Predicting detection performance with model observers

Fourier domain or spatial domain?

Baiyu Chen, Lifeng Yu, Shuai Leng, James Kofler, Christopher Favazza, Thomas Vrieze, Cynthia H McCollough

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

3 Citations (Scopus)

Abstract

The use of Fourier domain model observer is challenged by iterative reconstruction (IR), because IR algorithms are nonlinear and IR images have noise texture different from that of FBP. A modified Fourier domain model observer, which incorporates nonlinear noise and resolution properties, has been proposed for IR and needs to be validated with human detection performance. On the other hand, the spatial domain model observer is theoretically applicable to IR, but more computationally intensive than the Fourier domain method. The purpose of this study is to compare the modified Fourier domain model observer to the spatial domain model observer with both FBP and IR images, using human detection performance as the gold standard. A phantom with inserts of various low contrast levels and sizes was repeatedly scanned 100 times on a third-generation, dual-source CT scanner at 5 dose levels and reconstructed using FBP and IR algorithms. The human detection performance of the inserts was measured via a 2-alternative-forced-choice (2AFC) test. In addition, two model observer performances were calculated, including a Fourier domain non-prewhitening model observer and a spatial domain channelized Hotelling observer. The performance of these two mode observers was compared in terms of how well they correlated with human observer performance. Our results demonstrated that the spatial domain model observer correlated well with human observers across various dose levels, object contrast levels, and object sizes. The Fourier domain observer correlated well with human observers using FBP images, but overestimated the detection performance using IR images.

Original languageEnglish (US)
Title of host publicationMedical Imaging 2016: Physics of Medical Imaging
PublisherSPIE
Volume9783
ISBN (Electronic)9781510600188
DOIs
StatePublished - 2016
EventMedical Imaging 2016: Physics of Medical Imaging - San Diego, United States
Duration: Feb 28 2016Mar 2 2016

Other

OtherMedical Imaging 2016: Physics of Medical Imaging
CountryUnited States
CitySan Diego
Period2/28/163/2/16

Fingerprint

Computer-Assisted Image Processing
Image reconstruction
image reconstruction
Noise
inserts
dosage
Textures
scanners
textures

Keywords

  • channelized Hotelling observer (CHO)
  • Computed tomography (CT)
  • detectability index (d')
  • iterative reconstruction (IR)
  • model observer
  • non-prewhitening

ASJC Scopus subject areas

  • Atomic and Molecular Physics, and Optics
  • Electronic, Optical and Magnetic Materials
  • Biomaterials
  • Radiology Nuclear Medicine and imaging

Cite this

Chen, B., Yu, L., Leng, S., Kofler, J., Favazza, C., Vrieze, T., & McCollough, C. H. (2016). Predicting detection performance with model observers: Fourier domain or spatial domain? In Medical Imaging 2016: Physics of Medical Imaging (Vol. 9783). [978326] SPIE. https://doi.org/10.1117/12.2216962

Predicting detection performance with model observers : Fourier domain or spatial domain? / Chen, Baiyu; Yu, Lifeng; Leng, Shuai; Kofler, James; Favazza, Christopher; Vrieze, Thomas; McCollough, Cynthia H.

Medical Imaging 2016: Physics of Medical Imaging. Vol. 9783 SPIE, 2016. 978326.

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

Chen, B, Yu, L, Leng, S, Kofler, J, Favazza, C, Vrieze, T & McCollough, CH 2016, Predicting detection performance with model observers: Fourier domain or spatial domain? in Medical Imaging 2016: Physics of Medical Imaging. vol. 9783, 978326, SPIE, Medical Imaging 2016: Physics of Medical Imaging, San Diego, United States, 2/28/16. https://doi.org/10.1117/12.2216962
Chen B, Yu L, Leng S, Kofler J, Favazza C, Vrieze T et al. Predicting detection performance with model observers: Fourier domain or spatial domain? In Medical Imaging 2016: Physics of Medical Imaging. Vol. 9783. SPIE. 2016. 978326 https://doi.org/10.1117/12.2216962
Chen, Baiyu ; Yu, Lifeng ; Leng, Shuai ; Kofler, James ; Favazza, Christopher ; Vrieze, Thomas ; McCollough, Cynthia H. / Predicting detection performance with model observers : Fourier domain or spatial domain?. Medical Imaging 2016: Physics of Medical Imaging. Vol. 9783 SPIE, 2016.
@inproceedings{2214b6c9804240e191ce4847a698d58a,
title = "Predicting detection performance with model observers: Fourier domain or spatial domain?",
abstract = "The use of Fourier domain model observer is challenged by iterative reconstruction (IR), because IR algorithms are nonlinear and IR images have noise texture different from that of FBP. A modified Fourier domain model observer, which incorporates nonlinear noise and resolution properties, has been proposed for IR and needs to be validated with human detection performance. On the other hand, the spatial domain model observer is theoretically applicable to IR, but more computationally intensive than the Fourier domain method. The purpose of this study is to compare the modified Fourier domain model observer to the spatial domain model observer with both FBP and IR images, using human detection performance as the gold standard. A phantom with inserts of various low contrast levels and sizes was repeatedly scanned 100 times on a third-generation, dual-source CT scanner at 5 dose levels and reconstructed using FBP and IR algorithms. The human detection performance of the inserts was measured via a 2-alternative-forced-choice (2AFC) test. In addition, two model observer performances were calculated, including a Fourier domain non-prewhitening model observer and a spatial domain channelized Hotelling observer. The performance of these two mode observers was compared in terms of how well they correlated with human observer performance. Our results demonstrated that the spatial domain model observer correlated well with human observers across various dose levels, object contrast levels, and object sizes. The Fourier domain observer correlated well with human observers using FBP images, but overestimated the detection performance using IR images.",
keywords = "channelized Hotelling observer (CHO), Computed tomography (CT), detectability index (d'), iterative reconstruction (IR), model observer, non-prewhitening",
author = "Baiyu Chen and Lifeng Yu and Shuai Leng and James Kofler and Christopher Favazza and Thomas Vrieze and McCollough, {Cynthia H}",
year = "2016",
doi = "10.1117/12.2216962",
language = "English (US)",
volume = "9783",
booktitle = "Medical Imaging 2016: Physics of Medical Imaging",
publisher = "SPIE",

