Human Trunk Correction System to Avoid Bad Postures of Multiple Sclerosis Patients During Workout Sessions Using Image Processing Algorithms

摘要

Regular exercise may have beneficial effects on people with Multiple Sclerosis (MS) and does not have any harmful effects on their MS. However, poorly performed exercises can damage patient's health, primarily at muscle level. In this paper, we pro-pose a system which uses a time-of-flight camera to detect incorrect verticality of patient's trunk and corrects it in real time during exercise. An image processing algorithm has been developed using the depth information obtained from the time-of-flight (ToF) camera. For six months, 20 MS patients tested the system. The statistical analysis based on the Functional Independence Measure (FIM) and the Berg Balance Scale (BBS) shows an overall improvement, 7% and 6% respectively, relative to the control group.