Using Motion-Based Metrics to Objectively Classify Surgeon Skill and Assess Performance with Augmented Feedback
O'Malley, Marcia K
Master of Science
Objective evaluation of surgical skill based on assessment of tool movements made by the surgeon is a rapidly developing area of research. Motion-based performance metrics are computed from data recorded in real-time during surgical procedures. The approach offers advantages over the subjective observation based grading schemes typically used to characterize the expertise of the endovascular surgeon. In this thesis, these motion-based metrics are applied in two novel scenarios. First, the use of artificial neural network machine learning techniques demonstrates that these metrics that quantify tool movement smoothness and quality can be used to classify surgeons as experts, intermediates, or novices. This finding extends prior work that simply showed correlations between such metrics and expertise level. Second, motion-based metrics are used not to directly assess surgical expertise and skill, but to evaluate the benefits of new tool visualization technologies made possible with real-time electromagnetic (EM) sensing and tracking of endovascular tool tip movements. Visualizations based upon EM sensing technology are shown to increase motion smoothness while also decreasing radiation exposure during a procedure.
Motion-based Metrics, LVQ, SOM, Endovascular Surgery