomechanics , 25 (8), 857–867.

Jensen, R. H., & Davy, D. T. (1975). An investigation of muscle lines of action about the hip: a centroid line approach vs the straight line approach. Journal of Biomechanics , 8 (2), 103–110.

Blemker, S. S., & Delp, S. L. (2005). Three-dimensional representation of complex muscle architectures and geometries. Annals of Biomedical Engineering , 33 (5), 661–673.

Van der Helm, F. C., Veeger, H. E., Pronk, G. M., Van der Woude, L. H., & Rozendal, R. H. (1992). Geometry parameters for musculoskeletal modelling of the shoulder system. Journal of Biomechanics , 25 (2), 129–144.

Arnold, A. S., Salinas, S., Asakawa, D. J., & Delp, S. L. (2000). Accuracy of muscle moment arms estimated from MRI-based musculoskeletal models of the lower extremity. Computer Aided Surgery , 5 (2), 108–119.

Garner, B. A., & Pandy, M. G. (2000). The Obstacle-set method for representing muscle paths in musculoskeletal models. Computer Methods in Biomechanics and Biomedical Engineering , 3 (1), 1–30. doi: 10.1080/10255840008915251 .

Lieber, R. L., & Fridén, J. (2000). Functional and clinical significance of skeletal muscle architecture. Muscle and Nerve , 23 (11), 1647–1666.

Hill, A. V. (1938). The heat of shortening and the dynamic constants of muscle. Proceedings of the Royal Society , B126 , 136–195.

Zajac, F. E. (1989). Muscle and tendon: Properties, models, scaling, and application to biomechanics and motor control. Critical Reviews in Biomedical Engineering , 17 (4), 359.

Blemker, S. S., Asakawa, D. S., Gold, G. E., & Delp, S. L. (2007). Image-based musculoskeletal modeling: Applications, advances, and future opportunities. Journal of Magnetic Resonance Imaging , 25 (2), 441–451.

Veeger, D. H. (2011). “What if”: The use of biomechanical models for understanding and treating upper extremity musculoskeletal disorders. Manual Therapy , 16 (1), 48–50.

Stewart, C., & Shortland, A. P. (2010). The biomechanics of pathological gait-from muscle to movement. Acta of Bioengineering and Biomechanics , 12 (3), 3–12.

Anderson, A. E., Ellis, B. J., & Weiss, J. A. (2007). Verification, validation and sensitivity studies in computational biomechanics. Computer Methods in Biomechanics and Biomedical Engineering , 10 (3), 171–184.

Leardini, A., Chiari, L., Croce, U. D., Cappozzo, A., et al. (2005). Human movement analysis using stereophotogrammetry. Part 3. Soft tissue artifact assessment and compensation. Gait and Posture , 21 (2), 212.

Lu, T. W., & O’connor, J. J. (1999). Bone position estimation from skin marker coordinates using global optimisation with joint constraints. Journal of biomechanics , 32 (2), 129–134.

Högfors, C., Peterson, B., Sigholm, G., & Herberts, P. (1991). Biomechanical model of the human shoulder joint–II. The shoulder rhythm. Journal of Biomechanics , 24 (8), 699–709.

Komistek, R. D., Kane, T. R., Mahfouz, M., Ochoa, J. A., & Dennis, D. A. (2005). Knee mechanics: A review of past and present techniques to determine in vivo loads. Journal of Biomechanics , 38 , 215–228.

Anderst, W., Zauel, R., Bishop, J., Demps, E., & Tashman, S. (2009). Validation of three-dimensional model-based tiobio-femoral tracking during running. Medical Engineering and Physics , 31 , 10–16.

Nester, C. J., et al. (2009). Lessons from dynamic cadaver and invasive bone pin studies: Do we know how the foot really moves during gait. Journal of Foot and Ankle Research , 2 , 18.

Hurschler, C., Wülker, N., Windhagen, H., Hellmers, N., & Plumhoff, P. (2004). Evaluation of the lag sign tests for external rotator function of the shoulder. Journal of Shoulder and Elbow Surgery , 13 (3), 298–304.

Elias, J. J., Kirkpatrick, M. S., Saranathan, A., Mani, S., Smith, L. G., & Tanaka, M. J. (2011). Hamstrings loading contributes to lateral patellofemoral malalignment and elevated cartilage pressures: An in vitro study. Clinical Biomechanics , 26 (8), 841–846.

