Monkeypox detection and classification using multi-layer convolutional neural network from skin images

Abstract

The latest pandemic of monkeypox is a significant cause for worry for the public's health due to the rapidity with which it has spread to more than 40 nations outside of Africa. When monkeypox is so like both measles and chickenpox, making an accurate clinical diagnosis of the disease may be difficult. The monitoring and early identification of suspected cases of monkeypox may benefit from computer-assisted detection of lesions. This is particularly true in environments where confirmatory Polymerase Chain Reaction (PCR) assays are not easily accessible. It has been shown that it is possible to do automated skin lesion identification via deep learning (DL) approaches given sufficient training instances. However, it is expected that these procedures will be followed. However, there are currently no datasets of this sort available for monkeypox. Focusing on forecasting monkeypox disease from skin pictures, this study focuses on developing a transfer learning-based multi-layer convolutional neural network (MLCNN) algorithm. Through pre-processing, we can ensure that all the images are of the same quality and that any distracting sounds have been eliminated. The simulation results showed that the proposed MLCNN outperformed the conventional model, proving the validity of the proposed approach. The MLCNN resulted in an accuracy is 99.1, precision is 99.1%, recall is 99.1%, and F1-score is 99.1%.