A multi-modal multi-branch framework for retinal vessel segmentation using ultra-widefield fundus photographs

BackgroundVessel segmentation in fundus photography has become a cornerstone technique for disease analysis. Within this field, Ultra-WideField (UWF) fundus images offer distinct advantages, including an expansive imaging range, detailed lesion data, and minimal adverse effects. However, the high re...

Full description

Saved in:
Bibliographic Details
Main Authors: Qihang Xie, Xuefei Li, Yuanyuan Li, Jiayi Lu, Shaodong Ma, Yitian Zhao, Jiong Zhang
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-01-01
Series:Frontiers in Cell and Developmental Biology
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fcell.2024.1532228/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850150041775243264
author Qihang Xie
Qihang Xie
Xuefei Li
Yuanyuan Li
Yuanyuan Li
Jiayi Lu
Jiayi Lu
Shaodong Ma
Yitian Zhao
Yitian Zhao
Jiong Zhang
Jiong Zhang
author_facet Qihang Xie
Qihang Xie
Xuefei Li
Yuanyuan Li
Yuanyuan Li
Jiayi Lu
Jiayi Lu
Shaodong Ma
Yitian Zhao
Yitian Zhao
Jiong Zhang
Jiong Zhang
author_sort Qihang Xie
collection DOAJ
description BackgroundVessel segmentation in fundus photography has become a cornerstone technique for disease analysis. Within this field, Ultra-WideField (UWF) fundus images offer distinct advantages, including an expansive imaging range, detailed lesion data, and minimal adverse effects. However, the high resolution and low contrast inherent to UWF fundus images present significant challenges for accurate segmentation using deep learning methods, thereby complicating disease analysis in this context.MethodsTo address these issues, this study introduces M3B-Net, a novel multi-modal, multi-branch framework that leverages fundus fluorescence angiography (FFA) images to improve retinal vessel segmentation in UWF fundus images. Specifically, M3B-Net tackles the low segmentation accuracy caused by the inherently low contrast of UWF fundus images. Additionally, we propose an enhanced UWF-based segmentation network in M3B-Net, specifically designed to improve the segmentation of fine retinal vessels. The segmentation network includes the Selective Fusion Module (SFM), which enhances feature extraction within the segmentation network by integrating features generated during the FFA imaging process. To further address the challenges of high-resolution UWF fundus images, we introduce a Local Perception Fusion Module (LPFM) to mitigate context loss during the segmentation cut-patch process. Complementing this, the Attention-Guided Upsampling Module (AUM) enhances segmentation performance through convolution operations guided by attention mechanisms.ResultsExtensive experimental evaluations demonstrate that our approach significantly outperforms existing state-of-the-art methods for UWF fundus image segmentation.
format Article
id doaj-art-e63ceded157249c786f4010f7aa8376f
institution OA Journals
issn 2296-634X
language English
publishDate 2025-01-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Cell and Developmental Biology
spelling doaj-art-e63ceded157249c786f4010f7aa8376f2025-08-20T02:26:41ZengFrontiers Media S.A.Frontiers in Cell and Developmental Biology2296-634X2025-01-011210.3389/fcell.2024.15322281532228A multi-modal multi-branch framework for retinal vessel segmentation using ultra-widefield fundus photographsQihang Xie0Qihang Xie1Xuefei Li2Yuanyuan Li3Yuanyuan Li4Jiayi Lu5Jiayi Lu6Shaodong Ma7Yitian Zhao8Yitian Zhao9Jiong Zhang10Jiong Zhang11Cixi Biomedical Research Institute, Wenzhou Medical University, Ningbo, ChinaLaboratory of Advanced Theranostic Materials and Technology, Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo, ChinaLaboratory of Advanced Theranostic Materials and Technology, Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo, ChinaCixi Biomedical Research Institute, Wenzhou Medical University, Ningbo, ChinaLaboratory of Advanced Theranostic Materials and Technology, Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo, ChinaCixi Biomedical Research Institute, Wenzhou Medical University, Ningbo, ChinaLaboratory of Advanced Theranostic Materials and Technology, Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo, ChinaLaboratory of Advanced Theranostic Materials and Technology, Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo, ChinaCixi Biomedical Research Institute, Wenzhou Medical University, Ningbo, ChinaLaboratory of Advanced Theranostic Materials and Technology, Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo, ChinaCixi Biomedical Research Institute, Wenzhou Medical University, Ningbo, ChinaLaboratory of Advanced Theranostic Materials and Technology, Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo, ChinaBackgroundVessel segmentation in fundus photography has become a cornerstone technique for disease analysis. Within this