Novel machine learning-based hybrid strategy for severity assessment of Parkinson's disorders.

Parkinson's disease (PD) severity assessment in clinical settings largely depends on expertise level of clinicians which have inherent limitations and non-uniformity. Instrumented gait analysis plays a significant role in disease diagnosis and management. However, these are agonized from data dispersion due to demography, anthropometry, and self-selected walking speed of an individual. This research work aims to develop computationally efficient hybrid strategy using normalized features for PD severity evaluation. The relevance of each considered gait feature in demonstrating the outcomes is explained through feature importance and partial dependence plot (PDP) to build substantial insight for clinical needs. Gait, a biomarker, is used to access human mobility by utilizing vertical ground reaction force (VGRF) data of 72 healthy and 93 Parkinson's individuals. A multi-variate normalization approach identifies gait differences between PD and non-PD. The proposed hybrid model used is able to detect PD with accuracy of 99.39% and 99.9%, and its severity assessment based on MDS-UPDRS-III shows coefficient of determination (R2) as 97% and 98.7% using leave-one-out cross-validation (CV) and tenfold CV respectively. The significant features suitable for clinical implications are reported. Moreover, normalized gait parameters supplement capability to compare individuals with diverse physical properties, resulting in assistive system for evaluation of PD severity.

Role of machine learning in gait analysis: a review.

Human biomechanics and gait form an integral part of life. The gait analysis involves a large number of interdependent parameters that were difficult to interpret due to a vast amount of data and their inter-relations. To simplify evaluation, the integration of machine learning (ML) with biomechanics is a promising solution. The purpose of this review is to familiarise the readers with key directions of implementation of ML techniques for gait analysis and gait rehabilitation. An extensive literature survey was based on research articles from nine databases published from 1980 to 2019. With over 943 studies identified, finally, 43 studies met the inclusion criteria. The outcome reported illustrates that supervised ML techniques showed accuracies above 90% in the identified gait analysis domain. The statistical results revealed support vector machine (SVM) as the best classifier (mean-score = 0.87 ± 0.07) with remarkable generalisation capability even on small to medium datasets. It has also been analysed that the control strategies for gait rehabilitation are benefitted from reinforcement learning and (deep) neural-networks due to their ability to capture participants' variability. This review paper shows the success of ML techniques in detecting disorders, predicting rehabilitation length, and control of rehabilitation devices which make them suitable for clinical diagnosis.