Click here to


Are you sure ?

Yes, do it No, cancel

Four-Dimensional Cone-Beam CT Imaging Using Displacement Vector Fields Extracted From Deformable Image Registration: Phantom Study

N Alsbou1*, S Ahmad2, I Ali2, (1) Department of Engineering and Physics, University of Central Oklahoma, Edmond, OK, (2) University of Oklahoma Health Sciences, Oklahoma City, OK


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

Room: AAPM ePoster Library

Purpose: To extract the motion trajectory from the displacement vector fields (DVF) obtained from deformable image registration (DIR) from the average CT image to reconstruct 4D-cone-beam CT images (CBCT).

Methods: The CBCT images for a thorax mobile phantom with embedded inserts that moved with sinusoidal motion patterns with different amplitudes ranging from 0-20mm and frequencies 0-0.5Hz were acquired using an on-board imager on a Trilogy Varian machine. The CBCT images for the phantom moving with different motion patterns were registered with deformable image registration to the CBCT of the stationary phantom. An algorithm was developed to use the DVFs extracted from deformable image registration to build the motion trajectory of the voxels of the mobile targets in the phantom, which was then used to reconstruct 4D-CBCT images.

Results: 4D-CBCT images were reconstructed from the average CBCT images that covered all phases of the motion cycle. This algorithm eliminated image artifacts induced by sinusoidal cyclic motion in the reconstructed 4D-CBCT images. The DVFs obtained from the deformable image registration were used to reconstruct a motion model on a voxel-by-voxel basis in the mobile phantom. The maximal and minimal shifts calculated from the DFVs correlated with the motion amplitude applied on the mobile phantom. The motion amplitude and frequency were extracted from the distributions of the DVF. The CT-images can be represented on a certain motion phase, which provided more accurate volumes for the tumor and organs at risk, used for treatment planning of cancer patients. The algorithm produced accurate CT number values of the mobile targets in the CT images of the mobile phantom.

Conclusion: The motion artifacts and anatomical changes were reduced significantly in the reconstructed 4D-CBCT from average CBCT images. This approach provides an alternative for conventional 4D-CBCT imaging that is obtained by tracking an external marker.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email