I-Corps: System for Automatic Analysis of Surgical Skills

Georgia Tech Research Corporation, Atlanta GA

Investigators

Abstract

This project, based on research in the areas of robotics and intelligent systems, involves automated techniques for evaluating surgical skills. Routine evaluation of basic surgical skills in medical schools require considerable time and effort from the supervising faculty. For each surgical trainee, a supervisor has to observe the trainees in person or look at recorded videos. This manual assessment of surgical skills poses a significant resource problem to medical schools and teaching hospitals and results in complications in scheduling/executing their day-to-day activities. In addition to the extensive time requirements, manual assessments are often subjective and domain experts do not always agree on the assessment scores. Every surgical resident has to master basic suturing and knot tying tasks before they can move on to more complex surgeries. Considering the burden on supervising surgery faculty from performing actual surgeries and the amount of trainees that need to go through basic surgical skills training, a system for automated assessment of basic surgical tasks can benefit medical schools and teaching hospitals. The project addresses the potential commercialization of a system for automated assessment of basic surgical skills. The system consists of a central station with user interface and specialized instruments and gloves for the user. Data about movement of the gloves and instruments are captured using accelerometers and transmitted wirelessly to the central station. Using machine learning, the data will be processed and compared to the performance of a skilled surgeon on the same task. Based on that analysis, a score on the surgical skill will be displayed for the user. The users will be able to keep track of their previously obtained scores, which will allow them to see their progress over time. This will also be beneficial for the supervising surgeon who can monitor trainees' skill improvement before allowing them to move on to more complex procedures. Having such a system available in teaching hospitals and medical schools could help teaching surgeons save valuable time that they could spend doing more surgeries and also give students more opportunities for valuable feedback. This could result in a higher patient throughput for hospitals and improve the overall healthcare system. Through NSF I-Corps, the team will conduct interviews with expert surgeons and residents from different teaching hospitals to validate the need for an automated system for grading. Moreover, the team will also conduct interviews with people in dentistry to see if the proposed technology can be used in their domain as well. The team also plans to interview administrators of medical schools and teaching hospitals to understand their process for evaluating and purchasing training aids.