Quantitative and clinical analysis of SPECT image registration for epilepsy studies

Benjamin Brinkmann, Terence J. O'Brien, Shmuel Aharon, Michael K. O'Connor, Brian P. Mullan, Dennis P. Hanson, Richard A. Robb

Research output: Contribution to journalArticle

38 Citations (Scopus)

Abstract

This study reports quantitative measurements of the accuracy of two popular voxel-based registration algorithms-Woods' automated image registration algorithm and mutual information correlation-and compares these with conventional surface matching (SM) registration. Methods: The registration algorithms were compared (15 different matches each) for (a) three-dimensional brain phantom images, (b) an ictal SPECT image from a patient with partial epilepsy matched to itself after modification to simulate changes in the cerebral blood flow pattern and (c) ictal/interictal SPECT images from 15 patients with partial epilepsy. Blinded visual ranking and localization of the subtraction images derived from the patient images were also performed. Results: Both voxel-based registration methods were more accurate than SM registration (P < 0.0005). Automated image registration algorithm was more accurate than mutual information correlation for the computer-simulated ictal/interictal images and the patient ictal/interictal studies (P < 0.05). The subtraction SPECTs from SM were poorer in visual ranking more often than the voxel-based methods (P < 0.05). Conclusion: Voxel intensity-based registration algorithms provide significant improvement in ictal/interictal SPECT registration accuracy and result in a clinically detectable improvement in the subtraction SPECT images.

Original languageEnglish (US)
Pages (from-to)1098-1105
Number of pages8
JournalJournal of Nuclear Medicine
Volume40
Issue number7
StatePublished - Jul 1999

Fingerprint

Single-Photon Emission-Computed Tomography
Epilepsy
Stroke
Partial Epilepsy
Cerebrovascular Circulation
Brain

Keywords

  • Cerebral blood flow
  • Ictal/interictal SPECT
  • Image registration
  • Partial epilepsy

ASJC Scopus subject areas

  • Radiological and Ultrasound Technology

Cite this

Brinkmann, B., O'Brien, T. J., Aharon, S., O'Connor, M. K., Mullan, B. P., Hanson, D. P., & Robb, R. A. (1999). Quantitative and clinical analysis of SPECT image registration for epilepsy studies. Journal of Nuclear Medicine, 40(7), 1098-1105.

Quantitative and clinical analysis of SPECT image registration for epilepsy studies. / Brinkmann, Benjamin; O'Brien, Terence J.; Aharon, Shmuel; O'Connor, Michael K.; Mullan, Brian P.; Hanson, Dennis P.; Robb, Richard A.

In: Journal of Nuclear Medicine, Vol. 40, No. 7, 07.1999, p. 1098-1105.

Research output: Contribution to journalArticle

Brinkmann, B, O'Brien, TJ, Aharon, S, O'Connor, MK, Mullan, BP, Hanson, DP & Robb, RA 1999, 'Quantitative and clinical analysis of SPECT image registration for epilepsy studies', Journal of Nuclear Medicine, vol. 40, no. 7, pp. 1098-1105.
Brinkmann B, O'Brien TJ, Aharon S, O'Connor MK, Mullan BP, Hanson DP et al. Quantitative and clinical analysis of SPECT image registration for epilepsy studies. Journal of Nuclear Medicine. 1999 Jul;40(7):1098-1105.
Brinkmann, Benjamin ; O'Brien, Terence J. ; Aharon, Shmuel ; O'Connor, Michael K. ; Mullan, Brian P. ; Hanson, Dennis P. ; Robb, Richard A. / Quantitative and clinical analysis of SPECT image registration for epilepsy studies. In: Journal of Nuclear Medicine. 1999 ; Vol. 40, No. 7. pp. 1098-1105.
@article{cae85484452a49a98996006b9b068695,
title = "Quantitative and clinical analysis of SPECT image registration for epilepsy studies",
abstract = "This study reports quantitative measurements of the accuracy of two popular voxel-based registration algorithms-Woods' automated image registration algorithm and mutual information correlation-and compares these with conventional surface matching (SM) registration. Methods: The registration algorithms were compared (15 different matches each) for (a) three-dimensional brain phantom images, (b) an ictal SPECT image from a patient with partial epilepsy matched to itself after modification to simulate changes in the cerebral blood flow pattern and (c) ictal/interictal SPECT images from 15 patients with partial epilepsy. Blinded visual ranking and localization of the subtraction images derived from the patient images were also performed. Results: Both voxel-based registration methods were more accurate than SM registration (P < 0.0005). Automated image registration algorithm was more accurate than mutual information correlation for the computer-simulated ictal/interictal images and the patient ictal/interictal studies (P < 0.05). The subtraction SPECTs from SM were poorer in visual ranking more often than the voxel-based methods (P < 0.05). Conclusion: Voxel intensity-based registration algorithms provide significant improvement in ictal/interictal SPECT registration accuracy and result in a clinically detectable improvement in the subtraction SPECT images.",
keywords = "Cerebral blood flow, Ictal/interictal SPECT, Image registration, Partial epilepsy",
author = "Benjamin Brinkmann and O'Brien, {Terence J.} and Shmuel Aharon and O'Connor, {Michael K.} and Mullan, {Brian P.} and Hanson, {Dennis P.} and Robb, {Richard A.}",
year = "1999",
month = "7",
language = "English (US)",
volume = "40",
pages = "1098--1105",
journal = "Journal of Nuclear Medicine",
issn = "0161-5505",
publisher = "Society of Nuclear Medicine Inc.",
number = "7",

