Diego Pardo, Loris Fichera, Darwin G. Caldwell and Leonardo S. Mattos

Abstract — This paper presents a system to supervise tissue temperature during robotic laser surgery. The use of robotic systems for laser surgery provides mechanisms to control the motion of the laser beam and the exposure time of laser radiation, allowing the automatic generation of incisions. This work focuses on the perception side of the problem, developing a technology for the online verification of the thermal state of the tissue during robotic laser microsurgery. Obtaining this information is paramount to enable automatic control of laser incision quality, which is directly related to tissue temperature. A model learned from real data estimates the change in temperature given the exposure time and power of the laser. The model is implemented in the real system and validated during laser incisions on ex-vivo tissue. Results show that the model can reliably estimate the thermal state of the tissue in real-time, and thus is suitable to produce feedback for automatic control of laser incisions.