COVID DETECTION USING X-RAY

COVID-19 is spreading rapidly worldwide, under-scoring the critical need for early diagnosis and the isolation of patients to be mitigate as its spread. Deep learning (DL) strategies offer the effective and accessible for reliably detecting the COVID-19, particularly using the chest X-ray (CXR) images. This study introduces the two DL approaches which utilize a pre-trained ResNet-50 model that is tailored for COVID-19 detection. The pre-processing phase that includes augmenting, enhancing, normalizing, and resizing CXR images to a standard format. The proposed DL method classifies CXR images through an ensemble approach, leveraging multiple iterations of the modified ResNet-50 model. Evaluation against established benchmark datasets, such as COVID-19 Image Data Collection (IDC) and CXR Images (Pneumonia), demonstrates significant performance improvements over traditional methods like VGG or Dens-net, achieving metrics surpassing 99.63% in accuracy, precision, recall, F1-score, and Area under the curve (AUC).