This is a student project involving detection of Covid'19 infection among patients on the basis of their X-Ray reports. Essentially a binary classification problem using CNN
Amidst the pandemic of Coronavirus, testing is a key concern. Hence, it becomes essential to bring out an economic yet powerful method to detect the virus among patients. In this project, I've designed a CNN (Convolutional Neural Network) to classify X-Ray Reports of the chest (posteroanterior (PA) view) in patients, as Covid'19 positive or negative (Including both pneumonia patients and healthy ones). It performs well with an accuracy of 97% on the validation dataset.
The datasets essentially used have been referenced from Kaggle, including 150 training examples of the positive cases, as a consequence of low testing data currently available. This project aims at devising a cost-effective and accurate approach in the detection of the virus.