Ultrasound Image Analysis on Muscle Texture of Vastus Intermedius and Rectus Femoris Under Neuromuscular Electrical Stimulation

While neuromuscular electrical stimulation (NMES) can restore motor function in individuals with spinal cord injury (SCI), the structure of skeletal muscle that greatly determines muscle function under NMES is not well understood. The present study aims to measure the muscle texture under NMES-induced contraction using ultrasound images and to analyse the predicting factors of the muscle texture using multiple regression. This information will be useful in adjusting the stimulation pattern and intensity to achieve the regression of muscle characteristic parameters from NMES-induced contraction towards voluntary contraction. The quadriceps muscles of 7 normal young adult subjects were tested for 8 cycles of knee extension exercise with/without NMES. Gray level co-occurrence matrix (GLCM) was used for ultrasound image texture analysis. Angular second moment (ASM), contrast value (CON), homogeneity (HOM) and entropy (ENT) of GLCM were calculated as the muscle texture features. The changes in muscle texture were measured using ultrasound imaging, and the relationships between muscle texture features and knee joint angle were examined. The results demonstrated that the significant correlation between muscle texture features and knee joint angle (P < 0.05), which is stronger on Vastus Intermedius (VI) than that on Rectus Femoris (RF) (P < 0.05). Using stepwise multiple regressions, the identified predicting factors of muscle texture feature ratios between the NMES-induced movement and the voluntary movement would be useful for building a stimulation linear model and for accurately modulating stimulation patterns so that the NMES-induced contraction may better approach the voluntary contraction. This study on the muscle texture measurement using ultrasound images and the predicting factors of muscle thickness ratio may provide a potential way to more precisely control the NMES system.