Wünschel, M., Leichtle, U., Obloh, C., Wülker, N., & Müller, O. (2011). The effect of different quadriceps loading patterns on tibiofemoral joint kinematics and patellofemoral contact pressure during simulated partial weight-bearing knee flexion. Knee Surgery, Sports Traumatology, Arthroscopy , 19 (7), 1099–1106.

Wülker, N., Hurschler, C., & Emmerich, J. (2003). In vitro simulation of stance phase gait part II: Simulated anterior tibial tendon dysfunction and potential compensation. Foot and Ankle International , 24 , 623–629.

Wünschel, M., Leichtle, U., Lo, J., Wülker, N., & Müller, O. (2012). Differences in tibiofemoral kinematics between the unloaded robotic passive path and a weightbearing knee simulator. Orthopedic Reviews , 4 (1), e2. doi: 10.4081/or.2012.e2 .

Davoodi, Rahman, & Gerald, E. (2002). A software tool for faster development of complex models of musculoskeletal systems and sensorimotor controllers in simulinkTM. Journal of Applied Biomechanics , 18 (4), 357–365.

Delp, S. L., Frank, C., Chand, T., & Darryl, G. (2007). OpenSim: Open-source software to create and analyze dynamic simulations of movement. IEEE Transactions on Biomedical Engineering , 54 (11), 1940–1950.

Mansouri, M., & Reinbolt, J. A. (2012). A platform for dynamic simulation and control of movement based on OpenSim and MATLAB. Journal of Biomechanics , 45 (8), 1517–1521.

Higginson, J. S., Zajac, F. E., Neptune, R. R., Kautz, S. A., & Delp, S. L. (2006). Muscle contributions to support during gait in an individual with post-stroke hemiparesis. Journal of Biomechanics , 39 (10), 1769–1777.

Jonkers, I., Stewart, C., Desloovere, K., Molenaers, G., & Spaepen, A. (2006). Musculo-tendon length and lengthening velocity of rectus femoris in stiff knee gait. Gait and Posture , 23 (2), 222–229.

Crabtree, C. A., & Jill, S. (2009). Modeling neuromuscular effects of ankle foot orthoses (AFOs) in computer simulations of gait. Gait and Posture , 29 (1), 65–70.

Nair, P. M., Rooney, K. L. Kautz, S. A. Behrman, A. L. et al. (2010). Stepping with an ankle foot orthosis re-examined: A mechanical perspective for clinical decision making. Clinical Biomechanics (Bristol, Avon) , 25 (6), 618.

Sherman, M. A., & Seth, A. (2011). Simbody: Multibody dynamics for biomedical research. Procedia IUTAM , 2 , 241–261.

Silverman, A. K., & Neptune, R. R. (2010). Individual muscle function in below Knee amputee walking. In Conference Proceedings of the Annual Meeting of the American Society , p. 166.

Fey, N. P., Klute, G. K., & Neptune, R. R. (2013). Altering prosthetic foot stiffness influences foot and muscle function during below-knee amputee walking: A modeling and simulation analysis. Journal of Biomechanics , 46 (4), 637–644.

Worsley, P., Stokes, M., & Taylor, M. (2011). Predicted knee kinematics and kinetics during functional activities using motion capture and musculoskeletal modelling in healthy older people. Gait and Posture , 33 (2), 268–273.

Michiel, O., Telfezr, S., Tørholm, S., Carbes, S., van Rhijn, L., Ross, M. et al. (2011). Generation of subject-specific, dynamic, multisegment ankle and foot models to improve orthotic design: A feasibility study. BMC Musculoskeletal Disorders , 12 (1), 256.

Nolte, D., Andersen, M. S., Rasmussen, J., & Al-Munajjed, A. (2013). Development of a patient-specific musculoskeletal model of a healthy knee to analyze hard and soft tissue loading. In 21th Annual Symposium on Computational Methods in Orthoopeadic Biomechanics .

Alkjær, T., Wieland, M. R., Andersen, M. S., Simonsen, E. B., & John, R. (2012). Computational modeling of a forward lunge: Towards a better understanding of the function of the cruciate ligaments. Journal of Anatomy , 221 (6), 590–597.

Ali, Nicholas, Michael Skipper Andersen, John Rasmussen, D Gordon E Robertson, and Gholamreza Rouhi. 2013. “The application of musculoskeletal modeling to investigate gender bias in non-contact ACL injury rate during single-leg landings”. Computer methods in biomechanics and biomedical engineering (ahead-of-print): 1–15.

Cao, E., Inoue, Y. Liu, T. & Shibata, K. (2012). Estimation of lower limb muscle forces during human sit-to-stand process with a rehabilitation training system. In 2012 IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI) , pp. 1016–1019.

