A Conceptual Framework for Drift-Aware IMU-Based Gait Analysis Using Motion Capture References

Inertial measurement unit (IMU)-based gait analysis represents a promising modality for telemedical assessment; however, integration drift remains a fundamental limitation for reliable long-term motion estimation. In this perspective, we propose a conceptual framework for drift-aware IMU-based gait reconstruction that combines established drift-mitigation techniques with phase-dependent analysis of reconstruction residuals. The framework envisages comparing IMU-derived trajectories with motion-capture-derived reference trajectories to quantify reconstruction errors and characterize their temporal and gait-phase-dependent structure in future validation studies. It is hypothesized that these residuals may exhibit structured, gait-phase-dependent patterns arising from multiple sources, including sensor-related errors and biomechanical variability. Drift is therefore treated primarily as a technical error source, while any association between residual error patterns and gait stability or motor-control characteristics remains a hypothesis requiring future validation. The framework outlines a potential approach for integrating wearable inertial sensing, motion-capture-based reference data, and probabilistic filtering techniques. It may provide a conceptual foundation for future telemedical gait assessment systems incorporating phase-dependent analysis of reconstruction errors.