Directional view interpolation for compensation of sparse angular sampling in cone-beam CT

Deep Learning-Based Multi-View Projection Synthesis Approach for Improving the Quality of Sparse-View CBCT in Image-Guided Radiotherapy

While radiation hazards induced by cone-beam computed tomography (CBCT) in image-guided radiotherapy (IGRT) can be reduced by sparse-view sampling, the image quality is inevitably degraded. We propose a deep learning-based multi-view projection synthesis (DLMPS) approach to improve the quality of sparse-view low-dose CBCT images. In the proposed DLMPS approach, linear interpolation was first applied to sparse-view projections and the projections were rearranged into sinograms; these sinograms were processed with a sinogram restoration model and then rearranged back into projections. The sinogram restoration model was modified from the 2D U-Net by incorporating dynamic convolutional layers and residual learning techniques. The DLMPS approach was trained, validated, and tested on CBCT data from 163, 30, and 30 real patients respectively. Sparse-view projection datasets with 1/4 and 1/8 of the original sampling rate were simulated, and the corresponding full-view projection datasets were restored via the DLMPS approach. Tomographic images were reconstructed using the Feldkamp-Davis-Kress algorithm. Quantitative metrics including root-mean-square error (RMSE), peak signal-to-noise ratio (PSNR), structural similarity (SSIM), and feature similarity (FSIM) were calculated in both the projection and image domains to evaluate the performance of the DLMPS approach. The DLMPS approach was compared with 11 state-of-the-art (SOTA) models, including CNN and Transformer architectures. For 1/4 sparse-view reconstruction task, the proposed DLMPS approach achieved averaged RMSE, PSNR, SSIM, and FSIM values of 0.0271, 45.93 dB, 0.9817, and 0.9587 in the projection domain, and 0.000885, 37.63 dB, 0.9074, and 0.9885 in the image domain, respectively. For 1/8 sparse-view reconstruction task, the DLMPS approach achieved averaged RMSE, PSNR, SSIM, and FSIM values of 0.0304, 44.85 dB, 0.9785, and 0.9524 in the projection domain, and 0.001057, 36.05 dB, 0.8786, and 0.9774 in the image domain, respectively. The DLMPS approach outperformed all the 11 SOTA models in both the projection and image domains for 1/4 and 1/8 sparse-view reconstruction tasks. The proposed DLMPS approach effectively improves the quality of sparse-view CBCT images in IGRT by accurately synthesizing missing projections, exhibiting potential in substantially reducing imaging dose to patients with minimal loss of image quality.

Projection Image Synthesis Using Adversarial Learning Based Spatial Transformer Network For Sparse Angle Sampling CT

In sparse-angle X-ray tomography, CT images possess significant noise and artifacts in reconstruction. An important image reconstruction research effort in this area aims to remove these artifacts while preserving image features. Approaches to the problem can be divided into two main categories, post-processing on reconstructed images and pre-processing on the sinogram. Following the rise of deep learning methods, many data-driven models have been proposed in both categories. However, those methods require a large amount of fully sampled data for training. In this paper, inspired by the development of the video frame synthesis technique, we propose an adversarial learning-based spatial transformer network for projection image synthesis, which aims to improve the quality of reconstructed images by reasonably increasing the number of projection images based on the existing projection data. Simulation and experiments have shown that this data-driven model can give rise to competitive results compared with conventional algorithms.

PND-Net: Physics-Inspired Non-Local Dual-Domain Network for Metal Artifact Reduction

Metal artifacts caused by the presence of metallic implants tremendously degrade the quality of reconstructed computed tomography (CT) images and therefore affect the clinical diagnosis or reduce the accuracy of organ delineation and dose calculation in radiotherapy. Although various deep learning methods have been proposed for metal artifact reduction (MAR), most of them aim to restore the corrupted sinogram within the metal trace, which removes beam hardening artifacts but ignores other components of metal artifacts. In this paper, based on the physical property of metal artifacts which is verified via Monte Carlo (MC) simulation, we propose a novel physics-inspired non-local dual-domain network (PND-Net) for MAR in CT imaging. Specifically, we design a novel non-local sinogram decomposition network (NSD-Net) to acquire the weighted artifact component and develop an image restoration network (IR-Net) to reduce the residual and secondary artifacts in the image domain. To facilitate the generalization and robustness of our method on clinical CT images, we employ a trainable fusion network (F-Net) in the artifact synthesis path to achieve unpaired learning. Furthermore, we design an internal consistency loss to ensure the data fidelity of anatomical structures in the image domain and introduce the linear interpolation sinogram as prior knowledge to guide sinogram decomposition. NSD-Net, IR-Net, and F-Net are jointly trained so that they can benefit from one another. Extensive experiments on simulation and clinical data demonstrate that our method outperforms state-of-the-art MAR methods.

Learning CT projection denoising from adjacent views

BACKGROUND

Many learning-based low-dose (LD) computed tomography (CT) imaging methods require large paired full- and low-dose datasets for training, which are usually unavailable in clinic. Whereas models trained on simulated data often face the generalization problem on real clinical data.

PURPOSE

To develop an unsupervised learning technique to acquire clean CT projection from its adjacent LD projections.

METHODS

Given a sequential LD projection set, the method extracts out the middle projection as the target and treats the rest ones as the input. The model is trained with the mean absolute error with proposed inter-view gradient constraint term, which helps to suppress outliers and preserve edges in the denoised projection. The simulated low-dose CT grand challenge dataset and a real physical torso phantom dataset were employed for experiment. The peak signal to noise ratio (PSNR) and structural similarity index measure (SSIM) were calculated for quantitative evaluation.

RESULTS

In experiments with both the simulated and real datasets, visual comparisons reveal that the proposed method obtained images superior to unsupervised and supervised methods working in both image and projection domain. For numerical comparison, our method obtains larger SSIMs than other unsupervised methods at quarter and eighth dose levels. As for PSNR, our method obtains larger value at eighth dose whereas smaller value at quarter dose. The supervised models obtain better numerical results than all unsupervised models on simulated datasets.

CONCLUSION

The proposed method can reduce the noise in CT projections effectively, making it an attractive tool for practical LDCT pre-processing.

A dual-domain neural network based on sinogram synthesis for sparse-view CT reconstruction

OBJECTIVE

The dual-domain deep learning-based reconstruction techniques have enjoyed many successful applications in the field of medical image reconstruction. Applying the analytical reconstruction based operator to transfer the data from the projection domain to the image domain, the dual-domain techniques may suffer from the insufficient suppression or removal of streak artifacts in areas with the missing view data, when addressing the sparse-view reconstruction problems. In this work, to overcome this problem, an intelligent sinogram synthesis based back-projection network (iSSBP-Net) was proposed for sparse-view computed tomography (CT) reconstruction. In the iSSBP-Net method, a convolutional neural network (CNN) was involved in the dual-domain method to inpaint the missing view data in the sinogram before CT reconstruction.

METHODS

The proposed iSSBP-Net method fused a sinogram synthesis sub-network (SS-Net), a sinogram filter sub-network (SF-Net), a back-projection layer, and a post-CNN into an end-to-end network. Firstly, to inpaint the missing view data, the SS-Net employed a CNN to synthesize the full-view sinogram in the projection domain. Secondly, to improve the visual quality of the sparse-view CT images, the synthesized sinogram was filtered by a CNN. Thirdly, the filtered sinogram was brought into the image domain through the back-projection layer. Finally, to yield images of high visual sensitivity, the post-CNN was applied to restore the desired images from the outputs of the back-projection layer.

RESULTS

The numerical experiments demonstrate that the proposed iSSBP-Net is superior to all competing algorithms under different scanning condintions for sparse-view CT reconstruction. Compared to the competing algorithms, the proposed iSSBP-Net method improved the peak signal-to-noise ratio of the reconstructed images about 1.21 dB, 0.26 dB, 0.01 dB, and 0.37 dB under the scanning conditions of 360, 180, 90, and 60 views, respectively.

CONCLUSION

The promising reconstruction results indicate that involving the SS-Net in the dual-domain method is could be an effective manner to suppress or remove the streak artifacts in sparse-view CT images. Due to the promising results reconstructed by the iSSBP-Net method, this study is intended to inspire the further development of sparse-view CT reconstruction by involving a SS-Net in the dual-domain method.

