Inspiration In rural India survival rates for breast cancer are low due late detection as access to doctors is limited.
On the bright side, rural India has achieved excellent internet access in recent years.
There seemingly exists enormous opportunity to use technology particularly AI and realize mass screening programs.
What it does Sathi (Partner) is a mobile app that connects to x-ray machines and uses AI to screen mammograms. It is a low cost solution which is easy to use for health workers, who screen low risk cases, saving precious doctor time for high risk ones.
How we built it The heart of the solution is a deep neural network (FRCNN) machine learning model hosted on Amazon cloud. Being on cloud, it can easily scale to achieve massive screening coverage.
Challenges we ran into Finding the right algorithm - required in depth study of research literature and expert connects. Understanding ground realities in the field - learnt from experiences of doctors and radiologists among contacts.
Accomplishments that we're proud of High accuracy from the model and ease of use of the app. Brought it all together, fast.
What we learned Technology brings hope for a breast cancer free world in future.
What's next for Sathi (Partner) Train and test models on Indian datasets. Regulatory approvals and field trials.