Role of machine learning in gait analysis: a review

Temporal Variability in Stride Kinematics during the Application of TENS: A Machine Learning Analysis

INTRODUCTION

The purpose of our report was to use a Random Forest classification approach to predict the association between transcutaneous electrical nerve stimulation (TENS) and walking kinematics at the stride level when middle-aged and older adults performed the 6-min test of walking endurance.

METHODS

Data from 41 participants (aged 64.6 ± 9.7 yr) acquired in two previously published studies were analyzed with a Random Forest algorithm that focused on upper and lower limb, lumbar, and trunk kinematics. The four most predictive kinematic features were identified and utilized in separate models to distinguish between three walking conditions: burst TENS, continuous TENS, and control. SHAP analysis and linear mixed models were used to characterize the differences among these conditions.

RESULTS

Modulation of four key kinematic features-toe-out angle, toe-off angle, and lumbar range of motion (ROM) in coronal and sagittal planes-accurately predicted walking conditions for the burst (82% accuracy) and continuous (77% accuracy) TENS conditions compared with control. Linear mixed models detected a significant difference in lumbar sagittal ROM between the TENS conditions. SHAP analysis revealed that burst TENS was positively associated with greater lumbar coronal ROM, smaller toe-off angle, and less lumbar sagittal ROM. Conversely, continuous TENS was associated with less lumbar coronal ROM and greater lumbar sagittal ROM.

CONCLUSIONS

Our approach identified four kinematic features at the stride level that could distinguish between the three walking conditions. These distinctions were not evident in average values across strides.

Explainable Deep-Learning-Based Gait Analysis of Hip-Knee Cyclogram for the Prediction of Adolescent Idiopathic Scoliosis Progression

Accurate prediction of scoliotic curve progression is crucial for guiding treatment decisions in adolescent idiopathic scoliosis (AIS). Traditional methods of assessing the likelihood of AIS progression are limited by variability and rely on static measurements. This study developed and validated machine learning models for classifying progressive and non-progressive scoliotic curves based on gait analysis using wearable inertial sensors. Gait data from 38 AIS patients were collected using seven inertial measurement unit (IMU) sensors, and hip-knee (HK) cyclograms representing inter-joint coordination were generated. Various machine learning algorithms, including support vector machine (SVM), random forest (RF), and novel deep convolutional neural network (DCNN) models utilizing multi-plane HK cyclograms, were developed and evaluated using 10-fold cross-validation. The DCNN model incorporating multi-plane HK cyclograms and clinical factors achieved an accuracy of 92% in predicting curve progression, outperforming SVM (55% accuracy) and RF (52% accuracy) models using handcrafted gait features. Gradient-based class activation mapping revealed that the DCNN model focused on the swing phase of the gait cycle to make predictions. This study demonstrates the potential of deep learning techniques, and DCNNs in particular, in accurately classifying scoliotic curve progression using gait data from wearable IMU sensors.

Vision-based motion capture for the gait analysis of neurodegenerative diseases: A review

BACKGROUND

Developments in vision-based systems and human pose estimation algorithms have the potential to detect, monitor and intervene early on neurodegenerative diseases through gait analysis. However, the gap between the technology available and actual clinical practice is evident as most clinicians still rely on subjective observational gait analysis or objective marker-based analysis that is time-consuming.

RESEARCH QUESTION

This paper aims to examine the main developments of vision-based motion capture and how such advances may be integrated into clinical practice.

METHODS

The literature review was conducted in six online databases using Boolean search terms. A commercial system search was also included. A predetermined methodological criterion was then used to assess the quality of the selected articles.

RESULTS

A total of seventeen studies were evaluated, with thirteen studies focusing on gait classification systems and four studies on gait measurement systems. Of the gait classification systems, nine studies utilized artificial intelligence-assisted techniques, while four studies employed statistical techniques. The results revealed high correlations of gait features identified by classifier models with existing clinical rating scales. These systems demonstrated generally high classification accuracies and were effective in diagnosing disease severity levels. Gait measurement systems that extract spatiotemporal and kinematic joint information from video data generally found accurate measurements of gait parameters with low mean absolute errors, high intra- and inter-rater reliability.

SIGNIFICANCE

Low cost, portable vision-based systems can provide proof of concept for the quantification of gait, expansion of gait assessment tools, remote gait analysis of neurodegenerative diseases and a point of care system for orthotic evaluation. However, certain challenges, including small sample sizes, occlusion risks, and selection bias in training models, need to be addressed. Nevertheless, these systems can serve as complementary tools, equipping clinicians with essential gait information to objectively assess disease severity and tailor personalized treatment for enhanced patient care.

Optimizing Rare Disease Gait Classification through Data Balancing and Generative AI: Insights from Hereditary Cerebellar Ataxia

The interpretability of gait analysis studies in people with rare diseases, such as those with primary hereditary cerebellar ataxia (pwCA), is frequently limited by the small sample sizes and unbalanced datasets. The purpose of this study was to assess the effectiveness of data balancing and generative artificial intelligence (AI) algorithms in generating synthetic data reflecting the actual gait abnormalities of pwCA. Gait data of 30 pwCA (age: 51.6 ± 12.2 years; 13 females, 17 males) and 100 healthy subjects (age: 57.1 ± 10.4; 60 females, 40 males) were collected at the lumbar level with an inertial measurement unit. Subsampling, oversampling, synthetic minority oversampling, generative adversarial networks, and conditional tabular generative adversarial networks (ctGAN) were applied to generate datasets to be input to a random forest classifier. Consistency and explainability metrics were also calculated to assess the coherence of the generated dataset with known gait abnormalities of pwCA. ctGAN significantly improved the classification performance compared with the original dataset and traditional data augmentation methods. ctGAN are effective methods for balancing tabular datasets from populations with rare diseases, owing to their ability to improve diagnostic models with consistent explainability.

Emerging trends in biomedical trait-based human identification: A bibliometric analysis

Personal human identification is a crucial aspect of modern society with applications spanning from law enforcement to healthcare and digital security. This bibliometric paper presents a comprehensive analysis of recent advances in personal human identification methodologies focusing on biomedical traits. The paper examines a diverse range of research articles, reviews, and patents published over the last decade to provide insights into the evolving landscape of biometric identification techniques. The study categorizes the identified literature into distinct biomedical trait categories, including but not limited to, fingerprint and palmprint recognition, iris and retinal scanning, facial recognition, voice and speech analysis, gait recognition, and DNA-based identification. Through systematic analysis, the paper highlights key trends, emerging technologies, and interdisciplinary collaborations in each category, revealing the interdisciplinary nature of research in this field. Furthermore, the bibliometric analysis examines the geographical distribution of research efforts, identifying prominent countries and institutions contributing to advancements in personal human identification. Collaboration networks among researchers and institutions are visualized to depict the knowledge flow and collaborative dynamics within the field. Overall, this study serves as a valuable reference for researchers, practitioners, and policymakers, shedding light on the current status and potential future directions of personal human identification leveraging biomedical traits.

