Combining Static Stance and Dynamic Gait Personalization to Improve the Accuracy of Patient-Specific Finite-Element...

Finite element (FE) foot models are increasingly used to support clinical decision-making in hallux valgus (HV) surgery, yet most rely only on static standing simulations. Such calibration remains under-constrained, validating non-unique mechanical model parameters that fit static data but may fail to represent dynamic mechanics limiting clinical reliability. This study evaluated whether introducing dynamic gait loading and multi-metric validation improves parameter identifiability and increases patient-specific clinical validity.

Prior to surgery, a patient with HV underwent plantar pressure and motion analysis in static and walking. Patient-specific model was developed by morphing a standard FE foot mesh to the CT geometry. Then, thirteen biomechanical foot parameters were varied by Latin Hypercube Sampling. The 160 static simulations obtained were evaluated against fourteen targets from static experimental data. Sets with < 5% mean static error were finally used in FE time-dependent dynamic gait simulation driven by patient data. Predictions were compared for bone motion, center-of-pressure path, and plantar pressures at three instants of the gait.

Static study produced 9.9% average error across all models (eight models < 5% error) while input values differed widely. Dynamic validation split these candidates: one set kept mean errors < 13% whereas others rose to 25%, with substantial discrepancies in forefoot loading and kinematics.

Static plausibility does not guarantee dynamic validity. Dual loading combined with multi-metric validation proved to be essential to isolate physiologically coherent parameter sets. Resulting pipeline strengthens the credibility of HV digital twins for future clinical decision-making.