A Feasibility Trial of a Novel Robotic System for Retrograde Intrarenal Surgery

I. Introduction

Retrograde intrarenal surgery (RIRS) has become a preferred method for the diagnosis and treatment of urological diseases, such as kidney stone removal. However, the complex urinary and limited visibility of existing endoscope lead to inefficient manipulation of flexible ureteroscopes. Besides, conventional flexible ureteroscopy requires repetitive manual manipulation, which often results in surgeon fatigue, mucosa injury from respiratory motion, and variable stone clearance rates, particularly in complex calyceal anatomies.

The research focuses on the development of an novel robotic system for RIRS, currently dubbed "TaloStone T1000". The robotic system platform consists of a surgeon control console, a multi-functional video cart, and patient-side robotic arm with fiber-optic-sensitized flexible ureteroscopy as shown in Fig. 1. The surgeon console with optimized design of ergonomics is equipped with haptic master devices for smooth and precise control of the robotic arm to manipulate the flexible ureteroscope as well as instruments, e.g., stone baskets and laser fibers. The system also supports seamless integration of multiple modalities, including pre-operative CT scans, intra-operative endoscopic videos, and fiber-optic sensing. Besides, the self-developed flexible ureteroscope is embedded with fiber optic sensors for real-time shape sensing, force estimation, and simultaneous intrarenal pressure control and temperature monitoring. Shape sensing enables precise navigation of the ureteroscope within the renal collecting system, and force estimation provides accurate feedback of tip contact interaction to the master devices on the surgeon control.

Moreover, AI algorithms are incorporated to assist in diagnostics and higher level of supervised surgical autonomy, thereby improving safety and efficiency. The investigators developed AI-powered diagnostics for stone sensing, laser fiber recognition, depth awareness, and CT-to-endoscopy localization. Based on the sensing results from AI-powered diagnostics, the investigators proposed a supervised framework that can automate repetitive procedures throughout in-sheath and ureter navigation, laser approaching, and laser trajectory planning. The entire operation is under supervision of the surgeon, who can use one trigger on the master device or footswitch to enable or disable the supervised automated features. The foot pedal of laser device remains to trigger laser emission by the surgeon for stone fragmentation, dusting, and pop-corning. The basic safety and essential performance of both hardware and software in the robotic system were developed under clinical standards and medical device regulations.

To date, a total of three cadaveric studies have been conducted using the robotic system. In August 2024, the investigators performed the first cadaver study of the robotic system at Prince of Wales Hospital (PWH), where user study of ergonomic manners and tele-operation control of stone treatment was investigated. The second and third cadaver studies, focusing on the AI-powered features of the robotic system, were completed at PWH in June and December 2025. Synthetic renal stones of around 3mm were retrogradely inserted to the renal collecting systems, with successful fragmentation via the robotic RIRS system using Holmium:YAG laser. Over 10 doctors from PWH and the Chinese University of Hong Kong, participated in the cadaver studies. The current system response, motion speed of the robotic system, and operations with ergonomic control console can satisfy the requirements of the doctors. In addition to the cadaver studies, the investigators have conducted a set of laboratory testing and experiments, validating its robustness and stability of the system.