Multi-resolution extended-volume model for iterative reconstruction in cone beam CT

BackgroundIterative reconstruction for cone-beam computed tomography (CBCT) has been applied to improve image quality and reduce radiation dose. In a case where an object’s actual projection is larger than a flat panel detector, CBCT images contain truncated data or incomplete projections, which degrade image quality inside the field of view (FOV). In this work, we propose truncation effect reduction for fast iterative reconstruction in CBCT imaging.MethodsThe volume matrix size of the FOV and the height of projection images were extrapolated to a suitable size. These extended projections were reconstructed by fast iterative reconstruction. Moreover, a smoothing parameter for noise regularization in iterative reconstruction was modified to reduce the accumulated error while processing. The proposed work was evaluated by image quality measurements and compared with conventional filtered backprojection (FBP). To validate the proposed method, we used a head phantom for evaluation and preliminarily tested on a human dataset.ResultsIn the experimental results, the reconstructed images from the head phantom showed enhanced image quality. In addition, fast iterative reconstruction can be run continuously while maintaining a consistent mean-percentage-error value for many iterations. The contrast-to-noise ratio of the soft-tissue images was improved. Visualization of low contrast in the ventricle and soft-tissue images was much improved compared to those from FBP using the same dose index of 5 mGy.ConclusionsOur proposed method showed satisfactory performance to reduce the truncation effect, especially inside the FOV with better image quality for soft-tissue imaging. The convergence of fast iterative reconstruction tends to be stable for many iterations.

C-arm Cone-beam CT: General Principles and Technical Considerations for Use in Interventional Radiology

Purpose: This work is to develop a 3D dictionary learning based cone beam CT (CBCT) reconstruction algorithm on graphic processing units (GPU) to improve the quality of sparse-view CBCT reconstruction with high efficiency. Methods : A 3D dictionary containing 256 small volumes (atoms) of 3 × 3 × 3 was trained from a large number of blocks extracted from a high quality volume image. On the basis, we utilized cholesky decomposition based orthogonal matching pursuit algorithm to find the sparse representation of each block. To accelerate the time-consuming sparse coding in the 3D case, we implemented the sparse coding in a parallel fashion by taking advantage of the tremendous computational power of GPU. Conjugate gradient least square algorithm was adopted to minimize the data fidelity term. Evaluations are performed based on a head-neck patient case. FDK reconstruction with full dataset of 364 projections is used as the reference. We compared the proposed 3D dictionary learning based method with tight frame (TF) by performing reconstructions on a subset data of 121 projections. Results: Compared to TF based CBCT reconstruction that shows good overall performance, our experiments indicated that 3D dictionary learning based CBCT reconstruction is able to recover finer structures, remove more streaking artifacts and also induce less blocky artifacts. Conclusion : 3D dictionary learning based CBCT reconstruction algorithm is able to sense the structural information while suppress the noise, and hence to achieve high quality reconstruction under the case of sparse view. The GPU realization of the whole algorithm offers a significant efficiency enhancement, making this algorithm more feasible for potential clinical application. ------------------------------- Cite this article as: Bai T, Yan H, Shi F, Jia X, Lou Y, Xu Q, Jiang S, Mou X. 3D dictionary learning based iterative cone beam CT reconstruction. Int J Cancer Ther Oncol 2014; 2(2):020240. DOI: 10.14319/ijcto.0202.40

Evaluation of a Novel Metal Artifact Reduction Algorithm for Reconstruction of Cone Beam CT (CBCT) Images Acquired on a New Imaging System

Purpose: For iterative reconstruction (IR) in cone-beam CT (CBCT) imaging, data truncation along the superior-inferior (SI) direction causes severe cone artifacts in the reconstructed CBCT volume images. Not only does it reduce the effective SI coverage of the reconstructed volume, it also hinders the IR algorithm convergence. This is particular a problem for regularization based IR, where smoothing type regularization operations tend to propagate the artifacts to a large area. It is our purpose to develop a practical cone artifacts correction solution. Methods: We found it is the missing data residing in the truncated cone area that leads to inconsistency between the calculated forward projections and measured projections. We overcome this problem by using FDK type reconstruction to estimate the missing data and design weighting factors to compensate the inconsistency caused by the missing data. We validate the proposed methods in our multi-GPU low-dose CBCT reconstruction system on multiple patients' datasets. Results: Compared to the FDK reconstruction with full datasets, while IR is able to reconstruct CBCT images using a subset of projection data, the severe cone artifacts degrade overall image quality. For head-neck case under a full-fan mode, 13 out of 80 slices are contaminated. It is even moremore » severe in pelvis case under half-fan mode, where 36 out of 80 slices are affected, leading to inferior soft-tissue delineation. By applying the proposed method, the cone artifacts are effectively corrected, with a mean intensity difference decreased from ∼497 HU to ∼39HU for those contaminated slices. Conclusion: A practical and effective solution for cone artifacts correction is proposed and validated in CBCT IR algorithm. This study is supported in part by NIH (1R01CA154747-01)« less

