4D-CBCT Image Quality Research Articles

TH-CD-303-07: Evaluation of Four Data-Driven Respiratory Motion-Compensation Methods for Four-Dimensional Cone-Beam CT Registration and Reconstruction

Purpose: To develop and compare four 4D-CBCT reconstruction methods that correct for respiratory motion and enhance image quality. Methods: Four motion-compensation workflows were developed employing a combination of deformable image registration (DIR) and image reconstruction. Groupwise registration was used to simultaneously register all frames of a base Four-Dimensional Cone-Beam CT (4D-CBCT) FDK reconstruction to a reference frame, providing a 4D transformation. This 4D transformation expresses a respiratory motion model and was used to apply a deformation during the backprojection operation to create a motion-compensated reconstruction. Two variables were assessed: reference image of registration (fixed or mean frame) and inclusion or exclusion of the motion-compensated reconstruction step. Fixed frame registration refers to using an actual frame from the 4D image (end of inhalation) as the target image. Mean frame registration refers to using a per iteration-estimated average frame image instead. Image quality was assessed in one clinical case through tissue boundary sharpness, noise, and presence of view-aliasing artifact. Results: Diaphragm edge sharpness (DES) was defined as the slope of the intensity gradient along the lung-diaphragm boundary. DES relative to the base 4D-CBCT reconstruction improved by 26.8% using fixed-frame registration alone; 15.5% using fixed-frame with motion-compensated reconstruction; decreased by 2.9% using mean-frame registration alone; and improved by 12.2% using mean-frame with motion-compensated reconstruction. Noise was reduced in soft tissue by 8.7% and 75.8% for the fixed-frame registration and registration with motion-compensation methods, respectively, and by 8.8% and 77.7% for the corresponding mean-frame methods. Reductions in undersampling artifacts were visible for all four methods relative to the base reconstruction while retaining higher frequency spatial details (i.e. blood vessels). Conclusion: A data-driven approach combining groupwise registration and motion-compensation has the potential to improve 4D-CBCT image quality. Adding motion compensation after groupwise registration reduced sharpness somewhat but resulted in reduced noise and structural artifacts. This work was supported by the National Cancer Institute of the National Institutes of Health under Award Number R01CA166119. The authors have no conflicts of interest.

Image quality in thoracic 4D cone‐beam CT: A sensitivity analysis of respiratory signal, binning method, reconstruction algorithm, and projection angular spacing

Respiratory signal, binning method, and reconstruction algorithm are three major controllable factors affecting image quality in thoracic 4D cone-beam CT (4D-CBCT), which is widely used in image guided radiotherapy (IGRT). Previous studies have investigated each of these factors individually, but no integrated sensitivity analysis has been performed. In addition, projection angular spacing is also a key factor in reconstruction, but how it affects image quality is not obvious. An investigation of the impacts of these four factors on image quality can help determine the most effective strategy in improving 4D-CBCT for IGRT. Fourteen 4D-CBCT patient projection datasets with various respiratory motion features were reconstructed with the following controllable factors: (i) respiratory signal (real-time position management, projection image intensity analysis, or fiducial marker tracking), (ii) binning method (phase, displacement, or equal-projection-density displacement binning), and (iii) reconstruction algorithm [Feldkamp-Davis-Kress (FDK), McKinnon-Bates (MKB), or adaptive-steepest-descent projection-onto-convex-sets (ASD-POCS)]. The image quality was quantified using signal-to-noise ratio (SNR), contrast-to-noise ratio, and edge-response width in order to assess noise/streaking and blur. The SNR values were also analyzed with respect to the maximum, mean, and root-mean-squared-error (RMSE) projection angular spacing to investigate how projection angular spacing affects image quality. The choice of respiratory signals was found to have no significant impact on image quality. Displacement-based binning was found to be less prone to motion artifacts compared to phase binning in more than half of the cases, but was shown to suffer from large interbin image quality variation and large projection angular gaps. Both MKB and ASD-POCS resulted in noticeably improved image quality almost 100% of the time relative to FDK. In addition, SNR values were found to increase with decreasing RMSE values of projection angular gaps with strong correlations (r ≈ -0.7) regardless of the reconstruction algorithm used. Based on the authors' results, displacement-based binning methods, better reconstruction algorithms, and the acquisition of even projection angular views are the most important factors to consider for improving thoracic 4D-CBCT image quality. In view of the practical issues with displacement-based binning and the fact that projection angular spacing is not currently directly controllable, development of better reconstruction algorithms represents the most effective strategy for improving image quality in thoracic 4D-CBCT for IGRT applications at the current stage.

