Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
February 2, 2026European Heart Journal - Cardiovascular Imaging

R-GAN synthesizes high-fidelity T2 maps from OS-CMR images with ~100% correlation in edema detection.

View Full Paper
Ask AI
Bookmark
Share

Why the study?

T2-weighted MRI detects myocardial edema but is limited by long acquisition times and protocol variability.

Does a novel R-GAN model accurately synthesize T2 parametric maps from OS-CMR images for myocardial edema detection compared to original T2-weighted images?

Population

2,189 paired OS-CMR and T2-weighted images from two cohorts

Comparison

Novel R-GAN vs Pix2Pix

Design

Model development and validation study

Key result

A novel R-GAN successfully synthesized high-fidelity T2 parametric maps from OS-CMR images, achieving a peak signal-to-noise ratio of 21.56–24.95 dB and Pearson correlation of 0.862–0.923.

Authors

FLF LotfikazemiMFM G FriedrichMBM B Benovoy

Discussion

Loading...

Member takes

Overview

May enable contrast-free, time-efficient edema detection in CMR; extends GAN synthesis to quantitative T2 parametric mapping.

Key Points

  • The aim is to synthesize high-fidelity T2 parametric maps from native OS-CMR images for accurate detection of myocardial edema.
  • Developed a novel R-GAN using 2,189 paired OS-CMR and T2-weighted images.
  • Performed normalization, registration, and augmentation during preprocessing.
  • Compared R-GAN performance against Pix2Pix using PSNR, SSIM, and PCC.
  • Conducted clinical validation through signal intensity analysis and EPG simulations.
  • R-GAN achieved PSNR of 21.56–24.95 dB, SSIM of 0.672–0.739, and PCC of 0.862–0.923.
  • Signal intensities aligned closely between original and synthetic maps for both edema and healthy tissue.
  • EPG simulations showed perfect Pearson and Spearman correlations (1.0000) with minimal error metrics.

Structured PICO

Does a novel R-GAN model accurately synthesize T2 parametric maps from OS-CMR images for myocardial edema detection compared to original T2-weighted images?

P
Population
2,189 paired OS-CMR and T2-weighted images from two cohorts used to develop and validate a novel R-GAN for myocardial edema detection.
E
Exposure
Novel residual generative adversarial network (R-GAN) to synthesize T2 parametric maps from native OS-CMR images
C
Comparator
Pix2Pix model and original T2-weighted images
O
Outcome
Image synthesis fidelity measured by peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), and Pearson correlation coefficient (PCC), along with clinical validation of signal intensity and T2-curve evaluationssurrogate

A novel R-GAN model can synthesize high-fidelity T2 parametric maps from contrast-free OS-CMR images, potentially offering a time-efficient tool for myocardial edema detection.

Cite This Study

Lotfikazemi et al. (2026) studied Myocardial edema (n=2,189). R-GAN (synthetic T2 parametric maps from OS-CMR) vs. Pix2Pix and original T2-weighted images was evaluated on Image quality and correlation (PSNR, SSIM, and PCC). A novel R-GAN successfully synthesized high-fidelity T2 parametric maps from OS-CMR images, achieving a peak signal-to-noise ratio of 21.56–24.95 dB and Pearson correlation of 0.862–0.923.

synapsesocial.com/papers/6980ff19c1c9540dea811be1https://doi.org/10.1093/ehjci/jeaf367.391
View Full Paper
Ask AI
Bookmark
Share