Interfaz multimodal para un asistente robótico quirúrgico: uso de reconocimiento de maniobras quirúrgicas

Abstract

This paper proposes a methodology for the recognition of surgical maneuvers in laparoscopic surgical interventions. The aim is to create an interface between the surgeon and a surgical robotic assistant for two arms of minimally invasive surgery procedures. The proposed interface receives information about the positioning of surgical tools of the surgeon using 3D sensors and the recognition system facilitates the current maneuver is completed. Therefore, the recognition system maneuvers that supports this interface requires a library of models of maneuvers to work. The models chosen to represent the surgical maneuvers are Hidden Markov Models. To validate the proposed methodology, we have developed a series of in-vitro experiments.