Fluoroscopic 3D Image Generation from Patient-Specific PCA Motion Models Derived from 4D-CBCT Patient Datasets: A Feasibility Study

A method for generating fluoroscopic (time-varying) volumetric images using patient-specific motion models derived from four-dimensional cone-beam CT (4D-CBCT) images was developed. 4D-CBCT images acquired immediately prior to treatment have the potential to accurately represent patient anatomy and respiration during treatment. Fluoroscopic 3D image estimation is performed in two steps: (1) deriving motion models and (2) optimization. To derive motion models, every phase in a 4D-CBCT set is registered to a reference phase chosen from the same set using deformable image registration (DIR). Principal components analysis (PCA) is used to reduce the dimensionality of the displacement vector fields (DVFs) resulting from DIR into a few vectors representing organ motion found in the DVFs. The PCA motion models are optimized iteratively by comparing a cone-beam CT (CBCT) projection to a simulated projection computed from both the motion model and a reference 4D-CBCT phase, resulting in a sequence of fluoroscopic 3D images. Patient datasets were used to evaluate the method by estimating the tumor location in the generated images compared to manually defined ground truth positions. Experimental results showed that the average tumor mean absolute error (MAE) along the superior–inferior (SI) direction and the 95th percentile in two patient datasets were 2.29 and 5.79 mm for patient 1, and 1.89 and 4.82 mm for patient 2. This study demonstrated the feasibility of deriving 4D-CBCT-based PCA motion models that have the potential to account for the 3D non-rigid patient motion and localize tumors and other patient anatomical structures on the day of treatment.

Sinogram + image domain neural network approach for metal artifact reduction in low-dose cone-beam computed tomography

Purpose: Cone-beam computed tomography (CBCT) is commonly used in the operating room to evaluate the placement of surgical implants in relation to critical anatomical structures. A particularly problematic setting, however, is the imaging of metallic implants, where strong artifacts can obscure visualization of both the implant and surrounding anatomy. Such artifacts are compounded when combined with low-dose imaging techniques such as sparse-view acquisition. Approach: This work presents a dual convolutional neural network approach, one operating in the sinogram domain and one in the reconstructed image domain, that is specifically designed for the physics and setting of intraoperative CBCT to address the sources of beam hardening and sparse view sampling that contribute to metal artifacts. The networks were trained with images from cadaver scans with simulated metal hardware. Results: The trained networks were tested on images of cadavers with surgically implanted metal hardware, and performance was compared with a method operating in the image domain alone. While both methods removed most image artifacts, superior performance was observed for the dual-convolutional neural network (CNN) approach in which beam-hardening and view sampling effects were addressed in both the sinogram and image domain. Conclusion: The work demonstrates an innovative approach for eliminating metal and sparsity artifacts in CBCT using a dual-CNN framework which does not require a metal segmentation.

Sparse-view imaging of a fiber internal structure in holographic diffraction tomography via a convolutional neural network

Deep learning has recently shown great potential in computational imaging. Here, we propose a deep-learning-based reconstruction method to realize the sparse-view imaging of a fiber internal structure in holographic diffraction tomography. By taking the sparse-view sinogram as the input and the cross-section image obtained by the dense-view sinogram as the ground truth, the neural network can reconstruct the cross-section image from the sparse-view sinogram. It performs better than the corresponding filtered back-projection algorithm with a sparse-view sinogram, both in the case of simulated data and real experimental data.

Region-Based Segmentation and Wiener Pilot-Based Novel Amoeba Denoising Scheme for CT Imaging

