Computationally tuned dual-layer lattice pads adapted to gait-induced pressure distribution

Patients with Becker muscular dystrophy (BMD) often experience forefoot overload due to abnormal gait patterns, resulting in chronic pain, instability, and ulceration. Conventional orthotic devices frequently fail to accommodate the complex and individualized biomechanical needs of these patients. Here, we introduce a computational framework that integrates finite element analysis, machine learning, and Bayesian optimization to design dual-layer lattice pads adapted to gait-induced pressure distribution. A Gaussian Process Regression model accurately predicted structure–function relationships, enabling efficient multi-objective optimization within a high-dimensional design space. The optimized lattice configuration reduced peak plantar pressure by 51.36% during simulated gait, outperforming previous solutions. Unlike conventional uniform-stiffness insoles, the dual-layer architecture allows gradient mechanical tuning through parametric control, providing enhanced pressure offloading and structural adaptability. This work proposes a computationally scalable and theoretically generalizable framework for precision orthotic design and outlines a pathway for integrating computational biomechanics with AI-informed material engineering in future personalized rehabilitation technologies.