A novel hybrid layer-based encoder–decoder framework for 3D segmentation in congenital heart disease

Abstract The segmentation of cardiac anatomy represents a crucial stage in accurate diagnosis and subsequent treatment planning for patients with congenital heart disease (CHD). However, the current deep learning-based segmentation networks are ineffective when applied to 3D medical images of CHD be...

Full description

Saved in:
Bibliographic Details
Main Authors: Yaoxi Zhu, Hongbo Li, Bingxin Cao, Kun Huang, Jinping Liu
Format: Article
Language:English
Published: Nature Portfolio 2025-04-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-96251-9
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850202754529624064
author Yaoxi Zhu
Hongbo Li
Bingxin Cao
Kun Huang
Jinping Liu
author_facet Yaoxi Zhu
Hongbo Li
Bingxin Cao
Kun Huang
Jinping Liu
author_sort Yaoxi Zhu
collection DOAJ
description Abstract The segmentation of cardiac anatomy represents a crucial stage in accurate diagnosis and subsequent treatment planning for patients with congenital heart disease (CHD). However, the current deep learning-based segmentation networks are ineffective when applied to 3D medical images of CHD because of the limited availability of training datasets and the inherent complexity exhibited by the variability of cardiac and large vessel tissues. To address this challenge, we propose a novel hybrid layer-based encoder–decoder framework for 3D CHD image segmentation. The model incorporates a global volume mixing module and a local volume-based multihead attention module, which uses a self-attention mechanism to explicitly capture the local and global dependencies of the 3D image segmentation process. This enables the model to more effectively learn the shape boundary features of organs, thereby facilitating accurate segmentation of the whole heart (WH) and great vessels. We compare our method with several popular networks on the public ImageCHD and HVSMR-2.0 datasets. The experimental results show that the proposed model achieves excellent performance in WH and great vessel segmentation tasks with high Dice coefficients and IoU indices.
format Article
id doaj-art-bf5a61cd628a48edacb57316310c0da4
institution OA Journals
issn 2045-2322
language English
publishDate 2025-04-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-bf5a61cd628a48edacb57316310c0da42025-08-20T02:11:41ZengNature PortfolioScientific Reports2045-23222025-04-0115111010.1038/s41598-025-96251-9A novel hybrid layer-based encoder–decoder framework for 3D segmentation in congenital heart diseaseYaoxi Zhu0Hongbo Li1Bingxin Cao2Kun Huang3Jinping Liu4Department of Cardiovascular Surgery, Zhongnan Hospital of Wuhan UniversityDepartment of Clinical Medicine, HuanKui Academy, Nanchang UniversityDepartment of Cardiovascular Surgery, Union Hospital, Tongji Medical College, Huazhong University of Science and TechnologyDepartment of Cardiology, Union Hospital, Tongji Medical College, Huazhong University of Science and TechnologyDepartment of Cardiovascular Surgery, Zhongnan Hospital of Wuhan UniversityAbstract The segmentation of cardiac anatomy represents a crucial stage in accurate diagnosis and subsequent treatment planning for patients with congenital heart disease (CHD). However, the current deep learning-based segmentation networks are ineffective when applied to 3D medical images of CHD because of the limited availability of training datasets and the inherent complexity exhibited by the variability of cardiac and large vessel tissues. To address this challenge, we propose a novel hybrid layer-based encoder–decoder framework for 3D CHD image segmentation. The model incorporates a global volume mixing module and a local volume-based multihead attention module, which uses a self-attention mechanism to explicitly capture the local and global dependencies of the 3D image segmentation process. This enables the model to more effectively learn the shape boundary features of organs, thereby facilitating accurate segmentation of the whole heart (WH) and great vessels. We compare our method with several popular networks on the public ImageCHD and HVSMR-2.0 datasets. The experimental results show that the proposed model achieves excellent performance in WH and great vessel segmentation tasks with high Dice coefficients and IoU indices.https://doi.org/10.1038/s41598-025-96251-9Congenital heart diseaseCardiac segmentation3D CT imageDeep learningHybrid architectures
spellingShingle Yaoxi Zhu
Hongbo Li
Bingxin Cao
Kun Huang
Jinping Liu
A novel hybrid layer-based encoder–decoder framework for 3D segmentation in congenital heart disease
Scientific Reports
Congenital heart disease
Cardiac segmentation
3D CT image
Deep learning
Hybrid architectures
title A novel hybrid layer-based encoder–decoder framework for 3D segmentation in congenital heart disease
title_full A novel hybrid layer-based encoder–decoder framework for 3D segmentation in congenital heart disease
title_fullStr A novel hybrid layer-based encoder–decoder framework for 3D segmentation in congenital heart disease
title_full_unstemmed A novel hybrid layer-based encoder–decoder framework for 3D segmentation in congenital heart disease
title_short A novel hybrid layer-based encoder–decoder framework for 3D segmentation in congenital heart disease
title_sort novel hybrid layer based encoder decoder framework for 3d segmentation in congenital heart disease
topic Congenital heart disease
Cardiac segmentation
3D CT image
Deep learning
Hybrid architectures
url https://doi.org/10.1038/s41598-025-96251-9
work_keys_str_mv AT yaoxizhu anovelhybridlayerbasedencoderdecoderframeworkfor3dsegmentationincongenitalheartdisease
AT hongboli anovelhybridlayerbasedencoderdecoderframeworkfor3dsegmentationincongenitalheartdisease
AT bingxincao anovelhybridlayerbasedencoderdecoderframeworkfor3dsegmentationincongenitalheartdisease
AT kunhuang anovelhybridlayerbasedencoderdecoderframeworkfor3dsegmentationincongenitalheartdisease
AT jinpingliu anovelhybridlayerbasedencoderdecoderframeworkfor3dsegmentationincongenitalheartdisease
AT yaoxizhu novelhybridlayerbasedencoderdecoderframeworkfor3dsegmentationincongenitalheartdisease
AT hongboli novelhybridlayerbasedencoderdecoderframeworkfor3dsegmentationincongenitalheartdisease
AT bingxincao novelhybridlayerbasedencoderdecoderframeworkfor3dsegmentationincongenitalheartdisease
AT kunhuang novelhybridlayerbasedencoderdecoderframeworkfor3dsegmentationincongenitalheartdisease
AT jinpingliu novelhybridlayerbasedencoderdecoderframeworkfor3dsegmentationincongenitalheartdisease