Our inspiration was Adamya's grandmother who unfortunately died of lymphoma just a few years ago. She complained of pain but was never diagnosed until much later which likely contributed to her death.
What it does
It detects lymphoma using histopathologic or skin scans from a microscope and machine learning.
How we built it
We trained a model in Python using Keras using Convolutional Neural Networks. This model was then put on the web with Keras.js so that we could put images in and predict if they had lymphoma on a website.
Challenges we ran into
We had a lot of trouble getting the machine learning model to work because there were so many images and we did not have GPUs. We had to optimize our models so that they ran in enough time for this hackathon. We also had a lot of trouble getting keras to work on our website which took us a few hours to figure out.
Accomplishments that we're proud of
We are proud of creating a working product that could potentially save lives after future work and improvement.
What we learned
We learned a lot about Keras, Machine Learning, and Web Development.
What's next for Foma
We plan to use GPUs so we can train our models for longer and use data augmentation which will drastically increase the accuracy.