Generalised overlap measures for assessment of pairwise and groupwise image registration and segmentation

William R. Crum, Oscar Camara, Daniel Rueckert, Kanwal K. Bhatia, Mark Jenkinson, Derek L.G. Hill

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

26 Citations (Scopus)

Abstract

Effective validation techniques are an essential pre-requisite for segmentation and non-rigid registration techniques to enter clinical use. These algorithms can be evaluated by calculating the overlap of corresponding test and gold-standard regions. Common overlap measures compare pairs of binary labels but it is now common for multiple labels to exist and for fractional (partial volume) labels to be used to describe multiple tissue types contributing to a single voxel. Evaluation studies may involve multiple image pairs. In this paper we use results from fuzzy set theory and fuzzy morphology to extend the definitions of existing overlap measures to accommodate multiple fractional labels. Simple formulas are provided which define single figures of merit to quantify the total overlap for ensembles of pairwise or groupwise label comparisons. A quantitative link between overlap and registration error is established by defining the overlap tolerance. Experiments are performed on publicly available labeled brain data to demonstrate the new measures in a comparison of pairwise and groupwise registration.

LanguageEnglish
Title of host publicationMedical Image Computing and Computer-Assisted Intervention - MICCAI 2005 - 8th International Conference, Proceedings
Pages99-106
Number of pages8
DOIs
Publication statusPublished - 1 Dec 2005
Event8th International Conference on Medical Image Computing and Computer-Assisted Intervention - MICCAI 2005 - Palm Springs, CA, United States
Duration: 26 Oct 200529 Oct 2005

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3749 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other8th International Conference on Medical Image Computing and Computer-Assisted Intervention - MICCAI 2005
CountryUnited States
CityPalm Springs, CA
Period26/10/0529/10/05

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Crum, W. R., Camara, O., Rueckert, D., Bhatia, K. K., Jenkinson, M., & Hill, D. L. G. (2005). Generalised overlap measures for assessment of pairwise and groupwise image registration and segmentation. In Medical Image Computing and Computer-Assisted Intervention - MICCAI 2005 - 8th International Conference, Proceedings (pp. 99-106). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3749 LNCS). https://doi.org/10.1007/11566465_13
Crum, William R. ; Camara, Oscar ; Rueckert, Daniel ; Bhatia, Kanwal K. ; Jenkinson, Mark ; Hill, Derek L.G. / Generalised overlap measures for assessment of pairwise and groupwise image registration and segmentation. Medical Image Computing and Computer-Assisted Intervention - MICCAI 2005 - 8th International Conference, Proceedings. 2005. pp. 99-106 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{0ab52f059bff4b51b19d78d32d96a3c8,
title = "Generalised overlap measures for assessment of pairwise and groupwise image registration and segmentation",
abstract = "Effective validation techniques are an essential pre-requisite for segmentation and non-rigid registration techniques to enter clinical use. These algorithms can be evaluated by calculating the overlap of corresponding test and gold-standard regions. Common overlap measures compare pairs of binary labels but it is now common for multiple labels to exist and for fractional (partial volume) labels to be used to describe multiple tissue types contributing to a single voxel. Evaluation studies may involve multiple image pairs. In this paper we use results from fuzzy set theory and fuzzy morphology to extend the definitions of existing overlap measures to accommodate multiple fractional labels. Simple formulas are provided which define single figures of merit to quantify the total overlap for ensembles of pairwise or groupwise label comparisons. A quantitative link between overlap and registration error is established by defining the overlap tolerance. Experiments are performed on publicly available labeled brain data to demonstrate the new measures in a comparison of pairwise and groupwise registration.",
author = "Crum, {William R.} and Oscar Camara and Daniel Rueckert and Bhatia, {Kanwal K.} and Mark Jenkinson and Hill, {Derek L.G.}",
year = "2005",
month = "12",
day = "1",
doi = "10.1007/11566465_13",
language = "English",
isbn = "3540293272",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "99--106",
booktitle = "Medical Image Computing and Computer-Assisted Intervention - MICCAI 2005 - 8th International Conference, Proceedings",