Background and objective: In recent years, cone-beam computed tomography (CBCT) has become a key diagnostic tool in dentistry. CBCT can provide 3D images of the maxillofacial area to help dental practitioners in diagnosis and treatment, especially implant placement and treatment of pathogenic lesions. This study aimed to compare the Hounsfield Unit (HU) values obtained from CBCT images for bones of different densities with the corresponding HU values from MDCT images.
 Materials and methods: cube-shaped bone blocks of identical size were cut from the middle section of the cow ribs and femur area such that they had a layer of cortical bone in their buccal, lingual, and top surfaces and trabecular bone in the middle. MDCT scans were performed using a Somatom Sensation Ct Scanner. After determining HU from the results of these scans, nine suitable specimens from different ranges of HU were chosen for comparison. HU of the CBCT images was computed by the dedicated software of the CBCT machine. Finally, HU values obtained from MDCT and CBCT were compared. Data analysis was performed using SPSS version 25 at the 0.05 significance level.
 Results: The results showed a statistically significant difference between the mean HU from MDCT images and the mean HU from CBCT images (P<0.05). For similar specimens, CBCT produced higher mean HU values than MDCT. The Pearson correlation test detected a significant direct relationship between the HU values of specimens in MDCT and CBCT (P<0.05).
 Conclusion: For the tools and software used in this study, there was no significant difference between the HU values obtained from MDCT and CBCT, but the mean HU obtained from CBCT was higher than that from MDCT.

Cone-beam computed tomography (CBCT) is the main imaging tool for image-guided radiotherapy but its functionality is limited by a small imaging volume and restricted image position (imaged at the central instead of the treatment position for peripheral lesions to avoid collisions). In this paper, the authors present the concept of "panoramic CBCT," which can image patients at the treatment position with an imaging volume as large as practically needed. In this novel panoramic CBCT technique, the target is scanned sequentially from multiple view angles. For each view angle, a half scan (180° + θ(cone) where θ(cone) is the cone angle) is performed with the imaging panel positioned in any location along the beam path. The panoramic projection images of all views for the same gantry angle are then stitched together with the direct image stitching method (i.e., according to the reported imaging position) and full-fan, half-scan CBCT reconstruction is performed using the stitched projection images. To validate this imaging technique, the authors simulated cone-beam projection images of the Mathematical Cardiac Torso (MCAT) thorax phantom for three panoramic views. Gaps, repeated/missing columns, and different exposure levels were introduced between adjacent views to simulate imperfect image stitching due to uncertainties in imaging position or output fluctuation. A modified simultaneous algebraic reconstruction technique (modified SART) was developed to reconstruct CBCT images directly from the stitched projection images. As a gold standard, full-fan, full-scan (360° gantry rotation) CBCT reconstructions were also performed using projection images of one imaging panel large enough to encompass the target. Contrast-to-noise ratio (CNR) and geometric distortion were evaluated to quantify the quality of reconstructed images. Monte Carlo simulations were performed to evaluate the effect of scattering on the image quality and imaging dose for both standard and panoramic CBCT. Truncated images with artifacts were observed for the CBCT reconstruction using projection images of the central view only. When the image stitching was perfect, complete reconstruction was obtained for the panoramic CBCT using the modified SART with the image quality similar to the gold standard (full-scan, full-fan CBCT using one large imaging panel). Imperfect image stitching, on the other hand, lead to (streak, line, or ring) reconstruction artifacts, reduced CNR, and/or distorted geometry. Results from Monte Carlo simulations showed that, for identical imaging quality, the imaging dose was lower for the panoramic CBCT than that acquired with one large imaging panel. For the same imaging dose, the CNR of the three-view panoramic CBCT was 50% higher than that of the regular CBCT using one big panel. The authors have developed a panoramic CBCT technique and demonstrated with simulation data that it can image tumors of any location for patients of any size at the treatment position with comparable or less imaging dose and time. However, the image quality of this CBCT technique is sensitive to the reconstruction artifacts caused by imperfect image stitching. Better algorithms are therefore needed to improve the accuracy of image stitching for panoramic CBCT.