Simultaneous motion estimation and image reconstruction (SMEIR) for 4D cone‐beam CT

Image reconstruction and motion model estimation in four-dimensional cone-beam CT (4D-CBCT) are conventionally handled as two sequential steps. Due to the limited number of projections at each phase, the image quality of 4D-CBCT is degraded by view aliasing artifacts, and the accuracy of subsequent motion modeling is decreased by the inferior 4D-CBCT. The objective of this work is to enhance both the image quality of 4D-CBCT and the accuracy of motion model estimation with a novel strategy enabling simultaneous motion estimation and image reconstruction (SMEIR). The proposed SMEIR algorithm consists of two alternating steps: (1) model-based iterative image reconstruction to obtain a motion-compensated primary CBCT (m-pCBCT) and (2) motion model estimation to obtain an optimal set of deformation vector fields (DVFs) between the m-pCBCT and other 4D-CBCT phases. The motion-compensated image reconstruction is based on the simultaneous algebraic reconstruction technique (SART) coupled with total variation minimization. During the forward- and backprojection of SART, measured projections from an entire set of 4D-CBCT are used for reconstruction of the m-pCBCT by utilizing the updated DVF. The DVF is estimated by matching the forward projection of the deformed m-pCBCT and measured projections of other phases of 4D-CBCT. The performance of the SMEIR algorithm is quantitatively evaluated on a 4D NCAT phantom. The quality of reconstructed 4D images and the accuracy of tumor motion trajectory are assessed by comparing with those resulting from conventional sequential 4D-CBCT reconstructions (FDK and total variation minimization) and motion estimation (demons algorithm). The performance of the SMEIR algorithm is further evaluated by reconstructing a lung cancer patient 4D-CBCT. Image quality of 4D-CBCT is greatly improved by the SMEIR algorithm in both phantom and patient studies. When all projections are used to reconstruct a 3D-CBCT by FDK, motion-blurring artifacts are present, leading to a 24.4% relative reconstruction error in the NACT phantom. View aliasing artifacts are present in 4D-CBCT reconstructed by FDK from 20 projections, with a relative error of 32.1%. When total variation minimization is used to reconstruct 4D-CBCT, the relative error is 18.9%. Image quality of 4D-CBCT is substantially improved by using the SMEIR algorithm and relative error is reduced to 7.6%. The maximum error (MaxE) of tumor motion determined from the DVF obtained by demons registration on a FDK-reconstructed 4D-CBCT is 3.0, 2.3, and 7.1 mm along left-right (L-R), anterior-posterior (A-P), and superior-inferior (S-I) directions, respectively. From the DVF obtained by demons registration on 4D-CBCT reconstructed by total variation minimization, the MaxE of tumor motion is reduced to 1.5, 0.5, and 5.5 mm along L-R, A-P, and S-I directions. From the DVF estimated by SMEIR algorithm, the MaxE of tumor motion is further reduced to 0.8, 0.4, and 1.5 mm along L-R, A-P, and S-I directions, respectively. The proposed SMEIR algorithm is able to estimate a motion model and reconstruct motion-compensated 4D-CBCT. The SMEIR algorithm improves image reconstruction accuracy of 4D-CBCT and tumor motion trajectory estimation accuracy as compared to conventional sequential 4D-CBCT reconstruction and motion estimation.

