Automated beam model optimization

Daniel Ĺtourneau, Michael B. Sharpe, Amir Owrangi, David A. Jaffray

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Purpose: The beam model in a three dimensional treatment planning system (TPS) defines virtually the mechanical and dosimetric characteristics of a treatment unit. The manual optimization of a beam model during commissioning can be a time consuming task due to its iterative nature. Furthermore, the quality of the beam model commissioning depends on the user's ability to manage multiple parameters and assess their impact on the agreement between measured and calculated dose. The objective of this work is to develop and validate the performance of an automated beam model optimization system (ABMOS) based on intensity modulated radiotherapy (IMRT) beam measurements to improve beam model accuracy while streamlining the commissioning process. Methods: The ABMOS was developed to adjust selected TPS beam model parameters iteratively to maximize the agreement between measured and calculated 2D dose maps obtained for an IMRT beam pattern. A 2D diode array with high spatial resolution detectors was used to sample the entire IMRT beam pattern in a single dose measurement. The use of an IMRT beam pattern with large number of monitor units was selected to highlight the difference between planned and delivered dose and improve the signal to noise ratio in the low dose regions. ABMOS was applied to the optimization of a beam model for an Elekta Synergy S treatment unit. The optimized beam model was validated for two anatomical sites (25 paraspinal and 25 prostate cases) using two independent patient-specific IMRT quality control (QC) methods based on ion chamber and 2D diode array measurements, respectively. The conventional approach of comparing calculated and measured beam profiles and percent-depth dose curves was also used to assess improvement in beam model after ABMOS optimization. Elements of statistical process control were applied to the process of patient-specific QC performed with the ion chamber and the 2D array to complement the model comparison. Results: After beam model optimization with ABMOS, improvement in planned to delivered dose agreement was demonstrated with both patient-specific IMRT QC methods and the calculated to measured profile comparison. In terms of ion chamber measurements, the largest improvement was observed for the paraspinal cases with the mean measured to calculated dose difference at the low dose points decreasing from -13.8% to 2.0% with the optimized beam model. The 2D diode array patient-specific QC also demonstrated clearly the improvement in beam model for both paraspinal and prostate cases with, on average, more than 96% of the diodes satisfying tolerances of 3% of dose difference or 2 mm of distance to agreement after ABMOS optimization. The capability index (Cpk) for both patient-specific QC methods also increased with the optimized beam model. Conclusions: In this work, ABMOS was developed to use 2D diode array measurements of an IMRT beam pattern for the automated multivariable optimization of a TPS beam model. Based on the observed improvements in patient-specific QC results for 25 paraspinal and 25 prostate plans, optimization of the remaining clinical beam models using ABMOS is now ongoing in the institution.

Original languageEnglish (US)
Pages (from-to)2110-2120
Number of pages11
JournalMedical Physics
Volume37
Issue number5
DOIs
StatePublished - Jan 1 2010

Fingerprint

Intensity-Modulated Radiotherapy
Quality Control
Prostate
Ions
Therapeutics
Signal-To-Noise Ratio

Keywords

  • Beam model
  • Intensity-modulated radiation therapy
  • Treatment planning system

ASJC Scopus subject areas

  • Biophysics
  • Radiology Nuclear Medicine and imaging

Cite this

Ĺtourneau, D., Sharpe, M. B., Owrangi, A., & Jaffray, D. A. (2010). Automated beam model optimization. Medical Physics, 37(5), 2110-2120. https://doi.org/10.1118/1.3373519

Automated beam model optimization. / Ĺtourneau, Daniel; Sharpe, Michael B.; Owrangi, Amir; Jaffray, David A.

In: Medical Physics, Vol. 37, No. 5, 01.01.2010, p. 2110-2120.