Lemieux, P. O., Tétreault, P. Hagemeister, N. & Nuño, N. (2012). Influence of prosthetic humeral head size and medial offset on the mechanics of the shoulder with cuff tear arthropathy: A numerical study. Journal of Biomechanics , 46 (3), 806–812.

Jackson, M., Sylvestre, É. Bleau, J. Allard, P & Begon, M. (2012). Estimating optimal shoulder immobilization postures following surgical repair of massive rotator cuff tears. Journal of Biomechanics ,  46 (1), 179–182.

Galibarov, P. E., Dendorfer, S., & Rasmussen, J. (2011). Two computational models of the lumbar spine: Comparison and validation . Long Beach, CA: ORS—Orthopaedic Research Society.

Han, K. S., Zander, T., Taylor, W. R., & Rohlmann, A. (2012). An enhanced and validated generic thoraco-lumbar spine model for prediction of muscle forces. Medical Engineering and Physics , 34 (6), 709–716.

Ulrey, B. L., & Fathallah, F. A. (2012). Subject-specific, whole-body models of the stooped posture with a personal weight transfer device. Journal of Electromyography and Kinesiology .

Lundberg, H. J., Foucher, K. C., Andriacchi, T. P., & Wimmer, M. A. (2012). Direct comparison of measured and calculated total knee replacement force envelopes during walking in the presence of normal and abnormal gait patterns. Journal of Biomechanics , 45 (6), 990–996.

Taddei, F., Martelli, S., Valente, G., Leardini, A., Benedetti, M, G., Manfrini, M., et al. (2012). Femoral loads during gait in a patient with massive skeletal reconstruction. Clinical Biomechanics , 27 (3), 273–280.

Donnelly, C. J., & Lloyd, D. G. (2012). Optimizing whole-body kinematics to minimize valgus knee loading during sidestepping: Implications for ACL injury risk. Journal of Biomechanics , 45 (8), 1491–1497.

Crossley, K. M, Dorn, T. W. Ozturk, H., van den Noort, J., Schache, A. G., & Pandy, M. G. (2012). Altered hip muscle forces during gait in people with patellofemoral osteoarthritis. steoarthritis and Cartilage .

Mansouri, M., Clark, A. E., & Reinbolt, J. A. (2012). The use of a platform for dynamic simulation of movement: Application to balance recovery. Proceedings of the American Society of Biomechanics , Gainesville, FL. Aug. 2012 (Available at:  http://works.bepress.com/jeffrey_reinbolt/57 ).

Branemark, R., Branemark, P. I., Rydevik, B., & Myers, R. R. (2001). Osseointegration in skeletal reconstruction and rehabilitation: A review. Journal of rehabilitation research and development , 38 (2), 175–182.

Brånemark, P. I. (1983). Osseointegration and its experimental background. The Journal of prosthetic dentistry , 50 (3), 399–410.

Hagberg, K., Branemark, R., Guntorberg, B., et al. (2008). Osseointegrated trans-femoral amputation prostheses: Pro-spective results of general and condition-specific quality of life in 18 patients at 2-year follow-up. Prosthetics and Orthotics International , 32 , 29–41.

Grundei, H., Von Stein, T., Schulte-Bockhof, D., Kausch, C., Gollwitzer, H., & Gradinger, R. (2009). Die Endo-Exo-Femurprothese—Update eines Versorgungskonzeptes zur Rehabilitation von Oberschenkelamputierten. Orthopädie-Technik , 12 , 143–149.

Kadaba, M. P., Ramakrishnan, H. K., & Wootten, M. E. (1990). Measurement of lower extremity kinematics during level walking. Journal of Orthopaedic Research?: official publication of the Orthopaedic Research Society , 8 (3), 383–92. doi:10.1002/jor.1100080310.

Vicon Plug-in Gait Product Guide–Foundation Notes Revision 2.0 March 2010 For use with Plug-in Gait Version 2.0 in Vicon Nexus, 2010.

Tomaszewski, P. K., Verdonschot, N., Bulstra, S. K., & Verkerke, G. J. (2010). A comparative finite-element analysis of bone failure and load transfer of osseointegrated prostheses fixations. Annals of Biomedical Engineering , 38 (7), 2418–2427.

Download references

Acknowledgments

This study was funded by the German Federal Ministry of Education and Research (BMBF AZ: 01EZ0775). The authors like to thank TU Berlin and Otto Bock Healthcare GmbH, Duderstadt, Germany for cooperation in TExoPro and EU Marie Curie Actions-Marie Curie Research Training Networks/ Multi-scale Biological Modalities for Physiological Human articulation 289897 (FP7-PEOPLE-2011-ITN) for their funding

Author information

Authors and affiliations.