WE-G-134-06: Image Quality in Thoracic 4D Cone-Beam CT: A Sensitivity Analysis of Respiratory Signal Source, Binning Method, and Reconstruction Algorithm

Purpose: Respiratory signal source, binning method, and reconstruction algorithm are three major factors affecting thoracic 4D cone-beam CT (4D-CBCT) image quality. Various studies have investigated each of them individually, but no sensitivity analysis to these factors has been performed. This study quantitatively compares the impacts of the three factors on thoracic 4D-CBCT image quality. Methods: A 4D-CBCT patient projection dataset acquired using audio-visual biofeedback with known degree of baseline shift in respiratory motion was reconstructed with the following factors: (i) respiratory signal source (via real-time position management or projection image based method), (ii) binning method (phase, displacement, or equal-projection-density (EPD) displacement binning), and (iii) reconstruction algorithm (Feldkamp-Davis-Kress (FDK), McKinnon-Bates (MKB), or adaptive-steepest-descent projection-onto-convex-sets (ASD-POCS)). The image quality was quantified using signal-to-noise ratio (SNR), contrast ratio (CR), and edge-response width (ERW) to assess noise, contrast, and sharpness, respectively. Results: Respiratory signal sources had no significant effect on any of the three metrics. Displacement binning was found to produce slightly more inconsistent SNR values owing to unevenly distributed angular views. The binning method otherwise had minimal effects on image quality. For the reconstruction algorithms, MKB resulted in a slightly higher SNR than FDK, while ASD-POCS improved SNR by a factor of three. Conversely, degradation in CR was observed with MKB and ASD-POCS. FDK and MKB produced very similar ERW, while slightly higher values were observed with ASD-POCS, indicating a small degree of blurriness. Conclusion: Respiratory signal sources and binning methods had a negligible impact on image quality, and did not introduce apparent motion artifacts. The MKB and ASD-POCS reconstruction algorithms reduced the noise and streaking but produced poorer image contrast. Of the reconstruction factors tested, reconstruction algorithms play the most important role in improving 4D-CBCT image quality and therefore represent the highest priority for further development. This project is supported by an NHMRC Australia Fellowship, NHMRC project grant 1034060, US NCI P01CA116602, an International Postgraduate Research Scholarships, and an Australian Postgraduate Award.

SU-E-I-14: Improvement of Four Dimensional Cone Beam CT Image Quality with Iterative Reconstruction

Purpose: Four‐dimensional cone beam CT (4D‐CBCT) is well known as powerful modality for image guided radiation therapy. Conventionally, it is reconstructed by sorting the X‐ray projection images into each respiratory phase according to a breathing signal using the FDK algorithm. This usually leads to inadequate number of projections in each phase, resulting in low quality 4D‐CBCT images with obvious streaking artifacts. We tried to improve the image quality for 4D CBCT using iterative reconstruction method (ML‐convex) and compared with the images by FDK in SNR of the reconstructed images. Methods: We used ML‐convex algorithm, which has the ability to reconstruct CBCT images from a few number of noisy x‐ray projections by taking knowledge of x‐ray photon statistics into account. We compared the image quality of 4D CBCT by ML‐convex with that by FDK using projection images acquired with different rotational speeds by XVI 5.0 (Elekta). Projection images were sorted by 10 phases for both reconstruction algorithms. A phantom study and patient study were performed. The convergence rate and reconstruction accuracy were evaluated using Catphan‐600 physical phantom. Results: Image quality of 4D‐CBCT is substantially enhanced by the ML‐convex algorithm. Streaking artifacts, commonly presented in the image reconstructed by FDK algorithm, are almost suppressed in spite of the small number of projections. Quantitative evaluations indicated that, compared with the FDK results, ML‐convex method improves signal‐to‐noise‐ratio (SNR) and reconstructed more accurate CT value. Conclusion: The ML‐convex algorithm yielded images with higher SNR and more accurate CT value than those produced by FDK algorithm. This Result will enable us to apply 4D CBCT by ML‐convex algorithm to Image Guided Adaptive Radiation Therapy and patient daily CBCT‐based treatment localization in real clinical environments with the higher computational efficiency.

