Towards component-based validation of GATE: Aspects of the coincidence processor

Eder R. Moraes, Jonathan K. Poon, Karthikayan Balakrishnan, Wenli Wang, Ramsey D Badawi

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

GATE is public domain software widely used for Monte Carlo simulation in emission tomography. Validations of GATE have primarily been performed on a whole-system basis, leaving the possibility that errors in one sub-system may be offset by errors in others. We assess the accuracy of the GATE PET coincidence generation sub-system in isolation, focusing on the options most closely modeling the majority of commercially available scanners. Independent coincidence generators were coded by teams at Toshiba Medical Research Unit (TMRU) and UC Davis. A model similar to the Siemens mCT scanner was created in GATE. Annihilation photons interacting with the detectors were recorded. Coincidences were generated using GATE, TMRU and UC Davis code and results compared to "ground truth" obtained from the history of the photon interactions. GATE was tested twice, once with every qualified single event opening a time window and initiating a coincidence check (the "multiple window method"), and once where a time window is opened and a coincidence check initiated only by the first single event to occur after the end of the prior time window (the "single window method"). True, scattered and random coincidences were compared. Noise equivalent count rates were also computed and compared. The TMRU and UC Davis coincidence generators agree well with ground truth. With GATE, reasonable accuracy can be obtained if the single window method option is chosen and random coincidences are estimated without use of the delayed coincidence option. However in this GATE version, other parameter combinations can result in significant errors.

Original languageEnglish (US)
Pages (from-to)43-48
Number of pages6
JournalPhysica Medica
Volume31
Issue number1
DOIs
StatePublished - 2015

Fingerprint

GARP Atlantic Tropical Experiment
central processing units
Biomedical Research
Photons
Public Sector
ground truth
Noise
Software
History
Tomography
scanners
generators
photons
isolation
tomography
histories
computer programs
detectors

Keywords

  • GATE
  • Monte Carlo methods
  • Nuclear medicine
  • Positron emission tomography

ASJC Scopus subject areas

  • Biophysics
  • Radiology Nuclear Medicine and imaging
  • Physics and Astronomy(all)

Cite this

Towards component-based validation of GATE : Aspects of the coincidence processor. / Moraes, Eder R.; Poon, Jonathan K.; Balakrishnan, Karthikayan; Wang, Wenli; Badawi, Ramsey D.

In: Physica Medica, Vol. 31, No. 1, 2015, p. 43-48.

Research output: Contribution to journalArticle

Moraes, Eder R. ; Poon, Jonathan K. ; Balakrishnan, Karthikayan ; Wang, Wenli ; Badawi, Ramsey D. / Towards component-based validation of GATE : Aspects of the coincidence processor. In: Physica Medica. 2015 ; Vol. 31, No. 1. pp. 43-48.
@article{31d4c37997ca4519984de8385ea7f9fc,
title = "Towards component-based validation of GATE: Aspects of the coincidence processor",
abstract = "GATE is public domain software widely used for Monte Carlo simulation in emission tomography. Validations of GATE have primarily been performed on a whole-system basis, leaving the possibility that errors in one sub-system may be offset by errors in others. We assess the accuracy of the GATE PET coincidence generation sub-system in isolation, focusing on the options most closely modeling the majority of commercially available scanners. Independent coincidence generators were coded by teams at Toshiba Medical Research Unit (TMRU) and UC Davis. A model similar to the Siemens mCT scanner was created in GATE. Annihilation photons interacting with the detectors were recorded. Coincidences were generated using GATE, TMRU and UC Davis code and results compared to {"}ground truth{"} obtained from the history of the photon interactions. GATE was tested twice, once with every qualified single event opening a time window and initiating a coincidence check (the {"}multiple window method{"}), and once where a time window is opened and a coincidence check initiated only by the first single event to occur after the end of the prior time window (the {"}single window method{"}). True, scattered and random coincidences were compared. Noise equivalent count rates were also computed and compared. The TMRU and UC Davis coincidence generators agree well with ground truth. With GATE, reasonable accuracy can be obtained if the single window method option is chosen and random coincidences are estimated without use of the delayed coincidence option. However in this GATE version, other parameter combinations can result in significant errors.",
keywords = "GATE, Monte Carlo methods, Nuclear medicine, Positron emission tomography",
author = "Moraes, {Eder R.} and Poon, {Jonathan K.} and Karthikayan Balakrishnan and Wenli Wang and Badawi, {Ramsey D}",
year = "2015",
doi = "10.1016/j.ejmp.2014.09.001",
language = "English (US)",
volume = "31",
pages = "43--48",
journal = "Physica Medica",
issn = "1120-1797",
publisher = "Associazione Italiana di Fisica Medica",
number = "1",

