Parallel simulated annealing for emission tomography

K. A. Girodias, Harrison H Barrett, R. L. Shoemaker

Research output: Contribution to journalArticle

16 Citations (Scopus)

Abstract

A method for implementing simulated annealing in parallel to speed up the execution of emission tomography (ET) image reconstruction is presented. A high degree of parallelism can be attained by using a parallel-acceptance partitioning strategy, in which perturbations to subsets of the estimate are evaluated in parallel. However because the point spread function in ET imaging systems is globally dependent, processors cannot update the current estimate independently. Consequently, processors must be synchronized each time a perturbation is accepted to avoid introducing error. This can produce excessive communications overhead, especially when the acceptance rate is high. An energy function is constructed to reduce the synchronization requirements by using a reformulation of the log-likelihood function from the expectation maximization (EM) algorithm. The approach is to change the global dependence in the energy function from the current estimate to the estimate generated during the last iteration. The synchronization requirements for guaranteed convergence are then significantly reduced from once per acceptance to once per iteration. This parallel implementation on 54 Inmos T800 transputers connected in a ring topology resulted in execution times that were almost 50 times faster than on a VAX 8600.

Original languageEnglish (US)
Article number002
Pages (from-to)921-938
Number of pages18
JournalPhysics in Medicine and Biology
Volume36
Issue number7
DOIs
StatePublished - 1991

Fingerprint

simulated annealing
Simulated annealing
Tomography
tomography
acceptability
Likelihood Functions
Computer-Assisted Image Processing
Synchronization
estimates
iteration
central processing units
Transputers
synchronism
Optical transfer function
transputers
Image reconstruction
perturbation
Imaging systems
requirements
point spread functions

ASJC Scopus subject areas

  • Radiological and Ultrasound Technology
  • Radiology Nuclear Medicine and imaging
  • Physics and Astronomy (miscellaneous)
  • Biomedical Engineering

Cite this

Parallel simulated annealing for emission tomography. / Girodias, K. A.; Barrett, Harrison H; Shoemaker, R. L.

In: Physics in Medicine and Biology, Vol. 36, No. 7, 002, 1991, p. 921-938.

Research output: Contribution to journalArticle

Girodias, K. A. ; Barrett, Harrison H ; Shoemaker, R. L. / Parallel simulated annealing for emission tomography. In: Physics in Medicine and Biology. 1991 ; Vol. 36, No. 7. pp. 921-938.
@article{3a27444014f4483f9144fe79784867ab,
title = "Parallel simulated annealing for emission tomography",
abstract = "A method for implementing simulated annealing in parallel to speed up the execution of emission tomography (ET) image reconstruction is presented. A high degree of parallelism can be attained by using a parallel-acceptance partitioning strategy, in which perturbations to subsets of the estimate are evaluated in parallel. However because the point spread function in ET imaging systems is globally dependent, processors cannot update the current estimate independently. Consequently, processors must be synchronized each time a perturbation is accepted to avoid introducing error. This can produce excessive communications overhead, especially when the acceptance rate is high. An energy function is constructed to reduce the synchronization requirements by using a reformulation of the log-likelihood function from the expectation maximization (EM) algorithm. The approach is to change the global dependence in the energy function from the current estimate to the estimate generated during the last iteration. The synchronization requirements for guaranteed convergence are then significantly reduced from once per acceptance to once per iteration. This parallel implementation on 54 Inmos T800 transputers connected in a ring topology resulted in execution times that were almost 50 times faster than on a VAX 8600.",
author = "Girodias, {K. A.} and Barrett, {Harrison H} and Shoemaker, {R. L.}",
year = "1991",
doi = "10.1088/0031-9155/36/7/002",
language = "English (US)",
volume = "36",
pages = "921--938",
journal = "Physics in Medicine and Biology",
issn = "0031-9155",
publisher = "IOP Publishing Ltd.",
number = "7",

}

TY - JOUR

T1 - Parallel simulated annealing for emission tomography

AU - Girodias, K. A.

AU - Barrett, Harrison H

AU - Shoemaker, R. L.

PY - 1991

Y1 - 1991

N2 - A method for implementing simulated annealing in parallel to speed up the execution of emission tomography (ET) image reconstruction is presented. A high degree of parallelism can be attained by using a parallel-acceptance partitioning strategy, in which perturbations to subsets of the estimate are evaluated in parallel. However because the point spread function in ET imaging systems is globally dependent, processors cannot update the current estimate independently. Consequently, processors must be synchronized each time a perturbation is accepted to avoid introducing error. This can produce excessive communications overhead, especially when the acceptance rate is high. An energy function is constructed to reduce the synchronization requirements by using a reformulation of the log-likelihood function from the expectation maximization (EM) algorithm. The approach is to change the global dependence in the energy function from the current estimate to the estimate generated during the last iteration. The synchronization requirements for guaranteed convergence are then significantly reduced from once per acceptance to once per iteration. This parallel implementation on 54 Inmos T800 transputers connected in a ring topology resulted in execution times that were almost 50 times faster than on a VAX 8600.

AB - A method for implementing simulated annealing in parallel to speed up the execution of emission tomography (ET) image reconstruction is presented. A high degree of parallelism can be attained by using a parallel-acceptance partitioning strategy, in which perturbations to subsets of the estimate are evaluated in parallel. However because the point spread function in ET imaging systems is globally dependent, processors cannot update the current estimate independently. Consequently, processors must be synchronized each time a perturbation is accepted to avoid introducing error. This can produce excessive communications overhead, especially when the acceptance rate is high. An energy function is constructed to reduce the synchronization requirements by using a reformulation of the log-likelihood function from the expectation maximization (EM) algorithm. The approach is to change the global dependence in the energy function from the current estimate to the estimate generated during the last iteration. The synchronization requirements for guaranteed convergence are then significantly reduced from once per acceptance to once per iteration. This parallel implementation on 54 Inmos T800 transputers connected in a ring topology resulted in execution times that were almost 50 times faster than on a VAX 8600.

UR - http://www.scopus.com/inward/record.url?scp=0025848013&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0025848013&partnerID=8YFLogxK

U2 - 10.1088/0031-9155/36/7/002

DO - 10.1088/0031-9155/36/7/002

M3 - Article

C2 - 1886927

AN - SCOPUS:0025848013

VL - 36

SP - 921

EP - 938

JO - Physics in Medicine and Biology

JF - Physics in Medicine and Biology

SN - 0031-9155

IS - 7

M1 - 002

ER -