My current research lies at the intersection of computer vision, deep learning, and cardiac surgery. In particular, the focus of my PhD is to use the computational tools at hand, to tackle important challenges in minimally invasive mitral valve repair, a surgery of the mitral valve of the heart. My cumulative dissertation is titled Deep learning based image analysis for endoscopic minimally invasive mitral valve repair . To tackle the challenges of endoscopic mitral valve repair, the work of my PhD had three broad objectives:

Alt text
The image is adapted from my dissertation, and is provided with a CC-BY-NC-SA CC BY-NC-SA

For the references and a full list, check my publications page or my Google Scholar page. My dissertation will be published once my defense takes place.

Contributions to open science

  • Code for all my papers are made publicly available (check my publications page for individual repo links).
  • An in-house dataset for suture point detection was curated, containing 2,376 intraoperative images, 2,708 surgical simulator images, and 50,000+ suture points. This dataset was released to the community as part of a publicly organised challenge (AdaptOR 2021) at MICCAI 2021.
  • An extended clean stereo-endoscopic dataset from the intraoperative and surgical simulator domains was curated and publicly released to the community, for a novel view synthesis task as part of another publicly organised challenge (AdaptOR 2022) at MICCAI 2022.
  • I was part of various scientific outreach activities - I presented my research to a cumulative general audience of 5000+, in the form of science slams.