Identification of Coronary Atherosclerotic Plaques Causing type 1 Myocardial infarction with coronary computed tomography angiography and deep-learning algorithms

The objective is to examine whether advanced imaging techniques and AI-based plaque characterization with deep learning algorithms alone or in combination with other biomarkers, including targeted omics, can differentiate between coronary plaques that will cause a type 1 myocardial infarction (MI) and coronary plaques that will not cause a type-1 MI.SCAPIS included 30154 randomly selected men and women.

We will create a nested case control study withinSCAPIS with 400 cases that after undergoing CCTA have developed a type 1 MI and controls (n=400, matched forage, sex and time of CCTA) that after undergoing CCTA have not developed type 1 MI. All cases will be carefullyadjudicated and characterized regarding clinical and angiographic findings at the event.

The CCTAs performed atbaseline in cases and controls will be analyzed using deep learning algorithms, quantitativemeasurements from commercially available software and modern targeted omics techniques. Models will beoptimized to discriminate between cases and controls.

These models can then be validated in the whole SCAPIScohort but also in international CCTA cohorts.By developing methods that can discriminate between coronary plaques that will cause type 1 MI from those that willnot, and use more precise outcome definitions, we can improve understanding regarding underlying mechanisms anddevelop better and more precise methods for risk stratification and identify new treatment targets.