/HackTheU2020

HackTheU 2020

Primary LanguageCSSMIT LicenseMIT

Inspiration

We were inspired by the booming trends in big data, alongside the current ongoing pandeminc, and wanted to try to create a tool that would aid healthcare workers.

What it does

Our web app allows healthcare workers to upload Chest X-Rays taken of patients who are suspected of having COVID-19. We then tell the professionals the chance that the patient has COVID-19, or is suffering from pneumonia instead.

How we built it

We used Google's Cloud Computing Services as the main platform for hosting our ML model, trained using Google's Auto ML Vision tool. We then fed the model about 400 images of patients with either COVID-19 or Pneumonia.

Challenges we ran into

We would liked to have had more time to be able to train and fine tune our model further.

Accomplishments that we're proud of

Our model reached a 98% accuracy rate after testing, which we are extremely proud of.

What we've learned

We learned how to use many GCP tools, and a little background on ML.

What's next for CoVision

We would like to further the development of our image recognition model by feeding it images of healthy patients, or patients with other respiratory illnesses, in order to develop a more comprehensive and useful tool.