Robot based Transurethral Bladder Tumor Resection with automatic detection of tumor cells

Abstract

Thehandiness, perception, and risks of bladder wall perforation with Transurethral Bladder Tumor Resection (TBRD) have been challenging to process and diagnose due to muscle-invasive bladder tumors. Survey shows in the medical sector, TBTR is usually collected via a rigid resectoscope. Further, the deficiencies in mobility are increased when tumors in the bladder neck area are established. A TBTR Endoscopic Robotic System (ERS) is designed to overcome these limitations. This paper describes the challenges of designing, modeling, and controlling the first TBTR in the robot framework. A Master-Slave robot consists of a continuous cross-sectional robot that can control the ablation, the grasp, and fiber optics with a micro-snake-continuous robot. The strategy for limited telemanipulation, based on redundancy, is consistent with the degree of limitation where the robot experiences are transurethral. The Deep Learning assisted Multi-Scale Feature Fusion Algorithm (MSFFA) has been used to track hundreds of cells with a fully automated and efficient system. Finally, a retracing algorithm combines historical data from matching objects. The results are promising and show that the proposed approach is consistent with the other established methods, which means that potential applications are likely to exist in biomedical engineering.