Measles Detection Using Deep Learning

Measles is an infectious and potentially deadly viral disease. It continues to be a significant global health concern, necessitating accurate and efficient diagnostic tools for timely intervention. In this context, a deep learning-based solution utilizing the YOLOv5 architecture for automated measles detection in medical images is designed. Traditional methods for measles identification often rely on manual examination, which is time-consuming and prone to human error. The proposed YOLOv5-based model offers an approach to streamline this process, significantly enhancing the accuracy and efficiency of measles diagnosis. The novelty of the model lies in its ability to effectively distinguish between measles and non-measles cases in a wide range of medical images. To evaluate the proposed model, extensive experiments were conducted using MSID dataset. Experimental results demonstrate that YOLOv5-based model achieves an accuracy of 92%, with a balanced F1-Score of 0.92. Furthermore, the model’s precision-recall curve showcases its ability to balance precision and recall, giving healthcare professionals the flexibility to adapt the model's performance to specific clinical requirements.