}

TY - JOUR

T1 - Towards component-based validation of GATE

T2 - Aspects of the coincidence processor

AU - Moraes, Eder R.

AU - Poon, Jonathan K.

AU - Balakrishnan, Karthikayan

AU - Wang, Wenli

AU - Badawi, Ramsey D

PY - 2015

Y1 - 2015

N2 - GATE is public domain software widely used for Monte Carlo simulation in emission tomography. Validations of GATE have primarily been performed on a whole-system basis, leaving the possibility that errors in one sub-system may be offset by errors in others. We assess the accuracy of the GATE PET coincidence generation sub-system in isolation, focusing on the options most closely modeling the majority of commercially available scanners. Independent coincidence generators were coded by teams at Toshiba Medical Research Unit (TMRU) and UC Davis. A model similar to the Siemens mCT scanner was created in GATE. Annihilation photons interacting with the detectors were recorded. Coincidences were generated using GATE, TMRU and UC Davis code and results compared to "ground truth" obtained from the history of the photon interactions. GATE was tested twice, once with every qualified single event opening a time window and initiating a coincidence check (the "multiple window method"), and once where a time window is opened and a coincidence check initiated only by the first single event to occur after the end of the prior time window (the "single window method"). True, scattered and random coincidences were compared. Noise equivalent count rates were also computed and compared. The TMRU and UC Davis coincidence generators agree well with ground truth. With GATE, reasonable accuracy can be obtained if the single window method option is chosen and random coincidences are estimated without use of the delayed coincidence option. However in this GATE version, other parameter combinations can result in significant errors.

AB - GATE is public domain software widely used for Monte Carlo simulation in emission tomography. Validations of GATE have primarily been performed on a whole-system basis, leaving the possibility that errors in one sub-system may be offset by errors in others. We assess the accuracy of the GATE PET coincidence generation sub-system in isolation, focusing on the options most closely modeling the majority of commercially available scanners. Independent coincidence generators were coded by teams at Toshiba Medical Research Unit (TMRU) and UC Davis. A model similar to the Siemens mCT scanner was created in GATE. Annihilation photons interacting with the detectors were recorded. Coincidences were generated using GATE, TMRU and UC Davis code and results compared to "ground truth" obtained from the history of the photon interactions. GATE was tested twice, once with every qualified single event opening a time window and initiating a coincidence check (the "multiple window method"), and once where a time window is opened and a coincidence check initiated only by the first single event to occur after the end of the prior time window (the "single window method"). True, scattered and random coincidences were compared. Noise equivalent count rates were also computed and compared. The TMRU and UC Davis coincidence generators agree well with ground truth. With GATE, reasonable accuracy can be obtained if the single window method option is chosen and random coincidences are estimated without use of the delayed coincidence option. However in this GATE version, other parameter combinations can result in significant errors.

KW - GATE

KW - Monte Carlo methods

KW - Nuclear medicine

KW - Positron emission tomography

UR - http://www.scopus.com/inward/record.url?scp=84927597295&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84927597295&partnerID=8YFLogxK

U2 - 10.1016/j.ejmp.2014.09.001

DO - 10.1016/j.ejmp.2014.09.001

M3 - Article

C2 - 25240897

AN - SCOPUS:84927597295

VL - 31

SP - 43

EP - 48

JO - Physica Medica

JF - Physica Medica

SN - 1120-1797

IS - 1

ER -