High-quality four-dimensional cone-beam CT by deforming prior images

Due to a limited number of projections at each phase, severe view aliasing artifacts are present in four-dimensional cone beam computed tomography (4D-CBCT) when reconstruction is performed using conventional algorithms. In this work, we aim to obtain high-quality 4D-CBCT of lung cancer patients in radiation therapy by deforming the planning CT. The deformation vector fields (DVF) to deform the planning CT are estimated through matching the forward projection of the deformed prior image and measured on-treatment CBCT projection. The estimation of the DVF is formulated as an unconstrained optimization problem, where the objective function to be minimized is the sum of the squared difference between the forward projection of the deformed planning CT and the measured 4D-CBCT projection. A nonlinear conjugate gradient method is used to solve the DVF. As the number of the variables in the DVF is much greater than the number of measurements, the solution to such a highly ill-posed problem is very sensitive to the initials during the optimization process. To improve the estimation accuracy of DVF, we proposed a new strategy to obtain better initials for the optimization. In this strategy, 4D-CBCT is first reconstructed by total variation minimization. Demons deformable registration is performed to register the planning CT and the 4D-CBCT reconstructed by total variation minimization. The resulted DVF from demons registration is then used as the initial parameters in the optimization process. A 4D nonuniform rotational B-spline-based cardiac-torso (NCAT) phantom and a patient 4D-CBCT are used to evaluate the algorithm. Image quality of 4D-CBCT is substantially improved by using the proposed strategy in both NCAT phantom and patient studies. The proposed method has the potential to improve the temporal resolution of 4D-CBCT. Improved 4D-CBCT can better characterize the motion of lung tumors and will be a valuable tool for image-guided adaptive radiation therapy.

High Quality 4-dimensional Cone Beam CT by Deforming Prior Images

High Quality 4-dimensional Cone Beam CT by Deforming Prior Images

Accuracy and inter-observer variability of 3D versus 4D cone-beam CT based image-guidance in SBRT for lung tumors

BackgroundTo analyze the accuracy and inter-observer variability of image-guidance (IG) using 3D or 4D cone-beam CT (CBCT) technology in stereotactic body radiotherapy (SBRT) for lung tumors.Materials and methodsTwenty-one consecutive patients treated with image-guided SBRT for primary and secondary lung tumors were basis for this study. A respiration correlated 4D-CT and planning contours served as reference for all IG techniques. Three IG techniques were performed independently by three radiation oncologists (ROs) and three radiotherapy technicians (RTTs). Image-guidance using respiration correlated 4D-CBCT (IG-4D) with automatic registration of the planning 4D-CT and the verification 4D-CBCT was considered gold-standard. Results were compared with two IG techniques using 3D-CBCT: 1) manual registration of the planning internal target volume (ITV) contour and the motion blurred tumor in the 3D-CBCT (IG-ITV); 2) automatic registration of the planning reference CT image and the verification 3D-CBCT (IG-3D). Image quality of 3D-CBCT and 4D-CBCT images was scored on a scale of 1–3, with 1 being best and 3 being worst quality for visual verification of the IGRT results.ResultsImage quality was scored significantly worse for 3D-CBCT compared to 4D-CBCT: the worst score of 3 was given in 19 % and 7.1 % observations, respectively. Significant differences in target localization were observed between 4D-CBCT and 3D-CBCT based IG: compared to the reference of IG-4D, tumor positions differed by 1.9 mm ± 0.9 mm (3D vector) on average using IG-ITV and by 3.6 mm ± 3.2 mm using IG-3D; results of IG-ITV were significantly closer to the reference IG-4D compared to IG-3D. Differences between the 4D-CBCT and 3D-CBCT techniques increased significantly with larger motion amplitude of the tumor; analogously, differences increased with worse 3D-CBCT image quality scores. Inter-observer variability was largest in SI direction and was significantly larger in IG using 3D-CBCT compared to 4D-CBCT: 0.6 mm versus 1.5 mm (one standard deviation). Inter-observer variability was not different between the three ROs compared to the three RTTs.ConclusionsRespiration correlated 4D-CBCT improves the accuracy of image-guidance by more precise target localization in the presence of breathing induced target motion and by reduced inter-observer variability.

