Transformer-Based Framework for Motion Capture Denoising and Anomaly Detection in Medical Rehabilitation

2025-07-11

AI Summary

In medical rehabilitation, optical motion capture data are highly susceptible to occlusion, noise, and missing values, and existing methods lack real-time dynamic anomaly detection capability. To address these challenges, this paper proposes an end-to-end Transformer-based framework that jointly performs motion sequence denoising, missing value imputation, and online anomaly detection. Leveraging temporal masked modeling and long-range dependency modeling, the framework significantly enhances robustness against occlusion and noise. A lightweight anomaly discrimination module enables real-time safety alerts under low-supervision settings. Extensive experiments on stroke and orthopedic rehabilitation datasets demonstrate that our method outperforms state-of-the-art approaches, achieving a 21.3% reduction in mean absolute error (MAE) for reconstruction and an 18.7% improvement in F1-score for anomaly detection. The proposed framework is computationally efficient and well-suited for deployment in remote rehabilitation scenarios.

Technology Category

• Computer Vision: Motion & Tracking
• Intelligent Robots: State Estimation
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

This paper proposes an end-to-end deep learning framework integrating optical motion capture with a Transformer-based model to enhance medical rehabilitation. It tackles data noise and missing data caused by occlusion and environmental factors, while detecting abnormal movements in real time to ensure patient safety. Utilizing temporal sequence modeling, our framework denoises and completes motion capture data, improving robustness. Evaluations on stroke and orthopedic rehabilitation datasets show superior performance in data reconstruction and anomaly detection, providing a scalable, cost-effective solution for remote rehabilitation with reduced on-site supervision.

Problem

• Denoise motion capture data in medical rehabilitation
• Detect abnormal movements in real-time
• Handle missing data due to occlusion

Innovation

• Transformer-based model for motion capture denoising
• Real-time anomaly detection in rehabilitation movements
• Temporal sequence modeling for data completion