MSU-Net: A Synthesized U-Net for Exploiting Multi-Scale Features in OCT Image Segmentation

The U-Net architecture is widely recognized as a prominent algorithm for choroidal segmentation in optical coherence tomography (OCT) images. However, conventional U-Net implementations exhibit two critical limitations. First, the backbone employs uniform-sized convolutional kernels to process featu...

Full description

Saved in:
Bibliographic Details
Main Authors: Dejie Chen, Xiangping Chen, Hao Gu, Su Zhao, Hao Jiang
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10949143/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850141345680719872
author Dejie Chen
Xiangping Chen
Hao Gu
Su Zhao
Hao Jiang
author_facet Dejie Chen
Xiangping Chen
Hao Gu
Su Zhao
Hao Jiang
author_sort Dejie Chen
collection DOAJ
description The U-Net architecture is widely recognized as a prominent algorithm for choroidal segmentation in optical coherence tomography (OCT) images. However, conventional U-Net implementations exhibit two critical limitations. First, the backbone employs uniform-sized convolutional kernels to process feature maps across all channels within the same layer, resulting in homogeneous receptive fields and a single-scale bottleneck that impedes global contextual feature extraction. Second, the skip connections are restricted to same-scale feature maps between encoder and decoder, failing to exploit cross-semantic hierarchical feature interactions. To address these issues, this study introduces MSU-Net, a novel neural network for OCT-based choroidal segmentation. The proposed framework enhances performance through two innovations: 1) replacement of standard encoder blocks with a multi-branch module combining heterogeneous convolutions to achieve multi-scale receptive field diversification; 2) redesign of skip connections through a pyramid fusion module with spatial attention for adaptive multi-level feature weighting. This architecture enables progressive refinement of low-level features guided by high-level semantics, significantly improving feature discriminability. Experimental results demonstrate superior performance with metrics of 99.5% (accuracy), 96.7% (sensitivity), 94.7% (Dice), and 94.6% (MIoU), surpassing the baseline by 0.4%, 3.7%, 2.8%, and 2.9% respectively. Notably, the model shows consistent advantages in segmenting indistinct choroidal boundaries compared to state-of-the-art methods.
format Article
id doaj-art-b7b51caeb72b4de49028133190aab39b
institution OA Journals
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-b7b51caeb72b4de49028133190aab39b2025-08-20T02:29:27ZengIEEEIEEE Access2169-35362025-01-0113706637067510.1109/ACCESS.2025.355780010949143MSU-Net: A Synthesized U-Net for Exploiting Multi-Scale Features in OCT Image SegmentationDejie Chen0https://orcid.org/0009-0006-0157-9087Xiangping Chen1https://orcid.org/0000-0002-6064-4508Hao Gu2Su Zhao3Hao Jiang4School of Electrical Engineering, Guizhou University, Guiyang, ChinaSchool of Electrical Engineering, Guizhou University, Guiyang, ChinaDepartment of Ophthalmology, The Affiliated Hospital of Guizhou Medical University, Guiyang, ChinaDepartment of Ophthalmology, The Affiliated Hospital of Guizhou Medical University, Guiyang, ChinaDepartment of Ophthalmology, The Affiliated Hospital of Guizhou Medical University, Guiyang, ChinaThe U-Net architecture is widely recognized as a prominent algorithm for choroidal segmentation in optical coherence tomography (OCT) images. However, conventional U-Net implementations exhibit two critical limitations. First, the backbone employs uniform-sized convolutional kernels to process feature maps across all channels within the same layer, resulting in homogeneous receptive fields and a single-scale bottleneck that impedes global contextual feature extraction. Second, the skip connections are restricted to same-scale feature maps between encoder and decoder, failing to exploit cross-semantic hierarchical feature interactions. To address these issues, this study introduces MSU-Net, a novel neural network for OCT-based choroidal segmentation. The proposed framework enhances performance through two innovations: 1) replacement of standard encoder blocks with a multi-branch module combining heterogeneous convolutions to achieve multi-scale receptive field diversification; 2) redesign of skip connections through a pyramid fusion module with spatial attention for adaptive multi-level feature weighting. This architecture enables progressive refinement of low-level features guided by high-level semantics, significantly improving feature discriminability. Experimental results demonstrate superior performance with metrics of 99.5% (accuracy), 96.7% (sensitivity), 94.7% (Dice), and 94.6% (MIoU), surpassing the baseline by 0.4%, 3.7%, 2.8%, and 2.9% respectively. Notably, the model shows consistent advantages in segmenting indistinct choroidal boundaries compared to state-of-the-art methods.https://ieeexplore.ieee.org/document/10949143/Choroidmulti-scale featuresOCT imagereceptive fieldU-Net
spellingShingle Dejie Chen
Xiangping Chen
Hao Gu
Su Zhao
Hao Jiang
MSU-Net: A Synthesized U-Net for Exploiting Multi-Scale Features in OCT Image Segmentation
IEEE Access
Choroid
multi-scale features
OCT image
receptive field
U-Net
title MSU-Net: A Synthesized U-Net for Exploiting Multi-Scale Features in OCT Image Segmentation
title_full MSU-Net: A Synthesized U-Net for Exploiting Multi-Scale Features in OCT Image Segmentation
title_fullStr MSU-Net: A Synthesized U-Net for Exploiting Multi-Scale Features in OCT Image Segmentation
title_full_unstemmed MSU-Net: A Synthesized U-Net for Exploiting Multi-Scale Features in OCT Image Segmentation
title_short MSU-Net: A Synthesized U-Net for Exploiting Multi-Scale Features in OCT Image Segmentation
title_sort msu net a synthesized u net for exploiting multi scale features in oct image segmentation
topic Choroid
multi-scale features
OCT image
receptive field
U-Net
url https://ieeexplore.ieee.org/document/10949143/
work_keys_str_mv AT dejiechen msunetasynthesizedunetforexploitingmultiscalefeaturesinoctimagesegmentation
AT xiangpingchen msunetasynthesizedunetforexploitingmultiscalefeaturesinoctimagesegmentation
AT haogu msunetasynthesizedunetforexploitingmultiscalefeaturesinoctimagesegmentation
AT suzhao msunetasynthesizedunetforexploitingmultiscalefeaturesinoctimagesegmentation
AT haojiang msunetasynthesizedunetforexploitingmultiscalefeaturesinoctimagesegmentation