Machine learning applied to gait analysis data in cerebral palsy and stroke: A systematic review

BACKGROUND

Among neurological pathologies, cerebral palsy and stroke are the main contributors to walking disorders. Machine learning methods have been proposed in the recent literature to analyze gait data from these patients. However, machine learning methods still fail to translate effectively into clinical applications. This systematic review addressed the gaps hindering the use of machine learning data analysis in the clinical assessment of cerebral palsy and stroke patients.

RESEARCH QUESTION

What are the main challenges in transferring proposed machine learning methods to clinical applications?

METHODS

PubMed, Web of Science, Scopus, and IEEE databases were searched for relevant publications on machine learning methods applied to gait analysis data from stroke and cerebral palsy patients until February the 23rd, 2023. Information related to the suitability, feasibility, and reliability of the proposed methods for their effective translation to clinical use was extracted, and quality was assessed based on a set of predefined questions.

RESULTS

From 4120 resulting references, 63 met the inclusion criteria. Thirty-one studies used supervised, and 32 used unsupervised machine learning methods. Artificial neural networks and k-means clustering were the most used methods in each category. The lack of rationale for features and algorithm selection, the use of unrepresentative datasets, and the lack of clinical interpretability of the clustering outputs were the main factors hindering the clinical reliability and applicability of these methods.

SIGNIFICANCE

The literature offers numerous machine learning methods for clustering gait data from cerebral palsy and stroke patients. However, the clinical significance of the proposed methods is still lacking, limiting their translation to real-world applications. The design of future studies must take into account clinical question, dataset significance, feature and model selection, and interpretability of the results, given their criticality for clinical translation.

Classification of inertial sensor-based gait patterns of orthopaedic conditions using machine learning: A pilot study

Elderly patients often have more than one disease that affects walking behavior. An objective tool to identify which disease is the main cause of functional limitations may aid clinical decision making. Therefore, we investigated whether gait patterns could be used to identify degenerative diseases using machine learning. Data were extracted from a clinical database that included sagittal joint angles and spatiotemporal parameters measured using seven inertial sensors, and anthropometric data of patients with unilateral knee or hip osteoarthritis, lumbar or cervical spinal stenosis, and healthy controls. Various classification models were explored using the MATLAB Classification Learner app, and the optimizable Support Vector Machine was chosen as the best performing model. The accuracy of discrimination between healthy and pathologic gait was 82.3%, indicating that it is possible to distinguish pathological from healthy gait. The accuracy of discrimination between the different degenerative diseases was 51.4%, indicating the similarities in gait patterns between diseases need to be further explored. Overall, the differences between pathologic and healthy gait are distinct enough to classify using a classical machine learning model; however, routinely recorded gait characteristics and anthropometric data are not sufficient for successful discrimination of the degenerative diseases.

AI-based methodologies for exoskeleton-assisted rehabilitation of the lower limb: a review

Over the past few years, there has been a noticeable surge in efforts to design novel tools and approaches that incorporate Artificial Intelligence (AI) into rehabilitation of persons with lower-limb impairments, using robotic exoskeletons. The potential benefits include the ability to implement personalized rehabilitation therapies by leveraging AI for robot control and data analysis, facilitating personalized feedback and guidance. Despite this, there is a current lack of literature review specifically focusing on AI applications in lower-limb rehabilitative robotics. To address this gap, our work aims at performing a review of 37 peer-reviewed papers. This review categorizes selected papers based on robotic application scenarios or AI methodologies. Additionally, it uniquely contributes by providing a detailed summary of input features, AI model performance, enrolled populations, exoskeletal systems used in the validation process, and specific tasks for each paper. The innovative aspect lies in offering a clear understanding of the suitability of different algorithms for specific tasks, intending to guide future developments and support informed decision-making in the realm of lower-limb exoskeleton and AI applications.

Pathological Gait Analysis With an Open-Source Cloud-Enabled Platform Empowered by Semi-Supervised Learning-PathoOpenGait

We present PathoOpenGait, a cloud-based platform for comprehensive gait analysis. Gait assessment is crucial in neurodegenerative diseases such as Parkinson's and multiple system atrophy, yet current techniques are neither affordable nor efficient. PathoOpenGait utilizes 2D and 3D data from a binocular 3D camera for monitoring and analyzing gait parameters. Our algorithms, including a semi-supervised learning-boosted neural network model for turn time estimation and deterministic algorithms to estimate gait parameters, were rigorously validated on annotated gait records, demonstrating high precision and consistency. We further demonstrate PathoOpenGait's applicability in clinical settings by analyzing gait trials from Parkinson's patients and healthy controls. PathoOpenGait is the first open-source, cloud-based system for gait analysis, providing a user-friendly tool for continuous patient care and monitoring. It offers a cost-effective and accessible solution for both clinicians and patients, revolutionizing the field of gait assessment. PathoOpenGait is available at https://pathoopengait.cmdm.tw.

A global bibliometric and visualized analysis of gait analysis and artificial intelligence research from 1992 to 2022

Gait is an important basic function of human beings and an integral part of life. Many mental and physical abnormalities can cause noticeable differences in a person's gait. Abnormal gait can lead to serious consequences such as falls, limited mobility and reduced life satisfaction. Gait analysis, which includes joint kinematics, kinetics, and dynamic Electromyography (EMG) data, is now recognized as a clinically useful tool that can provide both quantifiable and qualitative information on performance to aid in treatment planning and evaluate its outcome. With the assistance of new artificial intelligence (AI) technology, the traditional medical environment has undergone great changes. AI has the potential to reshape medicine, making gait analysis more accurate, efficient and accessible. In this study, we analyzed basic information about gait analysis and AI articles that met inclusion criteria in the WoS Core Collection database from 1992-2022, and the VosViewer software was used for web visualization and keyword analysis. Through bibliometric and visual analysis, this article systematically introduces the research status of gait analysis and AI. We introduce the application of artificial intelligence in clinical gait analysis, which affects the identification and management of gait abnormalities found in various diseases. Machine learning (ML) and artificial neural networks (ANNs) are the most often utilized AI methods in gait analysis. By comparing the predictive capability of different AI algorithms in published studies, we evaluate their potential for gait analysis in different situations. Furthermore, the current challenges and future directions of gait analysis and AI research are discussed, which will also provide valuable reference information for investors in this field.

A Systematic Review of Gait Analysis in the Context of Multimodal Sensing Fusion and AI

BACKGROUND

Neurological diseases are a leading cause of disability and mortality. Gait, or human walking, is a significant predictor of quality of life, morbidity, and mortality. Gait patterns and other kinematic, kinetic, and balance gait features are accurate and powerful diagnostic and prognostic tools.

