A Refined Proprioceptive Feedback Framework for Spatiotemporal Mapping of Spinal Afferent Input During Gait in Female Knee Osteoarthritis Patients

Proprioceptive impairment in patients with knee osteoarthritis (KOA) is closely associated with abnormal neuromuscular control. This study aimed to establish a proprioceptive feedback framework to investigate the relationship between spinal proprioceptive input patterns and neuromuscular control strategies in KOA patients. Ten female KOA patients and ten healthy controls underwent gait test. Muscle spindle (MS) and Golgi tendon organ (GTO) models were developed, with MS parameters optimized using quantum genetic-particle swarm optimization (QGA-PSO) and subject-specific maximum isometric force for GTO calibrated via Electromyography (EMG). high-fidelity proprioceptor models were coupled with a musculoskeletal model to estimate MS and GTO afferent activities, projected onto spinal segments L1-S2 to construct spatiotemporal activation maps. The findings showed that QGA-PSO optimization reduced simulation error by 45.1%. In the symptomatic limb of patients, Ia afferents peaked 65.9% at L5 earlier than controls, with 11.5% lower amplitude; Ia afferents at S1-S2 showed a 27.6% reduction in peak and 25.7% in mean activation, whereas at L1-L4 exhibited increases of 22.6% and 9.8%. Ib afferents demonstrated the delayed peaks during late stance with significantly greater amplitudes (+54.5% at L5, +81.6% at L1-L4, p < 0.05) and elevated mean activation. These findings indicated that QGA-PSO enhanced proprioceptor-model accuracy and, combined with EMG calibration, enabled construction of a proprioceptive feedback framework and spinal mapping of afferent activities; together, these revealed segment-specific reorganization of spinal inputs in KOA. This framework provided evidence of altered sensory input and novel insights for understanding neuromuscular mechanisms and guiding rehabilitation strategies.