Refereed journal article or data article (A1)
Quality assurance measurements of geometric accuracy for magnetic resonance imaging-based radiotherapy treatment planning
List of Authors: Iiro Ranta, Reko Kemppainen, Jani Keyriläinen, Sami Suilamo, Samuli Heikkinen, Mika Kapanen, Jani Saunavaara
Publisher: Elsevier
Publication year: 2019
Journal: Physica Medica
Journal name in source: Physica Medica
Journal acronym: Phys. Med.
Volume number: 62
Start page: 47
End page: 52
Number of pages: 6
ISSN: 1120-1797
eISSN: 1724-191X
DOI: http://dx.doi.org/10.1016/j.ejmp.2019.04.022
Background:Using magnetic resonance imaging (MRI) as the only imaging method for radiotherapy treatmentplanning (RTP) is becoming more common as MRI-only RTP solutions have evolved. The geometric accuracy ofMR images is an essential factor of image quality when determining the suitability of MRI for RTP. The need istherefore clear for clinically feasible quality assurance (QA) methods for the geometric accuracy measurement.
Materials and methods:This work evaluates long-term stability of geometric accuracy and the validity of a 2Dgeometric accuracy QA method compared to a prototype 3D method and analysis software in routine QA. Thelong-term follow-up measurements were conducted on one of the 1.5 T scanners over a period of 19 monthsusing both methods. Inter-scanner variability of geometric distortions was also evaluated in three 1.5 T and three3 T MRI scanners from a single vendor by using the prototype 3D QA method.
Results:The geometric accuracy of the magnetic resonance for radiotherapy (MR-RT) platform remained stablewithin 2 mm at distances of < 250 mm from isocenter. All scanners achieved good geometric accuracy withmean geometric distortions of < 1 mm at < 150 mm and < 2 mm at < 250 mm from the isocenter. Both mea-surement methods provided relevant information about geometric distortions.
Conclusions:Geometric distortions are often considered a limitation of MRI-only RTP. Results indicate thatgeometric accuracy of modern scanners remain within acceptable limits by default even after many years ofclinical use based on the 3D QA evaluation.