Severe aortic stenosis can be readily detected by AI algorithm-assisted reading of echocardiograms
Australia: An artificial intelligence (AI) algorithm employed in the reading of routine echocardiograms can detect patients with aortic stenosis (AS) of sufficient severity that indicates a high risk of early mortality, says a study.
"The AI-DSA -- artificial intelligence decision support algorithm -- helped in automatically identifying patients with moderate to severe aortic stenosis tied to poor survival if not treated." Geoffrey A. Strange, faculty of medicine, University of Sydney, reported.
The training of AI-DSA was done on more than 1 million echocardiograms acquired from 630,000 patients in the National Echo Database (NEDA) of Australia. AI ENHANCED AS which was the testing phase of the study was conducted on 179,054 patients from the same database. In the testing phase, the researchers compared mortality for those determined by AI to have a low likelihood of clinically significant AS, severe AS, or moderate to severe AS.
Based on the study, the researchers found the following:
- In the nearly 200,000 patients evaluated from the database, the AI-DSA classified 2.5% as having moderate to severe AS and 1.4% as having severe AS.
- Relative to a 22.9% mortality at 5 years in the low-risk reference group, the rates were 56.2% and 67.9% in the moderate to severe and severe groups, respectively.
- When expressed as odds ratios, the mortality risk for the moderate to severe group (OR, 1.8) and severe group (HR, 2.8) were about two to three times higher than the low probability group.
Dr. Strange reported, "the algorithm selected all patients identified with severe AS in current guidelines and also identified patients that conventional definitions." It supports the idea that the AI algorithm could be utilized in clinical practice to alert physicians to patients who would be required to undergo further examinations to determine if they are eligible for aortic valve replacement.
According to Strange, this algorithm is not definitely not meant to take away from decision-making. However, there is an unmet need for doing better in AS detection. The data was presented to show that with regards to 5-year outcomes even moderate AS is not benign, and AI-DSA can allow clinicians to detect earlier significant diseases and intervene in a timelier manner.
He said, "it is time to revisit the practice of watchful waiting and consider more attempts to identify patients who are at risk." The next step would be to determine if AI-DSA makes a clinical difference.
"A study is now needed to determine if aortic valve replacement in patients identified as being at risk by AI-DSA betters survival and quality of life, especially in those who do not meet current guideline definitions of clinically significant disease," he concluded.
