ODFormer: Semantic fundus image segmentation using Transformer for optic nerve head detection

被引:0
作者
Wang, Jiayi [1 ,2 ,3 ,4 ]
Mao, Yi-An [5 ]
Ma, Xiaoyu [6 ]
Guo, Sicen [1 ,2 ,3 ,4 ]
Shao, Yuting [6 ]
Lv, Xiao [6 ]
Han, Wenting [6 ]
Christopher, Mark [7 ,8 ]
Zangwill, Linda M. [7 ,8 ]
Bi, Yanlong [6 ,9 ]
Fan, Rui [1 ,2 ,3 ,4 ]
机构
[1] Tongji Univ, Coll Elect & Informat Engn, Dept Control Sci & Engn, MIAS Grp, Shanghai 201804, Peoples R China
[2] Tongji Univ, Shanghai Inst Intelligent Sci & Technol, Shanghai 201804, Peoples R China
[3] Tongji Univ, Shanghai Res Inst Intelligent Autonomous Syst, State Key Lab Intelligent Autonomous Syst, Shanghai 201210, Peoples R China
[4] Tongji Univ, Frontiers Sci Ctr Intelligent Autonomous Syst, Shanghai 201210, Peoples R China
[5] Shanghai Univ, Sch Life Sci, 99 Shangda Ave, Shanghai 200444, Peoples R China
[6] Tongji Univ, Tongji Hosp, Sch Med, Dept Ophthalmol, Shanghai 200065, Peoples R China
[7] Univ Calif San Diego, Hamilton Glaucoma Ctr, La Jolla, CA 92037 USA
[8] Univ Calif San Diego, Shiley Eye Inst, Viterbi Family Dept Ophthalmol, Div Ophthalmol Informat & Data Sci, La Jolla, CA 92037 USA
[9] Tongji Univ, Tongji Eye Inst, Sch Med, Shanghai 200065, Peoples R China
基金
中国国家自然科学基金;
关键词
Optic nerve head; Fundus image; Semantic segmentation; Convolutional neural network; Transformer; FOVEA;
D O I
10.1016/j.inffus.2024.102533
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Optic nerve head (ONH) detection has been a crucial area of study in ophthalmology for years. However, the significant discrepancy between fundus image datasets, each generated using a single type of fundus camera, poses challenges to the generalizability of ONH detection approaches developed based on semantic segmentation networks. Despite the numerous recent advancements in general-purpose semantic segmentation methods using convolutional neural networks (CNNs) and Transformers, there is currently a lack of benchmarks for these state -of -the -art (SoTA) networks specifically trained for ONH detection. Therefore, in this article, we make contributions from three key aspects: network design, the publication of a dataset, and the establishment of a comprehensive benchmark. Our newly developed ONH detection network, referred to as ODFormer, is based upon the Swin Transformer architecture and incorporates two novel components: a multi-scale context aggregator and a lightweight bidirectional feature recalibrator. Our published large-scale dataset, known as TongjiU-DROD, provides multi -resolution fundus images for each participant, captured using two distinct types of cameras. Our established benchmark involves three datasets: DRIONS-DB, DRISHTI-GS1, and TongjiU-DROD, created by researchers from different countries and containing fundus images captured from participants of diverse races and ages. Extensive experimental results demonstrate that our proposed ODFormer outperforms other state -of -the -art (SoTA) networks in terms of performance and generalizability. Our dataset and source code are publicly available at https://mias.group/ODFormer.
引用
收藏
页数:10
相关论文
共 50 条
  • [41] Memory-Augmented Transformer for Remote Sensing Image Semantic Segmentation
    Zhao, Xin
    Guo, Jiayi
    Zhang, Yueting
    Wu, Yirong
    [J]. REMOTE SENSING, 2021, 13 (22)
  • [42] Detection of artificial spots in fundus images using modified U-Net based semantic segmentation
    Parashar, Anuj Kumar
    Kumar, Bambam
    [J]. COMPUTERS & ELECTRICAL ENGINEERING, 2024, 120
  • [43] Comparison of optic area measurement using fundus photography and Optical Coherence Tomography between optic nerve head drusen and control subjects
    Flores-Rodriguez, Patricia
    Gili, Pablo
    Dolores Martin-Rios, Maria
    Grifol-Clar, Eulalia
    [J]. OPHTHALMIC AND PHYSIOLOGICAL OPTICS, 2013, 33 (02) : 164 - 171
  • [44] Recognition of Glaucomatous Fundus Images Using Machine Learning Methods Based on Optic Nerve Head Topographic Features
    Wu, Chao-Wei
    Huang, Tzu-Yu
    Liou, Yeong-Cheng
    Chen, Shih-Hsin
    Wu, Kwou-Yeung
    Tseng, Han-Yi
    [J]. JOURNAL OF GLAUCOMA, 2024, 33 (08) : 601 - 606
  • [45] Automated Optic Nerve Head Detection Based on Different Retinal Vasculature Segmentation Methods and Mathematical Morphology
    Tavakoli, Meysam
    Nazar, Mahdieh
    Golestaneh, Alireza
    Kalantari, Faraz
    [J]. 2017 IEEE NUCLEAR SCIENCE SYMPOSIUM AND MEDICAL IMAGING CONFERENCE (NSS/MIC), 2017,
  • [46] Detection of Optic Disc Localization from Retinal Fundus Image Using Optimized Color Space
    Toptas, Buket
    Toptas, Murat
    Hanbay, Davut
    [J]. JOURNAL OF DIGITAL IMAGING, 2022, 35 (02) : 302 - 319
  • [47] Detection of Optic Disc Localization from Retinal Fundus Image Using Optimized Color Space
    Buket Toptaş
    Murat Toptaş
    Davut Hanbay
    [J]. Journal of Digital Imaging, 2022, 35 : 302 - 319
  • [48] Enhanced Pedestrian Detection using Deep Learning based Semantic Image Segmentation
    Liu, Tianrui
    Stathaki, Tania
    [J]. 2017 22ND INTERNATIONAL CONFERENCE ON DIGITAL SIGNAL PROCESSING (DSP), 2017,
  • [49] STransFuse: Fusing Swin Transformer and Convolutional Neural Network for Remote Sensing Image Semantic Segmentation
    Gao, Liang
    Liu, Hui
    Yang, Minhang
    Chen, Long
    Wan, Yaling
    Xiao, Zhengqing
    Qian, Yurong
    [J]. IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2021, 14 (14) : 10990 - 11003
  • [50] Semantic segmentation using deep learning to extract total extraocular muscles and optic nerve from orbital computed tomography images
    Zhu, Fubao
    Gao, Zhengyuan
    Zhao, Chen
    Zhu, Zelin
    Tang, Jinshan
    Liu, Yanyun
    Tang, Shaojie
    Jiang, Chengzhi
    Li, Xinhui
    Zhao, Min
    Zhou, Weihua
    [J]. OPTIK, 2021, 244