Anomalous aortic origin of the coronary arteries (AAOCA) is a rare congenital disease and one of the leading causes of sudden cardiac deaths (SCD) in young athletes but also has a lethal presentation in adult age with myocardial infarction, even if not related to obstructive coronary arteries. Unfortunately, diagnostic imaging techniques, invasive assessment, and provocative stress tests have shown low sensitivity and specificity in detecting inducible ischemia, and a multimodality assessment is then necessary.

Innovative tools have been developed in the medical field using computer-based simulation, 3-dimensional reconstruction, machine learning, and artificial intelligence (AI). With the application of such new technologies, we aim to fill the gap of knowledge and the diagnostic limitation regarding risk stratification for most subjects with AAOCA.

This work seeks to enhance, fasten, and personalize the clinical diagnosis of AAOCA by integrating anatomical measurements, clinical data, and biomechanical patient-specific features. The SMART study will set a system to automatically segment and classify coronary arteries with AAOCA from computerized tomography angiography (CTA) by artificial intelligence (AI). Segmentation will feed a 3D model of the aortic root and coronary artery for biomechanical assessment through finite element analysis (FEA). This will allow us to assess the location of possible coronary artery compression under an effort condition. These in-silico results, the anatomical features measured by AI, and the clinical data will be integrated into a risk model to estimate the hazard risk of adverse events such as SCD or myocardial infarction. This workflow will be framed in an IT system to allow a web-based remote diagnostic service.

Thanks to the proposed multidisciplinary approach, SMART aims to overcome the current diagnostic limitations related to the reduced ability of functional stress tests to detect ischemia. Potentially helping in patient-specific risk stratification, SMART is also thought to provide a way to get a first diagnostic indication about AAOCA being accessible from any hospital, fostering the diffusion of peripheral territorial support to the diagnosis and treatment of such rare disease.

The project aims to create a web-based platform that allows the uploading Computed Tomography Angiography (CTA) images, particularly cardio CTA, with contrast medium in anonymized form.

The CTA images will be processed by a neural network developed by the project, which will be able to segment CTA automatically, identify the presence or not of the anomalous coronary origin, and retrieve geometrical measurements of the anatomy of interest. The anatomical and geometrical measurements, automatically made by artificial intelligence, will be integrated with clinical data and computational simulations (Finite Element Structural Analysis) to understand the potential site of dynamic coronary compression under simulated stress conditions.

The final output of the platform will be a report that will integrate clinical data and geometrical and anatomical information to estimate the hazard risk of sudden cardiac deaths or major adverse ischemic events.

Inclusion Criteria:

• Adult and pediatric patients (age > 6 years) with anomalous coronary origin from the aorta (AAOCA).
• Patients in spontaneous sinus rhythm.
• Signed informed consent.

Exclusion Criteria:

• Patients with acute and chronic inflammatory conditions such as chronic liver disease, chronic kidney failure (creatinine > 1.5 mg/dl), and thyroid disorders.
• Patients with arrhythmias, absence of sinus rhythm.
• Contraindications to autonomic testing.
• Patients with a known allergy to materials in recording devices.
• Female patients who are pregnant.