This was a project I have worked on because medical imaging is one of my passions and I was interested in trying to build something.
The images are from a dataset named Covid19-Pneumonia-Normal Chest X-Ray Images and the purpose of this dataset is to help medical professionals classify pneuomonia or covid-19 using deep learning methods. The input images are 256x256. There are 1,6k x-ray images classified as COVID-19, 1,8k classified as Normal and 1,8k classified as Pneumonia thus this dataset is balanced and quite interesting to use directly
- PyTorch: Most of the code was written in Python using the Deep Learning library PyTorch
- Wandb: Used for logging, tracking and some visualizations