AI and the brAIn

ADITI JAIN

Sep 2022
from Shaastra :: vol 01 issue 05 :: Sep - Oct 2022

At Chitkara University, Prof Jaiteg Singh Khaira uses machine vision algorithms on photographs to discern a person's mental state.

Computation is helping to diagnose and treat brain disorders, providing relief to patients through deceptively simple products.

Rimjhim Agrawal first entered the gates of the National Institute of Mental Health and Neurosciences (Nimhans) in 2014 for an internship during her Master's programme in biotechnology. Attracted by the opportunities and challenges of neuroscience, she decided to stay there beyond her internship to learn about the human brain.

She finished a PhD from the Bengaluru institute, applying Artificial Intelligence (AI) to derive patterns from biological datasets that distinguish a healthy person from a person with schizophrenia.

After the PhD, however, she realised that her research would not translate into public good at scale unless she took the path of entrepreneurship.

Agrawal was soon to understand the challenges of entrepreneurship in this field. "It was a chicken-and-egg problem. We needed brain data to make our tool, and yet none of the hospitals was ready to share data till we could demonstrate the performance of the tool," Agrawal says.

She used data from the publicly available Human Connectome Project to build VoxelBox, a tool to diagnose brain disorders. And she set up BrainSightAI, a start-up based in Bengaluru.

VoxelBox processes brain scans to differentiate a healthy brain from a diseased one, using a variety of techniques such as computer vision, a subfield of AI, where the algorithm is trained to derive meaningful information and patterns from digital images.