AI-POWERED DERMATOLOGICAL ANALYSIS: COMPARATIVE STUDY ON DETECTING VITAMIN DEFICIENCY SIGNS FROM SKIN IMAGES

Abstract

Visible signs of vitamin deficiencies, such as B12, D, and Iron, often manifest on the skin, including dryness, changes in pigmentation, and nail discoloration. While early detection of these symptoms is important, it is frequently hindered by the reliance on clinical testing. This research introduces an AI-driven skin analysis system that identifies signs of vitamin deficiencies from skin images using deep learning techniques. A Convolutional Neural Network (CNN) model based on MobileNetV2 is utilized through transfer learning to categorize images into groups like Normal, Vitamin D Deficiency, and Vitamin B12 Deficiency. The dataset undergoes preprocessing with normalization and augmentation to enhance the model's accuracy. Experimental findings indicate that this method can successfully recognize visible patterns of vitamin deficiencies, providing a quick, cost-effective, and accessible diagnostic tool for healthcare providers.