Citation: Marquez Chavez, J.; Tang,
W. A Vision-Based System for Stage
Classification of Parkinsonian Gait
Using Machine Learning and
Synthetic Data. Sensors 2022, 22, 4463.
https://doi.org/10.3390/s22124463
Academic Editor: M. Jamal Deen,
Subhas Mukhopadhyay, Yangquan
Chen, Simone Morais, Nunzio
Cennamo and Junseop Lee
Received: 18 April 2022
Accepted: 18 May 2022
Published: 13 June 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Article
A Vision-Based System for Stage Classification of Parkinsonian
Gait Using Machine Learning and Synthetic Data
Jorge Marquez Chavez
1
and Wei Tang
2,
*
1
Department of Physics, New Mexico State University, Las Cruces, NM 88003, USA; jorgemar@nmsu.edu
2
Klipsch School of Electrical Engineering, New Mexico State University, Las Cruces, NM 88003, USA
* Correspondence: wtang@nmsu.edu
Abstract:
Parkinson’s disease is characterized by abnormal gait, which worsens as the condition
progresses. Although several methods have been able to classify this feature through pose-estimation
algorithms and machine-learning classifiers, few studies have been able to analyze its progression
to perform stage classification of the disease. Moreover, despite the increasing popularity of these
systems for gait analysis, the amount of available gait-related data can often be limited, thereby,
hindering the progress of the implementation of this technology in the medical field. As such, creating
a quantitative prognosis method that can identify the severity levels of a Parkinsonian gait with little
data could help facilitate the study of the Parkinsonian gait for rehabilitation. In this contribution,
we propose a vision-based system to analyze the Parkinsonian gait at various stages using linear
interpolation of Parkinsonian gait models. We present a comparison between the performance of a
k-nearest neighbors algorithm (KNN), support-vector machine (SVM) and gradient boosting (GB)
algorithms in classifying well-established gait features. Our results show that the proposed system
achieved 96–99% accuracy in evaluating the prognosis of Parkinsonian gaits.
Keywords: Parkinson’s disease; gait analysis; vision-based system
1. Introduction
Parkinson’s disease (PD) is a neurodegenerative disease that is characterized by motor
symptoms, such as tremor, rigidity and bradykinesia [
1
]. These movement impairments
are directly linked to a variety of abnormal gait patterns in PD patients, which can sub-
sequently increase the risk of injury and affect the overall quality of life [
2
]. The current
gold standard for the diagnosis and monitoring of the Parkinsonian gait (PG) generally
consists of clinical evaluation. However, the criteria used is often based on the examiner’s
expertise, and the implementation can therefore attach unwanted subjective components to
the analysis [
3
]. Moreover, since PD is a progressive condition, the variation introduced
by human interpretation makes clinical-based evaluations further unsuitable for the early
detection of PG.
To avoid the inaccuracy of these methods, several technological advances have made
it possible to perform gait analysis through specialized equipment. Previous studies,
for example, have consisted on the use of walkways or wearable sensors [
4
,
5
] to obtain
important features and perform gait analysis [
6
,
7
]. Still, although these techniques provide
a rich quantitative examination that can produce relevant data not observed by the eye,
they are generally considered impractical, as they often require costly additional equipment
that can be inconvenient to both the examiners and the patient [2,8,9].
Furthermore, the results obtained with similar body-worn technology were shown
to be affected by the patients changing their walking pattern upon acknowledging the
sensors themselves, leading to inaccurate results [
10
]. In this context, the development of
a vision-based system has gained increased attention because of its ability to objectively
quantify gait features through a practical camera setup and machine-learning algorithms.
Sensors 2022, 22, 4463. https://doi.org/10.3390/s22124463 https://www.mdpi.com/journal/sensors
