Encrypted login | home

Program Information

Interfraction Treatment Monitoring Using Integrated Invivo EPID Images


D Defoor

D Defoor1*, N Papanikolaou2 , S Stathakis3 , (1) University of Texas HSC SA, New Braunfles, TX, (2) University of Texas HSC SA, San Antonio, TX, (3) Cancer Therapy and Research Center, San Antonio, TX

Presentations

SU-G-TeP4-13 (Sunday, July 31, 2016) 5:30 PM - 6:00 PM Room: ePoster Theater


Purpose: To investigate inter-fraction differences of dose delivery by analyzing portal images acquired during treatment and implement an automated system to generate a report for each fraction. Large differences in images between fractions can alert the physicist of possible machine performance issues or patient set-up errors.

Methods: A Varian Novalis Tx equipped with a HD120 MLC and aS1000 electronic portal imaging device (EPID) was used in our study. EPID images are acquired in continuous acquisition mode for 32 volumetric arc therapy (VMAT) patients. The images are summed to create an image for each arc and a single image for each fraction. The first fraction is designated as the reference unless a machine error prevented acquisition of all images. The images for each beam as well as the fraction image are compared using gamma analysis at 1%/1mm, 2%/2mm and 3%/3mm. A report is then generated using an in house MatLab program containing the comparison for the current fraction as well as a history of previous fractions. The reports are automatically sent via email to the physicist for review. Fractions in which the total number of images was not within 5% of the reference number of images were not included in the results.

Results: 91 of the 182 fractions recorded an image count within 5% of the reference. Gamma averages over all fractions and patients were 96.2% ±0.8% at 3%/3mm, 92.9% ±1% at 2%/2mm and 80.6% ±1.8% at 1%/1mm. The SD between fractions for each patient ranged from .004% to 10.4%. Of the 91 fractions 3 flagged due to low gamma values. After further investigation no significant errors were found.

Conclusion: This toolkit can be used for in-vivo monitoring of treatment plan delivery an alert the physics staff of any inter-fraction discrepancies that may require further investigation.



Contact Email: