Artificial intelligence (AI) is more effective than humans at assessing heart ultrasound, the leading test of overall heart health.as detailed this Wednesday (04/05/2023) in a study published in the journal Nature.
While previous studies have demonstrated the potential of AI models for reading medical examinations, this study presents the results of the first blinded randomized clinical trial of heart health.
“We’re Ready for Prime Time”
“There is a lot of hype around AI, but rigorous testing is still needed,” said David Ouyang, lead author of the study. The success of this trial “really reinforces the argument that we are now ready for prime time,” added a cardiologist at Cedars-Sinai Medical Center in Los Angeles.
An ultrasound scan of the heart, also known as an echocardiogram, is performed on patients by ultrasound specialists.healthcare professionals, who usually do an initial assessment before passing it on to a cardiologist.
Sonographers versus artificial intelligence
A new study compared the AI model with sonographers to see who gave the most accurate initial estimate.. Both tested the so-called left ventricular ejection fraction, which measures the heart’s ability to pump blood around the body.
This test is the main way to measure the work of the heart.. It is used to find out if patients have had a heart attack or if they should undergo major treatment, such as a defibrillator implant.
AI has become faster and more accurate
Approximately 3,500 cardiac ultrasounds between sonographers and an AI model were randomized for the study.. Their scores were then evaluated by cardiologists, who did not know which scores were from humans and which were from the AI model.
According to the study, cardiologists have made significant changes to over 27% of sonographic evaluations and nearly 17% of evaluations performed with the AI model..
“AI was faster, more accurate and indistinguishable from cardiologists,” Ouyang said.
He added that there is a “huge shortage” of ultrasound specialists in the United States and around the world and this will save them valuable time.
Its use is being sought in the United States.
An artificial intelligence model called EchoNet-Dynamic has been trained on nearly 145,000 echocardiograms and uses what is known as deep learning to process huge amounts of data..
The researchers are currently applying for Federal Drug Administration (FDA) approval of the method and hope to do the same in the European Union and other countries soon, Ouyang said.
JU (efe, afp)
Source: La Opinion
Ashley Fitzgerald is an accomplished journalist in the field of technology. She currently works as a writer at 24 news breaker. With a deep understanding of the latest technology developments, Ashley’s writing provides readers with insightful analysis and unique perspectives on the industry.