Medic AI

Inspiration

Often we face the issue that we are unable to visit a doctor immediately and need a preliminary diagnosis for which we have built a disease prediction model. Initially, the model will not be too strong since we will be using a publicly available dataset. To strengthen the model we use Federated Learning. We have a reward-based system where hospitals supply real datasets and can train the model on their own and get the trained model as NFT. It solves the problem of maintaining the patients' anonymity and ensuring authentic models only reach the central server using blockchain.

Challenges we ran into

One of the biggest challenges we faced was running the Python-based machine learning model on the browser and then we came up with a solution of building a desktop downloadable application where the machine learning model runs locally.