Unencrypted login | home

Program Information

Evaluation of Mechanical Accuracy of Electronic Portal Imaging Devise On Its Use in Patient Specific IMRT QA

J Li

J Li1*, C Burman2, M Chan3, (1) Memorial Sloan-Kettering Cancer Center, Basking Ridge, NJ

SU-E-J-14 Sunday 3:00:00 PM - 6:00:00 PM Room: Exhibit Hall

Purpose: Electronic portal imaging devices (EPID) have been used for both in vivo dosimetry and in vitro dose verification in intensity modulated radiotherapy (IMRT). This study is to investigate the effect of EPID mechanical precision on the accuracy of measured dose distribution.

Methods: EPID energy fluences (dicom images) of H&N IMRT fields were collected daily on two Varian LINACs (Clinac-iX & Trilogy) over 4-week period. The energy fluences were converted to doses using EPIDoseTM (Sun Nuclear Corp). Mechanical deviations of EPIDs could be divided into two components: one with inherent detector center misalignment from the beam central axis, another caused by the 'sagging effect' from gantry rotation. The first component was detected by 'best matching' of the measured and calculated dose at zero gantry angle (G=0). The second component was computed by 'best matching' the 10x10cm field defined by MLC at G=0, 90,180, and 270, separately. A 'shift' was generated by the combination of these two components and then applied to correct the measured dose at the corresponding gantry angle for the IMRT field.

Results: Inherent misalignment of the detector's center and the 'sagging' deviation were found to be 1-2 mm and 1-5 mm, respectively for both LINACs. Each component was found very stable (change < 1mm) over the 4-week observational period. Using a Gamma index of 2%/2mm (DD/DTA), the 'shift' increased the average passing rate from 59% to more than 92%. On the other hand, blindly applying 'auto-shift' from commercially available software to obtain the best match would compound true QA issues with units' misalignments.

Conclusions: A false 'mismatch' between measured and calculated dose distribution caused by mechanical inaccuracies of EPID could be avoided by measuring the two components identified in this study. One should examine the mechanical precision of equipment prior to clinical use of EPID dosimetry.

Contact Email