[ad_1]
The algorithm interprets the images taken by the camera of a smartphone and indicates in the act whether the patient is afflicted with the disease or not
Anemia is the blood disease that affects most people in the world. Guilherme Eler
access_time
on December 5, 2018, 19:04 – published on December 5, 2018, with about two billion cases.
This is the result of the absence of healthy red blood cells or red blood cells. When such structures do not exist in sufficient quantity, the transport of hemoglobin, the blood protein responsible for the supply of oxygen to the body, is compromised. A body that does not have all the elements it needs can in turn suffer from symptoms such as fatigue, chest pain, dizziness or rapid heartbeat.
To detect if a person is anemic, the most common examination is blood count. After taking samples, the patient begins a wait that usually takes several days: the blood collected is returned to a laboratory that badyzes the material in different ways, including the amount of red blood cells, platelets and white blood cells, and produces a report. After searching for the documents, the patient must visit the doctor, who interprets the results and gives the diagnosis.
A new technology promises to simplify this path. Researchers from Emory University, in the United States, have been able to form an algorithm for badyzing the color of a person's nails and detecting the status of hemoglobin in his blood . Yes, nails can be good indicators of the level of hemoglobin in the blood, dispensing from blood collection. The idea is all the more interesting as you do not need a hand scanner or that sort of thing. Just a picture taken with an ordinary smartphone and the diagnosis is blamed on the spot. A study of this idea was published Tuesday (4) in the journal Nature .
To bring artificial intelligence to the perception of patterns in the colors of the nails, the researchers photographed the hands of 227 people. All had already had a CBC, which estimated total hemoglobin in their blood. By studying this photographic material, the algorithm has learned to relate certain nail colors to certain levels of red blood cells.
When the group repeated the test with 100 other people, the algorithm already knew with sufficient precision who had a problem with red blood cell concentration and who did not have it. According to the researchers, the effectiveness of the test is equal to or even greater than that of other diagnostic tools approved by the FDA, the federal government agency. The margin of error is only 0.92 gram / deciliter of blood . Healthy men have at least 13 grams / deciliter – the rate is 12 grams / deciliter in women.
Developers intend to launch the application until the middle of next year. If he takes revenge, it will be an interesting alternative – much more comfortable than a visit to the doctor, for example. Perhaps the anxiety of fasting for a blood test will give way to another, more aesthetic, in the next few years: let the cut nails look good in the photo.
Source link
