the artificial intelligence (AI) It may look like a cool robot system, but scientists at Osaka Metropolitan University (Japan) have shown that it can provide moving support or, more specifically, a “heart warning.” Classification of cardiac functions Point to Valvular heart disease with unprecedented accuracy.
This new finding, reported in The Lancet Digital Health, demonstrates continued progress in integrating the fields of medicine and technology to improve patient care. Valvular heart disease, one of Causes of heart failureIt is often diagnosed by echocardiography. However, this technology requires specialized skills, so there is a corresponding shortage of qualified technicians.
Meanwhile, a chest X-ray is one of the most common tests for identifying diseases, especially of the lungs. despite of The heart also shows up on a chest X-rayTo date, little has been known about the ability of chest X-rays to detect heart function or disease. Chest X-rays are performed in many hospitals and require very little time to perform, making them easily accessible and reproducible.
As a result, the research team led by Dr. Daigo Ueda of the Department of Diagnostic and Interventional Radiology, Osaka Metropolitan University College of Medicine, succeeded in developing A model using artificial intelligence to accurately classify heart functions And heart valve diseases from chest radiographs.
Improved diagnostic accuracy
Because AI trained on a single data set encounters potential bias, which leads to lower accuracy, the team targeted data from multiple institutions. Subsequently, A total of 22,551 chest radiographs were collected associated with 22,551 echocardiograms of 16,946 patients at four centers between 2013 and 2021. With chest radiographs set as input data and echocardiograms as output data, The AI model is trained to learn the features Connect the two data sets.
The AI model was able to accurately classify Six selected types of valvular heart disease, with the area under the curve, or AUC, ranging from 0.83 to 0.92 (AUC is a rating index that indicates the ability of an AI model and uses a range of values from 0 to 1, the closer to one, the better). The AUC was 0.92 with a 40 percent cutoff for detection of left ventricular ejection.
“It took us a long time to get to these results, but I think that’s it important research“In addition to improving physicians’ diagnostic efficiency, the system can also be used in areas where there are no specialists, in emergencies at night, and for patients who have difficulty undergoing echocardiograms,” says Dr. Ueda. .
Although it may contain data, data or observations from health institutions or professionals, the information contained in Redacción Médica is edited and prepared by journalists. We recommend the reader to consult a health professional with any health-related questions.
“Creator. Devoted pop culture specialist. Certified web fanatic. Unapologetic coffee lover.”