Laboratory for Biomechanics and Biomaterials (LBB) Orthopaedic Department, Hannover Medical School (MHH), Hannover, Germany

Karelia Tecante, Frank Seehaus, Bastian Welke, Gavin Olender, Michael Schwarze, Sean Lynch & Christoph Hurschler

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Karelia Tecante .

Editor information

Editors and affiliations.

Centre Universitaire Informatique, MIRALab - University of Geneva, Carouge, Geneve, Switzerland

Nadia Magnenat-Thalmann

Nuclear Medicine Division, University Hospital of Geneva, Geneva, Switzerland

Osman Ratib

Hon Fai Choi

Rights and permissions

Reprints and permissions

Copyright information

© 2014 Springer-Verlag London

About this chapter

Tecante, K. et al. (2014). Clinical Gait Analysis and Musculoskeletal Modeling. In: Magnenat-Thalmann, N., Ratib, O., Choi, H. (eds) 3D Multiscale Physiological Human. Springer, London. https://doi.org/10.1007/978-1-4471-6275-9_7

Download citation

DOI : https://doi.org/10.1007/978-1-4471-6275-9_7

Published : 24 December 2013

Publisher Name : Springer, London

Print ISBN : 978-1-4471-6274-2

Online ISBN : 978-1-4471-6275-9

eBook Packages : Computer Science Computer Science (R0)

Share this chapter

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Publish with us

Policies and ethics

• Find a journal
• Track your research

Gait Analysis in Neurorehabilitation: From Research to Clinical Practice

Affiliations.

• 1 IRCCS Centro Neurolesi "Bonino-Pulejo", Via Palermo, SS 113, C. da Casazza, 98123 Messina, Italy.
• 2 Department of Research and Development, LUNEX International University of Health, Exercise and Sports, Avenue du Parc des Sports, 50, 4671 Differdange, Luxembourg.
• PMID: 37508812
• PMCID: PMC10376523
• DOI: 10.3390/bioengineering10070785

When brain damage occurs, gait and balance are often impaired. Evaluation of the gait cycle, therefore, has a pivotal role during the rehabilitation path of subjects who suffer from neurological disorders. Gait analysis can be performed through laboratory systems, non-wearable sensors (NWS), and/or wearable sensors (WS). Using these tools, physiotherapists and neurologists have more objective measures of motion function and can plan tailored and specific gait and balance training early to achieve better outcomes and improve patients' quality of life. However, most of these innovative tools are used for research purposes (especially the laboratory systems and NWS), although they deserve more attention in the rehabilitation field, considering their potential in improving clinical practice. In this narrative review, we aimed to summarize the most used gait analysis systems in neurological patients, shedding some light on their clinical value and implications for neurorehabilitation practice.

Keywords: gait analysis; neurological disorders; neurorehabilitation; non-wearable sensors; wearable sensors.

Publication types

Grants and funding.

• This funding has not a grant number./Current Research funds 2023, Ministry of Health, Italy.

Study record managers: refer to the Data Element Definitions if submitting registration or results information.

Search for terms

• Advanced Search
• See Studies by Topic
• See Studies on Map
• How to Search
• How to Use Search Results
• How to Find Results of Studies
• How to Read a Study Record

• Learn About Studies
• Other Sites About Studies
• Glossary of Common Site Terms

• Submit Studies to ClinicalTrials.gov PRS
• Why Should I Register and Submit Results?
• FDAAA 801 and the Final Rule
• How to Apply for a PRS Account
• How to Register Your Study
• How to Edit Your Study Record
• How to Submit Your Results
• Frequently Asked Questions
• Support Materials
• Training Materials

• Selected Publications
• Clinical Alerts and Advisories
• Trends, Charts, and Maps
• Downloading Content for Analysis

• ClinicalTrials.gov Background
• About the Results Database
• History, Policies, and Laws
• ClinicalTrials.gov Modernization
• Media/Press Resources
• Linking to This Site
• Terms and Conditions
• Search Results
• Study Record Detail

Effect of Acupuncture on the Gait Disturbance and Hemodynamic Changes in the Prefrontal Cortex

• Study Details
• Tabular View
• No Results Posted

Background: Alzheimer's disease (AD) is characterized by cognitive impairment and behavioral impairment, and increasing attention is paid in the gait of AD patients. The aim of this randomized controlled trial (RCT) is to explore the effect of acupuncture on the cognitive function, gait performance, and hemodynamic changes in the prefrontal cortices.

