Insole-Based Falling Risk Evaluation for Patients With Parkinson's Disease Through Walking While Holding a Cup of Water

Motor functionality decline caused by Parkinson's disease (PD) can be represented by gait disorders, leading to high risks of falling and serious injuries. Most existing studies on falling risk evaluation mainly rely on gait analysis using wearable sensors but do not usually provide quantitative results to patients. In real-life scenarios, walking is always associated with other activities; thus, gait changes due to distracted attention. This letter aims to investigate how PD patients perform during walking while holding a full cup of water in hand compared to age-matched healthy controls (HC) with Monitoring Insoles (MONI) developed in our laboratory. Experiments, including two tasks, i.e., walking and walking with water in hand, are done with ten PD patients and eight HCs. Results are compared between groups and between tasks through statistical analysis of 63 gait and 864 acceleration-axis features. The results suggest acceleration-axis features can better differentiate PD from HC during walking with water in hand with an accuracy of 82.37% compared to walking (72.63%). To evaluate falling risks with the task of walking with water in hand, a Bayesian Ridge model developed by gait and acceleration-axis features provides falling risk scores compatible with the Fall Risk Questionnaire developed by the U.S. Centers for Disease Control and Prevention with an error ± standard deviation = 0.03 ± 0.38. This letter assesses and evaluates features of PD and HC in walking motion and walking while holding a cup of water to provide a more quantitative and understandable falling risk score to PD patients using MONI. The long-term score profile can provide references to physical therapists to help promote their performance.