Research output: Contribution to journalArticle

Ĺtourneau, D, Sharpe, MB, Owrangi, A & Jaffray, DA 2010, 'Automated beam model optimization', Medical Physics, vol. 37, no. 5, pp. 2110-2120. https://doi.org/10.1118/1.3373519
Ĺtourneau D, Sharpe MB, Owrangi A, Jaffray DA. Automated beam model optimization. Medical Physics. 2010 Jan 1;37(5):2110-2120. https://doi.org/10.1118/1.3373519
Ĺtourneau, Daniel ; Sharpe, Michael B. ; Owrangi, Amir ; Jaffray, David A. / Automated beam model optimization. In: Medical Physics. 2010 ; Vol. 37, No. 5. pp. 2110-2120.
@article{1e0dfae74e62447d9457bcacc96f6b10,
title = "Automated beam model optimization",
abstract = "Purpose: The beam model in a three dimensional treatment planning system (TPS) defines virtually the mechanical and dosimetric characteristics of a treatment unit. The manual optimization of a beam model during commissioning can be a time consuming task due to its iterative nature. Furthermore, the quality of the beam model commissioning depends on the user's ability to manage multiple parameters and assess their impact on the agreement between measured and calculated dose. The objective of this work is to develop and validate the performance of an automated beam model optimization system (ABMOS) based on intensity modulated radiotherapy (IMRT) beam measurements to improve beam model accuracy while streamlining the commissioning process. Methods: The ABMOS was developed to adjust selected TPS beam model parameters iteratively to maximize the agreement between measured and calculated 2D dose maps obtained for an IMRT beam pattern. A 2D diode array with high spatial resolution detectors was used to sample the entire IMRT beam pattern in a single dose measurement. The use of an IMRT beam pattern with large number of monitor units was selected to highlight the difference between planned and delivered dose and improve the signal to noise ratio in the low dose regions. ABMOS was applied to the optimization of a beam model for an Elekta Synergy S treatment unit. The optimized beam model was validated for two anatomical sites (25 paraspinal and 25 prostate cases) using two independent patient-specific IMRT quality control (QC) methods based on ion chamber and 2D diode array measurements, respectively. The conventional approach of comparing calculated and measured beam profiles and percent-depth dose curves was also used to assess improvement in beam model after ABMOS optimization. Elements of statistical process control were applied to the process of patient-specific QC performed with the ion chamber and the 2D array to complement the model comparison. Results: After beam model optimization with ABMOS, improvement in planned to delivered dose agreement was demonstrated with both patient-specific IMRT QC methods and the calculated to measured profile comparison. In terms of ion chamber measurements, the largest improvement was observed for the paraspinal cases with the mean measured to calculated dose difference at the low dose points decreasing from -13.8{\%} to 2.0{\%} with the optimized beam model. The 2D diode array patient-specific QC also demonstrated clearly the improvement in beam model for both paraspinal and prostate cases with, on average, more than 96{\%} of the diodes satisfying tolerances of 3{\%} of dose difference or 2 mm of distance to agreement after ABMOS optimization. The capability index (Cpk) for both patient-specific QC methods also increased with the optimized beam model. Conclusions: In this work, ABMOS was developed to use 2D diode array measurements of an IMRT beam pattern for the automated multivariable optimization of a TPS beam model. Based on the observed improvements in patient-specific QC results for 25 paraspinal and 25 prostate plans, optimization of the remaining clinical beam models using ABMOS is now ongoing in the institution.",
keywords = "Beam model, Intensity-modulated radiation therapy, Treatment planning system",
author = "Daniel Ĺtourneau and Sharpe, {Michael B.} and Amir Owrangi and Jaffray, {David A.}",
year = "2010",
month = "1",
day = "1",
doi = "10.1118/1.3373519",
language = "English (US)",
volume = "37",
pages = "2110--2120",
journal = "Medical Physics",
issn = "0094-2405",
publisher = "AAPM - American Association of Physicists in Medicine",
number = "5",

}

TY - JOUR

T1 - Automated beam model optimization

AU - Ĺtourneau, Daniel

AU - Sharpe, Michael B.

AU - Owrangi, Amir

AU - Jaffray, David A.

