Machine learning algorithms in spatiotemporal gait analysis can identify patients with Parkinsons disease.

Changes to spatiotemporal gait metrics in gait-altering conditions are characteristic of the pathology. This data can be interpreted by machine learning (ML) models which have recently emerged as an adjunct to clinical medicine. However, the literature is undecided regarding its utility in diagnosing pathological gait and is heterogeneous in its approach to applying ML techniques. This study aims to address these gaps in knowledge. This was a prospective observational study involving 32 patients with Parkinsons disease and 88 normative subjects. Spatiotemporal gait metrics were gathered from all subjects using the MetaMotionC inertial measurement unit and data obtained were used to train and evaluate the performance of 10 machine learning models. Principal component analysis and Genetic Algorithm were amongst the feature selection techniques used. Classification models included Logistic Regression, Support Vector Machine, Naive - Bayes, Random Forest, and Artificial Neural Networks. ML algorithms can accurately distinguish pathological gait in Parkinsons disease from that of normative controls. Two models which used the Random Forest classifier with Principal Component analysis and Genetic Algorithm feature selection techniques separately, were 100% accurate in its predictions and had an F1 score of 1. A third model using principal component analysis and Artificial neural networks was equally as successful (100% accuracy, F1 = 1). We conclude that ML algorithms can accurately distinguish pathological gait from normative controls in Parkinsons Disease. Random Forest classifiers, with Genetic Algorithm feature selection are the preferred ML techniques for this purpose as they produce the highest performing model. Author summaryThe way humans walk, are emblematic of their overall health status. These walking patterns, otherwise, can be captured as gait metrics from small and portable wearable sensors. Data gathered from these sensors can be interpreted by machine learning algorithms which can then be used to accurately distinguish healthy and non-healthy patients based on their gait or walking pattern. The applications of this technology are many and varied. Firstly, it can be used to simply aid in diagnosis as explored in this paper. In future, researchers may use their understanding of normal and pathological gait, and their differences to quantify how severely ones gait is affected in a disease state. This data can be used to track, and quantify, improvements or further deteriorations post treatment, whether these be medication-based or interventions like surgery. Retrospective analyses on data such as this can be used to judge the value of an intervention in reducing a patients disability, and advise health related expenditure.