Rapid Cervical Cancer Detection using Neuromorphic Hardware

Students of Indian Institute of Technology Delhi, Narayani Bhatia, Shridu Verma won the award for research work of Rapid Cervical Cancer Detection using Neuromorphic Hardware. She carried her project work under guidance of Dr. Manan Suri.

In a vast country like India, trained cytologists are constrained by paucity of time and outreach and manual screening of slides leads to slow detection of disease. This is inadequate, because overburdening of human resources leads to dangerously inaccurate diagnosis, and is not time and cost effective. Our solution is an intelligent cervical cancer diagnostic support system which is a first-in-the-series of experiments to follow. Our aim is to develop low-cost, fast and powerefficient healthcare diagnostic techniques so that no disease goes undetected. We train our neuromorphic hardware solution, which is inspired by the parallelism of the brain to mimic brain-like functionality, on existing datasets of classified slide images and use a machine-learning algorithm implemented on hardware for classification. In a scenario where 99% of the solutions are software-centric, consuming huge amounts of power and employing cloud based services, our custom hardware solution is based on true artificial intelligence resulting in drastic reductions in screening times, enhanced decision-making ability and lower operational costs. It is independent of the magnification at which the pap smear images may be viewed, and is invariant to staining variations that may arise due to laboratory testing procedures in medical labs.

Dr. Harsh Vardhan, Minister of Science and Technology, awarded the Gandhian Young Technological Innovation (GYTI) Appreciation to Narayani Bhatia, Shridu Verma, Indian Institute of Technology Delhi at the GYTI 2018 Awards function held at Rashtrapati Bhawan on March 19, 2018.

Leave a Reply

Your email address will not be published. Required fields are marked *