A robotic flexible endoscope with shared autonomy: a study of mockup cholecystectomy

Endoscope is the eye of surgeon in minimally invasive surgery (MIS). Prevailing handheld endoscopes are manually steered, which can cause endoscope-instrument fencing. Robotic endoscopes can reduce the fatigue but could not reduce collisions. Handheld endoscopes with a flexible bending tip can reduce the shaft pivoting and collisions. However, its steering is challenging. In this paper, we present a robotic flexible endoscope with auto-tracking function and compare it with the conventional rigid endoscopes.

A robotic flexible endoscope (RFE) with shared autonomy is developed. The RFE could either track the instruments automatically or be controlled by a foot pedal. A mockup cholecystectomy was designed to evaluate the performance. Five surgeons were invited to perform the mockup cholecystectomy in an abdominal cavity phantom with a manual rigid endoscope (MRE), a robotic rigid endoscope (RRE), and the RFE. Space occupation, time consumption, and questionnaires based on the NASA task load index were adopted to evaluate the performances and compare the three endoscope systems. An ex vivo experiment was conducted to demonstrate the feasibility of using the RFE in a biological tissue environment.

All surgeons completed the mockup cholecystectomy with the RFE independently. Failure occurred in the cases involving the RRE and the MRE. Inside the body cavity, the space occupied when using the RFE is 17.28% and 23.95% (p < 0.05) of that when using the MRE and the RRE, respectively. Outside the body cavity, the space occupied when using the RFE is 14.60% and 15.53% (p < 0.05) of that by using MRE and RRE. Time consumed in the operations with MRE, RRE, and RFE are 28.3 s, 93.2 s and 34.8 s, respectively. Questionnaires reveal that the performance of the RFE is the best among the three endoscope systems.

The RFE provides a wider field of view (FOV) and occupies less space than rigid endoscopes.