OBJECTIVE

This review article focuses on the applicability of gait analysis using fusion techniques and artificial intelligence (AI) models. The aim is to examine the significance of mixing several types of wearable and non-wearable sensor data and the impact of this combination on the performance of AI models.

METHOD

In this systematic review, 66 studies using more than two modalities to record and analyze gait were identified. 40 studies incorporated multiple gait analysis modalities without the use of artificial intelligence to extract gait features such as kinematic, kinetic, margin of stability, temporal, and spatial gait parameters, as well as cerebral activity. Similarly, 26 studies analyzed gait data using multimodal fusion sensors and AI algorithms.

RESULTS

The research summarized here demonstrates that the quality of gait analysis and the effectiveness of AI models can both benefit from the integration of data from many sensors. Meanwhile, the utilization of EMG signals in fusion data is especially advantageous.

CONCLUSION

The findings of this review suggest that a smart, portable, wearable-based gait and balance assessment system can be developed using multimodal sensing of the most cutting-edge, clinically relevant tools and technology available. The information presented in this article may serve as a vital springboard for such development.

Artificial Intelligence Approach in Biomechanics of Gait and Sport: A Systematic Literature Review

Background

Artificial neural network helps humans in a wide range of activities, such as sports.

Objective

This paper aims to investigate the effect of artificial intelligence on decision-making related to human gait and sports biomechanics, using computer-based software, and to investigate the impact of artificial intelligence on individuals' biomechanics during gait and sports performance.

Material and Methods

This review was conducted in compliance with the PRISMA guidelines. Abstracts and citations were identified through a search based on Science Direct, Google Scholar, PubMed, Elsevier, Springer Link, Web of Science, and Scopus search engines from 1995 up to 2023 to obtain relevant literature about the impact of artificial intelligence on biomechanics. A total of 1000 articles were found related to biomechanical characteristics of gait and sport and 26 articles were directly pertinent to the subject.

Results

The extent of the application of artificial intelligence in sports biomechanics in various fields. In addition, various variables in the fields of kinematics, kinetics, and the field of time can be investigated based on artificial intelligence. Conventional computational techniques are limited by the inability to process data in its raw form. Artificial Intelligence (AI) and Machine Learning (ML) techniques can handle complex and high-dimensional data.

Conclusion

The utilization of specialized systems and neural networks in gait analysis has shown great potential in sports performance analysis. Integrating AI into this field would be a significant advancement in sport biomechanics. Coaches and athletes can develop more precise training regimens with specialized performance prediction models.

Relationship between gait parameters and cognitive indexes in adult aging

PURPOSE

This study aimed to identify the most effective summary cognitive index predicted from spatio-temporal gait features (STGF) extracted from gait patterns.

METHODS

The study involved 125 participants, including 40 young (mean age: 27.65 years, 50% women), and 85 older adults (mean age: 73.25 years, 62.35% women). The group of older adults included both healthy adults and those with Mild Cognitive Impairment (MCI). Participant´s performance in various cognitive domains was evaluated using 12 cognitive measures from five neuropsychological tests. Four summary cognitive indexes were calculated for each case: 1) the z-score of Mini-Mental State Examination (MMSE) from a population norm (MMSE z-score); 2) the sum of the absolute z-scores of the patients' neuropsychological measures from a population norm (ZSum); 3) the first principal component scores obtained from the individual cognitive variables z-scores (PCCog); and 4) the Mahalanobis distance between the vector that represents the subject's cognitive state (defined by the 12 cognitive variables) and the vector corresponding to a population norm (MDCog). The gait patterns were recorded using a body-fixed Inertial Measurement Unit while participants executed four walking tasks (normal, fast, easy- and hard-dual tasks). Sixteen STGF for each walking task, and the dual-task costs for the dual tasks (when a subject performs an attention-demanding task and walks at the same time) were computed. After applied Principal Component Analysis to gait measures (96 features), a robust regression was used to predict each cognitive index and individual cognitive variable. The adjusted proportion of variance (adjusted-R2) coefficients were reported, and confidence intervals were estimated using the bootstrap procedure.

RESULTS

The mean values of adjusted-R2 for the summary cognitive indexes were as follows: 0.0248 for MMSE z-score, 0.0080 for ZSum, 0.0033 for PCCog, and 0.4445 for MDCog. The mean adjusted-R2 values for the z-scores of individual cognitive variables ranged between 0.0009 and 0.0693. Multiple linear regression was only statistically significant for MDCog, with the highest estimated adjusted-R2 value.

CONCLUSIONS

The association between individual cognitive variables and most of the summary cognitive indexes with gait parameters was weak. However, the MDCog index showed a stronger and significant association with the STGF, exhibiting the highest value of the proportion of the variance that can be explained by the predictor variables. These findings suggest that the MDCog index may be a useful tool in studying the relationship between gait patterns and cognition.

Emotion and motion: Toward emotion recognition based on standing and walking

Emotion recognition is key to interpersonal communication and to human-machine interaction. Body expression may contribute to emotion recognition, but most past studies focused on a few motions, limiting accurate recognition. Moreover, emotions in most previous research were acted out, resulting in non-natural motion, which is unapplicable in reality. We present an approach for emotion recognition based on body motion in naturalistic settings, examining authentic emotions, natural movement, and a broad collection of motion parameters. A lab experiment using 24 participants manipulated participants' emotions using pretested movies into five conditions: happiness, relaxation, fear, sadness, and emotionally-neutral. Emotion was manipulated within subjects, with fillers in between and a counterbalanced order. A motion capture system measured posture and motion during standing and walking; a force plate measured center of pressure location. Traditional statistics revealed nonsignificant effects of emotions on most motion parameters; only 7 of 229 parameters demonstrate significant effects. Most significant effects are in parameters representing postural control during standing, which is consistent with past studies. Yet, the few significant effects suggest that it is impossible to recognize emotions based on a single motion parameter. We therefore developed machine learning models to classify emotions using a collection of parameters, and examined six models: k-nearest neighbors, decision tree, logistic regression, and the support vector machine with radial base function and linear and polynomial functions. The decision tree using 25 parameters provided the highest average accuracy (45.8%), more than twice the random guess for five conditions, which advances past studies demonstrating comparable accuracies, due to our naturalistic setting. This research suggests that machine learning models are valuable for emotion recognition in reality and lays the foundation for further progress in emotion recognition models, informing the development of recognition devices (e.g., depth camera), to be used in home-setting human-machine interactions.

Gait Analysis in Neurorehabilitation: From Research to Clinical Practice

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.

Neurorehabilitation robotics: how much control should therapists have?

