AI-enhanced echography supports aortic stenosis patients

Echocardiography to assess aortic stenosis severity, supported by a novel AI algorithm, can better identify patients at high death risk who could benefit from treatment.

1,000,000 echocardiograms from over 630,000 patients contributed to training the AI

Prof. Geoffrey Strange (University of Notre Dame, Australia) pointed out that although echography can assess peak velocity, mean pressure gradient, and aortic valve area, the severity score determining the treatment plan varies from site to site. AI-ENHANCED AS examined whether an AI algorithm could methodologically support the accurate identification of moderate-to-severe and severe aortic stenosis phenotypes associated with high 5-year mortality. To that end, researchers developed the AI-Decision Support Algorithm (AI-DSA) and trained it using randomly selected data from 70% patients in the National Echo Database of Australia (NEDA), which contains more than 1,000,000 echocardiograms from over 630,000 patients and is linked to mortality data. The other 30% of NEDA data was used as a comparator. Out of 179,054 individuals, of the 2.5% in whom AI-DSA detected a severe phenotype, 77.2% of those also met guideline criteria for severe aortic stenosis.

In patients with the moderate-to-severe phenotype, the 5-year mortality rate was 56.2% compared with 67.9% for the severe phenotype versus 22.9% for those with neither of those phenotypes. All-cause mortality odds ratios were 1.82 (95% CI 1.63–2.02) and 2.80 (95% CI 2.57–3.06) for patients with the moderate-to-severe and severe phenotypes, respectively.

More proactivity recommended for identifying those at risk

Prof. Strange concluded: “This proprietary AI algorithm picks up patients with a high risk (and all patients within current guidelines) of dying within 5 years that may be missed by conventional definitions. The findings suggest that the AI algorithm could be used in clinical practice to alert physicians to patients who should undergo further investigations to determine if they qualify for aortic valve replacement. Given the rising prevalence of aortic stenosis and its impact on mortality, it is time to revisit the practice of watchful waiting and consider more proactive attempts to identify those at risk.

More research is needed to determine if aortic valve replacement improves survival and quality of life in patients identified by the AI-DSA as having a high risk of mortality, but who do not meet current guideline definitions.”

  1. Strange G, et al. AI-ENHANCED detection of Aortic Stenosis. Hot Line Session 6, ESC Congress 2022, Barcelona, Spain, 26–29 August.