Reconstructing the Gait Pattern of a Korean Cadaver with Bilateral Lower Limb Asymmetry Using a Virtual Humanoid Modeling Program.

Background and

Objective

This study presents a combined osteometric and biomechanical analysis of a Korean female cadaver exhibiting bilateral lower limb bone asymmetry with abnormal curvature and callus formation on the left femoral midshaft.

Methods

To investigate bilateral bone length differences, osteometric measurements were conducted at standardized landmarks. Additionally, we developed three gait models using Meta Motivo, an open-source reinforcement learning platform, to analyze how skeletal asymmetry influences stride dynamics and directional control.

Results

Detailed measurements revealed that the left lower limb bones were consistently shorter and narrower than their right counterparts. The calculated lower limb lengths showed a bilateral discrepancy ranging from 39 mm to 42 mm-specifically a 6 mm difference in the femur, 33 mm in the tibia, and 36 mm in the fibula. In the gait pattern analysis, the normal model exhibited a straight-line gait without lateral deviation. In contrast, the unbalanced, non-learned model demonstrated compensatory overuse and increased stride length of the left lower limb and a tendency to veer leftward. The unbalanced, learned model showed partial gait normalization, characterized by reduced limb dominance and improved right stride, although directional control remained compromised.

Conclusions

This integrative approach highlights the biomechanical consequences of lower limb bone discrepancy and demonstrates the utility of virtual agent-based modeling in elucidating compensatory gait adaptations.