Methods: In this RCT, a total of 108 AD patients will be randomly assigned into acupuncture group or control group for 8 weeks. The primary outcome will be Three-dimensional gait analysis and cerebral hemodynamics using functional near-infrared spectroscopy (fNIRS). Secondary outcomes include Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), and Barthel Index (BI).

Discussion: This trial is expected to explore the effect of acupuncture on cognitive function, gait performance, and hemodynamic changes in the prefrontal cortices for AD patients.

The inclusion criteria are as follows:

• Aged 40-85 years old;
• Meeting the mentioned-above diagnostic criteria from traditional Chinese medicine and Western medicine;
• With mild to moderate dementia (CDR=0.5, 1.0, or 2.0 points);
• Hachinski Ischemic Scale (HIS) ≤ 4 points;
• Hamilton Depression Rating Scale (HAMD) < 20 points;
• Without severe bone and joint diseases and able to walk independently;
• Able to conduct vision and hearing tests;
• Volunteering to participate in this trial and signing an informed consent form. Exclusion criteria

The exclusion criteria are as follows:

• Dementia caused by other systemic or neurological diseases, such as central nervous system infections, post-traumatic dementia, Parkinson's disease dementia;
• Suffering from acute illness, upper limb extrapyramidal stiffness, neurological or psychiatric disorders (except cognitive impairment);
• With medical history that interferes with cognitive function assessment, such as past history of psychiatric drug abuse, drug addiction within the past 5 years, and alcohol abuse;
• Afraid of acupuncture and cannot accept acupuncture treatment;
• Currently participating in other clinical trial that affects the outcome evaluation of this trial.
• For Patients and Families
• For Researchers
• For Study Record Managers
• Customer Support
• Accessibility
• Viewers and Players
• Freedom of Information Act
• HHS Vulnerability Disclosure
• U.S. National Library of Medicine
• U.S. National Institutes of Health
• U.S. Department of Health and Human Services

• International edition
• Australia edition
• Europe edition

Gender medicine ‘built on shaky foundations’, Cass review finds

Analysis finds most research underpinning clinical guidelines, hormone treatments and puberty blockers to be low quality

• Review of gender services has major implications for mental health services

The head of the world’s largest review into children’s care has said that gender medicine is “built on shaky foundations”.

Dr Hilary Cass, the paediatrician commissioned to conduct a review of the services provided by the NHS to children and young people questioning their gender identity, said that while doctors tended to be cautious in implementing new findings in emerging areas of medicine, “quite the reverse happened in the field of gender care for children”.

Cass commissioned the University of York to conduct a series of analyses as part of her review.

Two papers examined the quality and development of current guidelines and recommendations for managing gender dysphoria in children and young people. Most of the 23 clinical guidelines reviewed were not independent or evidence based, the researchers found.

A third paper on puberty blockers found that of 50 studies, only one was of high quality.

Similarly, of 53 studies included in a fourth paper on the use of hormone treatment, only one was of sufficiently high quality, with little or only inconsistent evidence on key outcomes.

Here are the main findings of the reviews:

Clinical guidelines

Increasing numbers of children and young people experiencing gender dysphoria are being referred to specialist gender services. There are various guidelines outlining approaches to the clinical care of these children and adolescents.

In the first two papers, the York researchers examined the quality and development of published guidelines or clinical guidance containing recommendations for managing gender dysphoria in children and young people up to the age of 18.

They studied a total of 23 guidelines published in different countries between 1998 and 2022. All but two were published after 2010.

Most of them lacked “an independent and evidence-based approach and information about how recommendations were developed”, the researchers said.

Few guidelines were informed by a systematic review of empirical evidence and they lack transparency about how their recommendations were developed. Only two reported consulting directly with children and young people during their development, the York academics found.

“Healthcare services and professionals should take into account the poor quality and interrelated nature of published guidance to support the management of children and adolescents experiencing gender dysphoria/incongruence,” the researchers wrote.

Writing in the British Medical Journal (BMJ) , Cass said that while medicine was usually based on the pillars of integrating the best available research evidence with clinical expertise, and patient values and preferences, she “found that in gender medicine those pillars are built on shaky foundations”.

She said the World Professional Association of Transgender Healthcare (WPATH) had been “highly influential in directing international practice, although its guidelines were found by the University of York’s appraisal to lack developmental rigour and transparency”.

In the foreword to her report, Cass said while doctors tended to be cautious in implementing new findings “quite the reverse happened in the field of gender care for children”.

