Retinal artery/vein classification via graph cut optimization

K.A.J. Eppenhof; E.J. Bekkers; T.T.J.M. Berendschot; J.P.W. Pluim; B.M. ter Haar Romenij

Retinal artery/vein classification via graph cut optimization

K.A.J. Eppenhof, E.J. Bekkers, T.T.J.M. Berendschot, J.P.W. Pluim, B.M. ter Haar Romenij

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

127 Downloads (Pure)

Abstract

In many diseases with a cardiovascular component, the geometry of microvascular blood vessels changes. These changes are specific to arteries and veins, and can be studied in the microvasculature of the retina using retinal photography. To facilitate large-scale studies of artery/vein-specific changes in the retinal vasculature, automated classification of the vessels is required. Here we present a novel method for artery/vein classification based on local and contextual feature analysis of retinal vessels. For each vessel, local information in the form of a transverse intensity profile is extracted. Crossings and bifurcations of vessels provide contextual information. The local and contextual features are integrated into a non-submodular energy function, which is optimized exactly using graph cuts. The method was validated on a ground truth data set of 150 retinal fundus images, achieving an accuracy of 88.0% for all vessels and 94.0% for the six arteries and six veins with highest caliber in the image.

Original language	English
Title of host publication	Proceedings of the Ophthalmic Medical Image Analysis Second International Workshop, OMIA 2015
Editors	X. Chen, M.K. Garvin, J. Liu, E. Trucco, Y. Xu
Pages	121-128
Publication status	Published - 2015
Event	2015 Ophthalmic Medical Image Analysis Second International Workshop (OMIA 2015) - Munich, Germany Duration: 9 Oct 2015 → 9 Oct 2015 Conference number: 2 http://ir.uiowa.edu/omia/2015_Proceedings/2015/16/

Conference

Conference	2015 Ophthalmic Medical Image Analysis Second International Workshop (OMIA 2015)
Abbreviated title	OMIA 2015
Country/Territory	Germany
City	Munich
Period	9/10/15 → 9/10/15
Internet address	http://ir.uiowa.edu/omia/2015_Proceedings/2015/16/

Access to Document

EppenhofRetinalartery2015Final published version, 1.13 MB

http://ir.uiowa.edu/omia/2015_Proceedings/2015/16

Cite this

Eppenhof, K. A. J., Bekkers, E. J., Berendschot, T. T. J. M., Pluim, J. P. W., & ter Haar Romenij, B. M. (2015). Retinal artery/vein classification via graph cut optimization. In X. Chen, M. K. Garvin, J. Liu, E. Trucco, & Y. Xu (Eds.), Proceedings of the Ophthalmic Medical Image Analysis Second International Workshop, OMIA 2015 (pp. 121-128) http://ir.uiowa.edu/omia/2015_Proceedings/2015/16

@inproceedings{54fc8567ca444046b4d86c78295d69e0,

title = "Retinal artery/vein classification via graph cut optimization",

abstract = "In many diseases with a cardiovascular component, the geometry of microvascular blood vessels changes. These changes are specific to arteries and veins, and can be studied in the microvasculature of the retina using retinal photography. To facilitate large-scale studies of artery/vein-specific changes in the retinal vasculature, automated classification of the vessels is required. Here we present a novel method for artery/vein classification based on local and contextual feature analysis of retinal vessels. For each vessel, local information in the form of a transverse intensity profile is extracted. Crossings and bifurcations of vessels provide contextual information. The local and contextual features are integrated into a non-submodular energy function, which is optimized exactly using graph cuts. The method was validated on a ground truth data set of 150 retinal fundus images, achieving an accuracy of 88.0% for all vessels and 94.0% for the six arteries and six veins with highest caliber in the image.",

author = "K.A.J. Eppenhof and E.J. Bekkers and T.T.J.M. Berendschot and J.P.W. Pluim and {ter Haar Romenij}, B.M.",

