CoroNet is a Deep Convolutional Neural Network model to automatically detect COVID-19 infection from chest X-ray images. It is based on Xception CNN architecture. The model can help in identifying the difference between three types of pneumonia infections; bacterial pneumonia, viral pneumonia and COVID-19 pneumonia and how COVID-19 is different from other infections. CoroNet was implemented in Keras on top of Tensorflow 2.0. There are three scenarios of the proposed model. First is the main multi-class model (4-class CoroNet) which is trained to classify chest X-ray images into four categories: COVID-19, Normal, Pneumonia-bacterial and Pneumonia-viral. The other two models 3-class CoroNet (COVID-19, Normal and Pneumonia) and binary 2-class CoroNet model (COVID-19, Normal and Pneumonia) are modifications of the main multi-class model. Model achieved an accuracy of 89.5%, 94.59% and 99% for 4-classes, 3-classes and binary class classification tasks respectively.
Input variables : Chest X-ray Images
Output Variables : Predicted class viz, COVID-19, Normal, Pneumonia-bacterial and Pneumonia-viral
Visit Model : github.com
Additional links : ncbi.nlm.nih.gov
| Model Category | : | Public |
| Date Published | : | June, 2020 |
| Healthcare Domain | : | Provider |
| Code | : | github.com |
| Health Risk Management |
| Disease Detection |