Computed tomography (CT) is one of the most common and beneficial medical imaging schemes, but the associated high radiation dose injurious to the patient is always a concern. Therefore, postprocessing-based enhancement of a CT reconstructed image acquired using a reduced dose is an active research area. Amoeba- (or spatially variant kernel-) based filtering is a strong candidate scheme for postprocessing of the CT image, which adapts its shape according to the image contents. In the reported research work, the amoeba filtering is customized for postprocessing of CT images acquired at a reduced X-ray dose. The proposed scheme modifies both the pilot image formation and amoeba shaping mechanism of the conventional amoeba implementation. The proposed scheme uses a Wiener filter-based pilot image, while region-based segmentation is used for amoeba shaping instead of the conventional amoeba distance-based approach. The merits of the proposed scheme include being more suitable for CT images because of the similar region-based and symmetric nature of the human body anatomy, image smoothing without compromising on the edge details, and being adaptive in nature and more robust to noise. The performance of the proposed amoeba scheme is compared to the traditional amoeba kernel in the image denoising application for CT images using filtered back projection (FBP) on sparse-view projections. The scheme is supported by computer simulations using fan-beam projections of clinically reconstructed and simulated head CT phantoms. The scheme is tested using multiple image quality matrices, in the presence of additive projection noise. The scheme implementation significantly improves the image quality visually and statistically, providing better contrast and image smoothing without compromising on edge details. Promising results indicate the efficacy of the proposed scheme.

Cubic-Spline Interpolation for Sparse-View CT Image Reconstruction With Filtered Backprojection in Dynamic Myocardial Perfusion Imaging

We investigated a projection interpolation method for reconstructing dynamic contrast-enhanced (DCE) heart images from undersampled x-ray projections with filtered backprojecton (FBP). This method may facilitate the application of sparse-view dynamic acquisition for ultralow-dose quantitative computed tomography (CT) myocardial perfusion (MP) imaging. We conducted CT perfusion studies on 5 pigs with a standard full-view acquisition protocol (984 projections). We reconstructed DCE heart images with FBP from all and a quarter of the measured projections evenly distributed over 360°. We interpolated the sparse-view (quarter) projections to a full-view setting using a cubic-spline interpolation method before applying FBP to reconstruct the DCE heart images (synthesized full-view). To generate MP maps, we used 3 sets of DCE heart images, and compared mean MP values and biases among the 3 protocols. Compared with synthesized full-view DCE images, sparse-view DCE images were more affected by streak artifacts arising from projection undersampling. Relative to the full-view protocol, mean bias in MP measurement associated with the sparse-view protocol was 10.0 mL/min/100 g (95%CI: −8.9 to 28.9), which was >3 times higher than that associated with the synthesized full-view protocol (3.3 mL/min/100 g, 95% CI: −6.7 to 13.2). The cubic-spline-view interpolation method improved MP measurement from DCE heart images reconstructed from only a quarter of the full projection set. This method can be used with the industry-standard FBP algorithm to reconstruct DCE images of the heart, and it can reduce the radiation dose of a whole-heart quantitative CT MP study to <2 mSv (at 8-cm coverage).

Quantifying day-to-day variations in 4DCBCT-based PCA motion models

The aim of this paper is to quantify the day-to-day variations of motion models derived from pre-treatment 4-dimensional cone beam CT (4DCBCT) fractions for lung cancer stereotactic body radiotherapy (SBRT) patients. Motion models are built by (1) applying deformable image registration (DIR) on each 4DCBCT image with respect to a reference image from that day, resulting in a set of displacement vector fields (DVFs), and (2) applying principal component analysis (PCA) on the DVFs to obtain principal components representing a motion model. Variations were quantified by comparing the PCA eigenvectors of the motion model built from the first day of treatment to the corresponding eigenvectors of the other motion models built from each successive day of treatment. Three metrics were used to quantify the variations: root mean squared (RMS) difference in the vectors, directional similarity, and an introduced metric called the Euclidean Model Norm (EMN). EMN quantifies the degree to which a motion model derived from the first fraction can represent the motion models of subsequent fractions. Twenty-one 4DCBCT scans from five SBRT patient treatments were used in this retrospective study. Experimental results demonstrated that the first two eigenvectors of motion models across all fractions have smaller RMS (0.00017), larger directional similarity (0.528), and larger EMN (0.678) than the last three eigenvectors (RMS: 0.00025, directional similarity: 0.041, and EMN: 0.212). The study concluded that, while the motion model eigenvectors varied from fraction to fraction, the first few eigenvectors were shown to be more stable across treatment fractions than others. This supports the notion that a pre-treatment motion model built from the first few PCA eigenvectors may remain valid throughout a treatment course. Future work is necessary to quantify how day-to-day variations in these models will affect motion reconstruction accuracy for specific clinical tasks.

