CBCT Image Quality Research Articles

GSOR6 Presentation Time: 9:25 AM: Integrating a Novel GYN Brachytherapy Workflow Using In-Room CBCT Images with MR-LINAC Scans

<h3>Purpose</h3> Imaging capabilities within Radiation Oncology departments are increasing. Departments now have access to MRI imaging with MR-Linacs and designated brachytherapy mobile CBCT scanners. Utilization of these within-department imaging capabilities have the potential to improve clinical brachytherapy workflow. The purpose of this report is to investigate the feasibility of utilizing an MR-LINAC and a mobile CBCT scanner for patient imaging during HDR GYN brachytherapy. <h3>Materials and Methods</h3> To compare the spatial integrity of these imaging modalities, an HDR ring and tandem applicator was placed in a water phantom and imaged using a CT simulator (SOMATOM go.Open Pro, Siemens), 0.35T MR-LINAC (MRIdian, ViewRay) and a mobile CBCT scanner (ImagingRing, Elekta). The images acquired included a MR T2 weighted scan, a pelvis image guided brachytherapy CBCT acquisition, and a CT simulation. The images from the MR-LINAC and mobile CBCT scanner were each rigidly registered to CT-SIM images focusing on the applicator. Deviations in ring diameter, angle, and the position of the tip of the applicator was measured on each fusion. Additionally, a patient was imaged with the applicator inserted using the CT-Sim and the MR-LINAC systems. Physician drawn contours using these two modalities were compared to one another to investigate the benefits of incorporating MR-LINAC scans in the clinical HDR treatment planning process. <h3>Results</h3> The applicator was visible on each of the imaging modalities investigated. The deviation in the position of the applicator dimensions were less than 1mm on the MR-CT and CBCT-CT fusions. The analysis of the multimodality image fusions for the applicator revealed no large geometric distortions or loss of image quality when comparing the MR-CBCT image sets to MR-CT image sets. In patient scans, The MR planning image provided superior soft tissue contrast compared to the image acquired using the CT simulator. In this particular case, the superior soft tissue contrast provided by the MR scan provided more information for the physician for target delineation and resulted in a clinical decision to reduce the volume of the GTV. <h3>Conclusion</h3> This work presents a novel clinical workflow for HDR GYN brachytherapy using a mobile CBCT scanner and MR-LINAC for patient imaging. Incorporating the MR T2 scans into the clinical HDR brachytherapy workflow may aide in GYN brachytherapy target and OAR delineation. Additionally, the mobile CBCT scanner can reduce the challenges associated with patient setup and transportation between imaging and treatment locations. We found that the image quality and spatial resolution of MR and mobile CBCT based applicator images were clinically acceptable. A potential clinical workflow could involve using the MRIdian to obtain a planning image and contours the first fraction of treatment. Then on subsequent fractions the mobile CBCT scan could be fused to MR scans and used for daily applicator verification. For future work we will investigate a modified T2 sequence for MR guided brachytherapy to improve catheter reconstruction. Our work validates the clinical feasibility of using new technology for an MR-CBCT clinical workflow for HDR brachytherapy treatment planning.

Development and validation of a scatter-corrected CBCT image-guided method for cervical cancer brachytherapy.

Multiple patient transfers have a nonnegligible impact on the accuracy of dose delivery for cervical cancer brachytherapy. We consider using on-site cone-beam CT (CBCT) to resolve this problem. However, CBCT clinical applications are limited due to inadequate image quality. This paper implements a scatter correction method using planning CT (pCT) prior to obtaining high-quality CBCT images and evaluates the dose calculation accuracy of CBCT-guided brachytherapy for cervical cancer. The CBCT of a self-developed female pelvis phantom and five patients was first corrected using empirical uniform scatter correction in the projection domain and further corrected in the image domain. In both phantom and patient studies, the CBCT image quality before and after scatter correction was evaluated with registered pCT (rCT). Model-based dose calculation was performed using the commercial package Acuros®BV. The dose distributions of rCT-based plans and corrected CBCT-based plans in the phantom and patients were compared using 3D local gamma analysis. A statistical analysis of the differences in dosimetric parameters of five patients was also performed. In both phantom and patient studies, the HU error of selected ROIs was reduced to less than 15 HU. Using the dose distribution of the rCT-based plan as the baseline, the γ pass rate (2%, 2mm) of the corrected CBCT-based plan in phantom and patients all exceeded 98% and 93%, respectively, with the threshold dose set to 3, 6, 9, and 12 Gy. The average percentage deviation (APD) of D90 of HRCTV and D2cc of OARs was less than 1% between rCT-based and corrected CBCT-based plans. Scatter correction using a pCT prior can effectively improve the CBCT image quality and CBCT-based cervical brachytherapy dose calculation accuracy, indicating promising prospects in both simplified brachytherapy processes and accurate brachytherapy dose delivery.

Improving image quality and reducing dose with 2.5 MV diamond target volume-of-interest cone beam CT imaging.

