Classification of Monkeypox Skin Lesion using the Explainable Artificial Intelligence Assisted Convolutional Neural Networks

The World Health Organization insert ignore into journalissuearticles values(WHO); has given people various protective warnings for Monkeypox. If monkeypox spreads rapidly, it becomes a serious public health problem. In this case, it creates a serious congestion in hospitals. Therefore, auxiliary systems can be needed in hospitals. In this study, explainable artificial intelligence insert ignore into journalissuearticles values(xAI); assisted convolutional neural networks insert ignore into journalissuearticles values(CNNs); based a decision support system was proposed. The data set was used for this task consists of 572 images in two classes, such as Monkeypox and Normal. 12 different CNN models were used for Monkeypox and Normal skin classification. MobileNet V2 model achieved best performance with the accuracy of 98.25%, sensitivity of 96.55%, specificity of 100.00% and F1-Score of 98.25%. This model was supported by explainable AI methods. As a result, an artificial intelligence insert ignore into journalissuearticles values(AI); assisted auxiliary diagnosis system has been proposed for Monkeypox skin lesion.