Muscle Movement Tracking from Ultrasound Image Sequences with Optical Flow

Understanding the correlation between muscle movement speed and change in the muscle contraction force has a great application in rehabilitation treatment and muscle evaluation. In this study, six healthy subjects were recruited to take part in the elbow joint flexion and extension experiment. Then, we adapted the optical flow method to track the human muscle movement from ultrasound image sequences. The tracking results demonstrate the feasibility of the optical flow method which can be successfully used for tracking human muscle movement. Then, we compared horizon average velocity (HAV) extracted by the optical flow method from ultrasound image with the measured angular velocity of the elbow joint. The comparison shows that both of them have similar tendencies. Furthermore, we evaluated how HAV and the measured angular velocity of the elbow joint are correlated. The evaluation shows that the correlation coefficient (CC) which has an average of 0.8490 and a variance of 0.0593 varies in the range of 0.7483-0.9102, while root mean square error (RMSE) which has an average of 0.0792 and a variance of 0.0107 varies in the range of 0.0680-0.0989. Therefore, it is concluded that, the HAV and angular velocity of the elbow joint are highly correlated.