PY - 2010/1/1

Y1 - 2010/1/1

N2 - Purpose: The beam model in a three dimensional treatment planning system (TPS) defines virtually the mechanical and dosimetric characteristics of a treatment unit. The manual optimization of a beam model during commissioning can be a time consuming task due to its iterative nature. Furthermore, the quality of the beam model commissioning depends on the user's ability to manage multiple parameters and assess their impact on the agreement between measured and calculated dose. The objective of this work is to develop and validate the performance of an automated beam model optimization system (ABMOS) based on intensity modulated radiotherapy (IMRT) beam measurements to improve beam model accuracy while streamlining the commissioning process. Methods: The ABMOS was developed to adjust selected TPS beam model parameters iteratively to maximize the agreement between measured and calculated 2D dose maps obtained for an IMRT beam pattern. A 2D diode array with high spatial resolution detectors was used to sample the entire IMRT beam pattern in a single dose measurement. The use of an IMRT beam pattern with large number of monitor units was selected to highlight the difference between planned and delivered dose and improve the signal to noise ratio in the low dose regions. ABMOS was applied to the optimization of a beam model for an Elekta Synergy S treatment unit. The optimized beam model was validated for two anatomical sites (25 paraspinal and 25 prostate cases) using two independent patient-specific IMRT quality control (QC) methods based on ion chamber and 2D diode array measurements, respectively. The conventional approach of comparing calculated and measured beam profiles and percent-depth dose curves was also used to assess improvement in beam model after ABMOS optimization. Elements of statistical process control were applied to the process of patient-specific QC performed with the ion chamber and the 2D array to complement the model comparison. Results: After beam model optimization with ABMOS, improvement in planned to delivered dose agreement was demonstrated with both patient-specific IMRT QC methods and the calculated to measured profile comparison. In terms of ion chamber measurements, the largest improvement was observed for the paraspinal cases with the mean measured to calculated dose difference at the low dose points decreasing from -13.8% to 2.0% with the optimized beam model. The 2D diode array patient-specific QC also demonstrated clearly the improvement in beam model for both paraspinal and prostate cases with, on average, more than 96% of the diodes satisfying tolerances of 3% of dose difference or 2 mm of distance to agreement after ABMOS optimization. The capability index (Cpk) for both patient-specific QC methods also increased with the optimized beam model. Conclusions: In this work, ABMOS was developed to use 2D diode array measurements of an IMRT beam pattern for the automated multivariable optimization of a TPS beam model. Based on the observed improvements in patient-specific QC results for 25 paraspinal and 25 prostate plans, optimization of the remaining clinical beam models using ABMOS is now ongoing in the institution.

AB - Purpose: The beam model in a three dimensional treatment planning system (TPS) defines virtually the mechanical and dosimetric characteristics of a treatment unit. The manual optimization of a beam model during commissioning can be a time consuming task due to its iterative nature. Furthermore, the quality of the beam model commissioning depends on the user's ability to manage multiple parameters and assess their impact on the agreement between measured and calculated dose. The objective of this work is to develop and validate the performance of an automated beam model optimization system (ABMOS) based on intensity modulated radiotherapy (IMRT) beam measurements to improve beam model accuracy while streamlining the commissioning process. Methods: The ABMOS was developed to adjust selected TPS beam model parameters iteratively to maximize the agreement between measured and calculated 2D dose maps obtained for an IMRT beam pattern. A 2D diode array with high spatial resolution detectors was used to sample the entire IMRT beam pattern in a single dose measurement. The use of an IMRT beam pattern with large number of monitor units was selected to highlight the difference between planned and delivered dose and improve the signal to noise ratio in the low dose regions. ABMOS was applied to the optimization of a beam model for an Elekta Synergy S treatment unit. The optimized beam model was validated for two anatomical sites (25 paraspinal and 25 prostate cases) using two independent patient-specific IMRT quality control (QC) methods based on ion chamber and 2D diode array measurements, respectively. The conventional approach of comparing calculated and measured beam profiles and percent-depth dose curves was also used to assess improvement in beam model after ABMOS optimization. Elements of statistical process control were applied to the process of patient-specific QC performed with the ion chamber and the 2D array to complement the model comparison. Results: After beam model optimization with ABMOS, improvement in planned to delivered dose agreement was demonstrated with both patient-specific IMRT QC methods and the calculated to measured profile comparison. In terms of ion chamber measurements, the largest improvement was observed for the paraspinal cases with the mean measured to calculated dose difference at the low dose points decreasing from -13.8% to 2.0% with the optimized beam model. The 2D diode array patient-specific QC also demonstrated clearly the improvement in beam model for both paraspinal and prostate cases with, on average, more than 96% of the diodes satisfying tolerances of 3% of dose difference or 2 mm of distance to agreement after ABMOS optimization. The capability index (Cpk) for both patient-specific QC methods also increased with the optimized beam model. Conclusions: In this work, ABMOS was developed to use 2D diode array measurements of an IMRT beam pattern for the automated multivariable optimization of a TPS beam model. Based on the observed improvements in patient-specific QC results for 25 paraspinal and 25 prostate plans, optimization of the remaining clinical beam models using ABMOS is now ongoing in the institution.

KW - Beam model

KW - Intensity-modulated radiation therapy

KW - Treatment planning system

UR - http://www.scopus.com/inward/record.url?scp=77952374071&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77952374071&partnerID=8YFLogxK

U2 - 10.1118/1.3373519

DO - 10.1118/1.3373519

M3 - Article

C2 - 20527544

AN - SCOPUS:77952374071

VL - 37

SP - 2110

EP - 2120

JO - Medical Physics

JF - Medical Physics

SN - 0094-2405

IS - 5

ER -