3D Conditional Image Synthesis of Left Atrial LGE MRI from Composite Semantic Masks

TL;DR

This paper explores 3D conditional generative models to synthesize realistic LGE MRI images from semantic masks, aiming to augment training data and improve left atrial segmentation accuracy.

Contribution

It introduces a pipeline using three 3D conditional generators, notably SPADE-LDM, to produce high-fidelity synthetic MRI images that enhance segmentation performance.

Findings

SPADE-LDM achieves the lowest FID of 4.063, indicating high realism.

Synthetic images improve LA cavity segmentation Dice score from 0.908 to 0.936.

Augmentation with synthetic data yields statistically significant segmentation improvements.

Abstract

Segmentation of the left atrial (LA) wall and endocardium from late gadolinium-enhanced (LGE) MRI is essential for quantifying atrial fibrosis in patients with atrial fibrillation. The development of accurate machine learning-based segmentation models remains challenging due to the limited availability of data and the complexity of anatomical structures. In this work, we investigate 3D conditional generative models as potential solution for augmenting scarce LGE training data and improving LA segmentation performance. We develop a pipeline to synthesize high-fidelity 3D LGE MRI volumes from composite semantic label maps combining anatomical expert annotations with unsupervised tissue clusters, using three 3D conditional generators (Pix2Pix GAN, SPADE-GAN, and SPADE-LDM). The synthetic images are evaluated for realism and their impact on downstream LA segmentation. SPADE-LDM generates the most realistic and structurally accurate images, achieving an FID of 4.063 and surpassing GAN models, which have FIDs of 40.821 and 7.652 for Pix2Pix and SPADE-GAN, respectively. When augmented with synthetic LGE images, the Dice score for LA cavity segmentation with a 3D U-Net model improved from 0.908 to 0.936, showing a statistically significant improvement (p < 0.05) over the baseline.These findings demonstrate the potential of label-conditioned 3D synthesis to enhance the segmentation of under-represented cardiac structures.