KOCHI: Dr Karthik Soman, a Thrissur-origin research scientist at the University of California, San Francisco (UCSF), has drawn significant attention from both academic and industry circles, after a technology his team presented at a hackathon at Stanford recently showcased cutting-edge AI in an often overlooked area of healthcare.
This innovative project, which leverages generative AI technology similar to widely-known models like ChatGPT, was demonstrated at the Stanford Medicine Rare Disease AI Hackathon held at GitHub headquarters in San Francisco, USA, in June.
Titled Zebra-Llama, it is similar to generative AI models like ChatGPT. However, unlike ChatGPT, Zebra-Llama specifically targets rare diseases. At present, it is implemented for Ehlers-Danlos Syndrome (EDS), a hereditary connective tissue disorders disease, with phased expansion to other rare diseases.
After the presentation, Dr Karthik and his team were flooded with offers from various industry organisations. Greg Brockman, president of Open AI (the company behind ChatGPT), was among the audience as the team presented the technology.
“To our pleasant surprise, Greg Brockman walked in during the hackathon,” Dr Karthik told TNIE over the phone from the US. Brockman’s presence was crucial as his wife was diagnosed with the genetic disorder EDS, and Dr Karthik’s Zebra-Llama came up with more details and information about the rare disease than what ChatGPT could throw up. “Our cutting-edge AI would help doctors to identify rare diseases among the patients,” said Dr Karthik, 37, an alumnus of IIT Madras.
“Because the medical system is designed for individual specialties while Hypermobile Ehlers-Danlos Syndrome (hEDS) affects every system in her body (orthopaedics, cardiology, neurology, gastroenterology, dermatology, etc.), we spent five years seeing more doctors and specialists than in her whole life prior. Most doctors would only focus on what was relevant to their own specialty. We were lucky that her allergist put together the pieces after observing and hearing her full set of symptoms and issues,”
Brockman posted on X early this year. Dr Karthik said his AI model analyses vast amounts of information related to rare diseases, potentially reducing diagnosis delays and improving patient care for these often neglected conditions.