Proceedings of the ASME 2012 11th Biennial Conference On Engineering Systems Design And Analysis ESDA 2012 July 2-4, 2012, Nantes, France DETC2010/MECH-12345 AN OBJECTIVE ASSESSMENT OF SURGICAL SKILLS Jenny A. cifuentes Laboratoire Amp ` ere D´ epartement G ´ enie M´ ecanique Conception INSA de Lyon 69621 Villeurbanne Cedex Email: jacifuentesq@gmail.com Minh Tu Pham ∗ Richard Moreau Flavio Prieto Pierre Boulanger D´ epartement G ´ enie M ´ ecanique Conception INSA de Lyon 69621 Villeurbanne Cedex France ABSTRACT Minimally Invasive Surgery (MIS) has definitively changed the procedures performed in operating rooms. In many cases, MIS has become the recommended standard technique, replac- ing the traditional open surgery. Effective training and objec- tive assessment of surgeons in these techniques become a ma- jor concern in hospitals in recent years, encouraged primarily by patients and a society that demands safer surgical procedures, which is associated with better surgeons training. Currently, surgeon training includes the acquisition of motor skills outside the operating room. Mechanical simulators and vir- tual reality simulators have been developed over the last decades not only to learn and to transfer the knowledge, skills or attitudes but also to evaluate procedure performance. These tools define metrics to get an objective assessment of surgical skills. How- ever, these metrics are primarily related to simple data measure- ments (such as time, trajectories, velocity, acceleration, smooth- ness, etc.). In the framework of surgery, the difficulty of defining objec- tive metrics for performance evaluation lies in the strict depen- dency between tasks and the difficulty of defining the meaning of optimal performance, related to the characterization of gestures made by the experts. A global rating scheme based on a video analysis and gesture decomposition into several main tasks was developed in [1]. Hidden Markov Models (HMM) have been also ∗ minh-tu.pham@insa-lyon.fr. used by [2] to evaluate physician gesture. However, it is well known that HMM requires a huge set of data sample in order to compute the various HMM parameters. The main objective of our paper is to propose an objective way to evaluate the performance of a surgery task such as knot ty- ing between an expert and novice surgeons through multidimen- sional data analysis as the 3D position (x, y, z) of the instrument tip. Recent research use tools for the analysis of time series such as Multi-dimensional Dynamic Time Warping (MD-DTW) [3], Dynamic Time Warping (DTW) [4] and Longest Common Sub- Sequence (LCSS) [5]. However, in this paper, we will show that one can objectively compare the difference in gesture between an expert and novices in an invariant and time independent way only based on a 1D signal. A survey of different algorithms for surgical gestures analy- sis in time domain is carried out. These ones include the MD- DTW applied to the 3D position data, the DTW and LCSS. An experimental validation is given in a last section and com- pare with a variation of the MD-DTW called MD-DDTW (this acronym has to be defined). INTRODUCTION Minimally Invasive Surgery (MIS) is a new kind of surgery which gets more and more common nowadays. With this new technology, surgeries previously performed through large inci- 1 Copyright c  2012 by ASME