Tai Chi Expertise Classification in Older Adults Using Wrist Wearables and Machine Learning.

Tai Chi is a Chinese martial art that provides an adaptive and accessible exercise for older adults with varying functional capacity. While Tai Chi is widely recommended for its physical benefits, wider adoption in at-home practice presents challenges for practitioners, as limited feedback may hamper learning. This study examined the feasibility of using a wearable sensor, combined with machine learning (ML) approaches, to automatically and objectively classify Tai Chi expertise. We hypothesized that the combination of wrist acceleration profiles with ML approaches would be able to accurately classify practitioners' Tai Chi expertise levels. Twelve older active Tai Chi practitioners were recruited for this study. The self-reported lifetime practice hours were used to identify subjects in low, medium, or highly experienced groups. Using 15 acceleration-derived features from a wearable sensor during a self-guided Tai Chi movement and 8 ML architectures, we found multiclass classification performance to range from 0.73 to 0.97 in accuracy and F1-score. Based on feature importance analysis, the top three features were found to each result in a 16-19% performance drop in accuracy. These findings suggest that wrist-wearable-based ML models may accurately classify practice-related changes in movement patterns, which may be helpful in quantifying progress in at-home exercises.

A Deep Learning Approach for Automatic and Objective Grading of the Motor Impairment Severity in Parkinson's Disease for Use in Tele-Assessments.

Wearable sensors provide a tool for at-home monitoring of motor impairment progression in neurological conditions such as Parkinson's disease (PD). This study examined the ability of deep learning approaches to grade the motor impairment severity in a modified version of the Movement Disorders Society-sponsored revision of the Unified Parkinson's Disease Rating Scale (MDS-UPDRS) using low-cost wearable sensors. We hypothesized that expanding training datasets with motion data from healthy older adults (HOAs) and initializing classifiers with weights learned from unsupervised pre-training would lead to an improvement in performance when classifying lower vs. higher motor impairment relative to a baseline deep learning model (XceptionTime). This study evaluated the change in classification performance after using expanded training datasets with HOAs and transferring weights from unsupervised pre-training compared to a baseline deep learning model (XceptionTime) using both upper extremity (finger tapping, hand movements, and pronation-supination movements of the hands) and lower extremity (toe tapping and leg agility) tasks consistent with the MDS-UPDRS. Overall, we found a 12.2% improvement in accuracy after expanding the training dataset and pre-training using max-vote inference on hand movement tasks. Moreover, we found that the classification performance improves for every task except toe tapping after the addition of HOA training data. These findings suggest that learning from HOA motion data can implicitly improve the representations of PD motion data for the purposes of motor impairment classification. Further, our results suggest that unsupervised pre-training can improve the performance of motor impairment classifiers without any additional annotated PD data, which may provide a viable solution for a widely deployable telemedicine solution.

Deep Learning for Multiple Sclerosis Differentiation Using Multi-Stride Dynamics in Gait.

OBJECTIVE: Multiple sclerosis (MS) is a chronic neurological condition of the central nervous system leading to various physical, mental and psychiatric complexities. Mobility limitations are amongst the most frequent and early markers of MS. We evaluated the effectiveness of a DeepMS2G (deep learning (DL) for MS differentiation using multistride dynamics in gait) framework, which is a DL-based methodology to classify multi-stride sequences of persons with MS (PwMS) from healthy controls (HC), in order to generalize over newer walking tasks and subjects. METHODS: We collected single-task Walking and dual-task Walking-while-Talking gait data using an instrumented treadmill from a balanced collection of 20 HC and 20 PwMS. We utilized domain knowledge-based spatiotemporal and kinetic gait features along with two normalization schemes, namely standard size-based and multiple regression normalization strategies. To differentiate between multi-stride sequences of HC and PwMS, we compared 16 traditional machine learning and DL algorithms. Further, we studied the interpretability of our highest-performing models; and discussed the association between the lower extremity function of participants and our model predictions. RESULTS: We observed that residual neural network (ResNet) based models with regression-based normalization were the top performers across both task and subject generalization classification designs. Considering regression-based normalization, a multi-scale ResNet attained a subject classification accuracy and F 1-score of 1.0 when generalizing from single-task Walking to dual-task Walking-while-Talking; and a ResNet resulted in the top subject-wise accuracy and F 1 of 0.83 and 0.81 (resp.), when generalizing over unseen participants. CONCLUSION: We used advanced DL and dynamics across domain knowledge-based spatiotemporal and kinetic gait parameters to successfully classify MS gait across distinct walking trials and unseen participants. SIGNIFICANCE: Our proposed DL algorithms might contribute to efforts to automate MS diagnoses.

