Abstract

Monkeypox is a rare viral disease that can cause severe illness in humans. Early and accurate detection of monkeypox is critical for the timely treatment and prevention of its spread. In recent years, convolutional neural networks (CNNs) have emerged as powerful image recognition tools. This paper presents a study on the use of a lightweight CNN model for monkeypox detection based on skin lesion images. We ran our experiments on a monkeypox skin lesion images dataset, and precision, recall, F1-score, and accuracy were 93%, 92%, 92%, and 93.12%, respectively. Finally, we evaluated the monkeypox skin lesion dataset's performance using sensitivity, specificity, BAS (Balanced Accuracy Score), and MCC (Matthew's Correlation Coefficient) parametric and achieved 95%, 89%, 92.46 %, and 85.15 % respectively. These promising results, imply that the proposed model is appropriate for mass screening by health practitioners.