Robotic technologies for rehabilitating motor impairments from neurological injuries have been the focus of intensive research and capital investment for more than 30 years. However, these devices have failed to convincingly demonstrate greater restoration of patient function compared to conventional therapy. Nevertheless, robots have value in reducing the manual effort required for physical therapists to provide high-intensity, high-dose interventions. In most robotic systems, therapists remain outside the control loop to act as high-level supervisors, selecting and initiating robot control algorithms to achieve a therapeutic goal. The low-level physical interactions between the robot and the patient are handled by adaptive algorithms that can provide progressive therapy. In this perspective, we examine the physical therapist's role in the control of rehabilitation robotics and whether embedding therapists in lower-level robot control loops could enhance rehabilitation outcomes. We discuss how the features of many automated robotic systems, which can provide repeatable patterns of physical interaction, may work against the goal of driving neuroplastic changes that promote retention and generalization of sensorimotor learning in patients. We highlight the benefits and limitations of letting therapists physically interact with patients through online control of robotic rehabilitation systems, and explore the concept of trust in human-robot interaction as it applies to patient-robot-therapist relationships. We conclude by highlighting several open questions to guide the future of therapist-in-the-loop rehabilitation robotics, including how much control to give therapists and possible approaches for having the robotic system learn from therapist-patient interactions.

Accelerations Recorded by Simple Inertial Measurement Units with Low Sampling Frequency Can Differentiate between Individuals with and without Knee Osteoarthritis: Implications for Remote Health Care

Determining the presence and severity of knee osteoarthritis (OA) is a valuable application of inertial measurement units (IMUs) in the remote monitoring of patients. This study aimed to employ the Fourier representation of IMU signals to differentiate between individuals with and without knee OA. We included 27 patients with unilateral knee osteoarthritis (15 females) and 18 healthy controls (11 females). Gait acceleration signals were recorded during overground walking. We obtained the frequency features of the signals using the Fourier transform. The logistic LASSO regression was employed on the frequency domain features as well as the participant's age, sex, and BMI to distinguish between the acceleration data from individuals with and without knee OA. The model's accuracy was estimated by 10-fold cross-validation. The frequency contents of the signals were different between the two groups. The average accuracy of the classification model using the frequency features was 0.91 ± 0.01. The distribution of the selected features in the final model differed between patients with different severity of knee OA. In this study, we demonstrated that using logistic LASSO regression on the Fourier representation of acceleration signals can accurately determine the presence of knee OA.

Towards automated video-based assessment of dystonia in dyskinetic cerebral palsy: A novel approach using markerless motion tracking and machine learning

Introduction: Video-based clinical rating plays an important role in assessing dystonia and monitoring the effect of treatment in dyskinetic cerebral palsy (CP). However, evaluation by clinicians is time-consuming, and the quality of rating is dependent on experience. The aim of the current study is to provide a proof-of-concept for a machine learning approach to automatically assess scoring of dystonia using 2D stick figures extracted from videos. Model performance was compared to human performance. Methods: A total of 187 video sequences of 34 individuals with dyskinetic CP (8-23 years, all non-ambulatory) were filmed at rest during lying and supported sitting. Videos were scored by three raters according to the Dyskinesia Impairment Scale (DIS) for arm and leg dystonia (normalized scores ranging from 0-1). Coordinates in pixels of the left and right wrist, elbow, shoulder, hip, knee and ankle were extracted using DeepLabCut, an open source toolbox that builds on a pose estimation algorithm. Within a subset, tracking accuracy was assessed for a pretrained human model and for models trained with an increasing number of manually labeled frames. The mean absolute error (MAE) between DeepLabCut's prediction of the position of body points and manual labels was calculated. Subsequently, movement and position features were calculated from extracted body point coordinates. These features were fed into a Random Forest Regressor to train a model to predict the clinical scores. The model performance trained with data from one rater evaluated by MAEs (model-rater) was compared to inter-rater accuracy. Results: A tracking accuracy of 4.5 pixels (approximately 1.5 cm) could be achieved by adding 15-20 manually labeled frames per video. The MAEs for the trained models ranged from 0.21 ± 0.15 for arm dystonia to 0.14 ± 0.10 for leg dystonia (normalized DIS scores). The inter-rater MAEs were 0.21 ± 0.22 and 0.16 ± 0.20, respectively. Conclusion: This proof-of-concept study shows the potential of using stick figures extracted from common videos in a machine learning approach to automatically assess dystonia. Sufficient tracking accuracy can be reached by manually adding labels within 15-20 frames per video. With a relatively small data set, it is possible to train a model that can automatically assess dystonia with a performance comparable to human scoring.

Predicting the metabolic cost of exoskeleton-assisted squatting using foot pressure features and machine learning

Introduction: Recent studies found that wearable exoskeletons can reduce physical effort and fatigue during squatting. In particular, subject-specific assistance helped to significantly reduce physical effort, shown by reduced metabolic cost, using human-in-the-loop optimization of the exoskeleton parameters. However, measuring metabolic cost using respiratory data has limitations, such as long estimation times, presence of noise, and user discomfort. A recent study suggests that foot contact forces can address those challenges and be used as an alternative metric to the metabolic cost to personalize wearable robot assistance during walking. Methods: In this study, we propose that foot center of pressure (CoP) features can be used to estimate the metabolic cost of squatting using a machine learning method. Five subjects' foot pressure and metabolic cost data were collected as they performed squats with an ankle exoskeleton at different assistance conditions in our prior study. In this study, we extracted statistical features from the CoP squat trajectories and fed them as input to a random forest model, with the metabolic cost as the output. Results: The model predicted the metabolic cost with a mean error of 0.55 W/kg on unseen test data, with a high correlation (r = 0.89, p < 0.01) between the true and predicted cost. The features of the CoP trajectory in the medial-lateral direction of the foot (xCoP), which relate to ankle eversion-inversion, were found to be important and highly correlated with metabolic cost. Conclusion: Our findings indicate that increased ankle eversion (outward roll of the ankle), which reflects a suboptimal squatting strategy, results in higher metabolic cost. Higher ankle eversion has been linked with the etiology of chronic lower limb injuries. Hence, a CoP-based cost function in human-in-the-loop optimization could offer several advantages, such as reduced estimation time, injury risk mitigation, and better user comfort.

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

Background

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.

Methods

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.

