PhD Scientific Days 2025

Poster Session I. - T: Cardiovascular Medicine and Research

Utility of Artificial Intelligence Plaque Quantification to Detect Ischemia on Cardiac MRI

Előadó neve

Dr. Nagy Kristóf

Neptun code

H7BUPL

Előadó munkahelye

Semmelweis University, Medical Imaging Centre

Előadó telefonszáma

+36305504988

Előadó e-mail címe

kristof9852@gmail.com

Az előadás címe

Utility of Artificial Intelligence Plaque Quantification to Detect Ischemia on Cardiac MRI

Szerző(k) neve és munkahelye

Kristóf Nagy¹, Matthias Aurich², Philipp Fortner², Kati Jatsch³, Jochannes Goerich², Florian André⁴, Sebastian Buß², Pál Maurovich-Horvat¹

1: Semmelweis University, Medical Imaging Centre
2: MVZ-DRZ, Heidelberg, Germany
3: University of Heidelberg, Medical Faculty Mannheim, Heidelberg, Germany
4: University Hospital Heidelberg, Department of Cardiology, Angiology and Pneumology, Heidelberg, Germany

Bemutatás módja

Poszter

Szekció

Poster Session I. - T: Cardiovascular Medicine and Research

Language of the presentation

English

Preferred session

Cardiovascular Medicine and Research

Összefoglaló szövege

Introduction
Coronary CT angiography (cCTA) is an established modality for the anatomical assessment of coronary artery disease (CAD). In the case of cCTA showing intermediate stenosis, functional assessment, either non-invasively or invasively, is recommended prior to potential revascularisation. Cardiovascular magnetic resonance (CMR) with stress perfusion is a feasible functional modality for the assessment of myocardial ischemia.
Aims
We aimed to evaluate the association between coronary plaque characteristics derived from cCTA and myocardial ischemia detected by CMR.
Methods
Patients who underwent both cCTA and CMR examinations within a 6-month period at a high-volume cardiac imaging center between 2018 and 2024 were identified through a database query. Potentially obstructed vessels defined as at least 50% diameter stenosis were included in the final analysis. Myocardial segments with ischemia were assigned to the corresponding coronary artery territories by two experienced readers. A novel, automated AI prototype was used for plaque analysis. To determine the total plaque burden per vessel, the plaques in each vessel and its side branches were summed. Uni- and multivariate logistic regression models were used to identify plaque characteristics predictive of ischemia, in case of nominal variables chi2 test was performed.
Results
The final study population included 400 patients with 748 potentially obstructed vessels, of which 70 patients showed evidence of ischemia or LGE on stress perfusion CMR. Vessels with ischemia had higher plaque volume, calcified-, and non-calcified volume, plaque length, and vessel volume at the lesion, percentage area- and diameter stenosis and smaller minimal luminal area and diameter. In addition, high-risk plaque characteristics were significantly more prevalent (chi2 =6.8; p< 0.001). The multivariate model included minimal luminal diameter, plaque volume and the presence of high-risk plaques as relevant parameters for the prediction of ischemia on CMR.
Conclusion
Minimal luminal diameter, plaque volume, and the presence of high-risk plaque characteristics were significant predictors of ischemia in the respective vessel territory. Plaque characteristics improve prediction of ischemia beyond degree of stenosis.
Funding
Kristóf Nagy EFOP-3.6.3-VEKOP-16-2017-00009
Florian André research funding from Siemens Healthineers

University

Semmelweis University

Supervisor

Florian André; Pál Maurovich-Horvat

Publication of my abstract

I do not give consent to the publication of my abstract on the website of the congress.

phd.section.field

in doctoral studies before complex exam (PhD)

Kind

Szabad

Status

elfogadva

Accepted presentation method

poszter

Előadás fájl jóváhagyás

nem rendelkezett róla

Előadó

9089

Start

17:12

End

17:18