• Ken Yoshida

Description

Post-stroke rehabilitation aims to improve motor function to increase a patient's independence and ability to perform activities of daily living. However, assessment of function to determine the efficacy of therapies are currently limited to subjective clinical ratings, such as the Action Research Arm Test (ARAT) or modified Ashworth scale (MAS). Such tests take time away from therapy time at the clinic, subject to assessor variations, lack objective quantitative measures and relevance to the subject's activities of daily living. Quantitative objective measures, such as motion capture, while providing kinematics and quantitative measures take up setup time and require expensive dedicated facilities that are not generally available to most rehabilitation clinics. This project aims to develop non-obtrusive measures through the use of inertial measurement units (IMU) or phone/webcam video capture to enable at-home assessment of hand function as a non-obtrusive means to determine the efficacy of a course of therapy. Specifically, the project aims to 1) Develop the means to measure upper extremity kinematics during the performance of the ARAT or Drink-test, 2) Compare the captured kinematics against the gold-standard motion capture method, 3) establish a link between clinical assessment/rating and 4) develop objective metrics based on objective quantitative measures to determine the quality of hand function in stroke patients.

Relevant Technologies

Inertial Measurement Units, Motion Capture, Computer Vision, Signal Processing, Machine Learning, Biomechanics, Kinematics

Prerequisites

Programming in Matlab, Python

Linear Algebra