In one example, she said a single Dutch medical study, “suggesting puberty blockers may improve psychological wellbeing for a narrowly defined group of children with gender incongruence”, had formed the basis for their use to “spread at pace to other countries”. Subsequently, there was a “greater readiness to start masculinising/feminising hormones in mid-teens”.

She added: “Some practitioners abandoned normal clinical approaches to holistic assessment, which has meant that this group of young people have been exceptionalised compared to other young people with similarly complex presentations. They deserve very much better.”

Both papers repeatedly pointed to a key problem in this area of medicine: a dearth of good data.

She said: “Filling this knowledge gap would be of great help to the young people wanting to make informed choices about their treatment.”

Cass said the NHS should put in place a “full programme of research” looking at the characteristics, interventions and outcomes of every young person presenting to gender services, with consent routinely sought for enrolment in a research study that followed them into adulthood.

Gender medicine was “an area of remarkably weak evidence”, her review found, with study results also “exaggerated or misrepresented by people on all sides of the debate to support their viewpoint”.

Alongside a puberty blocker trial, which could be in place by December, there should be research into psychosocial interventions and the use of the masculinising and feminising hormones testosterone and oestrogen, the review found.

Hormone treatment

Many trans people who seek medical intervention in their transition opt to take hormones to masculinise or feminise their body, an approach that has been used in transgender adults for decades.

“It is a well-established practice that has transformed the lives of many transgender people,” the Cass review notes, adding that while these drugs are not without long-term problems and side-effects, for many they are dramatically outweighed by the benefits.

For birth-registered females, the approach means taking testosterone, which brings about changes including the growth of facial hair and a deepening of the voice, while for birth-registered males, it involves taking hormones including oestrogen to promote changes including the growth of breasts and an increase in body fat. Some of these changes may be irreversible.

However, in recent years a growing proportion of adolescents have begun taking these cross-sex, or gender-affirming, hormones, with the vast majority who are prescribed puberty blockers subsequently moving on to such medication.

This growing take-up among young people has led to questions over the impact of these hormones in areas ranging from mental health to sexual functioning and fertility.

Now researchers at the University of York have carried out a review of the evidence, comprising an analysis of 53 previously published studies, in an attempt to set out what is known – and what is not – about the risks, benefits and possible side-effects of such hormones on young people.

All but one study, which looked at side-effects, were rated of moderate or low quality, with the researchers finding limited evidence for the impact of such hormones on trans adolescents with respect to outcomes, including gender dysphoria and body satisfaction.

The researchers noted inconsistent findings around the impact of such hormones on growth, height, bone health and cardiometabolic effects, such as BMI and cholesterol markers. In addition, they found no study assessed fertility in birth-registered females, and only one looked at fertility in birth-registered males.

“These findings add to other systematic reviews in concluding there is insufficient and/or inconsistent evidence about the risks and benefits of hormone interventions in this population,” the authors write.

However, the review did find some evidence that masculinising or feminising hormones might help with psychological health in young trans people. An analysis of five studies in the area suggested hormone treatment may improve depression, anxiety and other aspects of mental health in adolescents after 12 months of treatment, with three of four studies reporting an improvement around suicidality and/or self-harm (one reported no change).

But unpicking the precise role of such hormones is difficult. “Most studies included adolescents who received puberty suppression, making it difficult to determine the effects of hormones alone,” the authors write, adding that robust research on psychological health with long-term follow-up was needed.

The Cass review has recommended NHS England should review the current policy on masculinising or feminising hormones, advising that while there should be the option to provide such drugs from age 16, extreme caution was recommended, and there should be a clear clinical rationale for not waiting until an individual reached 18.

Puberty blockers

Treatments to suppress puberty in adolescents became available through routine clinical practice in the UK a decade ago.

While the drugs have long been used to treat precocious puberty – when children start puberty at an extremely young age – they have only been used off-label in children with gender dysphoria or incongruence since the late 1990s. The rationale for giving puberty blockers, which originated in the Netherlands, was to buy thinking time for young people and improve their ability to smooth their transition in later life.

Data from gender clinics reported in the Cass review showed the vast majority of people who started puberty suppression went on to have masculinising or feminising hormones, suggesting that puberty blockers did not buy people time to think.

To understand the broader effects of puberty blockers, researchers at the University of York identified 50 papers that reported on the effects of the drugs in adolescents with gender dysphoria or incongruence. According to their systematic review, only one of these studies was high quality, with a further 25 papers regarded as moderate quality. The remaining 24 were deemed too weak to be included in the analysis.

