Click here to


Are you sure ?

Yes, do it No, cancel

Deformable Image Registration for Dose Mapping and Dose Accumulation: Techniques and Challenges

K Brock1*, M Rosu-Bubulac2*, J Swamidas3*, (1) UT MD Anderson Cancer Center, Houston, TX, (2) Virginia Commonwealth University, Richmond, VA, (3) Department of Medical Physics, Navi Mumbai, Maharashtra


(Wednesday, 7/17/2019) 7:30 AM - 9:30 AM

Room: 304ABC

With recent advances in automation of image registration process, application of adaptive radiation therapy (ART) incorporating images from IGRT, is emerging quickly in the clinical setting. AAPM Task Group No. 132 was formed with the aims of providing guidance on the use, validation, quality assurance (QA), and accuracy assessment of algorithms for registration and fusion of multi-modality medical image datasets. The report includes pitfalls on the use of image registration algorithms under a variety of clinical circumstances. One caveat is associated with dose accumulation using images that contain regions of homogeneous intensity. Intensity-based deformable image registration (DIR) algorithms produce most accurate results in regions of high image contrast, but they can be limited in uniform intensity regions. Large registration uncertainties within these regions will potentially produce large uncertainties in dose estimates to individual voxels. This is of particular concern for ART, where the dose computed using the daily image dataset is typically accumulated onto the reference image dataset (e.g. planning CT) for comparison with the planned dose. Dose computed from displacement vector fields within individual voxels of the tumor and/or OAR can be fraught with uncertainty. Tumor mass loss during RT further obfuscates this issue. How do we perform deformable dose accumulation (DDA) with reasonable levels of confidence under clinical situations, including circumstances where tumors and/or normal organs undergo mass loss? What are practical approaches towards managing DDA using commercially available software?

Learning Objectives:
1. Understand the basics of deformable image registration (DIR), and different algorithms used in the clinical setting, along with limitations and caveats.
2. Learn the main aspects of the AAPM Tg-132 report and the associated recommendations for use of DIR in the clinic.
3. Become familiar with the salient components of dose mapping and dose accumulation, along with practical examples of use of these methods in the clinic.
4. Understand the role of DIR and dose accumulation in the context of HDR treatment of GYN cancers, along with clinical examples and challenges.

Funding Support, Disclosures, and Conflict of Interest: KKB discloses funding from the NIH/NCI, sponsored research funding and licensing agreement with RaySearch Laboratories (Stockholm, Sweden). IJC discloses sponsored research support from Varian Medical Systems (Palo Alto, CA) and Philips Health Care (Best, Netherlands).



Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email