Covid/Pneumonia X-Ray Image Classifier

With an abundance of Covid-19 patients globally, a faster method for identifying infected patients is crucial. Using our model built with VGG16, we have a fast alternative to identifying the virus.

Comment

Project Inspiration

The inspiration behind this project stemmed from wanting to develop a model that deals with image classification. We settled on this dataset in particular because Covid-19 is still a prominent global pandemic and having the ability to quickly identify infected vs. normal patients would generate a significant advantage when it comes to global recovery.

Functionality

Our project is a simple image classification convolutional neural network (CNN) that deals with 3 classes of X-ray images (Covid-19 patients, Viral Pneumonia patients , and Normal patients)

How it was built

Our model classification model was built on VGG16 which is an efficient CNN architecture used for image classification. We used Python with TensorFlow and some Keras modules for data augmentation and model construction.

Accomplishments

Created an effective classification model for X-ray images of potential Covid-19 patients.

What we learned

We learned how to build a classification model with the VGG16 architecture in mind as well as learning how to implement different parameters of our CNN's layers.

What's next for Covid-19/Pneumonia X-Ray Image Classifier

Next we could integrate our model with a web framework such as tensorflow.js to allow users to quickly input an image and receive a classification result.

Built With

  • google-colab
  • keras
  • python
  • tensorflow
  • vgg16
Share this project:

Updates