Modeling, Estimation, and Control of Laser-Tissue Interactions for Robot-Assisted Laser Surgery
This dissertation focuses on the development of modeling, estimation, and control techniques for robot-assisted laser surgery. Lasers offer several advantages in surgical applications, including high precision, the ability to target delicate structures, and the potential for minimally invasive procedures. Despite these benefits, controlling the thermal effects of laser-tissue interactions remains challenging due to the complexity of heat transfer during irradiation and the variability of biological tissue.
To address these challenges, I propose a computational model based on the Finite Element Method (FEM) to simulate the thermal response of laser-irradiated tissue. This model incorporates realistic boundary conditions and accounts for effects such as convection which is critical for accurate modeling. I further propose an estimation approach using Ensemble Kalman Filter (EnKF) to identify tissue-specific thermal and optical properties from observed temperature data. These estimates can then be used in laser-tissue interaction models and improve predictive accuracy. In addition, I will investigate two feedback control strategies to regulate tissue temperature without knowledge of the tissue a priori: a model-free proportional-integral (PI) controller and a model-based adaptive controller, both using infrared thermal imaging for real-time temperature monitoring. The proposed work is validated through experiments on tissue phantoms and ex-vivo specimens.
Advisor: Professor Loris Fichera
Committee: Professor Andrea Arnold, Professor Gregory Fischer, Professor Haichong Zhang