Research

My work aims to enhance surgeon’s capabilities by developing estimation, perception, and control method- ologies for surgical robotics. My doctoral studies spanned a number of topics in these domains to develop a magnetic resonance imaging (MRI) actuated and MRI-guided continuum robot for beating heart proce- dures.

To use visual feedback from intraoperative MR images and enable image-guidance, we developed an active fiducial marker based registration approach to relate the continuum robot’s spatial space to the MR scanner’s image space. This work led to the development of unscented Kalman filter and particle fil- ter based frameworks to track the continuum robot from MR images. We proposed an efficient Bayesian approach to track anatomical structures from MR image streams, where we employed a parameterized model of the shape, motion, and deformation of the underlying anatomy. These work, along with our analysis of the dynamics model of the continuum robot, paved the way for the closed-loop control and image-guidance of the proposed MRI actuated and MRI-guided continuum robotic system.

More recently, I have been working on active sensing algorithms for surgical robotic applications. We have been investigating selecting optimal imaging parameters and the optimal image slice plane that maximize expected information gain during the anatomical target tracking.