SU‐E‐J‐152: Improving 4D CBCT Image Quality by Using Tumor Trajectory Based Rebinning with Orthogonal Dual Source KV Imaging of the Novel VERO System

Purpose: To improve image quality of 4D CBCT FDK reconstructions by rebinning using 3D tumor trajectories acquired with orthogonal dual source kV imaging. Methods: The Shepp‐Logan in‐silico phantom was used to generate artificial single and dual source projection images during 4D CBCT acquisition. The Shepp‐Logan phantom was modified by adding a 3 cm tissue equivalent sphere inside an air cavity. The sphere was moved inside the phantom by applying six minutes of tumor motion retrieved from a Cyberknife treatment. Images were generated with frame rates of 12 fps and an acquisition speed of 1 degree per second. Through detection of a marker inserted in the moving sphere, the 3D tumor motion trajectory was extracted from the projection data. The projection images were binned either based on the 1D marker motion in the cranial‐caudal direction for 8 phases or based on the 3D marker position using orthogonal images. The 1D rebinning technique was used for both the single and dual source acquisition. The dual source marker projections formed a 3D point cloud representing the tumor track. The track based rebinning used the 175 most proximal marker locations corresponding to 350 projections with minimal intra‐bin residual motion. Results: Differences between single and dual source CBCT images were small, although dual source CBCT used less breathing cycles. The contrast‐to‐noise ratio for the track based reconstruction was significantly higher compared to 1D rebinned reconstructions. An intra‐bin motion of 2.0 ±1.4 mm was found for 1D rebinned dual source images and 1.2 ± 0.4 mm for track based rebinning. Maximum intra‐bin motion after track rebinning was found to be 2.2 ±0.8 mm. Conclusions: Using track‐based rebinning for 4D CBCT in dual source acquisition, intra‐bin residual motion is reduced and image quality was improved for reconstructions with projection images from irregular breathing.

1180 poster EVALUATION OF VIEW INTERPOLATION ALGORITHMS TO IMPROVE 4D-CBCT IMAGE QUALITY

1180 poster EVALUATION OF VIEW INTERPOLATION ALGORITHMS TO IMPROVE 4D-CBCT IMAGE QUALITY

TH-D-BRC-07: Impact of Respiratory Biofeedback On Adaptively Sampled 4D-CBCT Image Quality: Initial Experiences

Purpose: Linac‐mounted 4D‐Cone Beam CT(CBCT)imaging is an important tool for IGRT. Our protocol uses an in‐house built audio‐video (AV) biofeedback device to regularize patient respiration during 4D‐CBCT data acquisition. In this study, the impact of the respiratory stabilization on phase‐correlated 4D‐CBCT image quality is evaluated in our full field‐of‐view (FOV) 4D‐CBCT procedure using the Varian OBI and RPM respiratory monitoring system. Materials and methods: We have implemented a 4D‐CBCT imaging procedure with a 450.0 mm diameter FOV using adaptive acquisition time and projection sampling frequency. An AV respiratory biofeedback device consisting of a LCD visual and audio feedback channels was used. It provides the patient with a real‐time visual and auditory cues whenever the RPM deviates from the selected reference breathing trace. During the 8∼10 minute data acquisition, the patient was instructed to duplicate the reference trace as closely as possible both in terms of displacement and time. The acquired projections were sorted into 10 groups by the associated phase recorded by the RPM system. The effects of the AV biofeedback on the respiratory regularity were studied in terms of average cycle length, baseline variation and displacement. Subsequently, the corresponding 4D‐CBCT images were compared with different acquisition AV modes, such as free breathing, audio only and AV regularization. Results: The patient respiratory track presents larger variation in the free‐breathing data acquisition. The auditory instruction to the patient exhibits control of the displacement but a noticeable baseline drift through the time. The AV biofeedback device improves the reproducibility of respiratory pattern. Conclusion: Our results demonstrated that the AV biofeedback device improved the reproducibility of patient respiratory pattern in our 4D‐CBCT data acquisition. The image quality benefits from the improved projection consistency within an extended time span in every phase bin. Supported by PPG NIH P01 116602.

