UBC Theses and Dissertations
Quantitative modelling and assessment of surgical motor actions in minimally invasive surgery Cristancho, Sayra Magnolia
The goal of this research was to establish a methodology for quantifying performance of surgeons and distinguishing skill levels during live surgeries. We integrated three physical measures (kinematics, time and movement transitions) into a modeling technique for quantifying performance of surgical trainees. We first defined a new hierarchical representation called Motor and Cognitive Modeling Diagram for laparoscopic procedures, which: (1) decomposes ‘tasks’ into ‘subtasks’ and at the very detailed level into individual movements ‘actions’; and (2) includes an explicit cognitive/motor diagrammatic representation that enables to take account of the operative variability as most intraoperative assessments are conducted at the ‘whole procedure’ level and do not distinguish between performance of trivial and complicated aspects of the procedure. Then, at each level of surgical complexity, we implemented specific mathematical techniques for providing a quantitative sense of how far a performance is located from a reference level: (1) The Kolgomorov-Smirnov statistic to describe the similarity between two empirical cumulative distribution functions (e.g., speed profiles) (2) The symmetric normalized Jensen-Shannon Divergence to compare transition probability matrices (3) The Principal Component Analysis to identify the directions of greatest variability in a multidimensional space and to reduce the dimensionality of the data using a weight space. Two experimental studies were completed in order to show feasibility of our proposed assessment methodology by monitoring movements of surgical tools while: (1) dissecting mandarin oranges, and (2) performing laparoscopic cholecystectomy procedures at the operating room to compare residents and expert surgeons when executing two surgical tasks: exposing Calot’s Triangle and dissecting the cystic duct and artery. Results demonstrated the ability of our methodology to represent selected tasks using the Motor and Cognitive Modeling Diagram and to differentiate skill levels. We aim to use our approach in future studies to establish correspondences between specific surgical tasks and the corresponding simulations of these tasks, which may ultimately enable us to do validated assessments in a simulated setting, and to test its reliability in differentiating skill levels at the operating room as the number of subjects and procedures increase.
Item Citations and Data
Attribution-NonCommercial-NoDerivatives 4.0 International