MonkeyNet: A robust deep convolutional neural network for monkeypox disease detection and classification

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Epidemiology

Background:

• Monkeypox presents a growing pandemic threat, necessitating early detection methods.
• Current diagnostic tools lack sufficient public datasets for training advanced artificial intelligence models.
• Deep learning shows potential in analyzing medical images for disease identification.

Purpose of the Study:

• To develop a publicly accessible dataset of monkeypox skin images for AI model training.
• To propose and evaluate a deep learning model for accurate monkeypox diagnosis.
• To aid clinicians in early detection and management of monkeypox cases.

Main Methods:

• Creation of the Monkeypox Skin Images Dataset (MSID) from diverse online sources.
• Development and evaluation of a modified DenseNet-201 convolutional neural network (CNN) model named MonkeyNet.
• Utilizing original and augmented datasets for model training and validation.

Main Results:

• The MonkeyNet model achieved high accuracy in identifying monkeypox: 93.19% on the original dataset and 98.91% on the augmented dataset.
• Grad-CAM visualization confirmed the model's effectiveness by highlighting infected skin regions.
• The MSID dataset provides a valuable resource for training and testing deep learning models.

Conclusions:

• The developed MSID dataset and MonkeyNet model offer a promising approach for early and accurate monkeypox diagnosis.
• This AI-driven solution can support public health efforts in controlling the spread of monkeypox.
• Accessible datasets and accurate deep learning models are crucial for addressing emerging infectious disease threats.