Deep Learning for Cardiac Image Analysis

doi:10.1016/j.jcmg.2026.03.007

Deep Learning for Cardiac Image Analysis

2026-05-01

0

OA

AI

J

Joske van der Zande *

Dimitrios Karkalousos

DOI：10.1016/j.jcmg.2026.03.007

Original

Save

View PDF

Abstract

En 中文

• DL is increasingly adopted for cardiac image analysis, with developments in model architectures providing new opportunities to improve patient care. • New DL architectures enhance the accuracy and efficiency of cardiac image analysis by capturing more complex spatial and temporal relationships. • Novel DL techniques improve cross-modality synthesis, noise reduction, and high-resolution imaging. • Large-scale models trained on diverse data sources enable more generalizable and adaptable solutions for cardiac imaging. • Although promising, these architectures face challenges including explainability, domain generalizability, and clinical validation—key focus areas for future cardiac imaging DL research.

Keywords:

artificial intelligence

cardiac imaging

deep learning

foundation models

generative adversarial networks

graph neural networks

implicit neural representations

transformers

AI

artificial intelligence

CAC

coronary artery calcium

CMR

cardiac magnetic resonance

CNN

convolutional neural network

DL

deep learning

GAN

generative adversarial network

GCN

graph convolutional network

GNN

graph neural network

INR

implicit neural representation

SAM

segment anything model

SPECT

single-photon emission computed tomography

TEE

transesophageal echocardiography

ViT

vision transformer

Deep Learning for Cardiac Image Analysis

Abstract

AI Summary

Journal

Organization