Gait Trajectory Prediction on an Embedded Microcontroller Using Deep Learning

Achieving a normal gait trajectory for an amputee's active prosthesis is challenging due to its kinematic complexity. Accordingly, lower limb gait trajectory kinematics and gait phase segmentation are essential parameters in controlling an active prosthesis. Recently, the most practiced algorithm in gait trajectory generation is the neural network. Deploying such a complex Artificial Neural Network (ANN) algorithm on an embedded system requires performing the calculations on an external computational device; however, this approach lacks mobility and reliability. In this paper, more simple and reliable ANNs are investigated to be deployed on a single low-cost Microcontroller (MC) and hence provide system mobility. Two neural network configurations were studied: Multi-Layered Perceptron (MLP) and Convolutional Neural Network (CNN); the models were trained on shank and foot IMU data. The data were collected from four subjects and tested on a fifth to predict the trajectory of 200 ms ahead. The prediction was made for two cases: with and without providing the current phase of the gait. Then, the models were deployed on a low-cost microcontroller (ESP32). It was found that with fewer data (excluding the current gait phase), CNN achieved a better correlation coefficient of 0.973 when compared to 0.945 for MLP; when including the current phase, both network configurations achieved better correlation coefficients of nearly 0.98. However, when comparing the execution time required for the prediction on the intended MC, MLP was much faster than CNN, with an execution time of 2.4 ms and 142 ms, respectively. In summary, it was found that when training data are scarce, CNN is more efficient within the acceptable execution time, while MLP achieves relative accuracy with low execution time with enough data.

Computer Vision and Machine Learning-Based Gait Pattern Recognition for Flat Fall Prediction

BACKGROUND

Gait recognition has been applied in the prediction of the probability of elderly flat ground fall, functional evaluation during rehabilitation, and the training of patients with lower extremity motor dysfunction. Gait distinguishing between seemingly similar kinematic patterns associated with different pathological entities is a challenge for the clinician. How to realize automatic identification and judgment of abnormal gait is a significant challenge in clinical practice. The long-term goal of our study is to develop a gait recognition computer vision system using artificial intelligence (AI) and machine learning (ML) computing. This study aims to find an optimal ML algorithm using computer vision techniques and measure variables from lower limbs to classify gait patterns in healthy people. The purpose of this study is to determine the feasibility of computer vision and machine learning (ML) computing in discriminating different gait patterns associated with flat-ground falls.

METHODS

We used the Kinect^® Motion system to capture the spatiotemporal gait data from seven healthy subjects in three walking trials, including normal gait, pelvic-obliquity-gait, and knee-hyperextension-gait walking. Four different classification methods including convolutional neural network (CNN), support vector machine (SVM), K-nearest neighbors (KNN), and long short-term memory (LSTM) neural networks were used to automatically classify three gait patterns. Overall, 750 sets of data were collected, and the dataset was divided into 80% for algorithm training and 20% for evaluation.

RESULTS

The SVM and KNN had a higher accuracy than CNN and LSTM. The SVM (94.9 ± 3.36%) had the highest accuracy in the classification of gait patterns, followed by KNN (94.0 ± 4.22%). The accuracy of CNN was 87.6 ± 7.50% and that of LSTM 83.6 ± 5.35%.

CONCLUSIONS

This study revealed that the proposed AI machine learning (ML) techniques can be used to design gait biometric systems and machine vision for gait pattern recognition. Potentially, this method can be used to remotely evaluate elderly patients and help clinicians make decisions regarding disposition, follow-up, and treatment.

Machine learning corroborates subjective ratings of walking and balance difficulty in multiple sclerosis

Machine learning can discern meaningful information from large datasets. Applying machine learning techniques to raw sensor data from instrumented walkways could automatically detect subtle changes in walking and balance. Multiple sclerosis (MS) is a neurological disorder in which patients report varying degrees of walking and balance disruption. This study aimed to determine whether machine learning applied to walkway sensor data could classify severity of self-reported symptoms in MS patients. Ambulatory people with MS (n = 107) were asked to rate the severity of their walking and balance difficulties, from 1-No problems to 5-Extreme problems, using the MS-Impact Scale-29. Those who scored less than 3 (moderately) were assigned to the “mild” group (n = 35), and those scoring higher were in the “moderate” group (n = 72). Three machine learning algorithms were applied to classify the “mild” group from the “moderate” group. The classification achieved 78% accuracy, a precision of 85%, a recall of 90%, and an F1 score of 87% for distinguishing those people reporting mild from moderate walking and balance difficulty. This study demonstrates that machine learning models can reliably be applied to instrumented walkway data and distinguish severity of self-reported impairment in people with MS.

Examination of a foot mounted IMU-based methodology for a running gait assessment

Gait assessment is essential to understand injury prevention mechanisms during running, where high-impact forces can lead to a range of injuries in the lower extremities. Information regarding the running style to increase efficiency and/or selection of the correct running equipment, such as shoe type, can minimize the risk of injury, e.g., matching a runner's gait to a particular set of cushioning technologies found in modern shoes (neutral/support cushioning). Awareness of training or selection of the correct equipment requires an understanding of a runner's biomechanics, such as determining foot orientation when it strikes the ground. Previous work involved a low-cost approach with a foot-mounted inertial measurement unit (IMU) and an associated zero-crossing-based methodology to objectively understand a runner's biomechanics (in any setting) to learn about shoe selection. Here, an investigation of the previously presented ZC-based methodology is presented only to determine general validity for running gait assessment in a range of running abilities from novice (8 km/h) to experienced (16 km/h+). In comparison to Vicon 3D motion tracking data, the presented approach can extract pronation, foot strike location, and ground contact time with good [ICC(2,1) > 0.750] to excellent [ICC(2,1) > 0.900] agreement between 8-12 km/h runs. However, at higher speeds (14 km/h+), the ZC-based approach begins to deteriorate in performance, suggesting that other features and approaches may be more suitable for faster running and sprinting tasks.

Gait Phase Detection in Walking and Stairs Using Machine Learning

Activity and gait phase recognition algorithms are used in powered motion assistive devices to inform control of motorized components. The objective of this study was to develop a machine learning-based algorithm using inertial measurement data from the thigh and shank to simultaneously detect activity and gait phase (stance, swing) in real-world walking, stair ascent, and stair descent, with the intent of such an algorithm to be used in the control of a motion assistive device local to the knee. Using data from 80 participants, two decision tree and ten long short-term memory (LSTM) models that each used different feature sets and input data were tested and evaluated using a novel performance metric: proportion of perfectly classified strides (PPCS). Separate models were developed to classify i) both activity and gait phase simultaneously (one model predicting six states), and ii) activity-specific models (three individual binary classifiers predicting stance/swing phases). The superior activity-specific model had an accuracy of 98.0% and PPCS of 55.7%. The superior six-phase model used filtered inertial measurement data as its features and a median filter on its predictions and had an accuracy of 92.1% and PPCS of 22.9%. Pooling stance and swing phases from all activities and treating this model as a binary classifier, this model had an accuracy of 97.1%, which may be acceptable for real-world lower limb exoskeleton control if only stance and swing gait phases must be detected.

