Smartphone-based Strep Throat Detection using MTCNN: A Deep Learning Approach for Non-Invasive Diagnosis

This research proposes a novel approach for strep throat detection utilizing the Multi-Task Cascaded Convolutional Neural Network (MTCNN) algorithm and a smartphone. By incorporating advanced facial detection techniques, MTCNN enhances the accuracy of throat image analysis. The proposed method involves capturing throat images using a smartphone camera with an added device to reduce light reflection. MTCNN is then applied to extract relevant information from the images, enabling a classification model to distinguish between healthy and strep-infected throats. Experimental results demonstrate the effectiveness of this approach, providing a non-invasive and efficient method for strep throat diagnosis. The integration of MTCNN holds significant potential for smartphone-based healthcare applications due to its high accuracy and enhanced feature extraction capabilities.