Validation of in House Developed Software Designed Ad Hoc for IMRT QA Analysis Using EPID Based Images
F Salinas Aranda1*, R Sansogne2, V Suarez3, M Franco4, J Escobar5, S Arbiser6, (1) Vidt Centro Medico, Buenos Aires, Buenos Aires, (2) Vidt Centro Medico, Buenos Aires, Buenos Aires, (3) Vidt Centro Medico, Buenos Aires, Buenos Aires, (4) Centro de Radioterapia San Juan SA, San Juan, San Juan, (5) Instituto Privado de Radioterapia Cuyo, Mendoza, Mendoza, (6) Vidt Centro Medico, Buenos Aires, Buenos AiresSU-E-T-17 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose: To validate a recently installed portal dosimetry system that uses an in-house developed analysis application for IMRT QA
Methods: 6 MV X-ray Intensity modulated radiation therapy fields were delivered by a Varian 6EX linear accelerator (Varian Associates, Palo Alto, CA, USA) equipped with Millennium 120 multileaf collimator. An aS1000 Portal Vision electronic portal imaging device (EPID) was installed on the accelerator. In-house gamma analysis software was designed to compare the dosimetryc distributions predicted by the XiO 4.50 treatment planning system (Elekta, Stockholm, Sweden) and those measured by the EPID system. The EPID was calibrated in absolute dose following the vendor recommended procedure. IMRT fields were acquired at a source to detector distance of 105 cm. To validate the results of the in-house software the same IMRT fields were also acquired using a commercial 2D array of 445 n-type diodes system arranged in a 22 cm side cm2 area subdivided in two density regions. A central 10 cm side region containing 221 diodes spaced 7.07 mm and an outer region containing 221 diodes spaced 14.14 mm. Gamma distribution map and percentage of points with gamma <1 using a 3% 3mm criterion were used in both cases as QA pass/fail criteria.
Results: For one given field, gamma distribution maps of both detectors were similar in the inner region of the field but EPID based images showed more discrepancies in the borders of the fields due to higher resolution of the EPID panel. Percentage of points with gamma < 1 was between 97.2 and 99.7 for EPID images and between 96.0 and 100% for 2D diode array.
Conclusion: The software developed ad hoc for EPID based IMRT QA proved to be a valuable and affordable solution. Similar results confirm soundness of the analysis performed by the software and validates its clinical use.