Dr. Mamatha Bhat is a staff hepatologist and clinician-scientist at UHN's Ajmera Transplant Program and the University of Toronto's Division of Gastroenterology. (Photo: Visual Services, UHN)
(Originally posted in the T-CAIREM Buzz newsletter, February 2021)
We recently caught up with T-CAIREM member Dr. Mamatha Bhat. She's a staff hepatologist and clinician-scientist at UHN's Ajmera Transplant Program and the University of Toronto's Division of Gastroenterology.
What inspired you to pursue medicine?
I saw what a difference one could make in another person’s life through medicine. This is particularly true for patients with end-stage liver diseases who are offered a chance to live for many years through a transplant. As a physician, it's amazing to contribute to and witness this miraculous transformation.
What's the best part of your job?
I'm fortunate to be a physician-scientist, where I can be an agent of change in creating exciting new knowledge that can make a positive impact in the lives of patients.
What excites you the most about the possibilities of AI in healthcare?
The ability to transform population-level guideline-based clinical practice into personalized practice based on an individual's clinical, laboratory and molecular data.
What do you see as the biggest challenge to the field of AI?
Transferring algorithms into clinical practice is a challenge right now. We need to increase awareness of the power of AI among physicians, and T-CAIREM will be pivotal in enhancing knowledge of AI tools and applications in the medical community.
What do you like to do when you aren’t working?
Play with my two little boys, either Legos or outside!
What projects have you been working on lately?
I worked with Bo Wang on an exciting project. We used post-transplant longitudinal data to develop artificial neural network-based algorithms that predict mortality secondary to the most common complications after liver transplantation. I also published a study with Anna Goldenberg where we identified modifiable predictors of long-term survival in liver transplant recipients with diabetes using a machine learning algorithm. Bo, Anna and I also coauthored the first review on the application of machine learning to hepatology and transplantation.
The above studies used the large Scientific Registry of Transplant Recipients as the training set, followed by validation in our institutional dataset. Then we generated algorithms that have the potential to be translated into a clinical setting. Prospective studies would be required in order to fully implement the algorithms in a clinic, but it is wonderful that these approaches resulted in much better performance when compared to traditional biostatistical methods. I look forward to ultimately bringing algorithms like this into the clinic to enhance patient care!