Cone beam computed tomography (CBCT) is the current 3D imaging modality of choice in dentistry and is widely used for presurgical implant planning. Implant sites can potentially be evaluated based on their CBCT-derived Hounsfield Units (HU) values, from which clinical decisions, including suitability for implant placement and surgical approach can be made. The current CBCT, however, is inaccurate and unreliable for quantifying the HU, and attempts to correlate CBCT-derived HU with implant primary stability have produced mixed results. This study demonstrated that the novel multisource CBCT (ms-CBCT) enabled by the distributed carbon nanotube (CNT) x-ray source array technology provides a stronger correlation between the cortical bone density of porcine femur approximated by the HU values and the implant primary stability measured by the implant insertion torque, when compared to the conventional single-source CBCT. Twelve planned dental implant sites were imaged with both ms-CBCT and conventional CBCT. The averaged bone HU value was obtained around each site, and insertion torque was measured intraoperatively. Linear regression demonstrated a strong and statistically significant correlation between insertion torque and ms-CBCT-derived HU (R² = 0.86, p = 0.00014), while conventional CBCT showed a weaker but still significant correlation (R² = 0.55, p = 0.0056). These findings suggest that by mitigating x-ray photon scatter and cone beam artifacts, the ms-CBCT derived HU better reflects local bone quality relevant to implant stability, highlighting its potential to improve patient-specific implant planning and improve overall dental implant success.

Background: The measurement of Hounsfield units (HU) during implant treatment planning is important. Currently, various manufacturers' implant planning software programs offer HU capabilities; however, their accuracy remains unverified. In this study, we aimed to validate the accuracy of HU values measured by implant planning software programs. Methods: This study used one type of multidetector computed tomography (MDCT), two types of cone-beam computed tomography (CBCT), and four implant planning software packages. Three specimens were prepared for the evaluation of HUs, and the standard values of the HUs were measured. Digital Imaging and Communications in Medicine (DICOM) data obtained from MDCT and CBCT were loaded into four implant planning software packages to measure the HU values. The HU and reference values of the four implant planning software programs obtained from MDCT and CBCT were compared. Additionally, the HU values between each implant planning software program were compared. Results: The HU values of the three specimens, as measured using the four implant planning software programs utilizing MDCT, did not exhibit a significant difference from the standard values. Conversely, those obtained from CBCT were significantly different. The measured HU values after the MDCT imaging of the specimens were not significantly different between the implant planning software programs; however, they differed after CBCT imaging. Conclusions: The results of this study indicate that it is not possible to measure HU values using CBCT with implant planning software programs. However, HU values can be measured by any implant planning software using MDCT.

AimTo use cone-beam computed tomography (CBCT) images for treatment planning, the Hounsfield unit (HU)-electron density (ED) calibration table for CBCT should be stable. The purpose of this study was to verify the stability of the HU values for the CBCT system over 1 year and to evaluate the effects of variation in HU-ED calibration curves on dose calculation.Materials and MethodsA tissue characterisation phantom was scanned with the field of view (FOV) of size S (FOV-S) and FOV of size M (FOV-M) using the CBCT system once a month for 1 year. A single field treatment plan was constructed on digital phantom images to validate the dose distribution using mean HU-ED calibration curves and possible variations.ResultsHU values for each material rod over the observation period varied with trend. The HU value of the cortical bone rod decreased by about 100 HU for the FOV-S and by about 300 HU for the FOV-M. Possible variation in the HU-ED calibration curves produced a ≤17·9% dose difference in the dose maximum in the treatment plan.ConclusionsThe CBCT system should be calibrated periodically for consistent dose calculation.

