A novel 3D-vision-based collaborative robot as a scope holding system for port surgery: a technical feasibility study
Ruochu Xiong,Shiyu Zhang,Zhichao Gan,Ziyu Qi,Minghang Liu,Xinghua Xu,Qun Wang,Jiashu Zhang,Fangye Li,Xiaolei Chen
DOI: https://doi.org/10.3171/2021.10.FOCUS21484
Abstract:Objective: A clear, stable, suitably located vision field is essential for port surgery. A scope is usually held by hand or a fixing device. The former yields fatigue and requires lengthy training, while the latter increases inconvenience because of needing to adjust the scope. Thus, the authors innovated a novel robotic system that can recognize the port and automatically place the scope in an optimized position. In this study, the authors executed a preliminary experiment to test this system's technical feasibility and accuracy in vitro. Methods: A collaborative robotic (CoBot) system consisting of a mechatronic arm and a 3D camera was developed. With the 3D camera and programmed machine vision, CoBot can search a marker attached to the opening of the surgical port, followed by automatic alignment of the scope's axis with the port's longitudinal axis so that optimal illumination and visual observation can be achieved. Three tests were conducted. In test 1, the robot positioned a laser range finder attached to the robot's arm to align the sheath's center axis. The laser successfully passing through two holes in the port sheath's central axis defined successful positioning. Researchers recorded the finder's readings, demonstrating the actual distance between the finder and the sheath. In test 2, the robot held a high-definition exoscope and relocated it to the setting position. Test 3 was similar to test 2, but a metal holder substituted the robot. Trained neurosurgeons manually adjusted the holder. The manipulation time was recorded. Additionally, a grading system was designed to score each image captured by the exoscope at the setting position, and the scores in the two tests were compared using the rank-sum test. Results: The CoBot system positioned the finder successfully in all rounds in test 1; the mean height errors ± SD were 1.14 mm ± 0.38 mm (downward) and 1.60 mm ± 0.89 mm (upward). The grading scores of images in tests 2 and 3 were significantly different. Regarding the total score and four subgroups, test 2 showed a more precise, better-positioned, and more stable vision field. The total manipulation time in test 2 was 20 minutes, and for test 3 it was 52 minutes. Conclusions: The CoBot system successfully acted as a robust scope holding system to provide a stable and optimized surgical view during simulated port surgery, providing further evidence for the substitution of human hands, and leading to a more efficient, user-friendly, and precise operation.