MaD GUI: An Open-Source Python Package for Annotation and Analysis of Time-Series Data

Developing machine learning algorithms for time-series data often requires manual annotation of the data. To do so, graphical user interfaces (GUIs) are an important component. Existing Python packages for annotation and analysis of time-series data have been developed without addressing adaptability, usability, and user experience. Therefore, we developed a generic open-source Python package focusing on adaptability, usability, and user experience. The developed package, Machine Learning and Data Analytics (MaD) GUI, enables developers to rapidly create a GUI for their specific use case. Furthermore, MaD GUI enables domain experts without programming knowledge to annotate time-series data and apply algorithms to it. We conducted a small-scale study with participants from three international universities to test the adaptability of MaD GUI by developers and to test the user interface by clinicians as representatives of domain experts. MaD GUI saves up to 75% of time in contrast to using a state-of-the-art package. In line with this, subjective ratings regarding usability and user experience show that MaD GUI is preferred over a state-of-the-art package by developers and clinicians. MaD GUI reduces the effort of developers in creating GUIs for time-series analysis and offers similar usability and user experience for clinicians as a state-of-the-art package.

Machine Learning Approach to Support the Detection of Parkinson's Disease in IMU-Based Gait Analysis

The aim of this study was to determine which supervised machine learning (ML) algorithm can most accurately classify people with Parkinson’s disease (pwPD) from speed-matched healthy subjects (HS) based on a selected minimum set of IMU-derived gait features. Twenty-two gait features were extrapolated from the trunk acceleration patterns of 81 pwPD and 80 HS, including spatiotemporal, pelvic kinematics, and acceleration-derived gait stability indexes. After a three-level feature selection procedure, seven gait features were considered for implementing five ML algorithms: support vector machine (SVM), artificial neural network, decision trees (DT), random forest (RF), and K-nearest neighbors. Accuracy, precision, recall, and F1 score were calculated. SVM, DT, and RF showed the best classification performances, with prediction accuracy higher than 80% on the test set. The conceptual model of approaching ML that we proposed could reduce the risk of overrepresenting multicollinear gait features in the model, reducing the risk of overfitting in the test performances while fostering the explainability of the results.

Machine Learning Strategies for Low-Cost Insole-Based Prediction of Center of Gravity during Gait in Healthy Males

Whole-body center of gravity (CG) movements in relation to the center of pressure (COP) offer insights into the balance control strategies of the human body. Existing CG measurement methods using expensive measurement equipment fixed in a laboratory environment are not intended for continuous monitoring. The development of wireless sensing technology makes it possible to expand the measurement in daily life. The insole system is a wearable device that can evaluate human balance ability by measuring pressure distribution on the ground. In this study, a novel protocol (data preparation and model training) for estimating the 3-axis CG trajectory from vertical plantar pressures was proposed and its performance was evaluated. Input and target data were obtained through gait experiments conducted on 15 adult and 15 elderly males using a self-made insole prototype and optical motion capture system. One gait cycle was divided into four semantic phases. Features specified for each phase were extracted and the CG trajectory was predicted using a bi-directional long short-term memory (Bi-LSTM) network. The performance of the proposed CG prediction model was evaluated by a comparative study with four prediction models having no gait phase segmentation. The CG trajectory calculated with the optoelectronic system was used as a golden standard. The relative root mean square error of the proposed model on the 3-axis of anterior/posterior, medial/lateral, and proximal/distal showed the best prediction performance, with 2.12%, 12.97%, and 12.47%. Biomechanical analysis of two healthy male groups was conducted. A statistically significant difference between CG trajectories of the two groups was shown in the proposed model. Large CG sway of the medial/lateral axis trajectory and CG fall of the proximal/distal axis trajectory is shown in the old group. The protocol proposed in this study is a basic step to have gait analysis in daily life. It is expected to be utilized as a key element for clinical applications.

Behavioral Ethics Ecologies of Human-Artificial Intelligence Systems

Historically, evolution of behaviors often took place in environments that changed little over millennia. By contrast, today, rapid changes to behaviors and environments come from the introduction of artificial intelligence (AI) and the infrastructures that facilitate its application. Behavioral ethics is concerned with how interactions between individuals and their environments can lead people to questionable decisions and dubious actions. For example, interactions between an individual’s self-regulatory resource depletion and organizational pressure to take non-ethical actions. In this paper, four fundamental questions of behavioral ecology are applied to analyze human behavioral ethics in human–AI systems. These four questions are concerned with assessing the function of behavioral traits, how behavioral traits evolve in populations, what are the mechanisms of behavioral traits, and how they can differ among different individuals. These four fundamental behavioral ecology questions are applied in analysis of human behavioral ethics in human–AI systems. This is achieved through reference to vehicle navigation systems and healthcare diagnostic systems, which are enabled by AI. Overall, the paper provides two main contributions. First, behavioral ecology analysis of behavioral ethics. Second, application of behavioral ecology questions to identify opportunities and challenges for ethical human–AI systems.

Artificial intelligence in orthopaedic surgery: transforming technological innovation in patient care and surgical training

Artificial intelligence (AI) is an exciting field combining computer science with robust data sets to facilitate problem-solving. It has the potential to transform education, practice and delivery of healthcare especially in orthopaedics. This review article outlines some of the already used AI pathways as well as recent technological advances in orthopaedics. Additionally, this article further explains how potentially these two entities could be combined in the future to improve surgical education, training and ultimately patient care and outcomes.

Effects of Individualized Gait Rehabilitation Robotics for Gait Training on Hemiplegic Patients: Before-After Study in the Same Person

Background

Lower-limb exoskeleton robots are being widely used in gait rehabilitation training for patients with stroke. However, most of the current rehabilitation robots are guided by predestined gait trajectories, which are often different from the actual gait trajectories of specific patients. One solution is to train patients using individualized gait trajectories generated from the physical parameters of patients. Hence, we aimed to explore the effect of individual gaits on energy consumption situations during gait rehabilitation training for hemiplegic patients with lower-limb exoskeleton robots.

Methods

A total of 9 unilateral-hemiplegic patients were recruited for a 2-day experiment. On the first day of the experiment, the 9 patients were guided by a lower-limb exoskeleton robot, walking on flat ground for 15 min in general gait trajectory, which was gained by clinical gait analysis (CGA) method. On the other day, the same 9 patients wore the identical robot and walked on the same flat ground for 15 min in an individualized gait trajectory. The main physiological parameters including heart rate (HR) and peripheral capillary oxygen saturation (SpO2) were acquired via cardio tachometer and oximeter before and after the walking training. The energy consumption situation was indicated by the variation of the value of HR and SpO2 after walking training compared to before.

Results

Between-group comparison showed that the individualized gait trajectory training resulted in an increase in HR levels and a decrease in SpO2 levels compared to the general gait trajectory training. The resulting difference had a statistical significance of p < 0.05.

Conclusion

Using individualized gait guidance in rehabilitation walking training can significantly improve energy efficiency for hemiplegic patients with stroke.

