<Anchor> You

can check whether anemia or not by drawing blood, but a domestic research team has successfully trained an artificial intelligence that looks at the condition of the heart to artificial intelligence to predict anemia.

Cho Dong-chan is a medical journalist.


Mr. Song was recently diagnosed with anemia at a medical examination.

[Davin Song/Minor Anemia Diagnosis: Suddenly, or I felt a little tingling.]

To find out if anemia, I had to draw blood because I had to measure the amount of hemoglobin in the blood.

Domestic research team succeeded in predicting anemia without drawing blood.

[Jeon Jeon-hyeon/Cardiologist (thesis author): Anemia becomes one of the important causes of heart failure and dilated heart failure if it continues for a long period of time.] If

you suffer from anemia for a long time, it will interfere with the supply of oxygen to every corner of the body. That increases the burden on the heart.

Obviously, if the heart condition changes, there will be some change in the electrocardiogram, but it was too small to discern in the human eye.

The team trained in artificial intelligence to record about 1,700 ECG records of over 40,000 anemia patients without anemia, and found an algorithm to predict anemia with approximately 90% accuracy.

I tried putting the ECG of an emergency room patient into this algorithm.

[Kwon Jun-myeong/Emergency Medical Specialist (author of thesis): I am currently analyzing. There is a high likelihood of anemia, with 87% (coming out)]

Actually, I did a blood test.

[Kwon Jun-myeong/Emergency Medical Specialist (thesis author): (in blood test), the anemia level was 7 (anemia).] The

research team says it can be applied to various household products.

[Gwonjunmyeong / Emergency Medicine specialist (authors): a massage chair that can be used in everyday life, and the automatic handle yiradeunga ECG is in contact only two points on the skin can be measured -

The Heart Institute officials about the right of anemia diagnostic Although it is difficult to use it as an evaluation, it was meaningful to confirm that artificial intelligence could broaden the scope of diagnosis of electrocardiograms.

(Video coverage: Park Hyun-cheol, Video editing: Park Ki-deok)