Image-Based Sensing of Leukonychia for Early Diagnosis of Anemia Using a Smartphone Application

Mobile health has gained significant attention due to its low cost, portability, and ease of access. Given that more than 27% of the population suffers from iron deficiency anemia, an ample amount of research is being done to find point-of-care ways to diagnose this disease. The traditional methods...

Full description

Saved in:
Bibliographic Details
Published in:IEEE sensors letters 2022-11, Vol.6 (11), p.1-4
Main Authors: Das, Subham, Bhattacharjee, Mitradip
Format: Article
Language:English
Subjects:
Algorithms
Anemia
blob detection
Computer vision
Electromagnetic wave sensors
Image analysis
Image color analysis
Image processing
Iron
leukonychia
Medical imaging
Nails
Sensitivity
Sensors
Skin
smartphone
Smartphones
Thumb
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Mobile health has gained significant attention due to its low cost, portability, and ease of access. Given that more than 27% of the population suffers from iron deficiency anemia, an ample amount of research is being done to find point-of-care ways to diagnose this disease. The traditional methods are invasive and time-consuming. In this letter, we demonstrate a smartphone application that employs image processing techniques like contour and blob detection, morphological opening and closing, and pixel analysis to detect anemia. An algorithm for detecting leukonychia, which is the appearance of white spots or streaks, is developed; leukonychia is an indicator of iron deficiency anemia, and this is detected using an image analysis technique. We can detect anemia with an accuracy of 89% just by analyzing the photograph of nails from an individual; one can diagnose the early onset of anemia. This method demonstrates a low margin of error and high sensitivity of 96% compared to prior arts indicating its usability as an onsite detection technique. The method is employed in available datasets illustrating 89% accuracy overall. Furthermore, this technique is available in the form of an application and hence is very affordable and large-scale deployable in rural set-ups as well.
ISSN:2475-1472
2475-1472
DOI:10.1109/LSENS.2022.3217010