Over the past decades, continuous efforts have been made to improve megavoltage (MV) image quality versus dose characteristics, including the implementation of low atomic number (Z) targets in MV beamlines and the development of more efficient detectors. Recently, a diamond target beam within a commercial radiotherapy treatment platform demonstrated improved planar contrast-to-noise-ratio (CNR) per unit dose using a novel 2.5 MV sintered diamond target beam, which enabled image acquisition on the order of mGy. The present work assesses cone beam CT (CBCT) image quality characteristics for the novel 2.5 MV diamond target beam and the effects of volume-of-interest (VOI) collimation on the image quality and imaging dose distribution. A sintered diamond target was incorporated into the target arm of the linear accelerator, replacing the 2.5 MV commercial copper imaging target. CBCT image quality was evaluated against the commercial imaging beam with regard to spatial resolution and CNR versus dose. In addition to full-field acquisitions, we investigated VOI techniques that collimate the imaging beam to preselected anatomy, to determine potential image quality improvements and dose sparing capacity. Using an anthropomorphic phantom, VOI regions were defined to encompass the maxillary and ethmoid sinuses and ranged in dimension from 3 cm to 4.85 cm equivalent radius. The MLC was fit to each VOI structure throughout a full CBCT arc and the corresponding MLC sequences were produced as XML scripts for acquisition. Calibrated radiochromic film was used in phantom to measure cumulative axial dose distributions during each CBCT acquisition. In full-field CBCT, the 2.5 MV diamond target beam demonstrated improved CNR versus dose compared to the commercial imaging beam, by factors of up to 1.7. The calculated modulation transfer function (MTF) displayed an increase of nearly 30% in f50 for the 2.5 MV diamond target beam compared to the commercial beam. Using VOI techniques, CNR increased monotonically as a function of equivalent radius at the bone-tissue interface. At the bone-sinus interface, the CNR for the full-field case was slightly decreased compared to the largest VOI case. Imaging dose in the anteroposterior direction increased with increasing VOI equivalent radius. The novel 2.5 MV sintered diamond target beam presents a simple modification to the commercial imaging beam which provides improved image quality in full-field CBCT and the potential for simultaneous dose sparing and CNR improvement at high-contrast interfaces using VOI acquisition techniques.

Generation and Evaluation of Synthetic Computed Tomography (CT) from Cone-Beam CT (CBCT) by Incorporating Feature-Driven Loss into Intensity-Based Loss Functions in Deep Convolutional Neural Network.

Simple SummaryDespite numerous benefits of cone-beam computed tomography (CBCT), its applications to radiotherapy were limited mainly due to degraded image quality. Recently, enhancing the CBCT image quality by generating synthetic CT image by deep convolutional neural network (CNN) has become frequent. Most of the previous works, however, generated synthetic CT with simple, classical intensity-driven loss in network training, while not specifying a full-package of verifications. This work trained the network by combining feature- and intensity-driven losses and attempted to demonstrate clinical relevance of the synthetic CT images by assessing both image similarity and dose calculating accuracy throughout a commercial Monte-Carlo.Deep convolutional neural network (CNN) helped enhance image quality of cone-beam computed tomography (CBCT) by generating synthetic CT. Most of the previous works, however, trained network by intensity-based loss functions, possibly undermining to promote image feature similarity. The verifications were not sufficient to demonstrate clinical applicability, either. This work investigated the effect of variable loss functions combining feature- and intensity-driven losses in synthetic CT generation, followed by strengthening the verification of generated images in both image similarity and dosimetry accuracy. The proposed strategy highlighted the feature-driven quantification in (1) training the network by perceptual loss, besides L1 and structural similarity (SSIM) losses regarding anatomical similarity, and (2) evaluating image similarity by feature mapping ratio (FMR), besides conventional metrics. In addition, the synthetic CT images were assessed in terms of dose calculating accuracy by a commercial Monte-Carlo algorithm. The network was trained with 50 paired CBCT-CT scans acquired at the same CT simulator and treatment unit to constrain environmental factors any other than loss functions. For 10 independent cases, incorporating perceptual loss into L1 and SSIM losses outperformed the other combinations, which enhanced FMR of image similarity by 10%, and the dose calculating accuracy by 1–2% of gamma passing rate in 1%/1mm criterion.

Technical note: Characterization of novel iterative reconstructed cone beam CT images for dose tracking and adaptive radiotherapy on L-shape linacs.