Existing predictive methods applied to gait analysis of patients with diabetes: study protocol for a systematic review

INTRODUCTION

Type 2 diabetes can lead to gait abnormalities, including a longer stance phase, shorter steps and improper foot pressure distribution. Quantitative data from objective methods for evaluating gait patterns are accurate and cost-effective. In addition, it can also help predictive methods to forecast complications and develop early strategies to guide treatments. To date, no research has systematically summarised the predictive methods used to assess type 2 diabetic gait. Therefore, this protocol aims to identify which predictive methods have been employed to assess the diabetic gait.

METHODS AND ANALYSIS

This protocol will follow the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocol (PRISMA-P) statement. Electronic searches of articles from inception to January 2022 will be performed, from May 2021 to 31 January 2022, in the Web of Science, MEDLINE, Embase, IEEE Xplore Digital Library, Scopus, CINAHL, Google Scholar, APA PsycInfo, the Cochrane Library and in references of key articles and grey literature without language restrictions. We will include studies that examined the development and/or validation of predictive methods to assess type 2 diabetic gait in adults aged >18 years without amputations, use of assistive devices, ulcers or neuropathic pain. Two independent reviewers will screen the included studies and extract the data using a customised charting form. A third reviewer will resolve any disagreements. A narrative synthesis will be performed for the included studies. Risk of bias and quality of evidence will be assessed using the Prediction Model Risk of Bias Assessment Tool and the Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis.

ETHICS AND DISSEMINATION

Ethical approval is not required because only available secondary published data will be analysed. The findings will be disseminated through peer-reviewed journals and/or presentations at relevant conferences and other media platforms.

PROSPERO REGISTRATION NUMBER

CDR42020199495.

Insole-Based Systems for Health Monitoring: Current Solutions and Research Challenges

Wearable health monitoring devices allow for measuring physiological parameters without restricting individuals' daily activities, providing information that is reflective of an individual's health and well-being. However, these systems need to be accurate, power-efficient, unobtrusive and simple to use to enable a reliable, convenient, automatic and ubiquitous means of long-term health monitoring. One such system can be embedded in an insole to obtain physiological data from the plantar aspect of the foot that can be analyzed to gain insight into an individual's health. This manuscript provides a comprehensive review of insole-based sensor systems that measure a variety of parameters useful for overall health monitoring, with a focus on insole-based PPD measurement systems developed in recent years. Existing solutions are reviewed, and several open issues are presented and discussed. The concept of a fully integrated insole-based health monitoring system and considerations for future work are described. By developing a system that is capable of measuring parameters such as PPD, gait characteristics, foot temperature and heart rate, a holistic understanding of an individual's health and well-being can be obtained without interrupting day-to-day activities. The proposed device can have a multitude of applications, such as for pathology detection, tracking medical conditions and analyzing gait characteristics.

The application of artificial intelligence and custom algorithms with inertial wearable devices for gait analysis and detection of gait-altering pathologies in adults: A scoping review of literature

Background

The purpose of this scoping review was to explore the current applications of objective gait analysis using inertial measurement units, custom algorithms and artificial intelligence algorithms in detecting neurological and musculoskeletal gait altering pathologies from healthy gait patterns.

Methods

Literature searches were conducted of four electronic databases (Medline, PubMed, Embase and Web of Science) to identify studies that assessed the accuracy of these custom gait analysis models with inputs derived from wearable devices. Data was collected according to the preferred reporting items for systematic reviews and meta-analysis statement guidelines.

Results

A total of 23 eligible studies were identified for inclusion in the present review, including 10 custom algorithms articles and 13 artificial intelligence algorithms articles. Nine studies evaluated patients with Parkinson’s disease of varying severity and subtypes. Support vector machine was the commonest adopted artificial intelligence algorithm model, followed by random forest and neural networks. Overall classification accuracy was promising for articles that use artificial intelligence algorithms, with nine articles achieving more than 90% accuracy.

Conclusions

Current applications of artificial intelligence algorithms are reasonably effective discrimination between pathological and non-pathological gait. Of these, machine learning algorithms demonstrate the additional capacity to handle complicated data input, when compared to other custom algorithms. Notably, there has been increasing application of machine learning algorithms for conducting gait analysis. More studies are needed with unsupervised methods and in non-clinical settings to better reflect the community and home-based usage.

Artificial intelligence, machine learning, and deep learning for clinical outcome prediction

AI is a broad concept, grouping initiatives that use a computer to perform tasks that would usually require a human to complete. AI methods are well suited to predict clinical outcomes. In practice, AI methods can be thought of as functions that learn the outcomes accompanying standardized input data to produce accurate outcome predictions when trialed with new data. Current methods for cleaning, creating, accessing, extracting, augmenting, and representing data for training AI clinical prediction models are well defined. The use of AI to predict clinical outcomes is a dynamic and rapidly evolving arena, with new methods and applications emerging. Extraction or accession of electronic health care records and combining these with patient genetic data is an area of present attention, with tremendous potential for future growth. Machine learning approaches, including decision tree methods of Random Forest and XGBoost, and deep learning techniques including deep multi-layer and recurrent neural networks, afford unique capabilities to accurately create predictions from high dimensional, multimodal data. Furthermore, AI methods are increasing our ability to accurately predict clinical outcomes that previously were difficult to model, including time-dependent and multi-class outcomes. Barriers to robust AI-based clinical outcome model deployment include changing AI product development interfaces, the specificity of regulation requirements, and limitations in ensuring model interpretability, generalizability, and adaptability over time.

Comparison of three methods for reconstructing 3D motion from 2D video recordings for low cost gait analysis systems

Purpose: The aim of this study was to quantify the accuracy of 3D trajectory reconstructions performed from two planar video recordings, using three different reconstruction methods. Additionally, the recordings were carried out using easily available equipment, like built-in cellphone cameras, making the methods suitable for low-cost applications.

Methods: A setup for 3D motion tracking was constructed and used to acquire 2D video recordings subsequently used to reconstruct the 3D trajectories by 1) merging appropriate coordinates, 2) merging coordinates with proportional scaling, and 3) calculating the 3D position based on markers’ projections on the viewing plane. As experimental verification, two markers moving at a fixed distance of 98.9 cm were used to assess the consistency of results. Next, gait analysis in five volunteers was carried out to quantify the differences resulting from different reconstruction methods.

Results: Quantitative evaluation of the investigated 3D trajectories reconstruction methods showed significant differences between those methods, with the worst reconstruction approach resulting in a maximum error of 50% (standard deviation 13%), while the best resulting in a maximum error of 1% (standard deviation 0.44%). The gait analysis results showed differences in mean angles obtained with each reconstruction method reaching only 2°, which can be attributed to the limited measurement volume.

