Image-Guided Teleoperation for Soft Bronchoscopy Robot

摘要

Currently, the majority of traditional bronchoscopes require skilled doctors to manually insert them through the patient's nose or mouth into the airway, where a slight mistake can lead to severe consequences like tracheal perforation or pneumothorax. In order to address this issue, This paper introduces a soft bronchoscopy robot designed to align with human intuition during operation. By employing U-Net to process the image of the airway, a method based on artificial potential fields is proposed, which introduces a force feedback mechanism to intervene in the operator's actions, thereby enhancing the safety and reliability of bronchoscopy procedures performed remotely. The presented robotic system and approach are validated for safety and reliability on a human lung bronchial model.