Deformable Model and Four-Dimensional Reconstruction for Onboard CBCT

Respiration-correlated or 4D cone-beam CT imaging has recently been shown clinically feasible, and highly desirable for target localization in stereotactic radiotherapy of lung cancer. Due to the slow rotation speed of the Linac gantry and limited clinical time, the image quality of current 4DCBCT is significantly inferior to the diagnostic 4DCT. This work investigates a new technique to improve 4DCBCT imaging.

TH‐D‐M100J‐03: High‐Quality Four‐Dimensional CBCT Reconstruction with Virtual Projections

Purpose: Due to significantly reduced number of projections per phase, the quality of 4DCBCT images is often degraded by view‐aliasing artifacts. Acquisitions using slow‐gantry‐rotation or multiple‐gantry‐rotations can improve the 4D images, but at the cost of extra scan time, which may render them clinically impractical. Here we report a new progress in developing high‐quality 4DCBCT without prolonging the acquisition time. Method and Materials: The technique developed here is to reduce the view‐aliasing artifacts resulted from insufficient sampling by properly “borrowing” projections from other phases. To do so, a motion model linking the data of different phases is derived from deformable registrations of coarse (conventional) 4DCBCT phases with consideration of patient's simulation 4DCT. This 4D patient model allows us to properly transform projections from other phases onto the phase under reconstruction, leading to increased sampling in 4DCBCT. The proposed approach is quantitatively evaluated with motion phantoms and two clinical lung cases for a number of metrics, including RMSE of CT numbers, image uniformity, and contrast‐to‐noise‐ratio (CNR). Results: An important finding of this work is that, by the aid of “virtual projection” derived from coarse 4DCBCT and 4D (or breath‐hold 3D) simulation CT images, the 4DCBCT view‐aliasing artifacts can be dramatically reduced, resulting in greatly improved CT‐number accuracy, image uniformity and CNR in all testing cases. Compared with the conventional 4DCBCT, the overall reduction in CT number fluctuation is ∼38% and the CNR increase is ∼65%. Conclusion: Improving the trade‐off between image quality and scan time is the key to making 4D onboard imaging practical and clinically useful. A novel strategy for enhancing 4DCBCT images without increasing scan time and radiation dose has been developed for onboard CBCT imaging system. The method should find valuable applications in patient setup, dose verification in 4D, as well as adaptive radiation therapy in the future.

Optimizing 4D cone-beam CT acquisition protocol for external beam radiotherapy

Four-dimensional cone-beam computed tomography (4D-CBCT) imaging is sensitive to parameters such as gantry rotation speed, number of gantry rotations, X-ray pulse rate, and tube current, as well as a patient's breathing pattern. The aim of this study is to optimize the image acquisition on a patient-specific basis while minimizing the scan time and the radiation dose. More than 60 sets of 4D-CBCT images, each with a temporal resolution of 10 phases, were acquired using multiple-gantry rotation and slow-gantry rotation techniques. The image quality was quantified with a relative root mean-square error (RE) and correlated with various acquisition settings; specifically, varying gantry rotation speed, varying both the rotation speed and the number of rotations, and varying both the rotation speed and tube current to keep the radiation exposure constant. These experiments were repeated for three different respiratory periods. With similar radiation dose, 4D-CBCT images acquired with low current and low rotation speed have better quality over images obtained with high current and high rotation speed. In general, a one-rotation low-speed scan is superior to a two-rotation double-speed scan, even though they provide the same number of projections. Furthermore, it is found that the image quality behaves monotonically with the relative speed as defined by the gantry rotation speed and the patient respiratory period. The RE curves established in this work can be used to predict the 4D-CBCT image quality before a scan. This allows the acquisition protocol to be optimized individually to balance the desired quality with the associated scanning time and patient radiation dose.