Harnessing Artificial Intelligence and Deep Learning to Determine a Coronary Artery Calcium Estimate in Patients with No History of AtheroSclerotic CardioVascular Disease (HIDDEN-ASCVD) Study

Summary

This study aims to develop and validate a deep learning algorithm for estimating coronary artery calcium (CAC) from non-cardiac, contrast-enhanced CT scans and to identify optimal patient and provider notification strategies to increase statin prescription rates within an integrated health system.

What they want

The project involves validating an existing deep learning (DL) algorithm for CAC quantification and adapting it to contrast-enhanced CT scans using self-supervised DL. It will describe the epidemiology and outcomes of opportunistic CAC. A factorial randomized controlled trial (RCT) will be conducted with 5,760 adults to test multiple notification strategies based on MINDSPACE behavior change principles and the Multiphase Optimization Strategy (MOST) framework, aiming to maximize statin therapy and minimize patient anxiety. Co-design focus groups will inform strategy adaptation, followed by a single-arm validation study to estimate the intervention's net effect on statin initiation and patient acceptability. The RCTs will be enriched for historically marginalized racial and ethnic groups.

Deliverables

Validated DL-CAC algorithm adapted for contrast-enhanced CT scans
Description of the epidemiology and outcomes of opportunistic CAC
Identification of the most effective opportunistic CAC notification strategy
Unbiased estimate of the net effect of the intervention on statin initiation and patient acceptability
Generalizable knowledge regarding health disparities in ASCVD prevention

Technical requirements

Deep learning (DL) algorithms
Self-supervised DL
Electrocardiogram (ECG)-gated computed tomography (CT) scans
Non-gated chest CTs
Non-contrast chest CTs
Contrast-enhanced CT scans
Factorial Randomized Controlled Trial (RCT) design
MINDSPACE behavior change principles
Multiphase Optimization Strategy (MOST) framework