A Vision-Based Framework for Predicting Multiple Sclerosis and Parkinson's Disease Gait Dysfunctions-A Deep Learning Approach.

This study examined the effectiveness of a vision-based framework for multiple sclerosis (MS) and Parkinson's disease (PD) gait dysfunction prediction. We collected gait video data from multi-view digital cameras during self-paced walking from MS, PD patients and age, weight, height and gender-matched healthy older adults (HOA). We then extracted characteristic 3D joint keypoints from the collected videos. In this work, we proposed a data-driven methodology to classify strides in persons with MS (PwMS), persons with PD (PwPD) and HOA that may generalize across different walking tasks and subjects. We presented a comprehensive quantitative comparison of 16 diverse traditional machine and deep learning (DL) algorithms. When generalizing from comfortable walking (W) to walking-while-talking (WT), multi-scale residual neural network achieved perfect accuracy and AUC for classifying individuals with a given gait disorder; for subject generalization in W trials, residual neural network resulted in the highest accuracy and AUC of 78.1% and 0.87 (resp.), and 1D convolutional neural network (CNN) had highest accuracy of 75% in WT trials. Finally, when generalizing over new subjects in different tasks, again 1D CNN had the top classification accuracy and AUC of 79.3% and 0.93 (resp.). This work is the first attempt to apply and demonstrate the potential of DL with a multi-view digital camera-based gait analysis framework for neurological gait dysfunction prediction. This study suggests the viability of inexpensive vision-based systems for diagnosing certain neurological disorders.

A Deep Learning Approach for Grading of Motor Impairment Severity in Parkinson's Disease.

Objective and quantitative monitoring of movement impairments is crucial for detecting progression in neurological conditions such as Parkinson's disease (PD). This study examined the ability of deep learning approaches to grade motor impairment severity in a modified version of the Movement Disorders Society-sponsored revision of the Unified Parkinson's Disease Rating Scale (MDS-UPDRS) using low-cost wearable sensors. A convolutional neural network architecture, XceptionTime, was used to classify lower and higher levels of motor impairment in persons with PD, across five distinct rhythmic tasks: finger tapping, hand movements, pronation-supination movements of the hands, toe tapping, and leg agility. In addition, an aggregate model was trained on data from all tasks together for evaluating bradykinesia symptom severity in PD. The model performance was highest in the hand movement tasks with an accuracy of 82.6% in the hold-out test dataset; the accuracy for the aggregate model was 79.7%, however, it demonstrated the lowest variability. Overall, these findings suggest the feasibility of integrating low-cost wearable technology and deep learning approaches to automatically and objectively quantify motor impairment in persons with PD. This approach may provide a viable solution for a widely deployable telemedicine solution.

Application of Aligned-UMAP to longitudinal biomedical studies.

High-dimensional data analysis starts with projecting the data to low dimensions to visualize and understand the underlying data structure. Several methods have been developed for dimensionality reduction, but they are limited to cross-sectional datasets. The recently proposed Aligned-UMAP, an extension of the uniform manifold approximation and projection (UMAP) algorithm, can visualize high-dimensional longitudinal datasets. We demonstrated its utility for researchers to identify exciting patterns and trajectories within enormous datasets in biological sciences. We found that the algorithm parameters also play a crucial role and must be tuned carefully to utilize the algorithm's potential fully. We also discussed key points to remember and directions for future extensions of Aligned-UMAP. Further, we made our code open source to enhance the reproducibility and applicability of our work. We believe our benchmarking study becomes more important as more and more high-dimensional longitudinal data in biomedical research become available.

Identification and prediction of Parkinson's disease subtypes and progression using machine learning in two cohorts.