A novel electromagnetic tracking configuration was characterized and implemented for image-guided surgery incorporating C-arm fluoroscopy and/or cone-beam CT (CBCT). The tracker employed a field generator (FG) with an open rectangular aperture and a frame enclosure with two essentially hollow sides, yielding a design that presents little or no X-ray attenuation across the C-arm orbit. The "Window" FG (WFG) was characterized in comparison with a conventional "Aurora" FG (AFG), and a configuration in which the WFG was incorporated directly into the operating table was investigated in preclinical phantom studies. The geometric accuracy and field of view (FOV) of the WFG and AFG were evaluated in terms of target registration error (TRE) using an acrylic phantom on an (electromagnetic compatible) experimental bench. The WFG design was incorporated in a prototype operating table featuring a carbon fiber top beneath, which the FG could be translated for positioning under the patient. The X-ray compatibility was evaluated using a prototype mobile C-arm for fluoroscopy and CBCT in an anthropomorphic chest phantom. The susceptibility to EM field distortion associated with surgical tools (e.g., spine screws) and the C-arm itself was investigated in terms of TRE, and calibration methods were tested to provide robust image-world registration with minimal perturbation from the rotational C-arm. The WFG demonstrated mean TRE of 1.28 ± 0.79mm compared to 1.13 ± 0.72mm for the AFG, with no statistically significant difference between the two (p=0.32 and n=250). The WFG exhibited a deeper field of view by ~10cm providing an equivalent degree of geometric accuracy to a depth of z ~55cm, compared to z ~45cm for the AFG. Although the presence of a small number of spine screws did not degrade tracker accuracy, the mobile C-arm perturbed the electromagnetic field sufficiently to degrade TRE; however, a calibration method was identified to mitigate the effect. Specifically, the average calibration between posterior-anterior and lateral orientations of the C-arm was found to yield fairly robust registration for any C-arm pose with only a slight reduction in geometric accuracy (1.43 ± 0.31mm in comparison with 1.28 ± 0.79mm, p=0.05). The WFG demonstrated reasonable X-ray compatibility, although the initial design of the window frame included suboptimal material and shape of the side bars that caused a level of streak artifacts in CBCT reconstructions. The streak artifacts were of sufficient magnitude to degrade soft-tissue visibility in CBCT but were negligible in the context of high-contrast imaging tasks (e.g., bone visualization). The open frame of the WFG offers a potentially valuable configuration for electromagnetic trackers in image-guided surgery applications that are based on X-ray fluoroscopy and/or CBCT. The geometric accuracy and FOV are comparable to the conventional AFG and offers increased depth (z-direction) FOV. Incorporation directly within the operating table offers a streamlined implementation in which the tracker is in place but "invisible," potentially simplifying tableside logistics, avoidance of the sterile field, and compatibility with X-ray imaging.

Purpose: The purpose of this study is to propose a new concept of four-dimensional (4D) cone-beam CT (CBCT) reconstruction for non-periodic organ motion using the Time-ordered Chain Graph Model (TCGM), and to compare the reconstructed results with the previously proposed methods, the total variation-based compressed sensing (TVCS) and prior-image constrained compressed sensing (PICCS). Methods: CBCT reconstruction method introduced in this study consisted of maximum a posteriori (MAP) iterative reconstruction combined with a regularization term derived from a concept of TCGM, which includes a constraint coming from the images of neighbouring time-phases. The time-ordered image series were concurrently reconstructed in the MAP iterative reconstruction framework. Angular range of projections for each time-phase was 90 degrees for TCGM and PICCS, and 200 degrees for TVCS. Two kinds of projection data, an elliptic-cylindrical digital phantom data and two clinical patients’ data, were used for reconstruction. The digital phantom contained an air sphere moving 3 cm along longitudinal axis, and temporal resolution of each method was evaluated by measuring the penumbral width of reconstructed moving air sphere. The clinical feasibility of non-periodic time-ordered 4D CBCT reconstruction was also examined using projection data of prostate cancer patients. Results: The results of reconstructed digital phantom shows that the penumbral widths of TCGM yielded the narrowest result; PICCS and TCGM were 10.6% and 17.4% narrower than that of TVCS, respectively. This suggests that the TCGM has the better temporal resolution than the others. Patients’ CBCT projection data were also reconstructed and all three reconstructed results showed motion of rectal gas and stool. The result of TCGM provided visually clearer and less blurring images. Conclusion: The present study demonstrates that the new concept for 4D CBCT reconstruction, TCGM, combined with MAP iterative reconstruction framework enables time-ordered image reconstruction with narrower time-window.

Purpose: Compressed sensing‐based iterative cone beam CT (CBCT) reconstruction techniques can reconstruct CBCT from under‐sampled noisy projection data, allowing for imaging dose reduction. The long computation time prevents them from clinical applications. Although GPU dramatically improves computational efficiency, the computation time is still too long. The purpose of this project is to develop a reconstruction algorithm on a multi‐GPU platform. Methods: We have developed tight‐frame(TF) based CBCT reconstruction system on a workstation with 4 NVIDIA GTX590 GPUs. The algorithm iterates two steps: a conjugate gradient least square step (CGLS) enforcing projection condition and a regularization step improving image quality through TF domain. The first step involves frequent forward and backward x‐ray projections, which is accelerated by distributing tasks corresponding to different projection angles among GPUs. A parallel‐reduction algorithm is employed to accumulate data at all GPUs. The regularization step is achieved by having each GPU processing a sub‐volume. Boundary‐layer data between sub‐volumes are kept to maintain correct boundary conditions. A half‐fan reweighting technique is also invented to mitigate ring artifacts caused by imperfect scanning geometry. Results: Under the quad‐GPU system, the CGLS step, the regularization step are accelerated by 3.2∼3.6 times and 1.6∼2.6times compared to single‐GPU version, respectively. The overall speed‐up factor is 3.06∼3.51 times. As for the absolute time, it takes 0.41∼3.90 sec per iteration step depending on the image resolution and number of projections. Considering it usually takes about 10 iteration steps for the algorithm to achieve satisfactory image quality, the total reconstruction time ranges from a few seconds to up to 40 seconds. High quality CBCT images have been obtained in our system. The reweighing strategy also removes the ring artifacts in half‐fan cases. Conclusion: A TF‐based CBCT reconstruction on a multi‐GPU platform has been successfully developed. The achieved efficiency and image quality facilitates clinical implementations. This work is supported in part by NIH (1R01CA154747‐01), Varian Medical Systems through a Master Research Agreement, the Early Career Award from Thrasher Research Fund, and the University of California Lab Fees Research Program.

