Prediction of lower limb joint stiffness and optimization of anthropometric parameters in countermovement jump using an anthropometry-informed neural network.

Abstract

The Countermovement Jump (CMJ) test, widely used to assess athletes' musculoskeletal and neuromuscular readiness, hinges on the performance of the hip, knee, and ankle joints. Despite extensive research, there is no consensus on which joint is most critical for CMJ performance. This study aims to identify the primary lower limb joint contributing to CMJ execution by analyzing maximum energy production and peak stiffness. Additionally, a novel neural network model was developed to predict joint stiffness during CMJ based on jump height and detailed anthropometric parameters, including body fat mass, lower body mass, upper body mass, and skeletal muscle mass ratios. Finally, a genetic algorithm was employed to optimize these parameters, maximizing joint stiffness and energy output. Twelve male athletes performed CMJs, with data cleaning applied to their trials. Energy production and stiffness of the hip, knee, and ankle joints were calculated. The neural network, trained on joint stiffness data, facilitated two optimization problems solved via a genetic algorithm to determine optimal anthropometric parameters for maximizing joint peak stiffness and energy. The hip joint was identified as the primary energy contributor (4.75 ± 1.71 J/kg), while the knee exhibited the highest peak stiffness (0.37 ± 0.04 N.m/°kg). The knee outperformed the hip (0.29 ± 0.02 N.m/°kg) and ankle (0.25 ± 0.04 N.m/°kg) in stiffness. The hip generates the most energy during CMJ, while knee stiffness is crucial. Jump height, body fat, and skeletal muscle mass ratios significantly influence joint stiffness.