Development of a system for quantitative evaluation of motor function using Kinect v2 sensor

Quantitative evaluation of human motor function is a key technology for diagnosing and evaluating movement disorders such as cerebellar ataxia, Parkinson's disease, or stroke. In addition, it is ideal to make the evaluation anywhere, anytime (e.g. at home) for wider clinical applications. Thus, a number of studies have been done to achieve this goal. In our previous study, we developed a novel system for quantitative evaluation of wrist movement using a simple tracking task with a manipulandum. Although the evaluation system was effective, its custom-made input device was expensive. Therefore it was difficult to use it at home to capture daily or weekly changes of patients' conditions repeatedly. To overcome these problems, we have developed a new system for quantitative evaluation of motor function using a Kinect v2 sensor (Microsoft Inc.), which is easily available with low cost. A smooth tracking task employed in our previous system was also used in this study. We examined whether the new system with Kinect v2 sensor had sufficient accuracy and quality of data. We also tested the same analytical methods proposed in our previous study. Preliminary test showed promising performance of the new system for quantitative evaluation of motor function. We are going to connect the system with a cloud database system (e.g., Microsoft Azure) so that the system will be available for patients anywhere, anytime.