}

TY - GEN

T1 - Predicting detection performance with model observers

T2 - Fourier domain or spatial domain?

AU - Chen, Baiyu

AU - Yu, Lifeng

AU - Leng, Shuai

AU - Kofler, James

AU - Favazza, Christopher

AU - Vrieze, Thomas

AU - McCollough, Cynthia H

PY - 2016

Y1 - 2016

N2 - The use of Fourier domain model observer is challenged by iterative reconstruction (IR), because IR algorithms are nonlinear and IR images have noise texture different from that of FBP. A modified Fourier domain model observer, which incorporates nonlinear noise and resolution properties, has been proposed for IR and needs to be validated with human detection performance. On the other hand, the spatial domain model observer is theoretically applicable to IR, but more computationally intensive than the Fourier domain method. The purpose of this study is to compare the modified Fourier domain model observer to the spatial domain model observer with both FBP and IR images, using human detection performance as the gold standard. A phantom with inserts of various low contrast levels and sizes was repeatedly scanned 100 times on a third-generation, dual-source CT scanner at 5 dose levels and reconstructed using FBP and IR algorithms. The human detection performance of the inserts was measured via a 2-alternative-forced-choice (2AFC) test. In addition, two model observer performances were calculated, including a Fourier domain non-prewhitening model observer and a spatial domain channelized Hotelling observer. The performance of these two mode observers was compared in terms of how well they correlated with human observer performance. Our results demonstrated that the spatial domain model observer correlated well with human observers across various dose levels, object contrast levels, and object sizes. The Fourier domain observer correlated well with human observers using FBP images, but overestimated the detection performance using IR images.

AB - The use of Fourier domain model observer is challenged by iterative reconstruction (IR), because IR algorithms are nonlinear and IR images have noise texture different from that of FBP. A modified Fourier domain model observer, which incorporates nonlinear noise and resolution properties, has been proposed for IR and needs to be validated with human detection performance. On the other hand, the spatial domain model observer is theoretically applicable to IR, but more computationally intensive than the Fourier domain method. The purpose of this study is to compare the modified Fourier domain model observer to the spatial domain model observer with both FBP and IR images, using human detection performance as the gold standard. A phantom with inserts of various low contrast levels and sizes was repeatedly scanned 100 times on a third-generation, dual-source CT scanner at 5 dose levels and reconstructed using FBP and IR algorithms. The human detection performance of the inserts was measured via a 2-alternative-forced-choice (2AFC) test. In addition, two model observer performances were calculated, including a Fourier domain non-prewhitening model observer and a spatial domain channelized Hotelling observer. The performance of these two mode observers was compared in terms of how well they correlated with human observer performance. Our results demonstrated that the spatial domain model observer correlated well with human observers across various dose levels, object contrast levels, and object sizes. The Fourier domain observer correlated well with human observers using FBP images, but overestimated the detection performance using IR images.

KW - channelized Hotelling observer (CHO)

KW - Computed tomography (CT)

KW - detectability index (d')

KW - iterative reconstruction (IR)

KW - model observer

KW - non-prewhitening

UR - http://www.scopus.com/inward/record.url?scp=84978795000&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84978795000&partnerID=8YFLogxK

U2 - 10.1117/12.2216962

DO - 10.1117/12.2216962

M3 - Conference contribution

VL - 9783

BT - Medical Imaging 2016: Physics of Medical Imaging

PB - SPIE

ER -