DEEP LEARNING ON IMAGES FOR MICRONUTRIENT DEFICIENCY DETECTION: A COMPREHENSIVE SURVEY
Keywords:
Vitamin deficiency detection, Deep learning techniques, Medical image analysis, Convolutional Neural Network (CNN)Abstract
Micronutrient deficiencies, especially those involving Vitamin D, remain a significant public health concern worldwide due to their high prevalence and clinical implications. In recent years, deep learning has shown great potential as a non-invasive method for detecting deficiencies by analysing digital images. This paper offers a comprehensive overview of recent advancements in deep learning techniques applied to the identification of vitamin deficiencies using JPEG-compressed images. This discussion highlights how convolutional neural networks (CNNs), transfer learning techniques, attention mechanisms, and image enhancement methods contribute to increasing the accuracy of deficiency detection. In addition, the review explores the integration of multimodal data sources and addresses important considerations such as model interpretability, data integrity, and real-world clinical relevance. Drawing on advancements in related fields like skin disease classification, image segmentation in medical imaging, and histopathological evaluation, the paper highlights ongoing challenges and potential research directions. Overall, the study illustrates how deep learning methods could offer scalable, cost-effective, and accessible tools widely accessible tools for recognising micronutrient deficiencies during patient care.
