Publication detail

Registration of Bimodal Retinal Images improving modifications

KUBEČKA, L. JAN, J.

Original Title

Registration of Bimodal Retinal Images improving modifications

Czech Title

Registrace bimodálních obrazů sítnice - zlepšující změny

English Title

Registration of Bimodal Retinal Images improving modifications

Type

conference paper

Language

en

Original Abstract

The proper optical disc segmentation in images provided by confocal laser scanning ophthalmoscope and by color fundus-camera is a necessary step in early glaucoma or arteriosclerosis detection. Fusing information from both modalities into a vector-valued image is expected to improve the segmentation reliability. The paper describes a registration of these images using optimization based on mutual information criterion function extended with gradient-image mutual information. The controlled random search (CRS) has been found more robust optimization routine than the simulated annealing (SA) while tested on a set of 174 image pairs. Finally, the multi-resolution algorithm for bimodal retinal image registration achieving the success-rate of 94% is proposed.

Czech abstract

Správná segmentace optického disku v obrazech získaných z konfokálního laserového skenovacího oftalmoskopu a z barevné fotografie sítnice je nutný předpoklad pro brzkou detekci arteriosklezózy a glakomu. Očekáváme, že fúze obou modalit v jeden vektorový obraz podstatně zdvihne spolehlivost segmentace. Příspěvek popisuje registraci těchto multimodálních obrazů založenou na kritériu vzájemné informace rozšířeném o vzájemnou informaci gradientních obrazů. Jsou porovnávány robustní optimalizační algoritmy - řízené náhodné vyhledávání (CRS) a simulované žíhání (SA). Je ukázáno, že metoda CRS je robustnější při testování algoritmu na sérii 174 obrazových párů. Je navržen víceúrovňový algoritmus pro registraci multimodálních obrazů, dosahující spolehlivosti 94%.

English abstract

The proper optical disc segmentation in images provided by confocal laser scanning ophthalmoscope and by color fundus-camera is a necessary step in early glaucoma or arteriosclerosis detection. Fusing information from both modalities into a vector-valued image is expected to improve the segmentation reliability. The paper describes a registration of these images using optimization based on mutual information criterion function extended with gradient-image mutual information. The controlled random search (CRS) has been found more robust optimization routine than the simulated annealing (SA) while tested on a set of 174 image pairs. Finally, the multi-resolution algorithm for bimodal retinal image registration achieving the success-rate of 94% is proposed.

Keywords

image registration, ophthalmologic image processing, retina, mutual information, optimization, controlled random search, simulated annealing, simplex

RIV year

2004

Released

01.09.2004

Publisher

IEEE

Location

San Francisco

ISBN

0-7803-8440-7

Book

Proc. 26th Internat. Conf. IEEE-EMBS

Edition number

First

Pages from

1695

Pages to

1698

Pages count

4

BibTex


@inproceedings{BUT11607,
  author="Libor {Kubečka} and Jiří {Jan}",
  title="Registration of Bimodal Retinal Images improving modifications",
  annote="The proper optical disc segmentation in images provided by confocal laser scanning ophthalmoscope and by color fundus-camera is a necessary step in early glaucoma or arteriosclerosis detection. Fusing information from both modalities into a vector-valued image is expected to improve the segmentation reliability. The paper describes a registration of these images using optimization based on mutual information criterion function extended with gradient-image mutual information. The controlled random search (CRS) has been found more robust optimization routine than the simulated annealing (SA) while tested on a set of 174 image pairs. Finally, the multi-resolution algorithm for bimodal retinal image registration achieving the success-rate of 94% is proposed.",
  address="IEEE",
  booktitle="Proc. 26th Internat. Conf. IEEE-EMBS",
  chapter="11607",
  institution="IEEE",
  year="2004",
  month="september",
  pages="1695",
  publisher="IEEE",
  type="conference paper"
}