Automated Support for da Vinci Surgical System

摘要

While adoption of da Vinci systems has been rapid worldwide, there exists a wide variance in surgical procedure performance impacting care quality, cost and patient safety negatively, due in part to inefficient training practices and limited mechanisms for objectively assessing surgical performance. To address the demand for improved training, Mimic Technologies has developed da Vinci simulators, the dV-Trainer and da Vinci Skills Simulator, in collaboration with Intuitive Surgical, which collect and analyze diverse performance data, and then recommend steps for addressing identified weaknesses. Conversely, Johns Hopkins University's Surgical Assistant Workstation collects and analyzes data from da Vinci during training and surgery. In this Phase I feasibility study, we developed a proof-of-concept working prototype framework capable of providing continuous surgical skills assessment and decision support throughout initial training and thereafter during surgery with the ultimate goal of accelerating a surgeon's acquisition of da Vinci surgical skills. Preliminary analysis of training tasks (anastomosis and peg transfer) across simulation and phantom training laboratories performed by the developed prototype framework highlights the promise of cross-platform data collection and performance measurement across training curriculum, and the opportunity for designing dynamic, customized curricula to a trainee's needs. Developed prototype framework is also a first effort towards cross-platform assessment of robotic surgical performance. This work explored several new concepts including a distributed system for decision support, the role played by instrument and hand pose in planning surgical tasks and metrics assessing this role, distinction of man-machine skills from surgical and task skills, and refinement of simulation training environments to match the corresponding real world training.