VITAMIN DEFICIENCY DETECTION USING CONVOLUTIONAL NEURAL NETWORK WITH ADAM OPTIMIZATION

Machine Learning

Vitamin deficiency is very common worldwide that causes various types of diseases. Vitamins are an important part of our diet. Without vitamins, a deficiency occurs. In this paper we are introducing Deep Learning CNN model for diagnosing vitamin deficiency at early stage of deficiency. For this we had considered different types of input datasets such as nails, tongue, instead of Blood. In this paper, A wide spectrum of vitamin deficiencies can show one or more visually distinguishable symptoms and indications that appear in multiple locations in the human body. The platform also allows medical experts to assist in improving the range of detection and accuracy of the application through the contribution and verification of visual data of their patients allowing for more refined image analysis and feature extraction capabilities using Deep CNN (Convolutional neural network). Training is performed using a dataset comprising of images of finger nails, Tongue and eyes. Deep Learning is used to classify the images on the basis of color and textures and gives the probable cause for the same. In this proposed method we will develop and effective methodology for detecting vitamin deficiency using Adam optimizer with more accurate results with reduced MSE (Mean Square Error). The testing results of accuracy for all the cases are above 90%. This also shows the performance evaluation metric values are better that many state of art methods. This is a useful methodology for people to overcome a global problem that affects millions of people worldwide mainly as a result of inadequate nutritional awareness, and it will help healthcare workers in the long term in obtaining more accurate diagnoses in a simple way and is also cost effective without pain.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.