Recently I have been working on implementing semi-autonomous solutions for improving brachytherapy quality and reduce dependence on physician skill.
Brachytherapy is widely used treatment technique for a number of cancer types, namely prostate, gynaecological, oral etc. However it also accompanies a number of side effects as a result of current clinical practice.
I am exploring shortcomings current methods of treatment planning, procedure and dose delivery techniques. I aim to address these by leveraging advances in robotics, needle steering and rapid prototyping (3D printing).
In TASE-2013 paper, we have studied use of robot assisted needle implants to improve overall quality of prostate brachytherapy and also as a step towards de-skilling the process. We devised and demonstrated an end-to-end robot assisted brachytherapy treatment delivery system using automated dose and needle planners. Also refer CASE-2012
More recently we have extended the use of 3D printing in customised implants for brachtherapy procedures in gynaecoloical cancers. Details in CASE-2013 paper.
Surgical tasks,whether robotic or manual, are hard and often long. However automation of frequently repeated sub-tasks in a procedure would improve treatment quality by reducing surgeon fatigue and delegating the repetitive tasks to intelligent systems.
I am currently studying semi-supervised optimization frameworks for identification of these motions, and performing robust reconstruction of the required trajectories in a changed environment than observed before.
The same techniques would in principle be applicable to other high skill tasks like cooking.