Vita

Note: video submissions is the second link. Ignore the YouTube video.

Inspiration

My grandmother, Archana Ray, was misdiagnosed with pneumonia 2 years ago and was put on the wrong antibiotics. Only recently, we realized she actually had a different condition and was misdiagnosed. Her mistreatment has led to her undergoing further complications and being put in excessive pain. It has also raised the question of whether she will live.

What it does

Automatically diagnoses datasets of lung x-ray scans for pneumonia and generates a csv with all the predicted results.

How we built it

Using Python, Tensorflow, Flask, HTML, & CSS

Challenges we ran into

Installing python packages Learning flask syntax Learning HTML on the go Building a neural network from scratch

Accomplishments that we're proud of

Learning a lot Building a model almost from scratch Creating a good webpage

What we learned

CSS & HTML How to deal with python libraries

What's next for Vita

Boosting the program accuracy by implementing a new program to extract features from patient blood test microscopy and having a Random Forest Classifier drawing correlations between blood cell features and our CNN x-ray scan predictions

Built With

• bootstrap
• css
• flask
• html
• keras
• python
• tensorflow