Premium
Remote Cherenkov imaging‐based quality assurance of a magnetic resonance image‐guided radiotherapy system
Author(s) -
Andreozzi Jacqueline M.,
Mooney Karen E.,
Brůža Petr,
Curcuru Austen,
Gladstone David J.,
Pogue Brian W.,
Green Olga
Publication year - 2018
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 - 10.1002/mp.12919
Subject(s) - isocenter , dosimetry , quality assurance , cherenkov radiation , optics , ionization chamber , image quality , physics , nuclear medicine , radiation treatment planning , medical physics , radiation therapy , materials science , computer science , imaging phantom , medicine , ionization , radiology , detector , ion , external quality assessment , pathology , quantum mechanics , artificial intelligence , image (mathematics)
Purpose Tools to perform regular quality assurance of magnetic resonance image‐guided radiotherapy (MRIgRT) systems should ideally be independent of interference from the magnetic fields. Remotely acquired optical Cherenkov imaging‐based dosimetry measurements in water were investigated for this purpose, comparing measures of dose accuracy, temporal dynamics, and overall integrated IMRT delivery. Methods A 40 × 30.5 × 37.5 cm 3 water tank doped with 1 g/L of quinine sulfate was imaged using an intensified charge‐coupled device ( ICCD ) to capture the Cherenkov emission while being irradiated by a commercial MRIgRT system (ViewRay™). The ICCD was placed down‐bore at the end of the couch, 4 m from treatment isocenter and behind the 5‐Gauss line of the 0.35‐T MRI . After establishing optimal camera acquisition settings, square beams of increasing size (4.2 × 4.2 cm 2 , 10.5 × 10.5 cm 2 , and 14.7 × 14.7 cm 2 ) were imaged at 0.93 frames per second, from an individual cobalt‐60 treatment head, to develop projection measures related to percent depth dose ( PDD ) curves and cross beam profiles ( CPB ). These Cherenkov‐derived measurements were compared to ionization chamber ( IC ) and radiographic film dosimetry data, as well as simulation data from the treatment planning system ( TPS ). An intensity‐modulated radiotherapy ( IMRT ) commissioning plan from AAPM TG ‐119 (C4:C‐Shape) was also imaged at 2.1 frames per second, and the single linear sum image from 509 s of plan delivery was compared to the dose volume prediction generated by the TPS using gamma index analysis. Results Analysis of standardized test target images (1024 × 1024 pixels) yielded a pixel resolution of 0.37 mm/pixel. The beam width measured from the Cherenkov image‐generated projection CBP s was within 1 mm accuracy when compared to film measurements for all beams. The 502 point measurements (i.e., pixels) of the Cherenkov image‐based projection percent depth dose curves ( pPDD s) were compared to pPDD s simulated by the treatment planning system ( TPS ), with an overall average error of 0.60%, 0.56%, and 0.65% for the 4.2, 10.5, and 14.7 cm square beams, respectively. The relationships between pPDD s and central axis PDD s derived from the TPS were used to apply a weighting factor to the Cherenkov pPDD , so that the Cherenkov data could be directly compared to IC PDD s (average error of −0.07%, 0.10%, and −0.01% for the same sized beams, respectively). Finally, the composite image of the TG ‐119 C4 treatment plan achieved a 95.1% passing rate using 4%/4 mm gamma index agreement criteria between Cherenkov intensity and TPS dose volume data. Conclusions This is the first examination of Cherenkov‐generated pPDD s and pCBP s in an MR ‐ IGRT system. Cherenkov imaging measurements were fast to acquire, and minimal error was observed overall. Cherenkov imaging also provided novel real‐time data for IMRT QA . The strengths of this imaging are the rapid data capture ability providing real‐time, high spatial resolution data, combined with the remote, noncontact nature of imaging. The biggest limitation of this method is the two‐dimensional (2D) projection‐based imaging of three‐dimensional (3D) dose distributions through the transparent water tank.