BranchFusionNet: An energy-efficient lightweight framework for superior retinal vessel segmentation

被引：0

作者：

Qin, Jing ^{[1
]}

Qin, Zhiguang ^{[1
]}

Xiao, Peng ^{[1
]}

机构：

[1] Univ Elect Sci & Technol China, Network & Data Secur Key Lab Sichuan Prov, Jianshe North Rd, Chengdu 610054, Peoples R China

来源：

PEER-TO-PEER NETWORKING AND APPLICATIONS | 2024年 / 17卷 / 05期

基金：

中国国家自然科学基金;

关键词：

Retinal vessel segmentation; Energy-efficient deep learning; BranchFusionNet; Environmental sustainability; CONNECTIONS; NETWORK; IMAGES;

D O I：

10.1007/s12083-024-01738-3

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

In the rapidly advancing field of medical image analysis, accurate and efficient segmentation of retinal vessels is paramount for diagnosing ocular diseases, especially diabetic retinopathy. With the increasing emphasis on environmental sustainability, this paper presents BranchFusionNet, a novel lightweight neural network architecture tailored for retinal vessel segmentation. Embodying the principles of energy conservation, BranchFusionNet integrates multi-branch and lightweight dual-branch modules to optimize computational demands without sacrificing segmentation precision. This study not only contributes to the domain of retinal vessel segmentation but also showcases the potential of crafting energy-conscious deep learning methodologies in medical imaging applications.

引用

页码：3133 / 3145

页数：13

共 47 条

[1]

Adarsh R, 2020, INT CONF ADVAN COMPU, P280, DOI [10.1109/ICACCS48705.2020.9074172, 10.1109/icaccs48705.2020.9074172]

[2]

Alom MZ, 2018, ARXIV

[3] Prompt Deep Light-Weight Vessel Segmentation Network (PLVS-Net) [J].