Orthopedic Bone-Drilling Assessment Through Laplacian-Based Trajectory Noise Characterization

Abstract

Assessment techniques for orthopedics training are primarily subjective, and often based on qualitative metrics. In this paper, we propose an analytical approach for the quantitative assessment of orthopedic surgery training, specifically, bone drilling. Our goal in this paper is to help improve orthopedics training by providing a means to assess the resident training progress. To this end, we introduce a novel metric that assigns a unique signature to an individual’s drilling activity based on their drilling trajectory, and we compare it with the signatures of expert surgeons. We conduct a simple bone-drilling experiment with surgeons (experts) and novice users on a hybrid (physical - digital) setup consisting of 3D printed bone surrogates that emulate physical and perceptual properties of a human bone across the young and old age groups. We create expert models using our drilling signature metric to evaluate drilling performance for novice users with respect to expert orthopedic surgeons. Our preliminary analysis of drilling signatures across expert and novice users showcases a perceivable distinction across two different bone types highlighting some fundamental insights on the drilling setup, bone material, and user response to each bone type. Our results indicate that the drilling signature helps capture not only a novice user’s drilling behavior, but also their relative expertise as they progress with training.