Analysis of Conjunctiva for Screening of Anemia

Anemia is a condition related to the blood wherein the blood in the body has decreased red blood cells leading to insufficiency of hemoglobin content, an oxygen carrying molecule. This leads to reduced oxygen supply within the blood. It is a global concern worldwide where 40% of children who are 6-59 months old, 37% of women who are pregnant and 30% of menstruating women are majorly affected with anemia. The traditional methods involve invasive procedures of extracting blood and analyzing the blood sample to compute the hemoglobin content and detect anemia which could be time consuming and painful. To overcome the limitations noninvasive methods to diagnose the condition of anemia are under research. Through this study, a focus on noninvasive method of detecting anemia utilizing the idea of image processing of the conjunctiva region of the eye is made. The images acquired are preprocessed followed by segmentation to extract the conjunctiva region. Features of its color channels are extracted from the image and the multiple regression model is compared with machine learning models for classification. The multiple regression model performed better in detection of anemia providing an accuracy of 83.3%.