Program Information
Four-Dimensional Dose Calculation Algorithm Using Displacement Vector Fields From Deformable Image Registration
I Ali1*, N Alsbou2 , S Ahmad1 , (1) University of Oklahoma Health Sciences, Oklahoma City, OK, (2) University of Central Oklahoma, Edmond, OK
Presentations
WE-RAM1-GePD-JT-5 (Wednesday, August 2, 2017) 9:30 AM - 10:00 AM Room: Joint Imaging-Therapy ePoster Theater
Purpose: To develop a four-dimensional (4D) dose calculation algorithm that uses displacement-vector-fields (DVF) from deformable-image-registration algorithm (DIR) to modify the three-dimensional (3D) dose distributions considering patient motion.
Methods: The fluence maps of different treatment beams from 3D-conformal or IMRT optimized plans were modified to consider for phantom motion. DVF were used to modify the fluence distributions from the different beams from optimized 3D-conformal or intensity-modulated plans calculated on static simulation CT-images in order to account for phantom motion and dose variations in 4D. Different DIR-algorithms including: Demons, Fast-Demons, Horn-Schunck and Lucas-Kanade from DIRART-software were used to obtain the DVF. The variations in 4D-dose distributions were verified with a mobile phantom with a multiple-array-diode detector (MapCheck2) considering different cyclic motion patterns.
Results: This algorithm produced dose distributions that were optimized in 4D considering motion. The DVF extracted from different DIR-algorithm represented voxel-by-voxel motion which was used to modify the dose distributions optimized in 3D using static CT-images. This algorithm mainly shrunk the optimized fluence distributions in space in order to account accurately for spread-out of the dose induced by motion. This algorithm accounts variation in patient depth by modification of the fluence map in air before calculation in patient. One limitation of this algorithm is accuracy of the DVF in predicting the voxel-by-voxel motion obtained from different DIR-algorithms. DIR algorithms based on solving the optical-flow equation predicted more consistently the controlled motion patterns that were induced in the mobile phantom.
Conclusion: This algorithm provides a new approach to optimize dose calculation in 4D considering patient motion using the DVF obtained from DIR-algorithms. This approach can be employed to manage patient motion as an alternative technique for beam-gating based on motion tracking with external or internal markers. Furthermore, this algorithm has potential application to perform adaptive radiation therapy considering patient anatomical variations.
Contact Email: