Discovering the Language of Surgery
René Vidal. Department of Biomedical Engineering. Johns Hopkins University
Fecha: 28, Marzo 2012
Abstract: Recent technological advances have contributed to, and changed, the way in which surgery can be performed. One of them is Robotic Minimally Invasive Surgery (RMIS), which has several advantages over traditional surgery, such as better precision, smaller incisions and reduced recovery time. However, the steep learning curve together with the lack of fair, objective, and effective criteria for judging the skills acquired by a trainee may reduce the benefits of this technology. One alternative is to use RMIS recordings to model surgeon expertise and perform automatic skill assessment and gesture classification. A natural approach to modeling surgeon expertise is to decompose a surgical task into a series of pre-defined surgical gestures, such as "insert a needle", "grab a needle", "position a needle", etc., which should appear in some pattern, e.g., one gesture often follows another one, or several gestures form a motif. Different surgeons with different expertise will either execute different gestures differently or follow a slightly different sequence of gestures. This is analogous to what we see in natural language, where the grammar constrains the generation of words. The main difference is that, in the case of surgery, we know neither the words nor the grammar. In this talk, our describe our recent work on using both kinematic and video data to discover the "language of surgery" and using this language to automatically classify surgical gestures and skill.
Lugar: Sala de seminario física experimental