| dc.description.abstract | Ultrasound-guided fine needle biopsy and vacuum-assisted minimally invasive rotary cutting are favored in clinical applications due to their cost-effectiveness, minimal trauma, and lack of radiation. However, the existing spatial positioning technology for surgical instruments occupies operating room space and affects the doctor’s hand operation. This study aims to develop an integrated navigation mode by combining real-time ultrasound imaging, electromechanical guidance, and intelligent ultrasound image processing. By using a miniaturized angle detection sensor, we can accurately obtain the angle of the instrument. Combined with real-time segmentation of ultrasound images based on deep learning, the accuracy of lesion localization is significantly improved. We successfully integrated the ultrasound image acquisition, real-time lesion segmentation and navigation functions into a compact and portable embedded device, achieving the embedded platform deployment of the algorithm. This innovation not only reduces the impact of the ultrasound surgical guidance system on the surgeon’s surgical movements but also effectively reduces the space occupied in the operating room. In the future, sensors with higher precision performance can be replaced or models can be optimized to improve overall performance. The integrated navigation mode proposed in this project provides clinicians with a more comfortable, efficient, and convenient surgical assistant tool by integrating real-time ultrasound imaging, electromechanical guidance, and intelligent image processing technology. | |