The clinical manifestations of Parkinson's disease (PD) are characterized by heterogeneity in age at onset, disease duration, rate of progression, and the constellation of motor versus non-motor features. There is an unmet need for the characterization of distinct disease subtypes as well as improved, individualized predictions of the disease course. We used unsupervised and supervised machine learning methods on comprehensive, longitudinal clinical data from the Parkinson's Disease Progression Marker Initiative (n = 294 cases) to identify patient subtypes and to predict disease progression. The resulting models were validated in an independent, clinically well-characterized cohort from the Parkinson's Disease Biomarker Program (n = 263 cases). Our analysis distinguished three distinct disease subtypes with highly predictable progression rates, corresponding to slow, moderate, and fast disease progression. We achieved highly accurate projections of disease progression 5 years after initial diagnosis with an average area under the curve (AUC) of 0.92 (95% CI: 0.95 ± 0.01) for the slower progressing group (PDvec1), 0.87 ± 0.03 for moderate progressors, and 0.95 ± 0.02 for the fast-progressing group (PDvec3). We identified serum neurofilament light as a significant indicator of fast disease progression among other key biomarkers of interest. We replicated these findings in an independent cohort, released the analytical code, and developed models in an open science manner. Our data-driven study provides insights to deconstruct PD heterogeneity. This approach could have immediate implications for clinical trials by improving the detection of significant clinical outcomes. We anticipate that machine learning models will improve patient counseling, clinical trial design, and ultimately individualized patient care.

Predicting Multiple Sclerosis From Gait Dynamics Using an Instrumented Treadmill: A Machine Learning Approach.

OBJECTIVE: Multiple Sclerosis (MS) is a neurological condition which widely affects people 50-60 years of age. While clinical presentations of MS are highly heterogeneous, mobility limitations are one of the most frequent symptoms. This study examines a machine learning (ML) framework for identifying MS through spatiotemporal and kinetic gait features. METHODS: In this study, gait data during self-paced walking on an instrumented treadmill from 20 persons with MS and 20 age, weight, height, and gender-matched healthy older adults (HOA) were obtained. We explored two strategies to normalize data and minimize dependence on subject demographics; size-normalization (standard body size-based normalization) and regress-normalization (regression-based normalization using scaling factors derived by regressing gait features on multiple subject demographics); and proposed an ML based methodology to classify individual strides of older persons with MS (PwMS) from healthy controls. We generalized both across different walking tasks and subjects. RESULTS: We observed that regress-normalization improved the accuracy of identifying pathological gait using ML when compared to size-normalization. When generalizing from comfortable walking to walking while talking, gradient boosting machine achieved the optimal subject classification accuracy and AUC of 94.3 and 1.0, respectively and for subject generalization, a multilayer perceptron resulted in the best accuracy and AUC of 80% and 0.86, respectively, both with regression-normalized data. CONCLUSION: The integration of gait data and ML may provide a viable patient-centric approach to aid clinicians in monitoring MS. SIGNIFICANCE: The results of this study have future implications for the way regression normalized gait features may be clinically used to design ML-based disease prediction strategies and monitor disease progression in PwMS.

Feasibility of VR Technology in Eliciting State Anxiety Changes While Walking in Older Women.

Virtual reality (VR) technology offers an exciting way to emulate real-life walking conditions that may better elicit changes in emotional state. We aimed to determine whether VR technology is a feasible way to elicit changes in state anxiety during walking. Electrocardiogram data were collected for 18 older adult women while they navigated a baseline walking task, a dual walking task, and four walking VR environments. Using heart rate variability (HRV) analysis, we found that all four of the VR environments successfully elicited a significantly higher level of state anxiety as compared to the walking baseline, with 84% of participants eliciting a significantly lower HRV in each of the four VR conditions as compared to baseline walking. VR was also found to be a more reliable tool for increasing state anxiety as compared to a dual task, where only 47% of participants demonstrated a significantly lower HRV as compared to baseline walking. VR, therefore, could be promising as a tool to elicit changes in state anxiety and less limited in its ability to elicit changes as compared to a traditional dual task condition.

Automatic Identification of Brain Independent Components in Electroencephalography Data Collected while Standing in a Virtually Immersive Environment - A Deep Learning-Based Approach.

