Investigation of lesion detection in MAP reconstruction with non-Gaussian priors

Jinyi Qi

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

8 Citations (Scopus)

Abstract

Statistical image reconstruction methods based on maximum a posteriori (MAP) principle have been developed for emission tomography. The prior distribution of the unknown image plays an important role in MAP reconstruction. The most commonly used prior is the Gaussian prior, whose logarithm has a quadratic form. Gaussian priors are relatively easy to analyze. It has been shown that the effect of a Gaussian prior can be approximated by linear-filtering a maximum likelihood (ML) reconstruction. As a result, sharp edges in reconstructed images are not preserved. To preserve sharp transitions, non-Gaussian priors have been proposed. In this paper, we study the effect of non-Gaussian priors on lesion detection and region of interest quantification in MAP reconstructions using computer simulation. We compare three representative priors - Gaussian prior, Huber prior, and Geman-McClure prior. The results show that for detection and quantification of small lesions, using non-Gaussian priors is not beneficial.

Original languageEnglish (US)
Title of host publicationIEEE Nuclear Science Symposium Conference Record
Pages1704-1708
Number of pages5
Volume3
DOIs
StatePublished - 2005
EventNuclear Science Symposium Conference Record, 2005 IEEE - , Puerto Rico
Duration: Oct 23 2005Oct 29 2005

Other

OtherNuclear Science Symposium Conference Record, 2005 IEEE
CountryPuerto Rico
Period10/23/0510/29/05

Fingerprint

Image reconstruction
Maximum likelihood
Tomography
Computer simulation

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Industrial and Manufacturing Engineering

Cite this

Qi, J. (2005). Investigation of lesion detection in MAP reconstruction with non-Gaussian priors. In IEEE Nuclear Science Symposium Conference Record (Vol. 3, pp. 1704-1708). [1596648] https://doi.org/10.1109/NSSMIC.2005.1596648

Investigation of lesion detection in MAP reconstruction with non-Gaussian priors. / Qi, Jinyi.

IEEE Nuclear Science Symposium Conference Record. Vol. 3 2005. p. 1704-1708 1596648.

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

Qi, J 2005, Investigation of lesion detection in MAP reconstruction with non-Gaussian priors. in IEEE Nuclear Science Symposium Conference Record. vol. 3, 1596648, pp. 1704-1708, Nuclear Science Symposium Conference Record, 2005 IEEE, Puerto Rico, 10/23/05. https://doi.org/10.1109/NSSMIC.2005.1596648
Qi J. Investigation of lesion detection in MAP reconstruction with non-Gaussian priors. In IEEE Nuclear Science Symposium Conference Record. Vol. 3. 2005. p. 1704-1708. 1596648 https://doi.org/10.1109/NSSMIC.2005.1596648
Qi, Jinyi. / Investigation of lesion detection in MAP reconstruction with non-Gaussian priors. IEEE Nuclear Science Symposium Conference Record. Vol. 3 2005. pp. 1704-1708
@inproceedings{de4448b501434db4a6bb9ae3d6d28422,
title = "Investigation of lesion detection in MAP reconstruction with non-Gaussian priors",
abstract = "Statistical image reconstruction methods based on maximum a posteriori (MAP) principle have been developed for emission tomography. The prior distribution of the unknown image plays an important role in MAP reconstruction. The most commonly used prior is the Gaussian prior, whose logarithm has a quadratic form. Gaussian priors are relatively easy to analyze. It has been shown that the effect of a Gaussian prior can be approximated by linear-filtering a maximum likelihood (ML) reconstruction. As a result, sharp edges in reconstructed images are not preserved. To preserve sharp transitions, non-Gaussian priors have been proposed. In this paper, we study the effect of non-Gaussian priors on lesion detection and region of interest quantification in MAP reconstructions using computer simulation. We compare three representative priors - Gaussian prior, Huber prior, and Geman-McClure prior. The results show that for detection and quantification of small lesions, using non-Gaussian priors is not beneficial.",
author = "Jinyi Qi",
year = "2005",
doi = "10.1109/NSSMIC.2005.1596648",
language = "English (US)",
isbn = "0780392213",
volume = "3",
pages = "1704--1708",
booktitle = "IEEE Nuclear Science Symposium Conference Record",

}

TY - GEN

T1 - Investigation of lesion detection in MAP reconstruction with non-Gaussian priors

AU - Qi, Jinyi

PY - 2005

Y1 - 2005

N2 - Statistical image reconstruction methods based on maximum a posteriori (MAP) principle have been developed for emission tomography. The prior distribution of the unknown image plays an important role in MAP reconstruction. The most commonly used prior is the Gaussian prior, whose logarithm has a quadratic form. Gaussian priors are relatively easy to analyze. It has been shown that the effect of a Gaussian prior can be approximated by linear-filtering a maximum likelihood (ML) reconstruction. As a result, sharp edges in reconstructed images are not preserved. To preserve sharp transitions, non-Gaussian priors have been proposed. In this paper, we study the effect of non-Gaussian priors on lesion detection and region of interest quantification in MAP reconstructions using computer simulation. We compare three representative priors - Gaussian prior, Huber prior, and Geman-McClure prior. The results show that for detection and quantification of small lesions, using non-Gaussian priors is not beneficial.

AB - Statistical image reconstruction methods based on maximum a posteriori (MAP) principle have been developed for emission tomography. The prior distribution of the unknown image plays an important role in MAP reconstruction. The most commonly used prior is the Gaussian prior, whose logarithm has a quadratic form. Gaussian priors are relatively easy to analyze. It has been shown that the effect of a Gaussian prior can be approximated by linear-filtering a maximum likelihood (ML) reconstruction. As a result, sharp edges in reconstructed images are not preserved. To preserve sharp transitions, non-Gaussian priors have been proposed. In this paper, we study the effect of non-Gaussian priors on lesion detection and region of interest quantification in MAP reconstructions using computer simulation. We compare three representative priors - Gaussian prior, Huber prior, and Geman-McClure prior. The results show that for detection and quantification of small lesions, using non-Gaussian priors is not beneficial.

UR - http://www.scopus.com/inward/record.url?scp=33746834572&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33746834572&partnerID=8YFLogxK

U2 - 10.1109/NSSMIC.2005.1596648

DO - 10.1109/NSSMIC.2005.1596648

M3 - Conference contribution

AN - SCOPUS:33746834572

SN - 0780392213

SN - 9780780392212

VL - 3

SP - 1704

EP - 1708

BT - IEEE Nuclear Science Symposium Conference Record

ER -