Details

Autor(en) / Beteiligte

• Bedford, James L.
• Hanson, Ian M.
• Hansen, Vibeke N.

Titel

[OA151] Error sensitivity of forward- and back-projection EPID dosimetry for VMAT treatment of the prostate

Ist Teil von

• Physica medica, 2018-08, Vol.52, p.58-58

Ort / Verlag

Elsevier Ltd

Erscheinungsjahr

2018

Quelle

Access via ScienceDirect (Elsevier)

Beschreibungen/Notizen

• Volumetric modulated arc therapy (VMAT) treatments may be verified dosimetrically using an electronic portal imaging device (EPID), either using the forward-projection method to calculate integrated predicted images, which are compared with acquired images, or using the back-projection method to reconstruct delivered dose, for comparison with the planned dose. This study compares the error sensitivity of both methods for prostate treatments. VMAT treatment plans for prostate and seminal vesicles were created using AutoBeam in-house software, with final dose calculation carried out in Pinnacle3. Patients were treated using the 6 MV beam of a Versa HD accelerator. Portal images were acquired for 46 fractions in 13 patients, and EPID dosimetry was carried out using AutoBeam (forward-projection) and iViewDose (back-projection). Differences between measurements (M) and predictions (P) were evaluated on the central axis of the beam (forward-projection) and at the isocentre (back-projection). The plans were also retrospectively delivered to a Solid Water phantom and errors were deliberately introduced. Errors consisted of a 2–10% increase in monitor units, a 2–10 mm increase in aperture size, a 2–10 mm shift in aperture and introduction of a 10–50 mm air gap into the phantom. For the patient images, M-P = 3.1% ± 2.1% (1 SD) for forward-projection and 2.2% ± 2.2% for back-projection, with good correlation between forward-projection and back-projection. For the phantom images, M-P = 0.7% ± 2.0% for forward-projection and 3.0% ± 1.0% for back-projection. Monitor unit and aperture size errors are detected equally by both methods, aperture shift errors are mainly detected in forward-projection due to removal of shifts by the back-projection image processing software, while air gaps have a slightly larger impact on back-projection than forward-projection. Each method of EPID dosimetry has its own distinct sensitivity to errors, but overall, both methods can be used reliably for verification of delivered dose. The back-projection method, in clinical use at this centre, enables convenient visualisation of delivered dose to the patient. Other methods, such as cone-beam imaging, are required for verification of patient position, while further work is needed to investigate the potential of time-resolved EPID dosimetry.