Title: Can We Solve India’s Health Problems with AI/ML?

Description: Speaker: Dr. Surajit Nundy

Time: Friday, 11 October 2019, 11:00am
Venue: F. C. Kohli Auditorium Dept. of CSE, 01st Floor, Kanwal Rekhi (KReSIT) Bldg.

We have 1 doctor per 1700 patients in India with an extremely urban-focused distribution. Despite an increase in medical schools, we are never going to solve this problem hence we need low cost solutions to solve this problem. Here we present two approaches, data-driven and heuristic driven to solve this problem. In the first we aim to develop a fast, efficient, and automated way to assess the severity of complications after surgeries, so that medical professionals may compare different surgery and treatment options, as well as tailor the treatment plan for individual patients. The Clavien-Dindo (CD) classification is commonly used to categorize post-surgical complications into different grades. We used the data of 494 patients from Raxa, an Electronic Medical Record (EMR) provider in India, to build a model that automatically performs a binary classification of severity with 84% accuracy. We also report preliminary results of predicting the CD grade using this model. In the second we report preliminarily on an attempt to support the clinical decisions of community health practitioners in rural India with heuristics and the challenges faced. Our body of work uncovers many potential uses of AI/ML to tackle India’s health problems, especially involving techniques such as supervised and unsupervised learning, natural language processing, and medical image analysis.

Speaker Profile:
Dr. Surajit Nundy is Project Lead at Raxa. He was trained in Internal Medicine at Massachusetts General Hospital and Harvard Medical School in Boston and has an MPH from the Harvard School of Public Health. He holds an MD/PhD from Washington University in St. Louis/Duke University where he worked on deep-learning in cognitive neuroscience. In India, he has provided care amongst the rural and urban poor.


Prof. Shivaram Kalyanakrishnan

Date: Friday, 11 October, 2019
Time: 11:00am IST
Access: Public
Category: Talk*
Created by: Department Calendar
Updated: Sunday, 29 December, 2019 6:43am IST
Send Reminder: Yes  -  47 hours 20 minutes before start
Participants: Department Calendar
<office@cse.iitb.ac.in> (External User)
_NUC_department <all@cse.iitb.ac.in> (External User)