Digital Twin-Enabled Intent-Driven Control for Remote Robotic Sonography: A Neuroergonomic Approach with Hybrid Motion Analytics

This study proposes an intent-driven control paradigm for remote ultrasound robots, integrating ergonomic motion analysis with digital twin technology to enhance operational efficiency and safety. By capturing human arm motion data through a multi-sensor data glove, the system employs a hybrid feature extraction method combining spherical coordinate-based kinematics and time-domain statistical features, achieving 98.33% classification accuracy via a three-layer adaptive neural network. The control strategy discretizes complex ultrasound scanning into standardized sub-motions while enabling real-time fine adjustments through impedance-controlled trajectory planning. A digital twin framework optimizes virtual-reality fusion teaching and collision prediction, validated through experiments on breast phantoms and human subjects. Results demonstrate sub-7.23 mm tracking errors, 95.52% image similarity to manual scans, and stable 3D reconstruction capabilities, offering a scalable solution for remote diagnostic standardization.