Nosos

About
Demo
Articles

Inspiration

People in rural areas and third-world countries do not have immediate access to affordable healthcare. Often times people will avoid the doctor to save money. This is a problem because these genetic diseases are treatable if they're caught in the early stages. Our solution streamlines the diagnosis process by utilizing machine learning to bring disease detection to your fingertips.

What it does

Our model uses ML to find correlations in fingerprints to determine which disease you may have. It then recommends specialized doctors near your location.

How we built it

We used CoreML and Google Map's API and several other python packages to build our solution.

Challenges we ran into

The datasets that have the fingertips and labels of which diseases they correlate to are not publically available. We had to create a custom data set with the limited information out there. CoreML is only available to macOS users, so only one of our members was able to work on the ML model.

Accomplishments that we're proud of

Proof of concept was shown for our idea. We ran into many roadblocks (as most projects do) and almost quit at one point, but we persevered throughout the night and implemented everything we planned to.

What we learned

Our team was pretty familiar with ML and API requests, but we had minimal experience with front-end development. It was a struggle to implement everything on our easy-to-use website.

What's next for Nosos

The next step would be to acquire better data sets. Our current data set is minimal. With more in-depth data sets we would be able to further our model and increase its accuracy.