A neural network for predicting knee contact forces from clinic-friendly data

Physics-based neuromusculoskeletal models have been widely used to assess biomechanics not readily measured in laboratory or clinical settings (e.g., joint contact forces). However, these physics-based models require extensive input data and expertise to operate. Artificial intelligence (AI) is a promising alternative to physics-based modelling, as AI can learn complex input–output relationships directly from laboratory data, thus enabling scalable estimation of internal body biomechanics (e.g., joint contact loading) using clinic-friendly data (e.g., video, body measures, etc). In this study, we developed a bi-directional long short-term memory (BiLSTM) model to estimate lower limb joint kinematics, moments and knee contact force using clinic-friendly data (e.g., height, weight, sex, sparse positions of anatomical key points and surface electromyography (EMG)). We also explored the influence of the number and specific combinations of EMG on the BiLSTM prediction accuracy. The BiLSTM model accurately predicted hip, knee, and ankle kinematics (coefficient of determination (R 2 ) = 0.94 ± 0.04; root mean squared error (RMSE) = 3.60 ± 0.92°) and moments (R 2 = 0.91 ± 0.03; RMSE = 0.15 ± 0.02 Nm/kg) as well as knee contact force (R 2 = 0.81 ± 0.11; RMSE = 0.39 ± 0.17 bodyweights (BW)) using an optimal EMG configuration (n = 3). Across all possible EMG numbers and combinations (n = 255), four-quadrant and pareto analyses confirmed vastus medialis, gastrocnemius medialis, and gastrocnemius lateralis muscles consistently contributed to accurate knee contact force prediction. This 3-EMG configuration offered the best trade-off between BiLSTM prediction accuracy and practicality when predicting knee contact force. Results indicate high-performance, minimal-EMG AI models can reduce cost and complexity without compromising accuracy. These findings suggest minimal-EMG AI models could support future adoption of AI-based biomechanical analysis in clinical settings, by reducing cost and complexity without sacrificing accuracy.