Sparse-view tomography via displacement function interpolation

Sparse-view tomography has many applications such as in low-dose computed tomography (CT). Using under-sampled data, a perfect image is not expected. The goal of this paper is to obtain a tomographic image that is better than the naïve filtered backprojection (FBP) reconstruction that uses linear interpolation to complete the measurements. This paper proposes a method to estimate the un-measured projections by displacement function interpolation. Displacement function estimation is a non-linear procedure and the linear interpolation is performed on the displacement function (instead of, on the sinogram itself). As a result, the estimated measurements are not the linear transformation of the measured data. The proposed method is compared with the linear interpolation methods, and the proposed method shows superior performance.

[Sparse-view helical CT reconstruction based on tensor total generalized variation minimization]

OBJECTIVE

We propose a sparse-view helical CT iterative reconstruction algorithm based on projection of convex set tensor total generalized variation minimization (TTGV-POCS) to reduce the X-ray dose of helical CT scanning.

METHODS

The three-dimensional volume data of helical CT reconstruction was viewed as the third-order tensor. The tensor generalized total variation (TTGV) was used to describe the structural sparsity of the three-dimensional image. The POCS iterative reconstruction framework was adopted to achieve a robust result of sparse-view helical CT reconstruction. The TTGV-POCS algorithm fully used the structural sparsity of first-order and second-order derivation and the correlation between the slices of helical CT image data to effectively suppress artifacts and noise in the image of sparse-view reconstruction and better preserve image edge information.

RESULTS

The experimental results of XCAT phantom and patient scan data showed that the TTGVPOCS algorithm had better performance in reducing noise, removing artifacts and maintaining edges than the existing reconstruction algorithms. Comparison of the sparse-view reconstruction results of XCAT phantom data with 144 exposure views showed that the TTGV-POCS algorithm proposed herein increased the PSNR quantitative index by 9.17%-15.24% compared with the experimental comparison algorithm; the FSIM quantitative index was increased by 1.27%-9.30%.

CONCLUSIONS

The TTGV-POCS algorithm can effectively improve the image quality of helical CT sparse-view reconstruction and reduce the radiation dose of helical CT examination to improve the clinical imaging diagnosis.

A convolutional neural network for fast upsampling of undersampled tomograms in X-ray CT time-series using a representative highly sampled tomogram

X-ray computed tomography and, specifically, time-resolved volumetric tomography data collections (4D datasets) routinely produce terabytes of data, which need to be effectively processed after capture. This is often complicated due to the high rate of data collection required to capture at sufficient time-resolution events of interest in a time-series, compelling the researchers to perform data collection with a low number of projections for each tomogram in order to achieve the desired `frame rate'. It is common practice to collect a representative tomogram with many projections, after or before the time-critical portion of the experiment without detrimentally affecting the time-series to aid the analysis process. For this paper these highly sampled data are used to aid feature detection in the rapidly collected tomograms by assisting with the upsampling of their projections, which is equivalent to upscaling the θ-axis of the sinograms. In this paper, a super-resolution approach is proposed based on deep learning (termed an upscaling Deep Neural Network, or UDNN) that aims to upscale the sinogram space of individual tomograms in a 4D dataset of a sample. This is done using learned behaviour from a dataset containing a high number of projections, taken of the same sample and occurring at the beginning or the end of the data collection. The prior provided by the highly sampled tomogram allows the application of an upscaling process with better accuracy than existing interpolation techniques. This upscaling process subsequently permits an increase in the quality of the tomogram's reconstruction, especially in situations that require capture of only a limited number of projections, as is the case in high-frequency time-series capture. The increase in quality can prove very helpful for researchers, as downstream it enables easier segmentation of the tomograms in areas of interest, for example. The method itself comprises a convolutional neural network which through training learns an end-to-end mapping between sinograms with a low and a high number of projections. Since datasets can differ greatly between experiments, this approach specifically develops a lightweight network that can easily and quickly be retrained for different types of samples. As part of the evaluation of our technique, results with different hyperparameter settings are presented, and the method has been tested on both synthetic and real-world data. In addition, accompanying real-world experimental datasets have been released in the form of two 80 GB tomograms depicting a metallic pin that undergoes corruption from a droplet of salt water. Also a new engineering-based phantom dataset has been produced and released, inspired by the experimental datasets.

