Dr. Oguzhan Oruc Research Focus
Research Focus: Dr. Oguzhan Oruc
Assistant Professor, Mechanical Engineering
The Citadel
Digital Twin for Gait Rehabilitation
Every year, nearly 800,000 Americans suffer a stroke, and 80% of survivors experience gait impairment that significantly limits their mobility and independence. Traditional rehabilitation approaches rely on therapists’ subjective assessments and generic treatment protocols, leaving a critical gap in personalized care. Researchers at The Citadel, in collaboration with the Medical University of South Carolina’s Locomotion and Energetics Lab, are developing a breakthrough solution: a digital twin system that creates a personalized computational model of each patient’s walking pattern, enabling clinicians to predict how individuals will respond to specific interventions before applying them.
The system employs NARMAX (Nonlinear AutoRegressive Moving Average with eXogenous inputs) modeling, a proven nonlinear system identification technique that captures the complex dynamics of human gait. Raw data from force plates and motion capture systems undergoes preprocessing including marker interpolation, body-weight normalization, and feature engineering to extract center of mass position and velocity before being fed into the modeling pipeline. The NARMAX framework uses Error Reduction Ratio (ERR) for model structure detection and Orthogonal Forward Regression (OFR) for efficient parameter estimation, automatically identifying which biomechanical factors most strongly predict each patient’s foot placement. The model outputs anteroposterior and mediolateral heel coordinates with 95% confidence intervals, and incorporates Recursive Extended Least Squares (RELS) for adaptive retraining as patients progress. This allows clinicians to simulate “what-if” scenarios such as changes in treadmill speed or gait perturbations and predict stability or fall risk before testing interventions on actual patients. Unlike hardware-dependent solutions costing $50,000 to $150,000, this software-only approach integrates with existing clinical motion capture equipment and delivers per-patient personalization that generic devices cannot match.
The project, led by Principal Investigator Oguzhan Oruc, Ph.D. a mechanical engineering faculty of The Citadel, is currently at Technology Readiness Level 4, with clinical validation ongoing at MUSC. For rehabilitation facilities seeking to improve patient outcomes while increasing operational efficiency, this digital twin approach represents the future of data-driven, personalized neurological rehabilitation. Those interested in learning more about collaboration opportunities can contact the research team at The Citadel.
