Vitamin Deficiency Detection System

Janhavi Avinash Khune; Saniya Ajay Shiradkar; Tejas Pravin Admane; Vidhi Dilip Kumar Nimje; Dr. Nitin S. More

doi:10.5281/zenodo.15718974

Authors

Janhavi Avinash Khune Department of Computer Science and Engineering, MIT Art Design and Technology University, Pune, India.
Saniya Ajay Shiradkar Department of Computer Science and Engineering, MIT Art Design and Technology University, Pune, India.
Tejas Pravin Admane Department of Computer Science and Engineering, MIT Art Design and Technology University, Pune, India.
Vidhi Dilip Kumar Nimje Department of Computer Science and Engineering, MIT Art Design and Technology University, Pune, India.
Dr. Nitin S. More Department of Computer Science and Engineering, MIT Art Design and Technology University, Pune, India.

DOI:

https://doi.org/10.5281/zenodo.15718974

Abstract

Vitamin deficiency has become a rampant worldwide health problem, associated with life-threatening complications like cardiovascular conditions, cancer, and immune disease. Conventional diagnosis is costly, invasive, and needs the expertise of the diagnostician. This paper presents a new, automated vitamin deficiency diagnostic system utilizing image processing and deep learning technology. Our method employs a CNN model trained from a database of annotated facial, skin, nail, and eye images to identify indicators of deficiencies. A minimalist web app permits users to upload images and provide real-time diagnostic feedback. The solution is inexpensive, scalable, and available, with controlled trials. The findings confirm the capability of AI-based image diagnosis as an addition to conventional methods and an advancement in accessible preventive healthcare.