Few-view CT reconstruction via a novel non-local means algorithm

PURPOSE

Non-local means (NLM) based reconstruction method is a promising algorithm for few-view computed tomography (CT) reconstruction, but often suffers from over-smoothed image edges. To address this problem, an adaptive NLM reconstruction method based on rotational invariance (ART-RIANLM) is proposed.

METHODS

The method consists of four steps: 1) Initializing parameters; 2) ART reconstruction using raw data; 3) Positivity constraint of the reconstructed image; 4) Image updating by RIANLM filtering. In RIANLM, two kinds of rotational invariance measures which are average gradient (AG) and region homogeneity (RH) are proposed to calculate the distance between two patches and a novel NLM filter is developed to avoid over-smoothed image. Moreover, the parameter h in RIANLM which controls the decay of the weights is adaptive to avoid over-smoothness, while it is constant in NLM during the whole reconstruction process. The proposed method is validated on two digital phantoms and real projection data.

RESULTS

In our experiments, the searching neighborhood size is set as 15×15 and the similarity window is set as 3×3. For the simulated case of Shepp-Logan phantom, ART-RIANLM produces higher SNR (36.23dB>24.00dB) and lower MAE (0.0006<0.0024) reconstructed images than ART-NLM. The visual inspection demonstrated that the proposed method could suppress artifacts or noises more effectively and recover image edges better. The result of real data case is also consistent with the simulation result.

CONCLUSIONS

A RIANLM based reconstruction method for few-view CT is presented. Compared to the traditional ART-NLM method, SNR and MAE from ART-RIANLM increases 51% and decreases 75%, respectively.

Volumetric CT with sparse detector arrays (and application to Si-strip photon counters)

Novel x-ray medical imaging sensors, such as photon counting detectors (PCDs) and large area CCD and CMOS cameras can involve irregular and/or sparse sampling of the detector plane. Application of such detectors to CT involves undersampling that is markedly different from the commonly considered case of sparse angular sampling. This work investigates volumetric sampling in CT systems incorporating sparsely sampled detectors with axial and helical scan orbits and evaluates performance of model-based image reconstruction (MBIR) with spatially varying regularization in mitigating artifacts due to sparse detector sampling. Volumetric metrics of sampling density and uniformity were introduced. Penalized-likelihood MBIR with a spatially varying penalty that homogenized resolution by accounting for variations in local sampling density (i.e. detector gaps) was evaluated. The proposed methodology was tested in simulations and on an imaging bench based on a Si-strip PCD (total area 5 cm  ×  25 cm) consisting of an arrangement of line sensors separated by gaps of up to 2.5 mm. The bench was equipped with translation/rotation stages allowing a variety of scanning trajectories, ranging from a simple axial acquisition to helical scans with variable pitch. Statistical (spherical clutter) and anthropomorphic (hand) phantoms were considered. Image quality was compared to that obtained with a conventional uniform penalty in terms of structural similarity index (SSIM), image uniformity, spatial resolution, contrast, and noise. Scan trajectories with intermediate helical width (~10 mm longitudinal distance per 360° rotation) demonstrated optimal tradeoff between the average sampling density and the homogeneity of sampling throughout the volume. For a scan trajectory with 10.8 mm helical width, the spatially varying penalty resulted in significant visual reduction of sampling artifacts, confirmed by a 10% reduction in minimum SSIM (from 0.88 to 0.8) and a 40% reduction in the dispersion of SSIM in the volume compared to the constant penalty (both penalties applied at optimal regularization strength). Images of the spherical clutter and wrist phantoms confirmed the advantages of the spatially varying penalty, showing a 25% improvement in image uniformity and 1.8  ×  higher CNR (at matched spatial resolution) compared to the constant penalty. The studies elucidate the relationship between sampling in the detector plane, acquisition orbit, sampling of the reconstructed volume, and the resulting image quality. They also demonstrate the benefit of spatially varying regularization in MBIR for scenarios with irregular sampling patterns. Such findings are important and integral to the incorporation of a sparsely sampled Si-strip PCD in CT imaging.