}

TY - JOUR

T1 - Quantitative and clinical analysis of SPECT image registration for epilepsy studies

AU - Brinkmann, Benjamin

AU - O'Brien, Terence J.

AU - Aharon, Shmuel

AU - O'Connor, Michael K.

AU - Mullan, Brian P.

AU - Hanson, Dennis P.

AU - Robb, Richard A.

PY - 1999/7

Y1 - 1999/7

N2 - This study reports quantitative measurements of the accuracy of two popular voxel-based registration algorithms-Woods' automated image registration algorithm and mutual information correlation-and compares these with conventional surface matching (SM) registration. Methods: The registration algorithms were compared (15 different matches each) for (a) three-dimensional brain phantom images, (b) an ictal SPECT image from a patient with partial epilepsy matched to itself after modification to simulate changes in the cerebral blood flow pattern and (c) ictal/interictal SPECT images from 15 patients with partial epilepsy. Blinded visual ranking and localization of the subtraction images derived from the patient images were also performed. Results: Both voxel-based registration methods were more accurate than SM registration (P < 0.0005). Automated image registration algorithm was more accurate than mutual information correlation for the computer-simulated ictal/interictal images and the patient ictal/interictal studies (P < 0.05). The subtraction SPECTs from SM were poorer in visual ranking more often than the voxel-based methods (P < 0.05). Conclusion: Voxel intensity-based registration algorithms provide significant improvement in ictal/interictal SPECT registration accuracy and result in a clinically detectable improvement in the subtraction SPECT images.

AB - This study reports quantitative measurements of the accuracy of two popular voxel-based registration algorithms-Woods' automated image registration algorithm and mutual information correlation-and compares these with conventional surface matching (SM) registration. Methods: The registration algorithms were compared (15 different matches each) for (a) three-dimensional brain phantom images, (b) an ictal SPECT image from a patient with partial epilepsy matched to itself after modification to simulate changes in the cerebral blood flow pattern and (c) ictal/interictal SPECT images from 15 patients with partial epilepsy. Blinded visual ranking and localization of the subtraction images derived from the patient images were also performed. Results: Both voxel-based registration methods were more accurate than SM registration (P < 0.0005). Automated image registration algorithm was more accurate than mutual information correlation for the computer-simulated ictal/interictal images and the patient ictal/interictal studies (P < 0.05). The subtraction SPECTs from SM were poorer in visual ranking more often than the voxel-based methods (P < 0.05). Conclusion: Voxel intensity-based registration algorithms provide significant improvement in ictal/interictal SPECT registration accuracy and result in a clinically detectable improvement in the subtraction SPECT images.

KW - Cerebral blood flow

KW - Ictal/interictal SPECT

KW - Image registration

KW - Partial epilepsy

UR - http://www.scopus.com/inward/record.url?scp=0033014676&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033014676&partnerID=8YFLogxK

M3 - Article

C2 - 10405126

AN - SCOPUS:0033014676

VL - 40

SP - 1098

EP - 1105

JO - Journal of Nuclear Medicine

JF - Journal of Nuclear Medicine

SN - 0161-5505

IS - 7

ER -