Progressive Learning with Anatomical Priors for Reliable Left Atrial Scar Segmentation from Late Gadolinium Enhancement MRI

Jing Zhang; Bastien Bergere; Emilie Bollache; Jonas Leite; Mikaël Laredo; Alban Redheuil; Nadjia Kachenoura

cs.CVMar 27, 2026

Progressive Learning with Anatomical Priors for Reliable Left Atrial Scar Segmentation from Late Gadolinium Enhancement MRI

Authors: Jing Zhang, Bastien Bergere, Emilie Bollache, Jonas Leite, Mikaël Laredo, Alban Redheuil, Nadjia Kachenoura

Organizations: Laboratoire d’Imagerie Biomédicale (LIB), Sorbonne Université, INSERM, CNRS, Paris, France · APHP, Institut de Cardiologie, Hôpital Pitié-Salpêtrière, Paris, France · APHP, Imagerie Cardiovasculaire et thoracique, Institut de Cardiologie, Hôpital Pitié Salpêtrière, Paris, France

Abstract

Cardiac MRI late gadolinium enhancement (LGE) enables non-invasive identification of left atrial (LA) scar, whose spatial distribution is strongly associated with atrial fibrillation (AF) severity and recurrence. However, automatic LA scar segmentation remains challenging due to low contrast, annotation variability, and the lack of anatomical constraints, often leading to non-reliable predictions. Accordingly, our aim was to propose a progressive learning strategy to segment LA scar from LGE images inspired from a clinical workflow. A 3-stage framework based on SwinUNETR was implemented, comprising: 1) a first LA cavity pre-learning model, 2) dual-task model which further learns spatial relationship between LA geometry and scar patterns, and 3) fine-tuning on precise segmentation of the scar. Furthermore, we introduced an anatomy-aware spatially weighted loss that incorporates prior clinical knowledge by constraining scar predictions to anatomically plausible LA wall regions while mitigating annotation bias. Our preliminary results obtained on validation LGE volumes from LASCARQS public dataset after 5-fold cross validation, LA segmentation had Dice score of 0.94, LA scar segmentation achieved Dice score of 0.50, Hausdorff Distance of 11.84 mm, Average Surface Distance of 1.80 mm, outperforming only a one-stage scar segmentation with 0.49, 13.02 mm, 1.96 mm, repectively. By explicitly embedding clinical anatomical priors and diagnostic reasoning into deep learning, the proposed approach improved the accuracy and reliability of LA scar segmentation from LGE, revealing the importance of clinically informed model design.

Figures & tables

Operation		Degree	P	Rationale
$T_{\text{bias}}$	Bias Field	[0.02, 0.06]	0.15	Magnetic field inhomogeneity
$T_{\text{elastic}}$	3D Elastic	$\sigma$ : [3, 5]; $\Delta^{*}$	0.10	Anatomical deformation
$T_{\text{rot}}$	Rotation	[-15 ∘ , 15 ∘ ]	0.50	Rigid transformation
$T_{\text{zoom}}$	Zoom	[0.9, 1.1]	0.30	Rigid transformation
$T_{\text{blur}}$	Gaussian Blurring	[0.5, 1.5]	0.20	Reconstruction effect
$T_{\text{contrast}}$	Contrast	[0.8, 1.3]	0.30	Reconstruction effect

Table 1: Data augmentation operations along with rationale. Each operation has a probability (P) and range of variation.

Figure 1: Progressive learning for LA scar segmentation from LGE MRI. In Stage I, the model learns to identify the easy LA structure; in Stage II, the model inherited from Stage I, and learns to identify LA and scar at the same time with the use of two decoders as well as LA wall constraint for scar; in Stage III, the model inherited from Stage II, and keeps fine-tuning scar segmentation.

Panel of Figure 2 — Figure 2: Examples of cropped axial images from LGE MRI volumes along with LA cavity (green) as well as scar (red) expert annotations from LASCARQS public dataset.

Figure 3: Examples of LA cavity and scar segmentation from LASCARQS public dataset. From left to right columns: input LGE MRI images, LA ground truth (GT), LA predicitons from Stage I+II. LA scar GT, scar predictions from Stage I+II and Stage I+II+III respectively. The Dice scores are indicated in red.