3D fluoroscopic image estimation using patient-specific 4DCBCT-based motion models

3D fluoroscopic images represent volumetric patient anatomy during treatment with high spatial and temporal resolution. 3D fluoroscopic images estimated using motion models built using 4DCT images, taken days or weeks prior to treatment, do not reliably represent patient anatomy during treatment. In this study we developed and performed initial evaluation of techniques to develop patient-specific motion models from 4D cone-beam CT (4DCBCT) images, taken immediately before treatment, and used these models to estimate 3D fluoroscopic images based on 2D kV projections captured during treatment. We evaluate the accuracy of 3D fluoroscopic images by comparison to ground truth digital and physical phantom images. The performance of 4DCBCT-based and 4DCT-based motion models are compared in simulated clinical situations representing tumor baseline shift or initial patient positioning errors. The results of this study demonstrate the ability for 4DCBCT imaging to generate motion models that can account for changes that cannot be accounted for with 4DCT-based motion models. When simulating tumor baseline shift and patient positioning errors of up to 5 mm, the average tumor localization error and the 95th percentile error in six datasets were 1.20 and 2.2 mm, respectively, for 4DCBCT-based motion models. 4DCT-based motion models applied to the same six datasets resulted in average tumor localization error and the 95th percentile error of 4.18 and 5.4 mm, respectively. Analysis of voxel-wise intensity differences was also conducted for all experiments. In summary, this study demonstrates the feasibility of 4DCBCT-based 3D fluoroscopic image generation in digital and physical phantoms and shows the potential advantage of 4DCBCT-based 3D fluoroscopic image estimation when there are changes in anatomy between the time of 4DCT imaging and the time of treatment delivery.

Metal artifact reduction in CT using fusion based prior image

PURPOSE

In computed tomography, metallic objects in the scanning field create the so-called metal artifacts in the reconstructed images. Interpolation-based methods for metal artifact reduction (MAR) replace the metal-corrupted projection data with surrogate data obtained from interpolation using the surrounding uncorrupted sinogram information. Prior-based MAR methods further improve interpolation-based methods by better estimating the surrogate data using forward projections from a prior image. However, the prior images in most existing prior-based methods are obtained from segmented images and misclassification in segmentation often leads to residual artifacts and tissue structure loss in the final corrected images. To overcome these drawbacks, the authors propose a fusion scheme, named fusion prior-based MAR (FP-MAR).

METHODS

The FP-MAR method consists of (i) precorrect the image by means of an interpolation-based MAR method and an edge-preserving blur filter; (ii) generate a prior image from the fusion of this precorrected image and the originally reconstructed image with metal parts removed; (iii) forward project this prior image to guide the estimation of the surrogate data using well-developed replacement techniques.

RESULTS

Both simulations and clinical image tests are carried out to show that the proposed FP-MAR method can effectively reduce metal artifacts. A comparison with other MAR methods demonstrates that the FP-MAR method performs better in artifact suppression and tissue feature preservation.

CONCLUSIONS

From a wide range of clinical cases to which FP-MAR has been tested (single or multiple pieces of metal, various shapes, and sizes), it can be concluded that the proposed fusion based prior image preserves more tissue information than other segmentation-based prior approaches and can provide better estimates of the surrogate data in prior-based MAR methods.

Directional interpolation for motion weighted 4D cone-beam CT reconstruction

Image quality of four dimensional cone-beam computed tomography (4D CBCT) is limited by streaking artifacts due to insufficient projections after respiratory sorting. In this paper, a framework is proposed to combine improved motion and stationary regions of CBCT together to enhance the final reconstructed image. Firstly, streaking artifacts are decreased in the 4D CBCT by directional interpolation for additional cone-beam projections. Secondly, motion is estimated through deformable image registration of the 4D CBCT and motion proportional weights are assigned to each voxel. Finally, the weighted combination of the 3D and an interpolated 4D image is calculated. The proposed method is validated by both phantom and clinical data. Experiments demonstrate this method decreases streaking artifacts as well as image blur, and then improves image quality.

