Deep Learning Based Chest X-Ray Image as a Diagnostic Tool for COVID-19

摘要

The COVID-19 pandemic has a rapid spread across the globe, which has deployed life threatening complications ever since it started from China in December 2019. A quick detection of positive cases on corona virus will prevent the further community spread and initiates a earlier treatment to common man. In Recent findings, the images of Chest X ray and CT scan have shown salient features that illustrates the severity of corona virus in lungs. Scientific advancement of Artificial Intelligence in deploying a deep learning based medical field is remaining powerful to handle a huge data with accurate and fast results in medical imaging to diagnose diseases more accurately and efficiently with further assistance in the remote areas. Proposed method is developed for analyzing chest X ray images to detect COVID-19 for binary classes with an accuracy of 99% and validation accuracy of 98%, where the loss is approximately 0.15% by using convolution 2D techniques that are applied on the open source datasets of COVID-19 available at GitHub and Kaggle.