A proposal to analyze muscle fiber type composition in the soleus muscle of untrained subjects and sprinters using surface EMG signals.

Muscle fiber type proportion is a key determinant of fatigue, force generation, and functions of different skeletal muscles. Analysis of muscle fiber type composition aids in the assessment of athletic abilities and individualization of training methods. This study attempts to non-invasively analyze the muscle fiber type composition in the soleus (SOL) of untrained subjects (UT) and sprinters (SP) using surface electromyography-based time-frequency analysis. Signals are recorded from both groups during an isometric calf raise test with loads until fatigue. Filtered signals are segmented into epochs of 1-s duration and processed using a reassigned Morlet scalogram. Four time-frequency features namely averaged frequency, squared frequency bandwidth, averaged time, and squared time duration are extracted from the reassigned distribution and are subjected to linear regression analysis. A fiber-type-specific reassigned profile is noticed for UT and SP reflecting their distinct muscle composition during their non-fatigue and fatigue states. The regression parameters namely slope, intercept, and Adjusted R-square values are higher for the signals of SP indicating their fast-fatigue characteristics. Greater variation of features during fatigue is noticed in the signals of UT compared to SP. Among the features, the squared time duration exhibits the highest significance of p = 8.75E-07 in differentiating the signals of UT and SP during the non-fatigue state. Thus, the proposed approach is found suitable for analyzing the fiber type differences in both subject groups. This work may be further extended in sports biomechanics for studying the fiber-type transformations in muscles due to different athletic training strategies.