PD23-04 AN EXPERIMENTAL COMPARISON OF AUTONOMOUS ROBOTIC AND MANUAL LAPAROSCOPIC SUTURING

Abstract

INTRODUCTION AND OBJECTIVE: The inconsistency and difficulty of suturing in minimally invasive surgeries stem from the limitations in tool motion and visualization. In pelvic surgery, difficulty is amplified since the tool access is restricted by anatomy. Using an open surgical approach, we previously demonstrated superior suture precision and retention via the smart tissue anastomosis robot (STAR) guided by near infrared (NIR) and 3D imaging systems that reconstruct bowel. Here we demonstrate extensions to laparoscopic conditions in preparation for preclinical studies., METHODS: A KUKA LBR Med lightweight robot was equipped with a motorized commercial suture tool with pitch and roll controls. NIR and 3D cameras were used to reconstruct the 3D model of the tissue and enable robotic suture planning. We tested the STAR in two tasks, i) suturing on a 4-cm straight line on suture pads assuming a static target tissue with a one-time suture planning (STAR-S, n = 5), ii) and a vaginal cuff closure on synthetic tissues with 5 cm diameter and 5 mm wall thickness under dynamic motion and deformation and replanning after each stitch (STAR-D, n = 2). STAR-S/STAR-D includes 1/1 knot at the beginning and 3/10 running stitches. Based on the number of stiches and tissue size, a high-level planner determines ideal suture locations. A similar task as STAR-D but in a manual mode was completed by a surgeon in laparoscopic trainers (Manual, n= 4)., RESULTS: Standard deviation between consecutive sutures, and suture bite was 0.42mm/1.03mm (p < 0.005), and 0.84mm/1.63mm (p = 0.19) for STAR-S and STAR-D, respectively. STAR-S is 2.45 times more consistent in suture spacing since it represents an easy suturing task while STAR-D is closer to the clinical cases due to continuous deformations and irregular tissue shapes. The standard deviation between consecutive sutures and bite size for Manual are respectively 2.75 mm (STAR-D is 2.67 times more consistent, p <0.05) and 3.17 mm (STAR-D is 1.94 times better, p < 0.05). However, average task completion time for STAR-D is 719.5clip_image002.png'>12.5 sec (p < 0.001). This happened due to low motor speeds of the suture tool and safety velocity limits imposed on the robot by the control algorithms., CONCLUSIONS: Autonomous robotics is a promising solution for ensuring consistent laparoscopic suturing in spatially constrained surgeries. Using 3D imaging and improved planning methods, STAR performs laparoscopic suturing with higher consistency than an expert surgeon. Speed can be improved by relaxing safety speed limits. Future work will consider tests in preclinical studies on porcine animal models., Source of Funding: NIH R01EB020610, (C) 2020 by the American Urological Association, Inc.