LA cavity segmentation ( $n=12$ )
Stage	Dsc (%)	HD (mm)	ASD (mm)
I	$\textbf{0.94}\pm\textbf{0.02}$	$\textbf{3.05}\pm\textbf{0.96}$	$\textbf{0.78}\pm\textbf{0.32}$
I+II	$0.90\pm 0.01$	$8.30\pm 2.44$	$2.28\pm 1.13$
I+II+III	–	–	–
LA scar segmentation ( $n=12$ )
Stage	Dsc (%)	HD (mm)	ASD (mm)

Table 2: Performance for LA and scar segmentation on validation set in terms of Dice score (Dsc)

\uparrow

, Hausdorff distance (HD)

\downarrow

and Average Surface Distance (ASD)

\downarrow

Baseline	Stage	Dsc (%) $\uparrow$	HD (mm) $\downarrow$	ASD (mm) $\downarrow$
nnUNET	–	$0.50\pm 0.09$	$16.10\pm 24.33$	$1.51\pm 3.63$
B1	III	$0.49\pm 0.02$	$14.16\pm 4.26$	$2.12\pm 0.76$
B2	I+III	$0.50\pm 0.02$	$11.67\pm 2.48$	$1.72\pm 0.34$
B3 ∗	II	$0.45\pm 0.02$	$14.49\pm 9.86$	$2.98\pm 1.71$
B4 ∗	II+III	$0.49\pm 0.01$	$11.25\pm 0.80$	$1.84\pm 0.81$
Full	I+II+III	$0.50\pm 0.02$	$11.84\pm 3.11$	$1.80\pm 0.55$

Table 3: Scar segmentation performances on validation set according to different baselines (B), in terms of Dsc, HD and ASD.

Explore similar work

Curriculum-Guided Myocardial Scar Segmentation for Ischemic and Non-ischemic Cardiomyopathy
Mar 30, 2026Nivetha Jayakumar, Jonathan Pan, Shuo Wang +6Medical Image Segmentation Cardiac MRI
Department of Electrical and Computer Engineering, University of Virginia, USA · Division of Cardiovascular Medicine, University of Virginia Health System, USA · Department of Computer Science, University of Virginia, USA
CalcSeg: Confidence-aware 3D Latent Context Curriculum Learning For Myocardial Scar Segmentation From Single-Stack LGE-CMRs
Aug 20, 2026Nivetha Jayakumar, Hannah Kim, Amit R. Patel +13D Medical Image Segmentation Semi-Supervised Medical Image Segmentation
School of Engineering and Applied Science, University of Virginia, USA · School of Medicine, University of Virginia, USA
Learning from VAE Errors to support ECG-based Differential Diagnosis of Myocardial Scar
Sep 4, 2026Shayan Sharifi, Riccardo Treu, Ilaria Gandin +3Electrocardiogram Classification Variational Autoencoders
Department of Mathematics, Informatics, and Geosciences, University of Trieste, Italy · Centre of Diagnosis and Management of Cardiomyopathies; Azienda Sanitaria Universitaria Giuliano Isontina; University of Trieste, Italy; Member of Ern Guard-Heart · Department of Medical, Surgical and Health Sciences, University of Trieste, Italy

Mar 30, 2026J/K move · Enter open · S save

Curriculum-Guided Myocardial Scar Segmentation for Ischemic and Non-ischemic Cardiomyopathy

Nivetha Jayakumar, Jonathan Pan, Shuo Wang, Bishow Paudel+5

Department of Electrical and Computer Engineering, University of Virginia, USA · Division of Cardiovascular Medicine, University of Virginia Health System, USA · Department of Computer Science, University of Virginia, USA

Medical Image Segmentation Cardiac MRI

Open paper AskPDF arXiv

Identification and quantification of myocardial scar is important for diagnosis and prognosis of cardiovascular diseases. However, reliable scar segmentation from Late Gadolinium Enhancement Cardiac Magnetic Resonance (LGE-CMR) images remains a challenge due to variations in contrast enhancement across patients, suboptimal imaging conditions such as post contrast washout, and inconsistencies in ground truth annotations on diffuse scars caused by inter observer variability. In this work, we propose a curriculum learning-based framework designed to improve segmentation performance under these challenging conditions. The method introduces a progressive training strategy that guides the model from high-confidence, clearly defined scar regions to low confidence or visually ambiguous samples with limited scar burden. By structuring the learning process in this manner, the network develops robustness to uncertain labels and subtle scar appearances that are often underrepresented in conventional training pipelines. Experimental results show that the proposed approach enhances segmentation accuracy and consistency, particularly for cases with minimal or diffuse scar, outperforming standard training baselines. This strategy provides a principled way to leverage imperfect data for improved myocardial scar quantification in clinical applications. Our code is publicly available on GitHub.

Abstract

Explore similar work

Curriculum-Guided Myocardial Scar Segmentation for Ischemic and Non-ischemic Cardiomyopathy

CalcSeg: Confidence-aware 3D Latent Context Curriculum Learning For Myocardial Scar Segmentation From Single-Stack LGE-CMRs

Learning from VAE Errors to support ECG-based Differential Diagnosis of Myocardial Scar