Click here to


Are you sure ?

Yes, do it No, cancel

Biomechanical Modeling of Dose-Induced Volumetric Changes of the Parotid Glands for Deformable Image Registration

M McCulloch1*, G Cazoulat1 , A Ford1 , D Polan2 , B Elgohari1 , H Bahig1 , H Elhalawani1 , A Mohamed1 , C Peterson1 , J King1 , A Kim1 , A Ohrt1 , C Fuller1 , S Lai1 , K Brock1 , (1) The University of Texas MD Anderson Cancer Center, Houston, TX, (2) University of Michigan, Ann Arbor, MI


(Thursday, 7/18/2019) 7:30 AM - 9:30 AM

Room: 225BCD

Purpose: Early animal studies suggest that parotid gland (PG) toxicity prediction could be improved by considering the dose to sub-regions of the PG. Translation to clinical investigation requires voxel-level dose accumulation in this organ that responds volumetrically over the course of treatment. To date, deformable image registration (DIR) has been evaluated for the PG using only surface alignment. The goal of this study is to develop and evaluate an advanced DIR technique capable of modeling this complex motion over the course of radiation therapy (RT).

Methods: Planning and mid-treatment MRs from nineteen patients and CTs from ten patients treated for head and neck (HN) cancer with RT were retrospectively evaluated. A finite element model (FEM)-based DIR algorithm was applied between the two images, based on boundary conditions (BCs) on the PG surfaces only (FEM-std). To investigate an improvement in accuracy, a population model-based thermal expansion coefficient was added to simulate the dose distribution effect on the volume change distribution inside the glands (FEM-dbc). Model accuracy was quantified using target registration error (TRE) for MR cases where TRE points could be identified. The potential clinical impact was evaluated using differences in mean dose, median dose, D98 and D50 of the PGs.

Results: In the MR cohort, the average TRE improvement was statistically significantly for the FEM-dbc (p=0.01), with a mean 0.35±0.29mm. In the CT cohort, differences in mean dose, median dose, D98 and D50 of the PGs reached 2.9±0.8, 3.7, 4.1, and 3.8Gy, respectively between FEM-std and FEM-dbc demonstrating the potential for clinically meaningful results if more detailed models are to be pursued.

Conclusion: Differences between FEM-std and FEM-dbc may be impactful when considering high dose gradients in the PGs. The proposed DIR model can allow more accurate PG alignment, aid in better delivered dose estimation, and improve toxicity prediction modeling.

Funding Support, Disclosures, and Conflict of Interest: Dr. Fuller reports grants from Elekta AB, grants from National Institute for Dental and Craniofacial Research and grants from National Cancer Institute; Dr. Brock has a Licensing Agreement with RaySearch Laboratories; Dr. Lai is a Medical consultant for Cardinal Health.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email