Multi-Scale Regularized Deep Network for Retinal Vessel Segmentation

Venkateswararao Cherukuri, B. G. Vijay Kumar, Raja Bala, Vishal Monga

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Vessel segmentation of retinal images is a key diagnostic capability in ophthalmology. Early approaches addressing this problem employed hand-crafted filters to capture vessel structures, accompanied by morphological processing. More recently, deep learning techniques have been employed to significantly enhance segmentation accuracy. We propose a novel domain enriched deep network that consists of two components: 1) a representation network which learns geometric (specifically curvilinear) features that are tailored to retinal images, followed by 2) a task network that utilizes the features obtained from the representation layer to perform pixel-level segmentation. The representation and task networks are learned jointly for any given training set. To obtain effective representation filters, we develop a new orientation constraint that enables geometric diversity of curvilinear features. A multi-scale extension is further developed to enhance segmentation of thin vessels. Experiments performed on two challenging benchmark databases reveal that the proposed regularized deep network can outperform state of the art alternatives as measured by common evaluation metrics. Further, the proposed method exhibits a more graceful decay in performance as training data is reduced.

Original languageEnglish (US)
Title of host publication2019 IEEE International Conference on Image Processing, ICIP 2019 - Proceedings
PublisherIEEE Computer Society
Pages824-828
Number of pages5
ISBN (Electronic)9781538662496
DOIs
StatePublished - Sep 2019
Event26th IEEE International Conference on Image Processing, ICIP 2019 - Taipei, Taiwan, Province of China
Duration: Sep 22 2019Sep 25 2019

Publication series

NameProceedings - International Conference on Image Processing, ICIP
Volume2019-September
ISSN (Print)1522-4880

Conference

Conference26th IEEE International Conference on Image Processing, ICIP 2019
CountryTaiwan, Province of China
CityTaipei
Period9/22/199/25/19

Fingerprint

Ophthalmology
Pixels
Processing
Experiments
Deep learning

All Science Journal Classification (ASJC) codes

  • Software
  • Computer Vision and Pattern Recognition
  • Signal Processing

Cite this

Cherukuri, V., Vijay Kumar, B. G., Bala, R., & Monga, V. (2019). Multi-Scale Regularized Deep Network for Retinal Vessel Segmentation. In 2019 IEEE International Conference on Image Processing, ICIP 2019 - Proceedings (pp. 824-828). [8803762] (Proceedings - International Conference on Image Processing, ICIP; Vol. 2019-September). IEEE Computer Society. https://doi.org/10.1109/ICIP.2019.8803762
Cherukuri, Venkateswararao ; Vijay Kumar, B. G. ; Bala, Raja ; Monga, Vishal. / Multi-Scale Regularized Deep Network for Retinal Vessel Segmentation. 2019 IEEE International Conference on Image Processing, ICIP 2019 - Proceedings. IEEE Computer Society, 2019. pp. 824-828 (Proceedings - International Conference on Image Processing, ICIP).
@inproceedings{380372319bd9431c8ec8d7fd710d2bd3,
title = "Multi-Scale Regularized Deep Network for Retinal Vessel Segmentation",
abstract = "Vessel segmentation of retinal images is a key diagnostic capability in ophthalmology. Early approaches addressing this problem employed hand-crafted filters to capture vessel structures, accompanied by morphological processing. More recently, deep learning techniques have been employed to significantly enhance segmentation accuracy. We propose a novel domain enriched deep network that consists of two components: 1) a representation network which learns geometric (specifically curvilinear) features that are tailored to retinal images, followed by 2) a task network that utilizes the features obtained from the representation layer to perform pixel-level segmentation. The representation and task networks are learned jointly for any given training set. To obtain effective representation filters, we develop a new orientation constraint that enables geometric diversity of curvilinear features. A multi-scale extension is further developed to enhance segmentation of thin vessels. Experiments performed on two challenging benchmark databases reveal that the proposed regularized deep network can outperform state of the art alternatives as measured by common evaluation metrics. Further, the proposed method exhibits a more graceful decay in performance as training data is reduced.",
author = "Venkateswararao Cherukuri and {Vijay Kumar}, {B. G.} and Raja Bala and Vishal Monga",
year = "2019",
month = "9",
doi = "10.1109/ICIP.2019.8803762",
language = "English (US)",
series = "Proceedings - International Conference on Image Processing, ICIP",
publisher = "IEEE Computer Society",
pages = "824--828",
booktitle = "2019 IEEE International Conference on Image Processing, ICIP 2019 - Proceedings",
address = "United States",

