Regularization design in penalized maximum-likelihood image reconstruction for lesion detection in 3D PET

Li Yang, Jian Zhou, Andrea Ferrero, Ramsey D Badawi, Jinyi Qi

Research output: Contribution to journalArticle

20 Citations (Scopus)

Abstract

Detecting cancerous lesions is a major clinical application in emission tomography. In previous work, we have studied penalized maximum-likelihood (PML) image reconstruction for the detection task and proposed a method to design a shift-invariant quadratic penalty function to maximize detectability of a lesion at a known location in a two dimensional image. Here we extend the regularization design to maximize detectability of lesions at unknown locations in fully 3D PET. We used a multiview channelized Hotelling observer (mvCHO) to assess the lesion detectability in 3D images to mimic the condition where a human observer examines three orthogonal views of a 3D image for lesion detection. We derived simplified theoretical expressions that allow fast prediction of the detectability of a 3D lesion. The theoretical results were used to design the regularization in PML reconstruction to improve lesion detectability. We conducted computer-based Monte Carlo simulations to compare the optimized penalty with the conventional penalty for detecting lesions of various sizes. Only true coincidence events were simulated. Lesion detectability was also assessed by two human observers, whose performances agree well with that of the mvCHO. Both the numerical observer and human observer results showed a statistically significant improvement in lesion detection by using the proposed penalty function compared to using the conventional penalty function.

Original languageEnglish (US)
Pages (from-to)403-419
Number of pages17
JournalPhysics in Medicine and Biology
Volume59
Issue number2
DOIs
StatePublished - Jan 20 2014

Fingerprint

Computer-Assisted Image Processing
Tomography

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Radiological and Ultrasound Technology

Cite this

Regularization design in penalized maximum-likelihood image reconstruction for lesion detection in 3D PET. / Yang, Li; Zhou, Jian; Ferrero, Andrea; Badawi, Ramsey D; Qi, Jinyi.

In: Physics in Medicine and Biology, Vol. 59, No. 2, 20.01.2014, p. 403-419.

Research output: Contribution to journalArticle

@article{111cae7d862e4887ae26b2f62347590f,
title = "Regularization design in penalized maximum-likelihood image reconstruction for lesion detection in 3D PET",
abstract = "Detecting cancerous lesions is a major clinical application in emission tomography. In previous work, we have studied penalized maximum-likelihood (PML) image reconstruction for the detection task and proposed a method to design a shift-invariant quadratic penalty function to maximize detectability of a lesion at a known location in a two dimensional image. Here we extend the regularization design to maximize detectability of lesions at unknown locations in fully 3D PET. We used a multiview channelized Hotelling observer (mvCHO) to assess the lesion detectability in 3D images to mimic the condition where a human observer examines three orthogonal views of a 3D image for lesion detection. We derived simplified theoretical expressions that allow fast prediction of the detectability of a 3D lesion. The theoretical results were used to design the regularization in PML reconstruction to improve lesion detectability. We conducted computer-based Monte Carlo simulations to compare the optimized penalty with the conventional penalty for detecting lesions of various sizes. Only true coincidence events were simulated. Lesion detectability was also assessed by two human observers, whose performances agree well with that of the mvCHO. Both the numerical observer and human observer results showed a statistically significant improvement in lesion detection by using the proposed penalty function compared to using the conventional penalty function.",
author = "Li Yang and Jian Zhou and Andrea Ferrero and Badawi, {Ramsey D} and Jinyi Qi",
year = "2014",
month = "1",
day = "20",
doi = "10.1088/0031-9155/59/2/403",
language = "English (US)",
volume = "59",
pages = "403--419",
journal = "Physics in Medicine and Biology",
issn = "0031-9155",
publisher = "IOP Publishing Ltd.",
number = "2",

}

TY - JOUR

T1 - Regularization design in penalized maximum-likelihood image reconstruction for lesion detection in 3D PET

AU - Yang, Li

AU - Zhou, Jian

AU - Ferrero, Andrea

AU - Badawi, Ramsey D

AU - Qi, Jinyi

PY - 2014/1/20

Y1 - 2014/1/20

N2 - Detecting cancerous lesions is a major clinical application in emission tomography. In previous work, we have studied penalized maximum-likelihood (PML) image reconstruction for the detection task and proposed a method to design a shift-invariant quadratic penalty function to maximize detectability of a lesion at a known location in a two dimensional image. Here we extend the regularization design to maximize detectability of lesions at unknown locations in fully 3D PET. We used a multiview channelized Hotelling observer (mvCHO) to assess the lesion detectability in 3D images to mimic the condition where a human observer examines three orthogonal views of a 3D image for lesion detection. We derived simplified theoretical expressions that allow fast prediction of the detectability of a 3D lesion. The theoretical results were used to design the regularization in PML reconstruction to improve lesion detectability. We conducted computer-based Monte Carlo simulations to compare the optimized penalty with the conventional penalty for detecting lesions of various sizes. Only true coincidence events were simulated. Lesion detectability was also assessed by two human observers, whose performances agree well with that of the mvCHO. Both the numerical observer and human observer results showed a statistically significant improvement in lesion detection by using the proposed penalty function compared to using the conventional penalty function.

AB - Detecting cancerous lesions is a major clinical application in emission tomography. In previous work, we have studied penalized maximum-likelihood (PML) image reconstruction for the detection task and proposed a method to design a shift-invariant quadratic penalty function to maximize detectability of a lesion at a known location in a two dimensional image. Here we extend the regularization design to maximize detectability of lesions at unknown locations in fully 3D PET. We used a multiview channelized Hotelling observer (mvCHO) to assess the lesion detectability in 3D images to mimic the condition where a human observer examines three orthogonal views of a 3D image for lesion detection. We derived simplified theoretical expressions that allow fast prediction of the detectability of a 3D lesion. The theoretical results were used to design the regularization in PML reconstruction to improve lesion detectability. We conducted computer-based Monte Carlo simulations to compare the optimized penalty with the conventional penalty for detecting lesions of various sizes. Only true coincidence events were simulated. Lesion detectability was also assessed by two human observers, whose performances agree well with that of the mvCHO. Both the numerical observer and human observer results showed a statistically significant improvement in lesion detection by using the proposed penalty function compared to using the conventional penalty function.

UR - http://www.scopus.com/inward/record.url?scp=84896354249&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84896354249&partnerID=8YFLogxK

U2 - 10.1088/0031-9155/59/2/403

DO - 10.1088/0031-9155/59/2/403

M3 - Article

VL - 59

SP - 403

EP - 419

JO - Physics in Medicine and Biology

JF - Physics in Medicine and Biology

SN - 0031-9155

IS - 2

ER -