Verification of the linac isocenter for stereotactic radiosurgery using cine-EPID imaging and arc delivery

Authors

  • Rowshanfarzad Pejman,

    1. School of Mathematical and Physical Sciences, University of Newcastle, Newcastle, New South Wales 2308, Australia
    Search for more papers by this author
    • a)

      Author to whom correspondence should be addressed. Electronic mail: Pejman.Rowshanfarzad@uon.edu.au; Telephone: + 61249211869; Fax: + 61249211138.

  • Sabet Mahsheed,

    1. School of Mathematical and Physical Sciences, University of Newcastle, Newcastle, New South Wales 2308, Australia
    Search for more papers by this author
  • O’ Connor Daryl J.,

    1. School of Mathematical and Physical Sciences, University of Newcastle, Newcastle, New South Wales 2308, Australia
    Search for more papers by this author
  • Greer Peter B.

    1. Department of Radiation Oncology, Calvary Mater Newcastle Hospital, Newcastle, New South Wales 2310, Australia and School of Mathematical and Physical Sciences, University of Newcastle, Newcastle, New South Wales 2308, Australia
    Search for more papers by this author

Abstract

Purpose:Verification of the mechanical isocenter position is required as part of comprehensive quality assurance programs for stereotactic radiosurgery/radiotherapy (SRS/SRT) treatments. Several techniques have been proposed for this purpose but each of them has certain drawbacks. In this paper, a new efficient and more comprehensive method using cine-EPID images has been introduced for automatic verification of the isocenter with sufficient accuracy for stereotactic applications.Methods: Using a circular collimator fixed to the gantry head to define the field, EPID images of a Winston–Lutz phantom were acquired in cine-imaging mode during 360° gantry rotations. A robust matlab code was developed to analyze the data by finding the center of the field and the center of the ball bearing shadow in each image with sub-pixel accuracy. The distance between these two centers was determined for every image. The method was evaluated by comparison to results of a mechanical pointer and also by detection of a manual shift applied to the phantom position. The repeatability and reproducibility of the method were tested and it was also applied to detect couch and collimator wobble during rotation.Results:The accuracy of the algorithm was 0.03 ± 0.02 mm. The repeatability was less than 3 μm and the reproducibility was less than 86 μm. The time elapsed for the analysis of more than 100 cine images of Varian aS1000 and aS500 EPIDs were ∼65 and 20 s, respectively. Processing of images taken in integrated mode took 0.1 s. The output of the analysis software is printable and shows the isocenter shifts as a function of angle in both in-plane and cross-plane directions. It gives warning messages where the shifts exceed the criteria for SRS/SRT and provides useful data for the necessary adjustments in the system including bearing system and/or room lasers.Conclusions: The comprehensive method introduced in this study uses cine-images, is highly accurate, fast, and independent of the observer. It tests all gantry angles and is suitable for pretreatment QA of the isocenter for stereotactic treatments.

Ancillary