field, Ultra-WideField (UWF) fundus images offer distinct advantages, including an expansive imaging range, detailed lesion data, and minimal adverse effects. However, the high resolution and low contrast inherent to UWF fundus images present significant challenges for accurate segmentation using deep learning methods, thereby complicating disease analysis in this context.MethodsTo address these issues, this study introduces M3B-Net, a novel multi-modal, multi-branch framework that leverages fundus fluorescence angiography (FFA) images to improve retinal vessel segmentation in UWF fundus images. Specifically, M3B-Net tackles the low segmentation accuracy caused by the inherently low contrast of UWF fundus images. Additionally, we propose an enhanced UWF-based segmentation network in M3B-Net, specifically designed to improve the segmentation of fine retinal vessels. The segmentation network includes the Selective Fusion Module (SFM), which enhances feature extraction within the segmentation network by integrating features generated during the FFA imaging process. To further address the challenges of high-resolution UWF fundus images, we introduce a Local Perception Fusion Module (LPFM) to mitigate context loss during the segmentation cut-patch process. Complementing this, the Attention-Guided Upsampling Module (AUM) enhances segmentation performance through convolution operations guided by attention mechanisms.ResultsExtensive experimental evaluations demonstrate that our approach significantly outperforms existing state-of-the-art methods for UWF fundus image segmentation.https://www.frontiersin.org/articles/10.3389/fcell.2024.1532228/fullultra-widefieldfundus fluorescence angiographyretinal vessel segmentationmultimodal frameworkselective fusion
spellingShingle Qihang Xie
Qihang Xie
Xuefei Li
Yuanyuan Li
Yuanyuan Li
Jiayi Lu
Jiayi Lu
Shaodong Ma
Yitian Zhao
Yitian Zhao
Jiong Zhang
Jiong Zhang
A multi-modal multi-branch framework for retinal vessel segmentation using ultra-widefield fundus photographs
Frontiers in Cell and Developmental Biology
ultra-widefield
fundus fluorescence angiography
retinal vessel segmentation
multimodal framework
selective fusion
title A multi-modal multi-branch framework for retinal vessel segmentation using ultra-widefield fundus photographs
title_full A multi-modal multi-branch framework for retinal vessel segmentation using ultra-widefield fundus photographs
title_fullStr A multi-modal multi-branch framework for retinal vessel segmentation using ultra-widefield fundus photographs
title_full_unstemmed A multi-modal multi-branch framework for retinal vessel segmentation using ultra-widefield fundus photographs
title_short A multi-modal multi-branch framework for retinal vessel segmentation using ultra-widefield fundus photographs
title_sort multi modal multi branch framework for retinal vessel segmentation using ultra widefield fundus photographs
topic ultra-widefield
fundus fluorescence angiography
retinal vessel segmentation
multimodal framework
selective fusion
url https://www.frontiersin.org/articles/10.3389/fcell.2024.1532228/full
work_keys_str_mv AT qihangxie amultimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT qihangxie amultimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT xuefeili amultimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT yuanyuanli amultimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT yuanyuanli amultimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT jiayilu amultimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT jiayilu amultimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT shaodongma amultimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT yitianzhao amultimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT yitianzhao amultimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT jiongzhang amultimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT jiongzhang amultimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT qihangxie multimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT qihangxie multimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT xuefeili multimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT yuanyuanli multimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT yuanyuanli multimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT jiayilu multimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT jiayilu multimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT shaodongma multimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT yitianzhao multimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT yitianzhao multimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT jiongzhang multimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs
AT jiongzhang multimodalmultibranchframeworkforretinalvesselsegmentationusingultrawidefieldfundusphotographs