Research

Background

Aortic stenosis occurs when the heart’s aortic valve narrows, reducing blood flow from the heart to the body. Early stages of the disease—like aortic sclerosis or mild aortic stenosis—often have no symptoms, making it difficult to know who may progress to more serious disease.

The Yale CarDS Lab has developed an artificial intelligence (AI) tool designed to predict which patients with early-stage valve disease are at risk for progression. The Detect-AS Prognostic Study is a national research effort testing how well this AI tool can forecast disease progression over time—using follow-up heart ultrasounds and medical records to track outcomes.

About the study

Detect-AS Prognostic is a prospective cohort study funded by the National Institute on Aging. It evaluates how well an artificial intelligence (AI) tool can predict the progression of early-stage aortic valve disease in older adults. The study will enroll 210 adults aged 65 and older with aortic sclerosis or mild aortic stenosis across three U.S. health systems. Participants will receive a follow-up heart ultrasound and be monitored over time through their medical records to assess changes in their condition.

  • You may qualify if you are attending a participating clinic at an enrolling site and:

    • Are 65 or older

    • Have aortic sclerosis or mild aortic stenosis

    • Had an echocardiogram in the past 2-3 years

  • The study will be completed in a single visit. It includes:

    • brief questionnaire

    • cardiac ultrasound

    Participants agree to share their medical records with researchers for three years after their study visit.

  • Participants will be compensated $50 for completing this study.

Learn More

Enrollment is limited to select health clinics at enrolling sites. If you have been contacted regarding enrollment in the study, please reach out to the relevant study team below. If you are interested in learning more about the study design or results, please visit the study page on clinicaltrials.gov.

Enrolling Sites