Premium
SU‐D‐105‐01: Patient‐Specific Quality Assurance for Monte Carlo‐Calculated Lung SBRT On Cyberknife ‐ Is It Necessary?
Author(s) -
Fabien J,
Zhang Y,
Brindle J,
Dobbins D,
Podder T,
Wessels B
Publication year - 2013
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.4813997
Subject(s) - cyberknife , imaging phantom , nuclear medicine , monte carlo method , radiation treatment planning , quality assurance , medicine , ionization chamber , radiosurgery , computer science , medical physics , radiation therapy , radiology , physics , mathematics , ion , statistics , external quality assessment , pathology , quantum mechanics , ionization
Purpose: In transitioning from ray‐tracing (RT) to Monte Carlo (MC) dose computation algorithm for Cyberknife lung stereotactic body radiotherapy (SBRT), we realized large differences (10–20%) in PTV coverage to heterogeneous target regions which required dosimetric confirmation. Currently, our practice requires an independent second calculation for all plans. MuCheck software (Oncology Data Systems) is used in lieu of physical measurements for RT plans; however, there exists no commercial software for verifying MC plans. We determined all lung SBRT plans should utilize the MC algorithm and initially be confirmed by direct measurements. Methods: Lung treatment plans were first optimized with RT then recalculated using MC for final optimization and high‐resolution dose computation. The MC plan was superimposed on a heterogeneous thorax phantom CT (Standard Imaging 91230) using the phantom overlay tool of the MultiPlan 3.5.2 software. Isodose distributions were manually shifted to place the ion chamber (0.053cc Exradin‐A1SL) in a suitably homogeneous region within the CTV. The plan was then delivered to the thorax phantom with the ion chamber placed in the mediastinum insert location. Results: This methodology was used for 33 consecutive lung patients receiving Cyberknife SBRT with PTVs ranging from 10.7–185.9cm3. The mean deviation between measured and MC calculated doses was −2.31%±1.66%. The maximum deviation was −4.69%. Acceptable tolerance for patient‐specific QA was considered ±5%. Conclusion: The MC algorithm provides improved accuracy over RT for heterogeneous dose calculation, confirmed by direct point measurements. Patient‐specific QA using a heterogeneous lung phantom provided an acceptable anthropomorphic approximation of patient plans calculated with MC. Since the QA results establish satisfactory delivery of MC doses, patient‐specific plans calculated on a homogeneous phantom comparing RT and MC algorithms may provide a suitable second check without direct measurement. Conversely, many users may continue direct measurement as a second check until commercial MC verification software becomes available.