Advances in 4D radiation therapy for managing respiration: part I - 4D imaging

Techniques for managing respiration during imaging and planning of radiation therapy are reviewed, concentrating on free-breathing (4D) approaches. First, we focus on detailing the historical development and basic operational principles of currently-available "first generation" 4D imaging modalities: 4D computed tomography, 4D cone beam computed tomography, 4D magnetic resonance imaging, and 4D positron emission tomography. Features and limitations of these first generation systems are described, including necessity of breathing surrogates for 4D image reconstruction, assumptions made in acquisition and reconstruction about the breathing pattern, and commonly-observed artifacts. Both established and developmental methods to deal with these limitations are detailed. Finally, strategies to construct 4D targets and images and, alternatively, to compress 4D information into static targets and images for radiation therapy planning are described.

Partial-data interpolation method for arc handling in a computed tomography scanner

X-ray tube arcing in computed tomography scanners causes poor image quality. During the time that the X-ray tube recovers to full voltage after an arc, image data is being collected. Normally this data, acquired at less than full voltage, is discarded and interpolation is performed over the arc duration. We have developed an algorithm that corrects for improper tube voltage, allowing previously discarded data to be used for imaging. The use of voltage corrected data provides improved image quality compared to simple interpolation methods. This improvement is relevant today as the imaging field uses faster scanners with shorter sampling times.

Strategy of computed tomography sinogram inpainting based on sinusoid-like curve decomposition and eigenvector-guided interpolation

Projection incompleteness in x-ray computed tomography (CT) often relates to sparse sampling or detector gaps and leads to degraded reconstructions with severe streak and ring artifacts. To suppress these artifacts, this study develops a new sinogram inpainting strategy based on sinusoid-like curve decomposition and eigenvector-guided interpolation, where each missing sinogram point is considered located within a group of sinusoid-like curves and estimated from eigenvector-guided interpolation to preserve the sinogram texture continuity. The proposed approach is evaluated on real two-dimensional fan-beam CT data, for which the projection incompleteness, due to sparse sampling and symmetric detector gaps, is simulated. A Compute Unified Device Architecture (CUDA)-based parallelization is applied on the operations of sinusoid fittings and interpolations to accelerate the algorithm. A comparative study is then conducted to evaluate the proposed approach with two other inpainting methods and with a compressed sensing iterative reconstruction. Qualitative and quantitative performances demonstrate that the proposed approach can lead to efficient artifact suppression and less structure blurring.

Sparse angular CT reconstruction using non-local means based iterative-correction POCS

In divergent-beam computed tomography (CT), sparse angular sampling frequently leads to conspicuous streak artifacts. In this paper, we propose a novel non-local means (NL-means) based iterative-correction projection onto convex sets (POCS) algorithm, named as NLMIC-POCS, for effective and robust sparse angular CT reconstruction. The motivation for using NLMIC-POCS is that NL-means filtered image can produce an acceptable priori solution for sequential POCS iterative reconstruction. The NLMIC-POCS algorithm has been tested on simulated and real phantom data. The experimental results show that the presented NLMIC-POCS algorithm can significantly improve the image quality of the sparse angular CT reconstruction in suppressing streak artifacts and preserving the edges of the image.

Estimation of local orientations in fibrous structures with applications to the Purkinje system

The extraction of the cardiac Purkinje system (PS) from intensity images is a critical step toward the development of realistic structural models of the heart. Such models are important for uncovering the mechanisms of cardiac disease and improving its treatment and prevention. Unfortunately, the manual extraction of the PS is a challenging and error-prone task due to the presence of image noise and numerous fiber junctions. To deal with these challenges, we propose a framework that estimates local fiber orientations with high accuracy and reconstructs the fibers via tracking. Our key contribution is the development of a descriptor for estimating the orientation distribution function (ODF), a spherical function encoding the local geometry of the fibers at a point of interest. The fiber/branch orientations are identified as the modes of the ODFs via spherical clustering and guide the extraction of the fiber centerlines. Experiments on synthetic data evaluate the sensitivity of our approach to image noise, width of the fiber, and choice of the mode detection strategy, and show its superior performance compared to those of the existing descriptors. Experiments on the free-running PS in an MR image also demonstrate the accuracy of our method in reconstructing such sparse fibrous structures.