eng Oriental Scientific Publishing Company Biosciences Biotechnology Research Asia 0973-1245 2025-12-30 22 4 1712 1724 10.13005/bbra/3471 57121 article Venkatesh Puppala 1 Department: MBA – Finance, Osmania University, India We propose MpoxNet, a lightweight convolutional neural network for the multi- class classification of skin lesions associated with emerging infectious diseases. Built on six ConvNeXt blocks and a Dual Attention Block that jointly models spatial saliency and channel dependencies, MpoxNet enhances discriminative fea- ture learning while maintaining computational efficiency. The model is evaluated on the Mpox Skin Lesion Dataset Version 2.0 (755 images across six clinically annotated classes), achieving a test accuracy of 86.49%, a macro F1-score of 86.79%, and a Matthews Correlation Coefficient of 82.76%. Grad-CAM visual- izations confirm that MpoxNet focuses on pathologically meaningful regions such as umbilicated lesion centers (Monkeypox) or maculopapular spread (Measles), supporting clinical interpretability. In addition, cross-dataset experiments on BreastMNIST, OCTMNIST, and OrganAMNIST demonstrate strong generaliza- tion capability. These results indicate that MpoxNet provides a computationally efficient, interpretable, and robust diagnostic tool suitable for deployment in resource-constrained clinical and public health settings. https://www.biotech-asia.org/vol22no4/an-efficient-ai-model-for-the-classification-of-skin-lesions-in-emerging-infectious-diseases/ Deep Learning; Dual Attention; MpoxNet; Skin Lesion; ConvNext