Surgical Skill Training and Evaluation for a Peg Transfer Task of a Three Camera-Based Laparoscopic Box-Trainer System

Abstract

In laparoscopic surgery, surgeons should acquire additional skills before carrying out real operative procedures. The manual skills component of the Fundamentals of Laparoscopic Surgery exam is essential to measure the trainees’ technical skills. The peg transfer task is a hands-on exam in the FLS program. In this paper, a multi-object detection method is proposed to improve the performance of a laparoscopic box¬trainer-based skill assessment system from the top, side, and front cameras. Based on experimental results, the trained model could identify each instrument at a high score of fidelity and the train¬validation total loss for the SSD ResNet50 v1 FPN was about 0.06. In addition, this method could correctly identify the peg transfer time, the move, the carry and dropped states of each object from the top, side, and front cameras. This improved intelligent laparoscopic surgical box-trainer system helps in enhancing surgery residents’ laparoscopic skills. This project is a collaborative research effort between the Department of Electrical and Computer Engineering and the Department of Surgery, at Western Michigan University.