Premium
Development and testing of an improved dosimetry system using a backscatter shielded electronic portal imaging device
Author(s) -
King Brian W.,
Morf Daniel,
Greer Peter. B.
Publication year - 2012
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 - 10.1118/1.4709602
Subject(s) - shielded cable , dosimetry , detector , electromagnetic shielding , medical imaging , image guided radiation therapy , nuclear medicine , pixel , dose profile , backscatter (email) , biomedical engineering , optics , materials science , physics , computer science , medicine , artificial intelligence , wireless , telecommunications , composite material
Purpose: To investigate the properties of a modified backscatter shielded electronic portal imaging device (BSS‐EPID) and to develop a dose model to convert BSS‐EPID images to dose in water as part of an improved system for dosimetry using EPIDs.Methods: The effectiveness of the shielding of the BSS‐EPID was studied by comparing images measured with the BSS‐EPID mounted on the support arm to images measured with the BSS‐EPID removed from the support arm. A dose model was developed and optimized to reconstruct dose in water at different depths from measured BSS‐EPID images. The accuracy of the dose model was studied using BSS‐EPID images of 28 IMRT fields to reconstruct dose in water at depths of 2, 5, 10, and 20 cm and comparing to measured dose in water from a two‐dimensional diode array at the same depths. The ability of the BSS‐EPID system to operate independently of detector position was demonstrated by comparing the dose reconstruction of a 10 × 10 cm 2 field using different detector offsets to that measured by a two‐dimensional diode array.Results: The shielding of the BSS‐EPID was found to be effective, with more than 99% of pixels showing less than 0.5% change due to the presence of the support arm and at most a 0.2% effect on the central axis for 2 × 2 cm 2 fields to fully open 30 × 40 cm 2 images. The dose model was shown to accurately reconstruct measurements of dose in water using BSS‐EPID images with average γ pass rates (2%, 2 mm criteria) of 92.5%, 98.7%, 97.4%, and 97.2% at depths of 2, 5, 10, and 20 cm, respectively, when compared to two‐dimensional diode array measurements. When using 3%, 3 mm γ criteria, the average pass rate was greater than 97% at all depths. Reconstructed dose in water for a 10 × 10 cm 2 field measured with detector offsets as large as 10 cm agreed with each other and two‐dimensional diode array measurements within 0.9%.Conclusions: The modified BSS‐EPID and associated dose model provide an improved system for dosimetry measurements using EPIDs. Several important limitations of the current hardware and software are addressed by this system.