Encoder-Decoder Networks for Retinal Vessel Segmentation Using Large Multi-scale Patches

Björn Browatzki; Jörn Philipp Lies; Christian Wallraven

doi:10.1007/978-3-030-63419-3_5

Encoder-Decoder Networks for Retinal Vessel Segmentation Using Large Multi-scale Patches

Björn Browatzki, Jörn Philipp Lies, Christian Wallraven

Department of Brain and Cognitive Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

We propose an encoder-decoder framework for the segmentation of blood vessels in retinal images that relies on the extraction of large-scale patches at multiple image-scales during training. Experiments on three fundus image datasets demonstrate that this approach achieves state-of-the-art results and can be implemented using a simple and efficient fully-convolutional network with a parameter count of less than 0.8M. Furthermore, we show that this framework - called VLight - avoids overfitting to specific training images and generalizes well across different datasets, which makes it highly suitable for real-world applications where robustness, accuracy as well as low inference time on high-resolution fundus images is required.

Original language	English
Title of host publication	Ophthalmic Medical Image Analysis - 7th International Workshop, OMIA 2020, Held in Conjunction with MICCAI 2020, Proceedings
Editors	Huazhu Fu, Mona K. Garvin, Tom MacGillivray, Yanwu Xu, Yalin Zheng
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	42-52
Number of pages	11
ISBN (Print)	9783030634186
DOIs	https://doi.org/10.1007/978-3-030-63419-3_5
Publication status	Published - 2020
Event	6th International Workshop on Ophthalmic Medical Image Analysis, OMIA 2020, held in conjunction with 23rd International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2020 - Lima, Peru Duration: 2020 Oct 8 → 2020 Oct 8

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	12069 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	6th International Workshop on Ophthalmic Medical Image Analysis, OMIA 2020, held in conjunction with 23rd International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2020
Country/Territory	Peru
City	Lima
Period	20/10/8 → 20/10/8

Bibliographical note

Funding Information:
This work was supported by Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korean government (MSIT) (No. 2019-0-00079, Dept. of Artificial Intelligence, Korea Univer-sity).

Publisher Copyright:
© 2020, Springer Nature Switzerland AG.

Keywords

Fundus image
Residual networks
Retinal vessel detection
Semantic segmentation

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-030-63419-3_5

Cite this

Browatzki, B., Lies, J. P., & Wallraven, C. (2020). Encoder-Decoder Networks for Retinal Vessel Segmentation Using Large Multi-scale Patches. In H. Fu, M. K. Garvin, T. MacGillivray, Y. Xu, & Y. Zheng (Eds.), Ophthalmic Medical Image Analysis - 7th International Workshop, OMIA 2020, Held in Conjunction with MICCAI 2020, Proceedings (pp. 42-52). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12069 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-63419-3_5

Encoder-Decoder Networks for Retinal Vessel Segmentation Using Large Multi-scale Patches. / Browatzki, Björn; Lies, Jörn Philipp; Wallraven, Christian.
Ophthalmic Medical Image Analysis - 7th International Workshop, OMIA 2020, Held in Conjunction with MICCAI 2020, Proceedings. ed. / Huazhu Fu; Mona K. Garvin; Tom MacGillivray; Yanwu Xu; Yalin Zheng. Springer Science and Business Media Deutschland GmbH, 2020. p. 42-52 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12069 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Browatzki, B, Lies, JP & Wallraven, C 2020, Encoder-Decoder Networks for Retinal Vessel Segmentation Using Large Multi-scale Patches. in H Fu, MK Garvin, T MacGillivray, Y Xu & Y Zheng (eds), Ophthalmic Medical Image Analysis - 7th International Workshop, OMIA 2020, Held in Conjunction with MICCAI 2020, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12069 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 42-52, 6th International Workshop on Ophthalmic Medical Image Analysis, OMIA 2020, held in conjunction with 23rd International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2020, Lima, Peru, 20/10/8. https://doi.org/10.1007/978-3-030-63419-3_5

Browatzki B, Lies JP, Wallraven C. Encoder-Decoder Networks for Retinal Vessel Segmentation Using Large Multi-scale Patches. In Fu H, Garvin MK, MacGillivray T, Xu Y, Zheng Y, editors, Ophthalmic Medical Image Analysis - 7th International Workshop, OMIA 2020, Held in Conjunction with MICCAI 2020, Proceedings. Springer Science and Business Media Deutschland GmbH. 2020. p. 42-52. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-63419-3_5