Conclusions: Reconstructing 3D trajectory from 2D views without accounting for the “perspective error” results in significant reconstruction errors. The third method described in this study enables a significant reduction of this issue. Combined with the proposed setup, it provides a functional, low-cost gait analysis system.

Machine Learning-Based Classification of Dependence in Ambulation in Stroke Patients Using Smartphone Video Data

The goal of this study was to develop a framework to classify dependence in ambulation by employing a deep model in a 3D convolutional neural network (3D-CNN) using video data recorded by a smartphone during inpatient rehabilitation therapy in stroke patients. Among 2311 video clips, 1218 walk action cases were collected from 206 stroke patients receiving inpatient rehabilitation therapy (63.24 ± 14.36 years old). As ground truth, the dependence in ambulation was assessed and labeled using the functional ambulatory categories (FACs) and Berg balance scale (BBS). The dependent ambulation was defined as a FAC score less than 4 or a BBS score less than 45. We extracted patient-centered video and patient-centered pose of the target from the tracked target’s posture keypoint location information. Then, the extracted patient-centered video was input in the 3D-CNN, and the extracted patient-centered pose was used to measure swing time asymmetry. Finally, we evaluated the classification of dependence in ambulation using video data via fivefold cross-validation. When training the 3D-CNN based on FACs and BBS, the model performed with 86.3% accuracy, 87.4% precision, 94.0% recall, and 90.5% F1 score. When the 3D-CNN based on FACs and BBS was combined with swing time asymmetry, the model exhibited improved performance (88.7% accuracy, 89.1% precision, 95.7% recall, and 92.2% F1 score). The proposed framework for dependence in ambulation can be useful, as it alerts clinicians or caregivers when stroke patients with dependent ambulatory move alone without assistance. In addition, monitoring dependence in ambulation can facilitate the design of individualized rehabilitation strategies for stroke patients with impaired mobility and balance function.

Frailty Level Classification of the Community Elderly Using Microsoft Kinect-Based Skeleton Pose: A Machine Learning Approach

Frailty is one of the most important geriatric syndromes, which can be associated with increased risk for incident disability and hospitalization. Developing a real-time classification model of elderly frailty level could be beneficial for designing a clinical predictive assessment tool. Hence, the objective of this study was to predict the elderly frailty level utilizing the machine learning approach on skeleton data acquired from a Kinect sensor. Seven hundred and eighty-seven community elderly were recruited in this study. The Kinect data were acquired from the elderly performing different functional assessment exercises including: (1) 30-s arm curl; (2) 30-s chair sit-to-stand; (3) 2-min step; and (4) gait analysis tests. The proposed methodology was successfully validated by gender classification with accuracies up to 84 percent. Regarding frailty level evaluation and prediction, the results indicated that support vector classifier (SVC) and multi-layer perceptron (MLP) are the most successful estimators in prediction of the Fried’s frailty level with median accuracies up to 97.5 percent. The high level of accuracy achieved with the proposed methodology indicates that ML modeling can identify the risk of frailty in elderly individuals based on evaluating the real-time skeletal movements using the Kinect sensor.

Machine learning to extract muscle fascicle length changes from dynamic ultrasound images in real-time

BACKGROUND AND OBJECTIVE

Dynamic muscle fascicle length measurements through B-mode ultrasound have become popular for the non-invasive physiological insights they provide regarding musculoskeletal structure-function. However, current practices typically require time consuming post-processing to track muscle length changes from B-mode images. A real-time measurement tool would not only save processing time but would also help pave the way toward closed-loop applications based on feedback signals driven by in vivo muscle length change patterns. In this paper, we benchmark an approach that combines traditional machine learning (ML) models with B-mode ultrasound recordings to obtain muscle fascicle length changes in real-time. To gauge the utility of this framework for 'in-the-loop' applications, we evaluate accuracy of the extracted muscle length change signals against time-series' derived from a standard, post-hoc automated tracking algorithm.

METHODS

We collected B-mode ultrasound data from the soleus muscle of six participants performing five defined ankle motion tasks: (a) seated, constrained ankle plantarflexion, (b) seated, free ankle dorsi/plantarflexion, (c) weight-bearing, calf raises (d) walking, and then a (e) mix. We trained machine learning (ML) models by pairing muscle fascicle lengths obtained from standardized automated tracking software (UltraTrack) with the respective B-mode ultrasound image input to the tracker, frame-by-frame. Then we conducted hyperparameter optimizations for five different ML models using a grid search to find the best performing parameters for a combination of high correlation and low RMSE between ML and UltraTrack processed muscle fascicle length trajectories. Finally, using the global best model/hyperparameter settings, we comprehensively evaluated training-testing outcomes within subject (i.e., train and test on same subject), cross subject (i.e., train on one subject, test on another) and within/direct cross task (i.e., train and test on same subject, but different task).

RESULTS

Support vector machine (SVM) was the best performing model with an average r = 0.70 ±0.34 and average RMSE = 2.86 ±2.55 mm across all direct training conditions and average r = 0.65 ±0.35 and average RMSE = 3.28 ±2.64 mm when optimized for all cross-participant conditions. Comparisons between ML vs. UltraTrack (i.e., ground truth) tracked muscle fascicle length versus time data indicated that ML tracked images reliably capture the salient qualitative features in ground truth length change data, even when correlation values are on the lower end. Furthermore, in the direct training, calf raises condition, which is most comparable to previous studies validating automated tracking performance during isolated contractions on a dynamometer, our ML approach yielded 0.90 average correlation, in line with other accepted tracking methods in the field.

CONCLUSIONS

By combining B-mode ultrasound and classical ML models, we demonstrate it is possible to achieve real-time tracking of human soleus muscle fascicles across a number of functionally relevant contractile conditions. This novel sensing modality paves the way for muscle physiology in-the-loop applications that could be used to modify gait via biofeedback or unlock novel wearable device control techniques that could enable restored or augmented locomotion performance.

Artificial Neural Network Analyzing Wearable Device Gait Data for Identifying Patients With Stroke Unable to Return to Work

A potential dramatic effect of long-term disability due to stroke is the inability to return to work. An accurate prognosis and the identification of the parameters inflating the possibility of return to work after neurorehabilitation are crucial. Many factors may influence it, such as mobility and, in particular, walking ability. In this pilot study, two emerging technologies have been combined with the aim of developing a prognostic tool for identifying patients able to return to work: a wearable inertial measurement unit for gait analysis and an artificial neural network (ANN). Compared with more conventional statistics, the ANN showed a higher accuracy in identifying patients with respect to healthy subjects (90.9 vs. 75.8%) and also in identifying the subjects unable to return to work (93.9 vs. 81.8%). In this last analysis, the duration of double support phase resulted the most important input of the ANN. The potentiality of the ANN, developed also in other fields such as marketing on social networks, could allow a powerful support for clinicians that today should manage a large amount of instrumentally recorded parameters in patients with stroke.