Exploring AI as a Diagnostic Tool in Medical Imaging for Dermatopathological Diseases

Abstract Background: Oral health is critical to overall well-being, with abnormalities leading to various health issues. Pemphigus is an autoimmune disease characterized by painful blisters on the skin and mucous membranes, necessitating accurate diagnosis and timely treatment. Objective: This study aims to develop an artificial intelligence (AI) tool utilizing deep learning techniques for the detection of pemphigus from clinical and histological images, enhancing diagnostic accuracy. Methods: A supervised learning approach was employed, utilizing a pre-trained MobileNet deep learning model to classify histology images into healthy and pemphigus categories. The dataset was prepared through cropping, resizing, and labeling of images. Results: The MobileNet model achieved an accuracy of over 92% in distinguishing between normal mucosa and pemphigus lesions. The proposed AI tool can serve as an initial diagnostic check, facilitating faster treatment planning. Conclusion: The implementation of an AI-based diagnostic tool for pemphigus detection demonstrates significant potential in reducing the diagnostic burden in clinical settings. Future work will focus on real-time disease detection applications.