… on skipping two grades as a student growing up in India
Education was incredibly important to my parents. I was homeschooled until second grade, and then skipped third and fourth grade. My mom had hopes I would be a doctor, but I loved math. I knew I’d be an engineer from a young age. And today, I’m working on building a cognitive assistant for Watson Health!
… on moving to the US to attend MIT
The day I arrived at MIT, I barged into a professor’s (Dave McAllister) office and asked, “Do you have any interesting problems for me?” He must have thought I was crazy, but he drew a problem on the chalkboard which I later learned was an open problem in the theory of computation. He smiled and said, “Come back when you have solved this.” So, I went to my dorm room and worked on it all night. When I took it back to him the next day, I surprised him — I don’t think he had expected to ever see me again! But when I showed him my work, his jaw dropped! I had shown that the monotone inference relation was polynomial time decidable using mathematical induction on the length of a proof, a syntactic proving technique that was quite a departure from the model-theoretic proofs of those days and led to mechanizable theorem provers. It was the way I looked at that specific problem at the MIT AI lab — through a different lens — that I would carry with me throughout my career.
And MIT was also my first introduction to IBM. I received the IBM Graduate Fellowship that funded my education there. In January of 1988, the AI Lab organized a Robot Olympiad where each of the participants were given $1,000 to build something creative. My partner and I teamed up under the guidance of Prof. Rod Brooks (who later started iRobot) to turn a zoom broom into a motor car, and navigated it to clean around the room using an IR receiver and transmitter beacon. A spiral path was traced using the inverse square law of radiation to calibrate the distance from the beacon. This was an early prototype of the first vacuum cleaner robot whose current form in stores is the popular Roomba. We had a prototype in two weeks! And it was that competition that perhaps helped me earn the IBM Fellowship in 1990.
… on reinventing radiology
It is my hope that my work will help doctors worldwide. I was inspired to focus my research on healthcare after my father’s stroke was misdiagnosed. And I didn’t want other families to go through what mine went through, and I was confident there was a way to apply technology to advance patient care.
Simply put, my work centers on bringing the power of intelligent machines to help radiologists make quick, accurate and cost-effective diagnoses. This is important because, today, radiologists are asked to read large volumes of patient images without a sense of which images could be most relevant, and then to render a clinical diagnosis without the benefit of insight into an individuals’ complete health status.
In the future, I envision a radiologist has access to reports which include medical images integrated into the entirety of an individual’s health information and which suggest potential diagnostic options to consider, based on analysis of similar clinical cases culled from a trove of aggregated, de-identified data. This kind of application of intelligent machines could transform how doctors work and help doctors deliver more informed and personalized care.
… on her greatest influence
As a child, I was inspired by women scientists like Marie Curie. And I try to live by Thomas Edison’s quote, “I use my body to carry my mind.”
… on hobbies and the sands of time
I love poetry. Longfellow’s A Psalm of Life is a favorite. And I used to be a radio amateur (Ham operator), actively broadcasting from India using Morse and voice. My call sign was VU2STF.
Connect on Linkedin