<![CDATA[Monkeypox is a skin infection that has become a serious concern in Indonesia since the increase in cases in 2022. Diagnosis of monkeypox requires special expertise, laboratory tests, and clinical observations. Diagnosis generally uses PCR tests which are often not available in remote areas. This study aims to develop a deep learning-based mobile application for early detection of monkeypox through image classification of skin lesions. The CRISP-DM methodology is applied in developing this application, starting with collecting datasets from the Kaggle site consisting of 8,910 images and divided into 80% training groups, 10% validation, and 10% testing with augmentation techniques to improve model accuracy. The developed CNN model was implemented using Create ML on the iOS platform. The model evaluation uses several metrics such as accuracy, precision, recall, and F1 score, with the threshold being the highest probability of the model predicting model evaluation results show an accuracy of 81%, precision of 80.2%, recall of 76%, and F1 score of 0.78 for the test data. The resulting application allows rapid detection of monkeypox and is accessible to the wider community, thereby helping to reduce delays in diagnosis, especially in hard-to-reach areas. This study shows significant potential in supporting the health system in Indonesia through the application of artificial intelligence technology for infectious diseases.]]>
Copyrights © 2024
