The study of bioelectrodynamics, which examines the interplay between electromagnetism and mechanics in the human body, has gained considerable interdisciplinary attention in recent years. The human body continuously generates endogenous electromagnetic (EM) fields through various biophysical and biochemical processes, and these signals can be detected using advanced sensing and modelling techniques for diagnostic and therapeutic applications. Previous studies have employed computational methods such as KNN classifiers and image processing to segment electromagnetic (EM) radiation or visualize bio-fields. However, the use of a feature-selection-based stepwise logistic regression approach to distinguish gender-specific EM radiation patterns across four body segments remains unexplored. This study aims to develop and evaluate a stepwise logistic regression model to classify EM radiation characteristics between males and females and to identify the most influential predictors contributing to gender differentiation. With fifty-two participants were involved in the study, the results suggest that anthropometric measurements from both upper body and arm segments are reliable indicators for gender determination with p-values of 0.001 and 0.002, respectively, with the upper body (CF) providing marginally better performance. The logistic regression models for both upper and left body segments also demonstrated excellent fit and predictive accuracy, with high deviance R² values (63.37% and 71.56%) and strong discrimination performance (AUC = 0.9622 and 0.9754), effectively distinguishing gender-related EM radiation patterns. This finding confirms the stepwise logistic regression model successfully classified gender-specific EM radiation with upper and left body segments serve as reliable indicators for gender differentiation in bioelectrodynamic analysis.