Machine-Learning-Based Prediction of Gait Events From EMG in Cerebral Palsy Children

Abstract

Machine-learning techniques are suitably employed for gait-event prediction from only surface electromyographic (sEMG) signals in control subjects during walking. Nevertheless, a reference approach is not available in cerebral-palsy hemiplegic children, likely due to the large variability of foot-floor contacts. This study is designed to investigate a machine-learning-based approach, specifically developed to binary classify gait events and to predict heel-strike (HS) and toe-off (TO) timing from sEMG signals in hemiplegic-child walking. To this objective, sEMG signals are acquired from five hemiplegic-leg muscles in nearly 2500 strides from 20 hemiplegic children, acknowledged as Winters' group 1 and 2. sEMG signals, segmented in overlapping windows of 600 samples (pace = 5 samples), are used to train a multi-layer perceptron model. Intra-subject and inter-subject experimental settings are tested. The best-performing intrasubject approach is able to provide in the hemiplegic population a mean classification accuracy (±SD) of 0.97±0.01 and a suitable prediction of HS and TO events, in terms of average mean absolute error (MAE, 14.8±3.2 ms for HS and 17.6±4.2 ms for TO) and F1-score (0.95±0.03 for HS and 0.92±0.07 for TO). These results outperform previous sEMG-based attempts in cerebral-palsy populations and are comparable with outcomes achieved by reference approaches in control populations. In conclusion, the findings of the study prove the feasibility of neural networks in predicting the two main gait events using surface EMG signals, also in condition of high variability of the signal to predict as in hemiplegic cerebral palsy.

Predicting vertical ground reaction forces from 3D accelerometry using reservoir computers leads to accurate gait event detection

“…(multilayer) perceptron models 28,42,49,51,77 , as well as random forest classifiers 36,42,44 , K-nearest neighbours 42,54,78 , and other types of machine learning using, e.g., Bayesian models 31,32,73,76 , Gaussian mixture model 41 , and principal component analysis 39,40,51,73 . Echo state networks have the great advantage of low computational costs while still showing excellent performance.…”

Flexible Machine Learning Algorithms for Clinical Gait Assessment Tools

“…An alternative to laboratory-based 3D CGA are wearable sensor systems coupled with machine learning analytics [8][9][10][11][12]. These sensor systems use predictive machine learning methods to automatically partition and analyse gait (e.g., foot-contact and foot-off events) from sensor signals (e.g., inertial measurement units (IMUs)) [8][9][10][11][12][13].…”

“…An alternative to laboratory-based 3D CGA are wearable sensor systems coupled with machine learning analytics [8][9][10][11][12]. These sensor systems use predictive machine learning methods to automatically partition and analyse gait (e.g., foot-contact and foot-off events) from sensor signals (e.g., inertial measurement units (IMUs)) [8][9][10][11][12][13]. These sensor systems are mobile, cost-effective and allow the automation of some tasks that currently require laboratory equipment and skilled knowledge [8][9][10][11][12][13].…”

“…These sensor systems use predictive machine learning methods to automatically partition and analyse gait (e.g., foot-contact and foot-off events) from sensor signals (e.g., inertial measurement units (IMUs)) [8][9][10][11][12][13]. These sensor systems are mobile, cost-effective and allow the automation of some tasks that currently require laboratory equipment and skilled knowledge [8][9][10][11][12][13]. For example, partitioning gait into stance and swing phases is an important step in 3D CGA.…”

“…While force-plate-based gait partitioning is very accurate, it (1) requires expensive and stationary equipment, (2) does not usually allow partitioning of more than 2-3 steps within a trial, (3) significantly loses accuracy in severely impaired patients with very small step sizes or impaired foot clearances and (4) requires trained technical staff to manually control the partitioning accuracy and correct partitions if needed. Machine learning-based gait partitioning is a promising alternative because it can be performed with a single, wearable and cheap IMU sensor on the foot or pelvis, which allows the inclusion of more steps in the diagnostic process and can be performed in the patients' home environments [8][9][10][11][12].…”