MO‐F‐213CD‐10: PShield: An Algorithm for Optimization of PET/CT Shielding

Research output: Contribution to journalArticle

Abstract

Purpose: Calculation of radiation shielding requirements for high‐workload PET installations using the methods proposed by AAPM TG‐108 can be difficult. The principle challenge that makes PET shielding design more complex than other diagnostic imaging modalities, aside from the higher photon energy, is that it is a multisource problem for which no unique solution exists. The PShield algorithm incorporates three‐dimensional numerical methods to optimize PET shielding and deliver a cost‐optimized solution while making no approximations. Methods: PShield uses a sequential quadratic programming routine to optimize PET shielding by minimizing a cost function in 3‐dimensions using extrapolations of the TG‐108 formulas. PShield makes no approximations and accounts for the contribution of every radiation source to the dose rate at every location in the problem using a discrete mesh. We used two simple examples of shielding problems to compare PShield with the TG‐108 methods. Results: The benefit of applying an optimization routine to an indeterminate problem is the identification of the only solution to the problem that minimizes the desired cost function. Choosing a poorly optimized solution can Result in a shielding design that requires as much as 50% more shielding than an optimized design to reach the same dose rate at a given control point. The increased accuracy afforded by P Shield ensures that dose rates at every point in a control area never exceed the design dose, whereas a reasonable design based on the TG‐108 methods may have hot spots where the dose rate exceeds the design dose by a factor of 2 or more. Conclusions: PShield is an exact three‐dimensional numerical solution for optimal PET shielding which identifies a singular solution which is cost‐optimized. This is especially important for modern PET/CT suites, where increases in scanner capabilities have resulted in more complex shielding problems and the potential for high occupational doses.

Original languageEnglish (US)
Number of pages1
JournalMedical Physics
Volume39
Issue number6
DOIs
StatePublished - Jan 1 2012
Externally publishedYes

Fingerprint

Radiation
Costs and Cost Analysis
Diagnostic Imaging
Photons

All Science Journal Classification (ASJC) codes

  • Biophysics
  • Radiology Nuclear Medicine and imaging

Cite this

MO‐F‐213CD‐10 : PShield: An Algorithm for Optimization of PET/CT Shielding. / Pasciak, Alexander; Jones, A.

In: Medical Physics, Vol. 39, No. 6, 01.01.2012.

Research output: Contribution to journalArticle

@article{45f39a6ae32345e78525f52fe52b66ff,
title = "MO‐F‐213CD‐10: PShield: An Algorithm for Optimization of PET/CT Shielding",
abstract = "Purpose: Calculation of radiation shielding requirements for high‐workload PET installations using the methods proposed by AAPM TG‐108 can be difficult. The principle challenge that makes PET shielding design more complex than other diagnostic imaging modalities, aside from the higher photon energy, is that it is a multisource problem for which no unique solution exists. The PShield algorithm incorporates three‐dimensional numerical methods to optimize PET shielding and deliver a cost‐optimized solution while making no approximations. Methods: PShield uses a sequential quadratic programming routine to optimize PET shielding by minimizing a cost function in 3‐dimensions using extrapolations of the TG‐108 formulas. PShield makes no approximations and accounts for the contribution of every radiation source to the dose rate at every location in the problem using a discrete mesh. We used two simple examples of shielding problems to compare PShield with the TG‐108 methods. Results: The benefit of applying an optimization routine to an indeterminate problem is the identification of the only solution to the problem that minimizes the desired cost function. Choosing a poorly optimized solution can Result in a shielding design that requires as much as 50{\%} more shielding than an optimized design to reach the same dose rate at a given control point. The increased accuracy afforded by P Shield ensures that dose rates at every point in a control area never exceed the design dose, whereas a reasonable design based on the TG‐108 methods may have hot spots where the dose rate exceeds the design dose by a factor of 2 or more. Conclusions: PShield is an exact three‐dimensional numerical solution for optimal PET shielding which identifies a singular solution which is cost‐optimized. This is especially important for modern PET/CT suites, where increases in scanner capabilities have resulted in more complex shielding problems and the potential for high occupational doses.",
author = "Alexander Pasciak and A. Jones",
year = "2012",
month = "1",
day = "1",
doi = "10.1118/1.4735834",
language = "English (US)",
volume = "39",
journal = "Medical Physics",
issn = "0094-2405",
publisher = "AAPM - American Association of Physicists in Medicine",
number = "6",

}

TY - JOUR

T1 - MO‐F‐213CD‐10

T2 - PShield: An Algorithm for Optimization of PET/CT Shielding

AU - Pasciak, Alexander

AU - Jones, A.

