Footprints as Morphometric Evidence for Somatic Prediction and Body Proportion Reconstruction in Forensic Medicine

Background/Objectives: This study aimed to apply morphometric and multivariate analytical techniques to footprint evidence for forensic identity determination, focusing on sex estimation and the reconstruction of body proportions as components of the biological profile. By integrating detailed footprint metrics with body measurements, the research sought to develop discriminant and regression models to evaluate the predictive value of footprint metrics for sex estimation and selected somatic dimensions. Methods: Static bilateral footprints were obtained using charcoal powder impressions and digitized using ImageJ. Eleven footprint parameters (F1-F11) and eleven body measurements (B1-B11) were recorded. Sex-based differences were examined using appropriate parametric or non-parametric tests with effect sizes. Sex estimation was evaluated using discriminant function analysis and internally validated using leave-one-out and stratified 10-fold cross-validation. Regression models for stature and body dimension estimation were assessed with multicollinearity diagnostics and repeated 10-fold cross-validation, including RMSE, MAE, and cross-validated R². Results: The apparent discriminant classification accuracies were 74.0% for the right foot and 71.0% for the left foot. After internal validation, classification performance decreased to approximately 64-67%, indicating moderate discriminative ability. Reduced regression models showed the most stable validated performance for stature and arm span, although cross-validated R² values remained weak. Conclusions: Static footprint morphometry may provide supportive information for sex estimation and selected somatic dimensions in this Turkish adult sample. However, the validated performance indicates that these models should be interpreted as ancillary and exploratory tools rather than standalone forensic identification methods.