An algorithm developed by researchers in Atlanta was able to
accurately pick up signs of anemia just from the coloration of
people's nailbeds, the team reports in Nature Communications.
"The bottom line is that we have created a way for anyone to be able
to screen themselves for anemia anytime, anywhere, without the need
to draw blood," said senior study author Dr. Wilbur Lam, an
associate professor of biomedical engineering and pediatrics at the
Georgia Institute of Technology and Emory University.
Nearly 2 billion people in the world have anemia, according to the
World Health Organization. The condition is characterized by low
levels of hemoglobin, a molecule on red blood cells that carries
oxygen throughout the body. Anemia can be caused by nutritional
deficiencies or chronic illnesses like sickle cell disease and beta
thalassemia. Currently, diagnosis and monitoring require testing the
blood for hemoglobin levels.
The new app Lam and his colleagues are developing uses a form of
artificial intelligence to determine levels of hemoglobin by looking
at the color of a person's nailbeds.
"Essentially, our algorithm learns from every time we feed it
another smartphone image of someone's fingernails with a hemoglobin
level attached to it," Lam said. "We've created a large database in
my clinics. We enroll patients who are already getting their blood
drawn to measure hemoglobin levels. Every time we do that, the
algorithm is getting smarter and smarter."
The algorithm was developed by the study's lead author, Robert
Mannino, who has been one of Lam's patients since childhood. Now a
Ph.D. student at Georgia Tech and Emory, Mannino has a genetic
disorder that leads to chronic anemia and requires monthly
transfusions to keep his hemoglobin levels at a normal level. When
Mannino needed a dissertation topic, the choice seemed obvious.
"He's a brilliant computer programmer who is working on improving
the health of people with his own disease," Lam said.
To determine how accurately the new app could detect anemia, the
researchers rounded up 100 volunteers, some of whom had anemia from
a variety of causes, and some with healthy hemoglobin levels.
The volunteers downloaded the app and then took photos of their
fingernails. The app analyzed the images and compared them to the
ones it had "seen" before. Ultimately, the app was quite good at
detecting anemia, identifying 97 percent of the people who did have
the condition.
[to top of second column] |
The app could be even more accurate, Lam said, if it was given one
hemoglobin reading paired with a photo for an individual patient.
With this accuracy level, the app would allow people with chronic
anemia issues to regularly and instantaneously monitor their
hemoglobin levels.
The app would be especially useful for certain groups of people, Lam
said. For example, "pregnant women are always at risk for anemia and
they know how bad it is for their babies," he said. "Now they can
test whenever they want."
The app isn't ready for widespread use yet as the researchers are
continuing to refine it. But Lam thinks it might be available to the
general public by next spring.
We're going to see more and more of technology aiding in patient
care, said Daniel Barchi, senior vice president and chief
information officer at NewYork-Presbyterian Hospital in New York
City.
"I think we're generally going to find that technology, telemedicine
and artificial intelligence are going to replace many of the
functions we rely on physicians for today," said Barchi, who was not
involved in the new research. "And if technology can speed up
processes and relieve physicians from rote work and that allows them
to concentrate on higher processes, so much the better."
A smart phone app that can analyze a fingernail photo and "be able
to diagnose anemia fairly accurately is a movement in the right
direction," said Dr. Rasu Shrestha, chief innovation officer at the
University of Pittsburgh Medical Center in Pennsylvania. "This is
part of a trend, I think, of moving healthcare closer to the
consumer."
The new app may offer a window on the kind of medical tasks our
phones will be able to take over, said Shrestha, who was not
involved in the study. "There's a wealth of data just waiting to be
unlocked."
SOURCE: https://go.nature.com/2AQZyww Nature Communications, online
December 4, 2018.
[© 2018 Thomson Reuters. All rights
reserved.] Copyright 2018 Reuters. All rights reserved. This material may not be published,
broadcast, rewritten or redistributed.
Thompson Reuters is solely responsible for this content. |