To improve the accuracy of motion vector fields (MVFs) required for respiratory motion compensated (MoCo)
CT image reconstruction without increasing the computational complexity of the MVF estimation approach,
we propose a MVF upsampling method that is able to reduce the motion blurring in reconstructed 4D images.
While respiratory gating improves the temporal resolution, it leads to sparse view sampling artifacts. MoCo
image reconstruction has the potential to remove all motion artifacts while simultaneously making use of 100%
of the rawdata. However the MVF accuracy is still below the temporal resolution of the CBCT data acquisition.
Increasing the number of motion bins would increase reconstruction time and amplify sparse view artifacts, but
not necessarily the accuracy of MVF. Therefore we propose a new method to upsample estimated MVFs and
use those for MoCo. To estimate the MVFs, a modified version of the Demons algorithm is used. Our proposed
method is able to interpolate the original MVFs up to a factor that each projection has its own individual MVF.
To validate the method we use an artificially deformed clinical CT scan, with a breathing pattern of a real patient,
and patient data acquired with a TrueBeam<sup>TM</sup>4D CBCT system (Varian Medical Systems). We evaluate our
method for different numbers of respiratory bins, each again with different upsampling factors. Employing our
upsampling method, motion blurring in the reconstructed 4D images, induced by irregular breathing and the
limited temporal resolution of phase–correlated images, is substantially reduced.
Until today several algorithms have been developed that reduce or avoid artifacts caused by cardiac and respiratory motion in computed tomography (CT). The motion information is converted into so-called motion vector fields (MVFs) and used for motion compensation (MoCo) during the image reconstruction. To analyze these algorithms quantitatively there is the need for ground truth patient data displaying realistic motion. We developed a method to generate a digital ground truth displaying realistic cardiac and respiratory motion that can be used as a tool to assess MoCo algorithms. By the use of available MoCo methods we measured the motion in CT scans with high spatial and temporal resolution and transferred the motion information onto patient data with different anatomy or imaging modality, thereby reanimating the patient virtually. In addition to these images the ground truth motion information in the form of MVFs is available and can be used to benchmark the MVF estimation of MoCo algorithms. We here applied the method to generate 20 CT volumes displaying detailed cardiac motion that can be used for cone-beam CT (CBCT) simulations and a set of 8 MR volumes displaying respiratory motion. Our method is able to reanimate patient data virtually. In combination with the MVFs it serves as a digital ground truth and provides an improved framework to assess MoCo algorithms.
We propose an adapted method of our previously published five-dimensional (5D) motion compensation (MoCo)
algorithm1, developed for micro-CT imaging of small animals, to provide for the first time motion artifact-free
5D cone-beam CT (CBCT) images from a conventional flat detector-based CBCT scan of clinical patients. Image
quality of retrospectively respiratory- and cardiac-gated volumes from flat detector CBCT scans is deteriorated
by severe sparse projection artifacts. These artifacts further complicate motion estimation, as it is required for
MoCo image reconstruction. For high quality 5D CBCT images at the same x-ray dose and the same number of
projections as todays 3D CBCT we developed a double MoCo approach based on motion vector fields (MVFs)
for respiratory and cardiac motion. In a first step our already published four-dimensional (4D) artifact-specific
cyclic motion-compensation (acMoCo) approach is applied to compensate for the respiratory patient motion.
With this information a cyclic phase-gated deformable heart registration algorithm is applied to the respiratory
motion-compensated 4D CBCT data, thus resulting in cardiac MVFs. We apply these MVFs on double-gated
images and thereby respiratory and cardiac motion-compensated 5D CBCT images are obtained. Our 5D MoCo
approach processing patient data acquired with the TrueBeam 4D CBCT system (Varian Medical Systems). Our
double MoCo approach turned out to be very efficient and removed nearly all streak artifacts due to making use
of 100% of the projection data for each reconstructed frame. The 5D MoCo patient data show fine details and
no motion blurring, even in regions close to the heart where motion is fastest.
We present a new algorithm that allows for raw data-based automated cardiac and respiratory intrinsic gating
in cone-beam CT scans. It can be summarized in three steps: First, a median filter is applied to an initially
reconstructed volume. The forward projection of this volume contains less motion information and is subtracted
from the original projections. This results in new raw data that contain only moving and not static anatomy like
bones, that would otherwise impede the cardiac or respiratory signal acquisition. All further steps are applied to
these modified raw data. Second, the raw data are cropped to a region of interest (ROI). The ROI in the raw data
is determined by the forward projection of a binary volume of interest (VOI) that includes the diaphragm for
respiratory gating and most of the edge of the heart for cardiac gating. Third, the mean gray value in this ROI
is calculated for every projection and the respiratory/cardiac signal is acquired using a bandpass filter. Steps
two and three are carried out simultaneously for 64 or 1440 overlapping VOI inside the body for the respiratory
or cardiac signal respectively. The signals acquired from each ROI are compared and the most consistent one is
chosen as the desired cardiac or respiratory motion signal. Consistency is assessed by the standard deviation of
the time between two maxima. The robustness and efficiency of the method is evaluated using simulated and
measured patient data by computing the standard deviation of the mean signal difference between the ground
truth and the intrinsic signal.