Many of the reports looked at how well puberty was suppressed and the treatment’s side-effects, but fewer looked at whether the drugs had their intended benefits.

Of two studies that investigated gender dysphoria and body satisfaction, neither found a change after receiving puberty blockers. The York team found “very limited” evidence that puberty blockers improved mental health.

Overall, the researchers said “no conclusions” could be drawn about the impact on gender dysphoria, mental and psychosocial health or cognitive development, though there was some evidence bone health and height may be compromised during treatment.

Based on the York work, the Cass review finds that puberty blockers offer no obvious benefit in helping transgender males to help their transition in later life, particularly if the drugs do not lead to an increase in height in adult life. For transgender females, the benefits of stopping irreversible changes such as a deeper voice and facial hair have to be weighed up against the need for penile growth should the person opt for vaginoplasty, the creation of a vagina and vulva.

In March, NHS England announced that children with gender dysphoria would no longer receive puberty blockers as routine practice. Instead, their use will be confined to a trial that the Cass review says should form part of a broader research programme into the effects of masculinising and feminising hormones.

• Transgender
• Young people

Most viewed

ORIGINAL RESEARCH article

Clinical analysis of lupus miliaris disseminatus faciei: a cross-sectional study and literature review provisionally accepted.

• 1 The Second Affiliated Hospital of Xi'an Jiaotong University, China

The final, formatted version of the article will be published soon.

The clinical similarity of lupus miliaris disseminatus faciei (LMDF) and other papular granulomatous facial disorders often makes its correct diagnosis challenging. Diagnosis often requires the assistance of pathological examination, and dermoscopy can be used as an auxiliary and non-invasive examination method, however, the current findings remain incomplete. Objectives: This study aimed to summarize the clinical, histopathological and dermoscopic features of LMDF in the Chinese Han population and aiming to provide practical significance to correct diagnosis.. Methods: 109 patients of LMDF were collected in the Department of Dermatology, the Second Affiliated Hospital of Xi'an Jiaotong University from August 2015 to August 2023. Clinical and histopathological manifestation, as well as 44 cases of dermoscopic image features, including background, follicular findings, vessels, and other structures, were summarized and evaluated. The clinical and histopathological manifestations of all patients, as well as the dermoscopic image features of 44 cases, including background, follicular findings, vessels, and other structures, were summarized and evaluated.The most significant histopathological features of LMDF in 109 cases is epithelioid granulomatous infiltrate in the superficial dermis, with or without caseation.The most significant dermoscopic features of LMDF in all 44 cases were orange structureless background (30/44), follicular plug (32/44), follicular white scar-like area (32/44), unspecific linear vessels (24/44), linear vessels with branch (24/44) and white streaks (18/44).Conclusions: Histopathologically, LMDF is characterized by the presence of epithelioid granulomatous infiltrate in the superficial dermis, with or without caseation.Dermoscopically, it exhibits a distinctive orange structureless background, follicular plug, follicular white scar-like area, nonspecific linear vessels, linear vessels with branches, and white streaks.

Keywords: lupus miliaris disseminates faciei, Clinical analysis, histopathology, Image feature, Dermoscopy

Received: 05 Feb 2024; Accepted: 09 Apr 2024.

Copyright: © 2024 Wang, Li and Wang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

* Correspondence: Dr. Shuang Wang, The Second Affiliated Hospital of Xi'an Jiaotong University, Xi'an, 710049, Shaanxi Province, China

People also looked at

Maternal and Infant Research Electronic Data Analysis (MIREDA): A protocol for creating a common data model for federated analysis of UK birth cohorts and the life course.

• Find this author on Google Scholar
• Find this author on PubMed
• Search for this author on this site
• ORCID record for Michael Seaborne
• For correspondence: [email protected]
• ORCID record for Hope Jones
• ORCID record for Neil Cockburn
• ORCID record for Stevo Durbaba
• ORCID record for Tom Giles
• ORCID record for Arturo Gonzalez-Izquierdo
• ORCID record for Amy Hough
• ORCID record for Dan Mason
• ORCID record for Armando Mendez-Villalon
• ORCID record for Carlos Sanchez-Soriano
• ORCID record for Chris Orton
• ORCID record for David Ford
• ORCID record for Phillip Quinlan
• ORCID record for Krish Nirantharakumar
• ORCID record for Lucilla Poston
• ORCID record for Rebecca Reynolds
• ORCID record for Gillian Santorelli
• ORCID record for Sinead Brophy
• Info/History
• Preview PDF

