Deep Learning Based Non-Invasive Framework for Nutritional Deficiency Detection Using Hair and Nail Images

March 2026 in “ FMDB Transactions on Sustainable Health Science Letters ”

B. Shilpa Shree, D. Sumathi, Prem Kumar Ramesh, M. Sindhu, S. P. Indushri, S. Naga Lahari, Farrukh Arslan

Convolutional Neural Networks nutritional deficiencies hair images nail images iron deficiency zinc deficiency biotin deficiency vitamin deficiency CNNs

TLDR A deep learning method can detect nutritional deficiencies from hair and nail images with 89% accuracy.

The study presents a non-invasive framework using Convolutional Neural Networks (CNNs) to detect nutritional deficiencies through hair and nail images, achieving an 89% accuracy rate. This method offers a cost-effective alternative to traditional blood tests, particularly benefiting women, children, and socioeconomically disadvantaged groups. By analyzing high-resolution images of size 224 × 224 pixels and employing data augmentation, the model can identify deficiencies in iron, zinc, biotin, and vitamins, facilitating early diagnosis and improving healthcare accessibility.

View this study on fmdbpub.com →

Discuss this study in the Community →