}

Crum, WR, Camara, O, Rueckert, D, Bhatia, KK, Jenkinson, M & Hill, DLG 2005, Generalised overlap measures for assessment of pairwise and groupwise image registration and segmentation. in Medical Image Computing and Computer-Assisted Intervention - MICCAI 2005 - 8th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 3749 LNCS, pp. 99-106, 8th International Conference on Medical Image Computing and Computer-Assisted Intervention - MICCAI 2005, Palm Springs, CA, United States, 26/10/05. https://doi.org/10.1007/11566465_13

Generalised overlap measures for assessment of pairwise and groupwise image registration and segmentation. / Crum, William R.; Camara, Oscar; Rueckert, Daniel; Bhatia, Kanwal K.; Jenkinson, Mark; Hill, Derek L.G.

Medical Image Computing and Computer-Assisted Intervention - MICCAI 2005 - 8th International Conference, Proceedings. 2005. p. 99-106 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3749 LNCS).

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

TY - GEN

T1 - Generalised overlap measures for assessment of pairwise and groupwise image registration and segmentation

AU - Crum, William R.

AU - Camara, Oscar

AU - Rueckert, Daniel

AU - Bhatia, Kanwal K.

AU - Jenkinson, Mark

AU - Hill, Derek L.G.

PY - 2005/12/1

Y1 - 2005/12/1

N2 - Effective validation techniques are an essential pre-requisite for segmentation and non-rigid registration techniques to enter clinical use. These algorithms can be evaluated by calculating the overlap of corresponding test and gold-standard regions. Common overlap measures compare pairs of binary labels but it is now common for multiple labels to exist and for fractional (partial volume) labels to be used to describe multiple tissue types contributing to a single voxel. Evaluation studies may involve multiple image pairs. In this paper we use results from fuzzy set theory and fuzzy morphology to extend the definitions of existing overlap measures to accommodate multiple fractional labels. Simple formulas are provided which define single figures of merit to quantify the total overlap for ensembles of pairwise or groupwise label comparisons. A quantitative link between overlap and registration error is established by defining the overlap tolerance. Experiments are performed on publicly available labeled brain data to demonstrate the new measures in a comparison of pairwise and groupwise registration.

AB - Effective validation techniques are an essential pre-requisite for segmentation and non-rigid registration techniques to enter clinical use. These algorithms can be evaluated by calculating the overlap of corresponding test and gold-standard regions. Common overlap measures compare pairs of binary labels but it is now common for multiple labels to exist and for fractional (partial volume) labels to be used to describe multiple tissue types contributing to a single voxel. Evaluation studies may involve multiple image pairs. In this paper we use results from fuzzy set theory and fuzzy morphology to extend the definitions of existing overlap measures to accommodate multiple fractional labels. Simple formulas are provided which define single figures of merit to quantify the total overlap for ensembles of pairwise or groupwise label comparisons. A quantitative link between overlap and registration error is established by defining the overlap tolerance. Experiments are performed on publicly available labeled brain data to demonstrate the new measures in a comparison of pairwise and groupwise registration.

UR - http://www.scopus.com/inward/record.url?scp=33744823826&partnerID=8YFLogxK

U2 - 10.1007/11566465_13

DO - 10.1007/11566465_13

M3 - Conference contribution

SN - 3540293272

SN - 9783540293279

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

SP - 99

EP - 106

BT - Medical Image Computing and Computer-Assisted Intervention - MICCAI 2005 - 8th International Conference, Proceedings

ER -

Crum WR, Camara O, Rueckert D, Bhatia KK, Jenkinson M, Hill DLG. Generalised overlap measures for assessment of pairwise and groupwise image registration and segmentation. In Medical Image Computing and Computer-Assisted Intervention - MICCAI 2005 - 8th International Conference, Proceedings. 2005. p. 99-106. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/11566465_13