Electroencephalography (EEG) is a commonly used method for monitoring brain activity. Automating an EEG signal processing pipeline is imperative to the exploration of real-time brain computer interface (BCI) applications. EEG analysis demands substantial training and time for removal of distinct unwanted independent components (ICs), generated via independent component analysis, corresponding to artifacts. The considerable subject-wise variations across these components motivates defining a procedural way to identify and eliminate these artifacts. We propose DeepIC-virtual, a convolutional neural network (CNN) deep learning classifier to automatically identify brain components in the ICs extracted from the subject's EEG data gathered while they are being immersed in a virtual reality (VR) environment. This work examined the feasibility of DL techniques to provide automated ICs classification on noisy and visually engaging upright stance EEG data. We collected the EEG data for six subjects while they were standing upright in a VR testing setup simulating pseudo-randomized variations in height and depth conditions and induced perturbations. An extensive 1432 IC representation images data set was generated and manually labelled via an expert as brain components or one of the six distinct removable artifacts. The supervised CNN architecture was utilized to categorize good brain ICs and bad artifactual ICs via generated images of topographical maps. Our model categorizing good versus bad IC topographical maps resulted in a binary classification accuracy and area under curve of 89.20% and 0.93 respectively. Despite significant imbalance, only 1 out of the 57 present brain ICs in the withheld testing set was miss-classified as an artifact. These results will hopefully encourage clinicians to integrate BCI methods and neurofeedback to control anxiety and provide a treatment of acrophobia, given the viability of automatic classification of artifactual ICs.

Exploration of Machine Learning to Identify Community Dwelling Older Adults with Balance Dysfunction Using Short Duration Accelerometer Data.

The incidence of fall-related injuries in older adults is high. Given the significant and adverse outcomes that arise from injurious falls in older adults, it is of the utmost importance to identify older adults at greater risk for falls as early as possible. Given that balance dysfunction provides a significant risk factor for falls, an automated and objective identification of balance dysfunction in community dwelling older adults using wearable sensor data when walking may be beneficial. In this study, we examine the feasibility of using wearable sensors, when walking, to identify older adults who have trouble with balance at an early stage using state-of-the-art machine learning techniques. We recruited 21 community dwelling older women. The experimental paradigm consisted of two tasks: Normal walking with a self-selected comfortable speed on an instrumented treadmill and a test of reflexive postural response, using the motor control test (MCT). Based on the MCT, identification of older women with low or high balance function was performed. Using short duration accelerometer data from sensors placed on the knee and hip while walking, supervised machine learning was carried out to classify subjects with low and high balance function. Using a Gradient Boosting Machine (GBM) algorithm, we classified balance function in older adults using 60 seconds of accelerometer data with an average cross validation accuracy of 91.5% and area under the receiver operating characteristic curve (AUC) of 0.97. Early diagnosis of balance dysfunction in community dwelling older adults through the use of user friendly and inexpensive wearable sensors may help in reducing future fall risk in older adults through earlier interventions and treatments, and thereby significantly reduce associated healthcare costs.

Exploring Characteristic Features in Gait Patterns for Predicting Multiple Sclerosis.

Multiple Sclerosis (MS), an autoimmune and demyelinating disease, is one the most prevalent neurological disabilities in young adults. It results in damage of the central nervous system, disrupting communication between the patient's brain, spinal cord and body. Mobility limitations is one of the earliest symptoms and affects a majority of persons with Multiple Sclerosis. We are working towards an effort to characterize individuals with MS, from those without, on the basis of variations in the gait patterns. In the proposed work, statistical methods were used to identify differentiating gait data features for MS characterization. The prediction algorithms built upon these characteristic features will help clinicians develop effective and early cure and therapy designs for persons with Multiple Sclerosis.

Using Virtual Reality to Examine the Neural and Physiological Responses to Height and Perturbations in Quiet Standing.

We describe an experimental setup, which uses virtual reality to understand neural responses to height and perturbations in human postural control. This system could help clinicians develop better methods to alleviate symptoms from a significant fear of heights, especially in the elderly and those with movement disorders, such as Parkinson's disease. In our design, EEG and EKG systems monitor the participants' neural responses and heart activities respectively, while they try to maintain balance on a force plate in an induced virtual world, experiencing randomized height changes and perturbations. These responses are then analyzed to understand the participants' anxiety caused by height and postural challenges.

Virtual Reality, Visual Cliffs, and Movement Disorders.

We outline an experimental setup designed to dynamically understand neural responses to visual cliffs while walking. The goal of our work is understanding and mitigating fear of falling, particularly among the elderly. In our setup, an EEG cap monitors a subject's neural activity while the subject is immersed in a virtual world and walking on an instrumented treadmill. The subject's response to visual stimuli is measured by both the EEG cap and by speed and pressure data from the treadmill. Based on this data, we can dynamically alter the landscape in the virtual world. We hope that our setup may be useful in helping subjects develop mechanisms to compensate for significant fear of falling while walking.