XScanner

Inspiration

We wanted to use tensorflow in some sort of medical application, so when we saw that there was a lot of data for healthy and pneumonia infected lungs, we saw the opportunity to create something beneficial.

What it does

Our application tensorflow model takes in user input (an image in this case) and outputs the probability of it being healthy or unhealthy (pneumonia infected).

How we built it

To make the application we used a react frontend with typescript and firebase backend, which stored the tensorflow model and user inputted images.

Challenges we ran into

One of the main challenges we ran into was the hardware limitations of the google collab notebook and once we finished the model, loading the model and running data through it took awhile to connect to the database.

Accomplishments that we're proud of

Overall, we are proud of the challenges we overcame with the tensorflow model since it is one of the more challenging things we have dealt with in terms of development, as well as being proud of the high accuracy the model has with the images.

What we learned

In this project, we learned how to train, test, and deploy a machine learning model as well as connect it to a cloud storage database.

What's next for XScanner

Our next step with XScanner is to widen our use case being able to be more inclusive of different x-ray types and scanning for different ailments.

Built With

• firebase
• html
• nextjs
• node.js
• react
• tensorflow
• typescript