On the dosimetric effect and reduction of inverse consistency and transitivity errors in deformable image registration for dose accumulation Bender, Edward T.; Hardcastle, Nicholas; Tome, Wolfgang A. [Department of Human Oncology, University of Wisconsin-Madison, Madison, Wisconsin 53792 (United States)]; Department of Human Oncology, University of Wisconsin-Madison, Madison, Wisconsin 53792 and Centre for Medical Radiation Physics, University of Wollongong, Wollongong, New South Wales 2522 (Australia); Department of Human Oncology, University of Wisconsin-Madison, Madison, Wisconsin 53792 (United States); Department of Medical Physics, University of Wisconsin-Madison, Madison, Wisconsin 53792 (United States); Department of Biomedical Engineering, University of Wisconsin-Madison, Madison, Wisconsin 53792 and Centre for Medical Radiation Physics, University of Wollongong, Wollongong, New South Wales 2522 (Australia)] 62 RADIOLOGY AND NUCLEAR MEDICINE; 61 RADIATION PROTECTION AND DOSIMETRY; ACCURACY; ALGORITHMS; COMPUTERIZED TOMOGRAPHY; DATA ANALYSIS; DOSIMETRY; IMAGE PROCESSING; LUNGS; RADIATION DOSES; RADIOTHERAPY; SYMMETRY; VECTOR FIELDS Purpose: Deformable image registration (DIR) is necessary for accurate dose accumulation between multiple radiotherapy image sets. DIR algorithms can suffer from inverse and transitivity inconsistencies. When using deformation vector fields (DVFs) that exhibit inverse-inconsistency and are nontransitive, dose accumulation on a given image set via different image pathways will lead to different accumulated doses. The purpose of this study was to investigate the dosimetric effect of and propose a postprocessing solution to reduce inverse consistency and transitivity errors. Methods: Four MVCT images and four phases of a lung 4DCT, each with an associated calculated dose, were selected for analysis. DVFs between all four images in each data set were created using the Fast Symmetric Demons algorithm. Dose was accumulated on the fourth image in each set using DIR via two different image pathways. The two accumulated doses on the fourth image were compared. The inverse consistency and transitivity errors in the DVFs were then reduced. The dose accumulation was repeated using the processed DVFs, the results of which were compared with the accumulated dose from the original DVFs. To evaluate the influence of the postprocessing technique on DVF accuracy, the original and processed DVF accuracy was evaluated on the lung 4DCT data on which anatomical landmarks had been identified by an expert. Results: Dose accumulation to the same image via different image pathways resulted in two different accumulated dose results. After the inverse consistency errors were reduced, the difference between the accumulated doses diminished. The difference was further reduced after reducing the transitivity errors. The postprocessing technique had minimal effect on the accuracy of the DVF for the lung 4DCT images. Conclusions: This study shows that inverse consistency and transitivity errors in DIR have a significant dosimetric effect in dose accumulation; Depending on the image pathway taken to accumulate the dose, different results may be obtained. A postprocessing technique that reduces inverse consistency and transitivity error is presented, which allows for consistent dose accumulation regardless of the image pathway followed. United States 2012-01-15 English Journal Article Journal Name: Medical Physics; Journal Volume: 39; Journal Issue: 1; Other Information: (c) 2012 American Association of Physicists in Medicine; Country of input: International Atomic Energy Agency (IAEA) Medium: X; Size: page(s) 272-280 https://doi.org/10.1118/1.3666948 [] 1 39 Journal ID: ISSN 0094-2405; CODEN: MPHYA6; TRN: US12A7246056536 USN 2013-07-08 22098718