PY - 2012/1/1

Y1 - 2012/1/1

N2 - Purpose: Calculation of radiation shielding requirements for high‐workload PET installations using the methods proposed by AAPM TG‐108 can be difficult. The principle challenge that makes PET shielding design more complex than other diagnostic imaging modalities, aside from the higher photon energy, is that it is a multisource problem for which no unique solution exists. The PShield algorithm incorporates three‐dimensional numerical methods to optimize PET shielding and deliver a cost‐optimized solution while making no approximations. Methods: PShield uses a sequential quadratic programming routine to optimize PET shielding by minimizing a cost function in 3‐dimensions using extrapolations of the TG‐108 formulas. PShield makes no approximations and accounts for the contribution of every radiation source to the dose rate at every location in the problem using a discrete mesh. We used two simple examples of shielding problems to compare PShield with the TG‐108 methods. Results: The benefit of applying an optimization routine to an indeterminate problem is the identification of the only solution to the problem that minimizes the desired cost function. Choosing a poorly optimized solution can Result in a shielding design that requires as much as 50% more shielding than an optimized design to reach the same dose rate at a given control point. The increased accuracy afforded by P Shield ensures that dose rates at every point in a control area never exceed the design dose, whereas a reasonable design based on the TG‐108 methods may have hot spots where the dose rate exceeds the design dose by a factor of 2 or more. Conclusions: PShield is an exact three‐dimensional numerical solution for optimal PET shielding which identifies a singular solution which is cost‐optimized. This is especially important for modern PET/CT suites, where increases in scanner capabilities have resulted in more complex shielding problems and the potential for high occupational doses.

AB - Purpose: Calculation of radiation shielding requirements for high‐workload PET installations using the methods proposed by AAPM TG‐108 can be difficult. The principle challenge that makes PET shielding design more complex than other diagnostic imaging modalities, aside from the higher photon energy, is that it is a multisource problem for which no unique solution exists. The PShield algorithm incorporates three‐dimensional numerical methods to optimize PET shielding and deliver a cost‐optimized solution while making no approximations. Methods: PShield uses a sequential quadratic programming routine to optimize PET shielding by minimizing a cost function in 3‐dimensions using extrapolations of the TG‐108 formulas. PShield makes no approximations and accounts for the contribution of every radiation source to the dose rate at every location in the problem using a discrete mesh. We used two simple examples of shielding problems to compare PShield with the TG‐108 methods. Results: The benefit of applying an optimization routine to an indeterminate problem is the identification of the only solution to the problem that minimizes the desired cost function. Choosing a poorly optimized solution can Result in a shielding design that requires as much as 50% more shielding than an optimized design to reach the same dose rate at a given control point. The increased accuracy afforded by P Shield ensures that dose rates at every point in a control area never exceed the design dose, whereas a reasonable design based on the TG‐108 methods may have hot spots where the dose rate exceeds the design dose by a factor of 2 or more. Conclusions: PShield is an exact three‐dimensional numerical solution for optimal PET shielding which identifies a singular solution which is cost‐optimized. This is especially important for modern PET/CT suites, where increases in scanner capabilities have resulted in more complex shielding problems and the potential for high occupational doses.

UR - http://www.scopus.com/inward/record.url?scp=85024809717&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85024809717&partnerID=8YFLogxK

U2 - 10.1118/1.4735834

DO - 10.1118/1.4735834

M3 - Article

AN - SCOPUS:85024809717

VL - 39

JO - Medical Physics

JF - Medical Physics

SN - 0094-2405

IS - 6

ER -