}

Cherukuri, V, Vijay Kumar, BG, Bala, R & Monga, V 2019, Multi-Scale Regularized Deep Network for Retinal Vessel Segmentation. in 2019 IEEE International Conference on Image Processing, ICIP 2019 - Proceedings., 8803762, Proceedings - International Conference on Image Processing, ICIP, vol. 2019-September, IEEE Computer Society, pp. 824-828, 26th IEEE International Conference on Image Processing, ICIP 2019, Taipei, Taiwan, Province of China, 9/22/19. https://doi.org/10.1109/ICIP.2019.8803762

Multi-Scale Regularized Deep Network for Retinal Vessel Segmentation. / Cherukuri, Venkateswararao; Vijay Kumar, B. G.; Bala, Raja; Monga, Vishal.

2019 IEEE International Conference on Image Processing, ICIP 2019 - Proceedings. IEEE Computer Society, 2019. p. 824-828 8803762 (Proceedings - International Conference on Image Processing, ICIP; Vol. 2019-September).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Multi-Scale Regularized Deep Network for Retinal Vessel Segmentation

AU - Cherukuri, Venkateswararao

AU - Vijay Kumar, B. G.

AU - Bala, Raja

AU - Monga, Vishal

PY - 2019/9

Y1 - 2019/9

N2 - Vessel segmentation of retinal images is a key diagnostic capability in ophthalmology. Early approaches addressing this problem employed hand-crafted filters to capture vessel structures, accompanied by morphological processing. More recently, deep learning techniques have been employed to significantly enhance segmentation accuracy. We propose a novel domain enriched deep network that consists of two components: 1) a representation network which learns geometric (specifically curvilinear) features that are tailored to retinal images, followed by 2) a task network that utilizes the features obtained from the representation layer to perform pixel-level segmentation. The representation and task networks are learned jointly for any given training set. To obtain effective representation filters, we develop a new orientation constraint that enables geometric diversity of curvilinear features. A multi-scale extension is further developed to enhance segmentation of thin vessels. Experiments performed on two challenging benchmark databases reveal that the proposed regularized deep network can outperform state of the art alternatives as measured by common evaluation metrics. Further, the proposed method exhibits a more graceful decay in performance as training data is reduced.

AB - Vessel segmentation of retinal images is a key diagnostic capability in ophthalmology. Early approaches addressing this problem employed hand-crafted filters to capture vessel structures, accompanied by morphological processing. More recently, deep learning techniques have been employed to significantly enhance segmentation accuracy. We propose a novel domain enriched deep network that consists of two components: 1) a representation network which learns geometric (specifically curvilinear) features that are tailored to retinal images, followed by 2) a task network that utilizes the features obtained from the representation layer to perform pixel-level segmentation. The representation and task networks are learned jointly for any given training set. To obtain effective representation filters, we develop a new orientation constraint that enables geometric diversity of curvilinear features. A multi-scale extension is further developed to enhance segmentation of thin vessels. Experiments performed on two challenging benchmark databases reveal that the proposed regularized deep network can outperform state of the art alternatives as measured by common evaluation metrics. Further, the proposed method exhibits a more graceful decay in performance as training data is reduced.

UR - http://www.scopus.com/inward/record.url?scp=85076819118&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85076819118&partnerID=8YFLogxK

U2 - 10.1109/ICIP.2019.8803762

DO - 10.1109/ICIP.2019.8803762

M3 - Conference contribution

AN - SCOPUS:85076819118

T3 - Proceedings - International Conference on Image Processing, ICIP

SP - 824

EP - 828

BT - 2019 IEEE International Conference on Image Processing, ICIP 2019 - Proceedings

PB - IEEE Computer Society

ER -

Cherukuri V, Vijay Kumar BG, Bala R, Monga V. Multi-Scale Regularized Deep Network for Retinal Vessel Segmentation. In 2019 IEEE International Conference on Image Processing, ICIP 2019 - Proceedings. IEEE Computer Society. 2019. p. 824-828. 8803762. (Proceedings - International Conference on Image Processing, ICIP). https://doi.org/10.1109/ICIP.2019.8803762