LIME-based Explainable AI classifier to detect COVID-19 pandemic through X-ray images

The term “black box” is used in the field of artificial intelligence to describe the lack of transparency between the model and its users. This work proposes the Explainable AI classifier, a major model for detecting COVID-19 using X-ray pictures, a topic that is at the cutting edge of research in the area of medical image analysis. The purpose of this research is to look at the process by which a model diagnoses a disease based on X-ray pictures. Create a COVID-19 radiography dataset, which includes evaluated lung images of 3616 COVID-19-positive cases and 10192 normal cases that can be accessed by the public. Which features of the image influence the results are identified with the help of the LIME method. The superpixels are categorized by the model and then shown with the help of the LimeImageExplainer function. It's very apparent how the model arrives at its results. With 97% accuracy, 94% precision, and 99.9% recall, the proposed approach exceeds the state-of-the-art model.