Skip to content

Lucifergene/covid-prediction-classifier

Repository files navigation

COVID-Predictioned

COVID-19 Prediction Powered by AI

COVID CNN Docker Build Contributions welcome Live Demo License

DockerHub : avik6028/covid_predictor_flask_cnn:1.0.0

The project is a Computer-Aided Diagnostic System which is used to predict whether a person has been infected with COVID-19. Currently, the reference project does the only classification between COVID and non-COVID X-Rays using Modified DenseNet architectures. This project involves a Web Application, REST API, Dockerhub Container and automated pipeline to deploy the COVID Prediction model based on Chest Radiological Images. Containerized Deployment ensures platform independence during deployment.

Inter-Disciplinary Domains

  1. Deep Learning/ Image Processing
  2. Web Development
  3. DevOps
  4. Database Management

Technologies Used

  1. Flask
  2. Tensorflow/ Keras
  3. Docker/ DockerHub
  4. Heroku Container Service
  5. GitHub Action CI/CD
  6. MongoDB

Market Demand

In the year 2020-21, the most demanded medical test, is the test of the COVID-19 virus. The traditional methods involve collecting samples for RT-PCR tests. But this process often gives false negatives and also takes a long time to predict the result.

It has been proved by various research works by foreign scientists and researchers that X-Ray/CT images can be efficiently used for the detection of the COVID-19 virus. It is seen that patients suffering from COVID-19, developed some particular type of lesions in their lungs, which is medically termed as Ground-Glass Opacities (GGOs). These lesions can be easily be seen through Chest X-Ray or CT images.

Several researchers have developed various Deep Learning models to automatically classify the Radiological images of patients. There are various journal as well as conference papers in this context. Consecutive models have been developed, beating the previous ones in terms of Accuracy, Precision, F1 score, etc.

Screenshots



License

You can check out the full license here

This project is licensed under the terms of the MIT license.