Distinct tibialis anterior electromyography profiles in strength and endurance athletes revealed by principal component analysis

Recent data science advances have provided powerful methods to utilise biomechanical data sets through machine learning (ML) techniques. This includes classification of various populations from discriminating kinematic, kinetic, and electromyography (EMG) measures, particularly using support vector machines (SVM) [1]. Despite this, there is limited SVM literature on athletic characteristics. Since muscle fibre type is linked to exercise-specific adaptations [2], SVM classification of athlete types may provide insight into these adaptations. Our previous work used a forward selection approach in SVM models to distinguish between strength- and endurance-trained athletes, identifying the tibialis anterior (TA) as the most discriminatory muscle with 82.83% accuracy in the best performing model using only TA EMG features [3]. Building upon and extending these previous findings, this study used Principal Component Analysis (PCA) to examine differences in TA EMG stance phase patterns between these same strength- and endurance-trained athletes.