Case H: Triaging Patients with Suspected Cardiac Disease

Work Packages: WP1, WP3

Background:

The medical community is facing a growing demand for cardiac ultrasound examinations, while simultaneously experiencing a shortage of cardiologists and skilled cardiac sonographers capable of performing and interpreting these studies. The United States Bureau of Labor Statistics projects a 15% increase in employment of diagnostic medical sonographers between 2021 and 2031. At the same time, low-cost ultrasound equipment has become increasingly accessible to general practitioners, creating opportunities to expand ultrasound use in primary care settings.

Objective:

This project aims to develop AI-based models for automated classification of echocardiography exams as normal or abnormal. Subsequently, models for automated cardiac pathology classification will be developed, along with an investigation into the minimal set of cardiac views required for accurate classification. The proposed technology will support primary care physicians in triaging patients with suspected cardiovascular abnormalities and assist in determining whether referral to a specialized cardiac facility is necessary.