The Case Ordering Problem in Surgical Procedural Training through Process Mining

Recently, process perspective has been incorporated into the surgical training. These approaches use Process Mining to provide instructors with procedural reports of each student. However, reviewing those reports in a random order could be unnecessarily tedious: each report could be totally different from the previous one, preventing the instructors of any comparative aid, and forcing them to change their mindset every time. Complementary to other approaches such as clustering or tacit knowledge, we propose an approach to optimize, from a process perspective, the ordering of the students executions, such that the procedural distance between an execution and the next one is the lowest possible. The contribution of this paper is twofold: first, the problem is formalized; second, a method for solving the problem tailored to surgical training is presented. Preliminary results are provided, and the preliminary validity of the approach has been corroborated by a medical expert.