Postural instability is one of the main burdens of Parkinson's Disease as it increases the risk of falls and injuries. Monitoring any changes in postural stability, as a consequence of therapies and disease progression, is therefore highly desirable to preserve people's safety and quality of life. In this context, we present a vision-based system built around an optical RGB-Depth device for the automatic evaluation and home monitoring of postural stability. The system is able to track and evaluate body movements, and can be self-managed by people with impairment through an easy-to-use human-machine interaction. A set of static and dynamic balance tasks are delivered and analyzed by the system to estimate some postural and temporal parameters used for the objective and automatic assessment of postural stability. Compliance between automatic and clinical assessment is supported by a machine learning approach with supervised classifiers. Preliminary results of the study are presented and discussed.
A Vision-Based Approach for the at Home Assessment of Postural Stability in Parkinson's Disease
Claudia Ferraris;Roberto Nerino;Antonio Chimienti;
2021
Abstract
Postural instability is one of the main burdens of Parkinson's Disease as it increases the risk of falls and injuries. Monitoring any changes in postural stability, as a consequence of therapies and disease progression, is therefore highly desirable to preserve people's safety and quality of life. In this context, we present a vision-based system built around an optical RGB-Depth device for the automatic evaluation and home monitoring of postural stability. The system is able to track and evaluate body movements, and can be self-managed by people with impairment through an easy-to-use human-machine interaction. A set of static and dynamic balance tasks are delivered and analyzed by the system to estimate some postural and temporal parameters used for the objective and automatic assessment of postural stability. Compliance between automatic and clinical assessment is supported by a machine learning approach with supervised classifiers. Preliminary results of the study are presented and discussed.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.