Kinect based Upper Limb Performance Assessment in Daily Life Activities

Surviving stroke is a life changing experience. Rehabilitation therapies for post-stroke patients plays an important role to regain independence and improve their quality of life. However therapies are often costly and requires constant monitoring by a trained physician. In this paper we present a cost effective home based rehabilitation system for post-stroke patients using kinect sensor. Since upper limb movements are greatly impaired after stroke, a quantitative evaluation of the limb functions of the patients is analysed. Healthy subjects, without any form of impairments, were asked to perform predefined actions with their dominant limb. Limb motions of these subjects are captured using kinect sensor. Skeletal data obtained from kinect sensor are processed to compute various kinematic parameters and a database is then created with these computed parameters. These database can be used as reference to compare and asses the values computed from the data of stroke patients who are undergoing therapies for limb evaluation. Thus with these reference values, an automated cost effective, user-friendly instrumental evaluation of post-stroke patients can be made.