GPU-Assisted Generation of System Matrices for High-Resolution Imaging Systems

Alexander Lin, Matthew A. Kupinski, Xin Li, Lars R. Furenlid

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

Abstract

Traditional methods of point spread function (PSF) modeling of pinhole SPECT systems, such as fitting the PSF with a 2D Gaussian, are generally sufficient in characterizing imaging systems in which the detector is the main source of PSF blur. However, when modeling the PSF of a high-resolution imaging system, these methods are too simplistic and fail to capture important PSF features, resulting in errors in later simulation studies. In this work, we present a method for parameterizing point spread functions that is then used to rapidly generate system matrices of gamma-ray imaging systems using high-resolution detectors with sub-millimeter spatial resolution. Our algorithm, which utilizes a GPU-based ray tracer to simulate a system's true PSF, accounts for the PSF blur due to not only the detector's depth of interaction, but also the penetration through the pinhole and the finite size of of the source. By considering all three blurring sources, we are able to better model the PSF, capturing features like the flat-top in the center and the tilt and drop-off towards the edges. Comparisons to the 2D Gaussian fitting model are examined and we report that as the ratio between the source size and pinhole radius increases, the two models converge to one another.

Original languageEnglish (US)
Title of host publication2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781538684948
DOIs
StatePublished - Nov 2018
Event2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Sydney, Australia
Duration: Nov 10 2018Nov 17 2018

Publication series

Name2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings

Conference

Conference2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018
CountryAustralia
CitySydney
Period11/10/1811/17/18

Fingerprint

point spread functions
high resolution
matrices
Gamma Rays
pinholes
Single-Photon Emission-Computed Tomography
detectors
blurring
tracers
rays
penetration
spatial resolution
gamma rays
radii

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Nuclear and High Energy Physics

Cite this

Lin, A., Kupinski, M. A., Li, X., & Furenlid, L. R. (2018). GPU-Assisted Generation of System Matrices for High-Resolution Imaging Systems. In 2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings [8824756] (2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/NSSMIC.2018.8824756

GPU-Assisted Generation of System Matrices for High-Resolution Imaging Systems. / Lin, Alexander; Kupinski, Matthew A.; Li, Xin; Furenlid, Lars R.

2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2018. 8824756 (2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings).

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

Lin, A, Kupinski, MA, Li, X & Furenlid, LR 2018, GPU-Assisted Generation of System Matrices for High-Resolution Imaging Systems. in 2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings., 8824756, 2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings, Institute of Electrical and Electronics Engineers Inc., 2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018, Sydney, Australia, 11/10/18. https://doi.org/10.1109/NSSMIC.2018.8824756
Lin A, Kupinski MA, Li X, Furenlid LR. GPU-Assisted Generation of System Matrices for High-Resolution Imaging Systems. In 2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2018. 8824756. (2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings). https://doi.org/10.1109/NSSMIC.2018.8824756
Lin, Alexander ; Kupinski, Matthew A. ; Li, Xin ; Furenlid, Lars R. / GPU-Assisted Generation of System Matrices for High-Resolution Imaging Systems. 2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2018. (2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings).
@inproceedings{dfd56e6d21c643528c4f16eb87ea3491,
title = "GPU-Assisted Generation of System Matrices for High-Resolution Imaging Systems",
abstract = "Traditional methods of point spread function (PSF) modeling of pinhole SPECT systems, such as fitting the PSF with a 2D Gaussian, are generally sufficient in characterizing imaging systems in which the detector is the main source of PSF blur. However, when modeling the PSF of a high-resolution imaging system, these methods are too simplistic and fail to capture important PSF features, resulting in errors in later simulation studies. In this work, we present a method for parameterizing point spread functions that is then used to rapidly generate system matrices of gamma-ray imaging systems using high-resolution detectors with sub-millimeter spatial resolution. Our algorithm, which utilizes a GPU-based ray tracer to simulate a system's true PSF, accounts for the PSF blur due to not only the detector's depth of interaction, but also the penetration through the pinhole and the finite size of of the source. By considering all three blurring sources, we are able to better model the PSF, capturing features like the flat-top in the center and the tilt and drop-off towards the edges. Comparisons to the 2D Gaussian fitting model are examined and we report that as the ratio between the source size and pinhole radius increases, the two models converge to one another.",
author = "Alexander Lin and Kupinski, {Matthew A.} and Xin Li and Furenlid, {Lars R.}",
year = "2018",
month = "11",
doi = "10.1109/NSSMIC.2018.8824756",
language = "English (US)",
series = "2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
booktitle = "2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings",

}

TY - GEN

T1 - GPU-Assisted Generation of System Matrices for High-Resolution Imaging Systems

AU - Lin, Alexander

AU - Kupinski, Matthew A.

AU - Li, Xin

AU - Furenlid, Lars R.

PY - 2018/11

Y1 - 2018/11

N2 - Traditional methods of point spread function (PSF) modeling of pinhole SPECT systems, such as fitting the PSF with a 2D Gaussian, are generally sufficient in characterizing imaging systems in which the detector is the main source of PSF blur. However, when modeling the PSF of a high-resolution imaging system, these methods are too simplistic and fail to capture important PSF features, resulting in errors in later simulation studies. In this work, we present a method for parameterizing point spread functions that is then used to rapidly generate system matrices of gamma-ray imaging systems using high-resolution detectors with sub-millimeter spatial resolution. Our algorithm, which utilizes a GPU-based ray tracer to simulate a system's true PSF, accounts for the PSF blur due to not only the detector's depth of interaction, but also the penetration through the pinhole and the finite size of of the source. By considering all three blurring sources, we are able to better model the PSF, capturing features like the flat-top in the center and the tilt and drop-off towards the edges. Comparisons to the 2D Gaussian fitting model are examined and we report that as the ratio between the source size and pinhole radius increases, the two models converge to one another.

AB - Traditional methods of point spread function (PSF) modeling of pinhole SPECT systems, such as fitting the PSF with a 2D Gaussian, are generally sufficient in characterizing imaging systems in which the detector is the main source of PSF blur. However, when modeling the PSF of a high-resolution imaging system, these methods are too simplistic and fail to capture important PSF features, resulting in errors in later simulation studies. In this work, we present a method for parameterizing point spread functions that is then used to rapidly generate system matrices of gamma-ray imaging systems using high-resolution detectors with sub-millimeter spatial resolution. Our algorithm, which utilizes a GPU-based ray tracer to simulate a system's true PSF, accounts for the PSF blur due to not only the detector's depth of interaction, but also the penetration through the pinhole and the finite size of of the source. By considering all three blurring sources, we are able to better model the PSF, capturing features like the flat-top in the center and the tilt and drop-off towards the edges. Comparisons to the 2D Gaussian fitting model are examined and we report that as the ratio between the source size and pinhole radius increases, the two models converge to one another.

UR - http://www.scopus.com/inward/record.url?scp=85073120807&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85073120807&partnerID=8YFLogxK

U2 - 10.1109/NSSMIC.2018.8824756

DO - 10.1109/NSSMIC.2018.8824756

M3 - Conference contribution

AN - SCOPUS:85073120807

T3 - 2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings

BT - 2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

ER -