Gait Analysis for Parkinson’s Patients by using Wrist Acceleration

Gait analysis is an important method used to assess the bio-mechanics of human movement with various applications in bio-mechanics rehabilitation, and sports medicine. Traditional methods that implement force plates or motion capture systems are costly and require specialized equipment and expertise, which causes continuous medical check-ups, and therefore are unaffordable for individuals who have financial difficulties. Recently, wearable sensors like IMU sensors, smartwatches, and smart insoles have garnered more popularity in terms of gait analysis. The smartwatch is a practical device that can be used on a daily basis to quantify gait parameters from collected wrist acceleration data. The objective of this study is to document a systematic literature survey on the effectiveness of using wrist-wearable device to estimate PD patients’ gait motions. It is determined that Support Vector Machine (SVM), Long-short Term Memory (LSTM), and Convolutional Neural Network (CNN) are the potential algorithms that can provide promising results for Parkinson’s disease (PD) patients classifications and continuous lower limb motion estimation of PD patients.