Introduction Birth cohorts are valuable resources for studying early life, the determinants of health, disease, and development. They are essential for studying life course. Electronic cohorts are live, dynamic longitudinal cohorts using anonymised, routinely collected data. There is no selection bias through direct recruitment, but they are limited to health and administrative system data and may lack contextual information. The MIREDA (Maternal and Infant Research Electronic Data Analysis) partnership creates a UK-wide birth cohort by aligning existing electronic birth cohorts to have the same structure, content, and vocabularies, enabling UK-wide federated analyses. Objectives 1) Create a core dynamic, live UK-wide electronic birth cohort with approximately 100,000 new births per year using a common data model (CDM). 2) Provide data linkage and automation for long-term follow up of births from MuM-PreDiCT and the Born-in initiatives of Bradford, Wales, Scotland, and South London for comparable analyses. Methods We will establish core data content and collate linkable data. Use a suite of extraction, transformation, and load (ETL) tools will be used to transform the data for each birth cohort into the CDM. Transformed datasets will remain within each cohorts trusted research environment (TRE). Metadata will be uploaded for the public to the Health Data Research (HDRUK) Innovation Gateway. We will develop a single online data access request for researchers. A cohort profile will be developed for researchers to reference the resource. Ethics Each cohort has approval from their TRE through compliance with their project application processes and information governance. Dissemination We will engage with researchers in the field to promote our resource through partnership networking, publication, research collaborations, conferences, social media, and marketing communications strategies. Keywords: Birth Cohort, Life Course Perspective, Data Science, Data Curation, Routinely Collected Health Data, Electronic Health Records, Unified Medical Language System.

Competing Interest Statement

The authors have declared no competing interest.

Funding Statement

This work was supported by an MRC Partnership Grant [MR/X02055X/1], MatCHNet pump-priming [U20005/302873] and an MRC Programme Grant [MR/X009742/1].

Author Declarations

I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.

The details of the IRB/oversight body that provided approval or exemption for the research described are given below:

Access to data is granted according to the information governance requirements of each TRE. The Data Protection Act 2018 is not applicable to anonymised data and the OMOP CDM will be anonymised and provide aggregated data and statistics only. Each TRE has ethical approval for its operation and use, thus no additional ethical approval was required beyond the standard project approval by official channels.

I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.

I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).

I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.

There was an error in the order of authorship and a missing punctuation mark after the title abstract. Also needed to update the authors order in the file I uploaded as it did not match that of the paper.

Data Availability

Data will be available upon reasonable request through the Health Data Research (HDRUK) Innovation Gateway.

View the discussion thread.

Thank you for your interest in spreading the word about medRxiv.

NOTE: Your email address is requested solely to identify you as the sender of this article.

Citation Manager Formats

• EndNote (tagged)
• EndNote 8 (xml)
• RefWorks Tagged
• Ref Manager
• Tweet Widget
• Facebook Like
• Google Plus One
• Addiction Medicine (316)
• Allergy and Immunology (619)
• Anesthesia (160)
• Cardiovascular Medicine (2281)
• Dentistry and Oral Medicine (280)
• Dermatology (201)
• Emergency Medicine (370)
• Endocrinology (including Diabetes Mellitus and Metabolic Disease) (803)
• Epidemiology (11586)
• Forensic Medicine (10)
• Gastroenterology (680)
• Genetic and Genomic Medicine (3597)
• Geriatric Medicine (337)
• Health Economics (618)
• Health Informatics (2309)
• Health Policy (915)
• Health Systems and Quality Improvement (865)
• Hematology (335)
• HIV/AIDS (753)
• Infectious Diseases (except HIV/AIDS) (13167)
• Intensive Care and Critical Care Medicine (757)
• Medical Education (359)
• Medical Ethics (100)
• Nephrology (389)
• Neurology (3358)
• Nursing (191)
• Nutrition (507)
• Obstetrics and Gynecology (652)
• Occupational and Environmental Health (647)
• Oncology (1763)
• Ophthalmology (526)
• Orthopedics (209)
• Otolaryngology (284)
• Pain Medicine (223)
• Palliative Medicine (66)
• Pathology (441)
• Pediatrics (1008)
• Pharmacology and Therapeutics (422)
• Primary Care Research (406)
• Psychiatry and Clinical Psychology (3067)
• Public and Global Health (5999)
• Radiology and Imaging (1226)
• Rehabilitation Medicine and Physical Therapy (715)
• Respiratory Medicine (811)
• Rheumatology (367)
• Sexual and Reproductive Health (353)
• Sports Medicine (318)
• Surgery (389)
• Toxicology (50)
• Transplantation (171)
• Urology (142)