A framework for automated administration of post stroke assessment test

UnrestrictedIn this thesis, we undertake a technology feasibility study of automating the Wolf Motor Function Test(WMFT) using the sensor framework and algorithms that we have developed. WMFT is a standardized assessment test to rate the functional ability (FA) of post stroke individuals. It consists of 17 tasks, 15 of which are scored on the basis of both performance time and quality of movement. We utilize inertial and vision sensors to automate seven of these tasks. The concept of gesture time, which measures the duration for which the arm is actually in motion, is also introduced. We validate the timing accuracy of our prototype system on the seven tasks with two participants. To assess the quality of motion, we propose a novel measurement methodology which involves the sub-categorization of the FA score into individual scores for smoothness, precision, and velocity of movement.We validate the estimation accuracy of the framework with data from one stroke participant. We also discuss the technologies that can be utilized to extend this framework to encompass the remaining WMFT tasks. We believe that the technologies we highlight in this thesis have the capacity to inexpensively enhance the quality and ease of administration of the assessment processes that stroke survivors undergo after hospital discharge. Our long term goal is to be able to automatically: 1) Quantify long term motor functionality changes in real world settings; 2) Evaluate the intervention methodologies; and 3) Use a humanoid robot to provide motivational feedback