Abstract
Accurate characterization of dynamic joint stiffness (DJS) is essential for understanding human locomotion and clinical pathologies. However, traditional interfaces for ankle system identification, such as custom fiberglass boots, are limited by fabrication complexity, participant discomfort, and potential mechanical issues that degrade signal fidelity. To address these bottlenecks, we developed a 3D-printed, adjustable Sub-Malleolar Interface (SMI) designed to improve mechanical coupling and experimental efficiency. We validated the SMI against a participant-specific fiberglass legacy boot in a proof-of-concept study involving one healthy participant. Three experimental paradigms were employed: (1) estimation of ankle-foot moment of inertia, (2) identification of relaxed intrinsic and reflex stiffness, and (3) identification of stiffness during 15% maximal voluntary contraction. A parallel-cascade system identification approach was used to decouple intrinsic and reflex contributions to DJS based on torque responses to position perturbations. The SMI provided reliable kinematic and kinetic data. In the relaxed condition, stiffness models achieved high goodness-of-fit (variance-accounted-for ∼91-94%); under tonic contraction, total-torque VAF remained high at most positions but fell below 90% at some angles for both interfaces. Parameter estimates closely matched the legacy boot across all paradigms. Notably, the SMI's improved mechanical coupling was associated with higher-fidelity capture of reflex-mediated torques compared with the legacy boot, which may have attenuated these signals through mechanical compliance. The newly developed sub-malleolar interface functions as a robust, reusable interface for neuromechanical system identification. By providing high-fidelity torque transmission while reducing setup time from hours to minutes, this interface offers a scalable, comfortable alternative to traditional interfaces. This advancement supports more accurate characterization of ankle dynamics and facilitates the translation of system identification methods into broader clinical research settings.