Study On Muscle Morphology And Activity Under Functional Electrical Stimulation Based On Ultrasound Image Analysis

Nowadays, the incidence of paraplegia induced by spinal cord injury (SCI) has been increased significantly. The rebuilding of limb function is a key and difficult point in rehabilitation of paraplegic patients. Functional electrical stimulation (FES) is widely considered as an effective tool in this field.But lots of new technologies of FES have been employed only in laboratory, the stimulation patterns and the effects of FES in clinical application are very limited, one of the reasons is that the mechanism research of FES is still lacking, and there is few research on the macro-level description of how it affects muscle functional activities and status. The state of skeletal muscle contraction and relaxation is the basis of the human body movement, the structure of skeletal muscle is the primary determinant of its function and activity. Therefore, in order to achieve the purpose of the rebuilding of limb function and its rapid development and application, the structure and function of target skeletal muscle must be explored more deeply under FES.Same-angle contraction and same-torque contraction conditions were designed in this research. FES-induced movement and voluntary movement patterns were included in each contraction. B-mode ultrasound imaging system is used to image the target muscle under FES-induced pattern, the information of muscle activity including muscle thickness and texture features was obtained and at the same time joint angles were collected, then angle is as a feedback to collect muscle image under voluntary movement pattern.First, the correlation between muscles'characteristics parameters including thickness and texture characteristics and joint angle were calculated and the significant result indicates that the periodic FES is the direct reason for periodic muscle activity. Second, the muscle characteristic parameters differences between FES-induced contraction and voluntary contraction with the same movement effect were compared based on proportion, then the factors causing this difference were obtained through analyzing the subject individual differences. The multiple linear regression method was used to find the main factor. The main factor was adjusted through adjusting the stimulation pattern and intensity to achieve the regression of muscle characteristic parameters from FES-induced contraction towards voluntary contraction, which led to the regression of muscle activity and status.In this study, the machine learning methods including support vector machine and artificial neural network were used to predict the joint angle based on the default stimulation intensity, muscle thickness and texture information. The result shows that the method of combining stimulation intensity and muscle information is superior to stimulation intensity only as input. This indicates muscle thickness and texture information play an important role in reflecting muscle activity and status, and it provides a potential stimulation method in the precise control of FES system.