Anemia Detection May Now Be Possible With Smartphone Cameras
According to a recent study published in the open-access journal PLOS ONE, predicting anemia with a smartphone photo of a person’s inside the lower eyelid is roughly 72% accurate.
Anemia affects more than 5% of the population in the US and over 25% of the world’s population. The symptoms of this medical problem, which involves a low blood hemoglobin concentration, include headaches, fatigue, difficulties concentrating, dizziness, and shortness of breath.
Since anemia is typically diagnosed with a complete blood count utilizing sensitive lab equipment, anemia is more prevalent in rural areas with limited access to healthcare.
According to the study’s authors, there is a need for a low-cost, easily accessible, and noninvasive point-of-care technology that can detect anemia.
A two-phase study was done to determine the potential of utilizing smartphone cameras to aid in identifying anemia. The first phase involves using a smartphone to photograph the inner lower eyelids of 142 individuals in an emergency room.
Researchers created an algorithm that enhances color resolution and a predictive model that matches hemoglobin levels to the skin and whites of the eyes by zooming in on a small section in each photo.
The algorithm was then tested on smartphone photographs of 202 different emergency department patients in the second phase. The algorithm was 72.6% accurate in predicting anemia, according to the data. It had a better accuracy rate in predicting severe anemia that would demand a blood transfusion, ranging from 86% to 94.4%.
Dr. Selim Suner of Brown University and Rhode Island Hospital, the study’s lead author, explained that after a diagnosis of anemia, people only require iron supplements, which are inexpensive and simple to use. The problematic element, according to Suner, is making the diagnosis.
The study’s findings revealed that flash photography was not required to provide acceptable anemia-detection photos.
However, variable image quality was identified as a potential drawback by the researchers. But this might be the result of the person retracting their eyelid while the image was being recorded. Furthermore, the lighting was not uniform, and it is unknown whether varied brightness levels influenced image quality.
Photo by Artem Beliaikin on Unsplash