Click here to


Are you sure ?

Yes, do it No, cancel

Estimating the Dosimetric Impact of Patient Anatomy Changes with In-Vivo EPID Images, Using the Gradient Dose Segmented Analysis Technique

J Steers, J Zavala Bojorquez, C Bojechko*, University of California San Diego, La Jolla, CA


(Sunday, 7/12/2020)   [Eastern Time (GMT-4)]

Room: AAPM ePoster Library

Purpose: It has been shown that in vivo EPID images can detect changes in patient anatomy. However, for in vivo measurements there is no clear metric that quantitatively estimates the dosimetric effect of anatomy changes. Efforts to link in-vivo gamma metrics to dosimetric impacts have shown weak correlations that are highly dependent on treatment site and dose comparison techniques. We propose a metric to directly give a quantitative estimate of the change in PTV mean due to patient anatomy changes.

Methods: Anatomy changes such as patient shifts, gas bubbles, head-and-neck weight loss and a limb in the treatment field were simulated with a phantom. Both IMRT and VMAT treatment plans were delivered to the phantom with and without the anatomy changes and in-vivo EPID images were recorded and analyzed. In addition, in-vivo EPID images were analyzed for patient cases with observed weight loss, tumor shrinkage and changes in bowel filling. We utilized the gradient dose segmented analysis (GDSA) technique to select the high-dose, low-gradient region of an in-vivo image. Comparing to the first fraction image, the change in the GDSA mean was compared to the change in the PTV mean dose simulated in the treatment planning system.

Results: For anatomy changes simulated in the phantom, excluding shifts, the correlation between the change in GDSA mean and the simulated change in PTV mean dose was 0.95. For the patient cases the correlation between the change in GDSA mean and the simulated change in PTV mean dose was 0.87.

Conclusions: Using the GDSA method gives a metric for in-vivo EPID images that strongly correlates with changes in the PTV mean dose. Using in-vivo images we are able to estimate changes in the PTV mean dose due to anatomy changes which provides a quantitative dosimetric flag that can be used for plan adaptation.

Funding Support, Disclosures, and Conflict of Interest: Research supported by a grant from Varian Medical Systems


Not Applicable / None Entered.


TH- External Beam- Photons: portal dosimetry, in-vivo dosimetry and dose reconstruction

Contact Email