Boundary aware microscopic hyperspectral pathology image segmentation network guided by information entropy weight

IntroductionAccurate segmentation of lesion tissues in medical microscopic hyperspectral pathological images is crucial for enhancing early tumor diagnosis and improving patient prognosis. However, the complex structure and indistinct boundaries of lesion tissues present significant challenges in ac...

Full description

Saved in:
Bibliographic Details
Main Authors: Xueying Cao, Hongmin Gao, Ting Qin, Min Zhu, Ping Zhang, Peipei Xu
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-03-01
Series:Frontiers in Oncology
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fonc.2025.1549544/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850063215174615040
author Xueying Cao
Hongmin Gao
Ting Qin
Min Zhu
Ping Zhang
Ping Zhang
Peipei Xu
Peipei Xu
author_facet Xueying Cao
Hongmin Gao
Ting Qin
Min Zhu
Ping Zhang
Ping Zhang
Peipei Xu
Peipei Xu
author_sort Xueying Cao
collection DOAJ
description IntroductionAccurate segmentation of lesion tissues in medical microscopic hyperspectral pathological images is crucial for enhancing early tumor diagnosis and improving patient prognosis. However, the complex structure and indistinct boundaries of lesion tissues present significant challenges in achieving precise segmentation.MethodsTo address these challenges, we propose a novel method named BE-Net. It employs multi-scale strategy and edge operators to capture fine edge details, while incorporating information entropy to construct attention mechanisms that further strengthen the representation of relevant features. Specifically, we first propose a Laplacian of Gaussian operator convolution boundary feature extraction block, which encodes feature gradient information through the improved edge detection operators and emphasizes relevant boundary channel weights based on channel information entropy weighting. We further designed a grouped multi-scale edge feature extraction module to optimize the fusion process between the encoder and decoder, with the goal of optimize boundary details and emphasizing relevant channel representations. Finally, we propose a multi-scale spatial boundary feature extraction block to guide the model in emphasizing the most important spatial locations and boundary regions.ResultWe evaluate BE-Net on medical microscopic hyperspectral pathological image datasets of gastric intraepithelial neoplasia and gastric mucosal intestinal metaplasia. Experimental results demonstrate that BE-Net outperforms other state-of-the-art segmentation methods in terms of accuracy and boundary preservation.DiscussionThis advance has significant implications for the field of MHSIs segmentation. Our code is freely available at https://github.com/sharycao/BE-NET.
format Article
id doaj-art-0f958adc457c4982a8460c41e1cd4ef6
institution DOAJ
issn 2234-943X
language English
publishDate 2025-03-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Oncology
spelling doaj-art-0f958adc457c4982a8460c41e1cd4ef62025-08-20T02:49:43ZengFrontiers Media S.A.Frontiers in Oncology2234-943X2025-03-011510.3389/fonc.2025.15495441549544Boundary aware microscopic hyperspectral pathology image segmentation network guided by information entropy weightXueying Cao0Hongmin Gao1Ting Qin2Min Zhu3Ping Zhang4Ping Zhang5Peipei Xu6Peipei Xu7College of Computer Science and Software Engineering, Hohai University, Nanjing, ChinaCollege of Computer Science and Software Engineering, Hohai University, Nanjing, ChinaDepartment of Hematology, Nanjing Drum Tower Hospital Clinical College of Nanjing Medical University, Nanjing, ChinaCollege of Computer Science and Software Engineering, Hohai University, Nanjing, ChinaSchool of Information Management & Engineering, Shanghai University of Finance and Economics, Shanghai, ChinaCollege of International Exchange, Nanjing Normal University of Special Education, Nanjing, ChinaDepartment of Hematology, Nanjing Drum Tower Hospital Clinical College of Nanjing Medical University, Nanjing, ChinaDepartment of Hematology, Nanjing Drum Tower Hospital, Affiliated Hospital of Medical School, Nanjing University, Nanjing, ChinaIntroductionAccurate segmentation of lesion tissues in medical microscopic hyperspectral pathological images is crucial for enhancing early tumor diagnosis and improving patient prognosis. However, the complex structure and indistinct boundaries of lesion tissues present significant challenges in achieving precise segmentation.MethodsTo address these challenges, we propose a novel method named BE-Net. It employs multi-scale strategy and edge operators to capture fine edge details, while incorporating information entropy to construct attention mechanisms that further strengthen the representation of relevant features. Specifically, we first propose a Laplacian of Gaussian operator convolution boundary feature extraction block, which encodes feature gradient information through the improved edge detection operators and emphasizes relevant boundary channel weights based on channel information entropy weighting. We further designed a grouped multi-scale edge feature extraction module to optimize the fusion process between the encoder and decoder, with the goal of optimize boundary details and emphasizing relevant channel representations. Finally, we propose a multi-scale spatial boundary feature extraction block to guide the model in emphasizing the most important spatial locations and boundary regions.ResultWe evaluate BE-Net on medical microscopic hyperspectral pathological image datasets of gastric intraepithelial neoplasia and gastric mucosal intestinal metaplasia. Experimental results demonstrate that BE-Net outperforms other state-of-the-art segmentation methods in terms of accuracy and boundary preservation.DiscussionThis advance has significant implications for the field of MHSIs segmentation. Our code is freely available at https://github.com/sharycao/BE-NET.https://www.frontiersin.org/articles/10.3389/fonc.2025.1549544/fullmicroscopic hyperspectral imageboundary-awareinformation entropyattention mechanismmulti-scale
spellingShingle Xueying Cao
Hongmin Gao
Ting Qin
Min Zhu
Ping Zhang
Ping Zhang
Peipei Xu
Peipei Xu
Boundary aware microscopic hyperspectral pathology image segmentation network guided by information entropy weight
Frontiers in Oncology
microscopic hyperspectral image
boundary-aware
information entropy
attention mechanism
multi-scale
title Boundary aware microscopic hyperspectral pathology image segmentation network guided by information entropy weight
title_full Boundary aware microscopic hyperspectral pathology image segmentation network guided by information entropy weight
title_fullStr Boundary aware microscopic hyperspectral pathology image segmentation network guided by information entropy weight
title_full_unstemmed Boundary aware microscopic hyperspectral pathology image segmentation network guided by information entropy weight
title_short Boundary aware microscopic hyperspectral pathology image segmentation network guided by information entropy weight
title_sort boundary aware microscopic hyperspectral pathology image segmentation network guided by information entropy weight
topic microscopic hyperspectral image
boundary-aware
information entropy
attention mechanism
multi-scale
url https://www.frontiersin.org/articles/10.3389/fonc.2025.1549544/full
work_keys_str_mv AT xueyingcao boundaryawaremicroscopichyperspectralpathologyimagesegmentationnetworkguidedbyinformationentropyweight
AT hongmingao boundaryawaremicroscopichyperspectralpathologyimagesegmentationnetworkguidedbyinformationentropyweight
AT tingqin boundaryawaremicroscopichyperspectralpathologyimagesegmentationnetworkguidedbyinformationentropyweight
AT minzhu boundaryawaremicroscopichyperspectralpathologyimagesegmentationnetworkguidedbyinformationentropyweight
AT pingzhang boundaryawaremicroscopichyperspectralpathologyimagesegmentationnetworkguidedbyinformationentropyweight
AT pingzhang boundaryawaremicroscopichyperspectralpathologyimagesegmentationnetworkguidedbyinformationentropyweight
AT peipeixu boundaryawaremicroscopichyperspectralpathologyimagesegmentationnetworkguidedbyinformationentropyweight
AT peipeixu boundaryawaremicroscopichyperspectralpathologyimagesegmentationnetworkguidedbyinformationentropyweight