Automated Screening for Aortic Stenosis

Aortic stenosis (AS) affects over 12.6 million adults and causes an estimated 102,700 deaths annually. There is interest in novel approaches to identify valve disease earlier the disease course before symptoms occur. We previously used machine learning (ML) to develop a system for fully automated AS screening. Here we are validating the performance of this new method for identifying AS.

Inclusion Criteria

• Age > 65 years

• Referred for clinically indicated transthoracic echocardiography

Exclusion Criteria

• Age < 65 years

• Prior valve replacment procedure

Study Requirements

The study design involves 10 minutes of additional image acquisition after a routine transthoracic echocardiography. Point of care ultrasound (POCUS) data acquisition is non-invasive and by design appropriate for population screening. The acquisition will be similar to the imaging exam that was just completed (as part of routine care) but with a different device (Butterfly IQ+ device). This is a commercially available POCUS device. The POCUS imaging will be saved, encrypted, and uploaded to a HIPAA compliant cloud in standard fashion for automated analysis.