year = "2015",

language = "English",

pages = "121--128",

editor = "X. Chen and M.K. Garvin and J. Liu and E. Trucco and Y. Xu",

booktitle = "Proceedings of the Ophthalmic Medical Image Analysis Second International Workshop, OMIA 2015",

note = "2015 Ophthalmic Medical Image Analysis Second International Workshop (OMIA 2015), OMIA 2015 ; Conference date: 09-10-2015 Through 09-10-2015",

url = "http://ir.uiowa.edu/omia/2015_Proceedings/2015/16/",

}

Eppenhof, KAJ, Bekkers, EJ, Berendschot, TTJM , Pluim, JPW & ter Haar Romenij, BM 2015, Retinal artery/vein classification via graph cut optimization. in X Chen, MK Garvin, J Liu, E Trucco & Y Xu (eds), Proceedings of the Ophthalmic Medical Image Analysis Second International Workshop, OMIA 2015. pp. 121-128, 2015 Ophthalmic Medical Image Analysis Second International Workshop (OMIA 2015), Munich, Germany, 9/10/15. <http://ir.uiowa.edu/omia/2015_Proceedings/2015/16>

Retinal artery/vein classification via graph cut optimization. / Eppenhof, K.A.J.; Bekkers, E.J.; Berendschot, T.T.J.M. et al.

Proceedings of the Ophthalmic Medical Image Analysis Second International Workshop, OMIA 2015. ed. / X. Chen; M.K. Garvin; J. Liu; E. Trucco; Y. Xu. 2015. p. 121-128.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Retinal artery/vein classification via graph cut optimization

AU - Eppenhof, K.A.J.

AU - Bekkers, E.J.

AU - Berendschot, T.T.J.M.

AU - Pluim, J.P.W.

AU - ter Haar Romenij, B.M.

N1 - Conference code: 2

PY - 2015

Y1 - 2015

N2 - In many diseases with a cardiovascular component, the geometry of microvascular blood vessels changes. These changes are specific to arteries and veins, and can be studied in the microvasculature of the retina using retinal photography. To facilitate large-scale studies of artery/vein-specific changes in the retinal vasculature, automated classification of the vessels is required. Here we present a novel method for artery/vein classification based on local and contextual feature analysis of retinal vessels. For each vessel, local information in the form of a transverse intensity profile is extracted. Crossings and bifurcations of vessels provide contextual information. The local and contextual features are integrated into a non-submodular energy function, which is optimized exactly using graph cuts. The method was validated on a ground truth data set of 150 retinal fundus images, achieving an accuracy of 88.0% for all vessels and 94.0% for the six arteries and six veins with highest caliber in the image.

AB - In many diseases with a cardiovascular component, the geometry of microvascular blood vessels changes. These changes are specific to arteries and veins, and can be studied in the microvasculature of the retina using retinal photography. To facilitate large-scale studies of artery/vein-specific changes in the retinal vasculature, automated classification of the vessels is required. Here we present a novel method for artery/vein classification based on local and contextual feature analysis of retinal vessels. For each vessel, local information in the form of a transverse intensity profile is extracted. Crossings and bifurcations of vessels provide contextual information. The local and contextual features are integrated into a non-submodular energy function, which is optimized exactly using graph cuts. The method was validated on a ground truth data set of 150 retinal fundus images, achieving an accuracy of 88.0% for all vessels and 94.0% for the six arteries and six veins with highest caliber in the image.

M3 - Conference contribution

SP - 121

EP - 128

BT - Proceedings of the Ophthalmic Medical Image Analysis Second International Workshop, OMIA 2015

A2 - Chen, X.

A2 - Garvin, M.K.

A2 - Liu, J.

A2 - Trucco, E.

A2 - Xu, Y.

T2 - 2015 Ophthalmic Medical Image Analysis Second International Workshop (OMIA 2015)

Y2 - 9 October 2015 through 9 October 2015

ER -

Retinal artery/vein classification via graph cut optimization

Abstract

Conference

Access to Document

Fingerprint

Cite this