Evaluation of Shoulder Functionality Based on Accelerometer and Bending Sensor

• Main Authors: Hao-Yuan Chen, 陳皓遠
• Other Authors: Hao-Hua Chu
• Format: Others
• Language: en_US
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/37339095217563308979

碩士 === 國立臺灣大學 === 資訊工程學研究所 === 97 === Persistent rehabilitation can help post-operation patients maintain functionality of shoulder motion. It prevents body from more severe symptoms such as lymphedema and keeps well capability of doing activities of daily life. However, lots of patients always ignore the importance of persistent rehabilitation due to the deficiency of self-awareness about body status. It makes patients do not spontaneously continue rehabilitation for at least one year. Therefore, it is necessary to provide a mechanism that makes patients gain much incentive to motivate them to do rehabilitation more frequently. Based on the signals from accelerometer and bending sensor, this thesis adopts the supervised learning technique to implement the shoulder functionality evaluation system with linear regression model. The system identifies the most stable shoulder range of motion according to the six shoulder evaluation exercises performed by patients. It also shows the assessment for capability of activities of daily life in order to instantly boost self-awareness about shoulder health status. Not only the feasibility study shows that the mean square error in prediction angle of shoulder range of motion is below 12 degree, but also the system obtains the positive affirmation from real-world patients and physical therapist after they try out the system.