Browatzki, Björn ; Lies, Jörn Philipp ; Wallraven, Christian. / Encoder-Decoder Networks for Retinal Vessel Segmentation Using Large Multi-scale Patches. Ophthalmic Medical Image Analysis - 7th International Workshop, OMIA 2020, Held in Conjunction with MICCAI 2020, Proceedings. editor / Huazhu Fu ; Mona K. Garvin ; Tom MacGillivray ; Yanwu Xu ; Yalin Zheng. Springer Science and Business Media Deutschland GmbH, 2020. pp. 42-52 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{0dd49468faf244f6bfbbcf8dfe963a56,

title = "Encoder-Decoder Networks for Retinal Vessel Segmentation Using Large Multi-scale Patches",

abstract = "We propose an encoder-decoder framework for the segmentation of blood vessels in retinal images that relies on the extraction of large-scale patches at multiple image-scales during training. Experiments on three fundus image datasets demonstrate that this approach achieves state-of-the-art results and can be implemented using a simple and efficient fully-convolutional network with a parameter count of less than 0.8M. Furthermore, we show that this framework - called VLight - avoids overfitting to specific training images and generalizes well across different datasets, which makes it highly suitable for real-world applications where robustness, accuracy as well as low inference time on high-resolution fundus images is required.",

keywords = "Fundus image, Residual networks, Retinal vessel detection, Semantic segmentation",

author = "Bj{\"o}rn Browatzki and Lies, {J{\"o}rn Philipp} and Christian Wallraven",

note = "Funding Information: This work was supported by Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korean government (MSIT) (No. 2019-0-00079, Dept. of Artificial Intelligence, Korea Univer-sity). Publisher Copyright: {\textcopyright} 2020, Springer Nature Switzerland AG.; 6th International Workshop on Ophthalmic Medical Image Analysis, OMIA 2020, held in conjunction with 23rd International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2020 ; Conference date: 08-10-2020 Through 08-10-2020",

year = "2020",

doi = "10.1007/978-3-030-63419-3_5",

language = "English",

isbn = "9783030634186",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "42--52",

editor = "Huazhu Fu and Garvin, {Mona K.} and Tom MacGillivray and Yanwu Xu and Yalin Zheng",

booktitle = "Ophthalmic Medical Image Analysis - 7th International Workshop, OMIA 2020, Held in Conjunction with MICCAI 2020, Proceedings",

address = "Germany",

}

TY - GEN

T1 - Encoder-Decoder Networks for Retinal Vessel Segmentation Using Large Multi-scale Patches

AU - Browatzki, Björn

AU - Lies, Jörn Philipp

AU - Wallraven, Christian

N1 - Funding Information: This work was supported by Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korean government (MSIT) (No. 2019-0-00079, Dept. of Artificial Intelligence, Korea Univer-sity). Publisher Copyright: © 2020, Springer Nature Switzerland AG.

PY - 2020

Y1 - 2020

N2 - We propose an encoder-decoder framework for the segmentation of blood vessels in retinal images that relies on the extraction of large-scale patches at multiple image-scales during training. Experiments on three fundus image datasets demonstrate that this approach achieves state-of-the-art results and can be implemented using a simple and efficient fully-convolutional network with a parameter count of less than 0.8M. Furthermore, we show that this framework - called VLight - avoids overfitting to specific training images and generalizes well across different datasets, which makes it highly suitable for real-world applications where robustness, accuracy as well as low inference time on high-resolution fundus images is required.

AB - We propose an encoder-decoder framework for the segmentation of blood vessels in retinal images that relies on the extraction of large-scale patches at multiple image-scales during training. Experiments on three fundus image datasets demonstrate that this approach achieves state-of-the-art results and can be implemented using a simple and efficient fully-convolutional network with a parameter count of less than 0.8M. Furthermore, we show that this framework - called VLight - avoids overfitting to specific training images and generalizes well across different datasets, which makes it highly suitable for real-world applications where robustness, accuracy as well as low inference time on high-resolution fundus images is required.

KW - Fundus image

KW - Residual networks

KW - Retinal vessel detection

KW - Semantic segmentation

UR - http://www.scopus.com/inward/record.url?scp=85097397058&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-63419-3_5

DO - 10.1007/978-3-030-63419-3_5

M3 - Conference contribution

AN - SCOPUS:85097397058

SN - 9783030634186

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 42

EP - 52

BT - Ophthalmic Medical Image Analysis - 7th International Workshop, OMIA 2020, Held in Conjunction with MICCAI 2020, Proceedings

A2 - Fu, Huazhu

A2 - Garvin, Mona K.

A2 - MacGillivray, Tom

A2 - Xu, Yanwu

A2 - Zheng, Yalin

PB - Springer Science and Business Media Deutschland GmbH

T2 - 6th International Workshop on Ophthalmic Medical Image Analysis, OMIA 2020, held in conjunction with 23rd International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2020

Y2 - 8 October 2020 through 8 October 2020

ER -

Encoder-Decoder Networks for Retinal Vessel Segmentation Using Large Multi-scale Patches

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this