Purpose: Cone-beam computed tomography (CBCT) is the main imaging tool for image-guided radiotherapy but its functionalities is limited by small imaging volume and restricted image position (imaged at the central instead of treatment position for lateral targets to avoid collisions). In this paper we present the concept of “panoramic CBCT” which can image patients at the treatment position with a volume as large as practically needed. Methods: This novel “panoramic CBCT” technique scans the target panoramically for different imager positions, stitch together the panoramic projection images of the same gantry angle to form a larger projection image, and perform CBCT reconstruction using the stitched projection images. To study the potential reconstruction artifacts of this imaging technique, we calculated cone-beam projections of the MCAT thorax phantom for one imager in three adjacent positions. Different gaps and exposure levels were introduced between adjacent imager positions to simulate imperfect stitching. Simultaneous algebraic reconstruction technique (SART) was used to reconstruct CBCT images using the stitched projection images for half-scan (180-degree+cone-anlge/2 gantry rotation). As a gold standard, CBCT reconstruction using projection images big enough to encompass the target were also performed for full scan (360-degree gantry rotation). Results: Incomplete reconstruction with artifacts was observed for reconstruction using projections from the central imager position only. When stitching is perfect, complete reconstruction was obtained from the stitched projection images with quality similar to the gold standard. Streak and ring artifacts were observed when stitching is imperfect and the severity of artifacts increases with the gap size and exposure level difference. Conclusions: Since half-scan can be achieved for most treatment positions without collisions, the proposed “panoramic CBCT” can image tumors of any location for patients of any size at the treatment position. Good image stitching algorithms are needed to eliminate the reconstruction artifacts from imperfect stitching.

Cone beam computed tomography (CBCT) images obtained from linac-based kV imagers are typically used for image-guided radiotherapy, in particular to perform three-dimensional image matching. CBCT image sets can also be used for adaptive radiotherapy where the treatment plan is modified on the basis of periodic imaging throughout the treatment course. CBCT images provide both anatomical information and Hounsfield unit (HU) values, which are required for dose calculations. This study evaluates treatment plans based on CBCT datasets calibrated using the Catphan 504 phantom to investigate the feasibility of using CBCT for adaptive replanning. The CBCT images were acquired from a Varian On-Board Imager system. Conventional planning CT (PCT) images obtained from a Philips Brilliance Big Bore CT scanner were used as reference images. The HU-density calibration curves of CBCT were obtained using a Catphan 504 phantom and a CIRS density phantom and compared with the clinical PCT calibration curve (obtained using the CIRS density phantom). Treatment plans created using the different calibration curves were compared. Identical targets were delineated on CBCT and PCT images on four different-sized phantoms and planar dose maps were generated. The dose-volume histograms of PCT-and CBCT-based plans were compared and evaluated by gamma analysis. To extend the study to a typical clinical situation, two prostate cases were included. The dose distribution comparison between PCT-and CBCT-based plans for patients yielded similar results to those obtained using phantoms. The study also analyzed the effect of phantom dimensions on HU values and its impact on dose calculations. The isodose distributions computed based on PCT and CBCT using the Catphan calibration curve agree to within ± 1% compared to that based on CBCT using the density phantom calibration curve. However, for phantoms of larger diameter, there is a pronounced discrepancy in the 50% and 60% isodose lines, with the dose difference being about ± 3%. For phantoms whose thickness is less than the cone beam scan length (16 cm) and for phantoms whose diameter is less than that of the calibration phantom, the variation in HU values is high. The effect of a change in radial diameter has a larger impact on dose calculations. This study shows that the CIRS density phantom is not suitable for CBCT calibration and that individual calibration curves obtained using phantoms of appropriate dimensions should be used for planning individual treatment sites.