Cone-beam computed tomography (CBCT) allows for patient setup and positioning, and potentially dose verification or adaptive replanning prior to each treatment delivery. Poor CBCT image quality due to scatter artifacts and patient motion has been a major limiting factor. A new image reconstruction algorithm was recently clinically implemented for improving image quality through iterative reconstruction (iCBCT). This study aims to characterize iCBCT image quality, establish image value (HU)-to-relative electron density (RED) calibration curves for dose calculation, and assess the dosimetric accuracy for different anatomical sites. Both conventional CBCT and iCBCT scans were acquired from a Varian TrueBeam On-Board Imager system. A Catphan 604 phantom was scanned to compare image quality between the traditional Feldkamp-Davis-Kress (FDK) and novel iterative reconstruction techniques. Computerized Imaging Reference Systems (CIRS) electron density phantom was used to construct site-specific HU-RED curves corresponding to various scan settings. The CIRS Dynamic Thorax phantom, Rando pelvis phantom, and BrainLab head phantom were used for assessing dosimetric accuracy calculated on iCBCT images, compared to that on traditional FDK-based CBCT images. All phantoms were scanned on a computed tomography (CT) to obtain baseline HU values for comparison. Test results obtained from Catphan showed statistically significant improvement with iCBCT, compared to FDK CBCT. Average HU differences from the baseline CT values were improved to within ±30HU for iCBCT, compared to FDK CBCT for phantom studies. Dose calculated on iCBCT for both phantoms and patient cases directly using baseline HU-RED calibration from CT showed 0.5%-2.0% accuracy from the baseline dose calculated on CT, which is comparable to doses calculated using site-specific HU-RED calibration curves. iCBCT provides improved image quality, improved HU accuracy compared to CT baseline, and has potential to provide online dose verification as part of the adaptive radiotherapy workflow directly using the baseline HU-RED calibration curve from CT.

Image enhancement of ultra-low dose CBCT images using a deep generative model

<h3>Objective</h3> The aim of this study was to improve the image quality of CBCT images, obtained with an ultra-low dose (ULD) protocol, using a deep generative model called pix2pix image-to-image translation technique. <h3>Study Design</h3> CBCT images of a dry skull and mandible were obtained using preset standard and ULD imaging protocols. We combined the scans into a dataset with 1,412 ULD scans and normal-dose DICOM image pairs (2,824 scans in total), and randomly selected 10% (140) image pairs as the test set. We trained a pix2pix deep generative model by first converting the DICOM images to JPG images and then feeding the training set image pairs to the network. We tested the performance of the trained model on the previously unseen test set ULD images and compared the synthesized images with the corresponding normal dose and ULD images. A preliminary blind study on the ULD, standard, and synthesized images by an oral and maxillofacial radiologist was done and image quality and edge definition were assessed. <h3>Results</h3> Preliminary results indicated that in the three datasets used for review, all synthesized images were rated as having higher image quality and better edge definition compared to ULD images. Analysis also indicated that synthesized images were equivalent to standard images in image quality and edge definition. <h3>Conclusion</h3> CBCT usage is expected to grow globally by 9.7%. With the expected growth, reducing radiation dose and improving image quality is of paramount importance to reduce the radiation burden on the population. The current study focused on image enhancement of CBCT images acquired with the ULD protocol. Initial analysis indicated a marked improvement in image quality of the synthesized images when compared to ULD images. With further research and successful implementation, we have an opportunity to reduce radiation risks to patients while improving image quality. <b>Statement of Ethical Review</b> Ethical Review or exemption was not warranted for this study.

Concurrent kilovoltage CBCT imaging and megavoltage beam delivery: suppression of cross-scatter with 2D antiscatter grids and grid-based scatter sampling

Objective. The concept of using kilovoltage (kV) and megavoltage (MV) beams concurrently has potential applications in cone beam computed tomography (CBCT) guided radiation therapy, such as single breath hold scans, metal artifact reduction, and simultaneous imaging during MV treatment delivery. However, MV cross-scatter generated during MV beam delivery degrades CBCT image quality. To address this, a 2D antiscatter grid and a cross-scatter correction method were investigated in the context of high dose MV treatment delivery. Approach. A 3D printed, tungsten 2D antiscatter grid prototype was utilized in kV CBCT scans to reduce MV cross-scatter fluence during concurrent MV beam delivery. Remaining cross-scatter in projections was corrected by using the 2D grid itself as a cross-scatter intensity sampling device, referred to as grid-based scatter sampling (GSS). To test this approach, kV CBCT acquisitions were performed while delivering 6 and 10 MV beams, mimicking high dose rate treatment delivery scenarios. kV and MV beam deliveries were not synchronized to eliminate MV beam delivery interruption. MV cross-scatter suppression performance of the proposed approach was evaluated in projections and CBCT images of phantoms. Main results. 2D grid reduced the intensity of MV cross-scatter in kV projections by a factor of 3 on the average, when compared to conventional antiscatter grid. Remaining cross scatter as measured by the GSS method was within 7% of measured reference intensity values, and subsequently corrected. CBCT image quality was improved substantially during concurrent kV–MV beam delivery. Median Hounsfield Unit (HU) inaccuracy was up to 182 HU without our methods, and it was reduced to a median 6.5 HU with our 2D grid and scatter correction approach. Our methods provided a factor of 2–6 improvement in contrast-to-noise ratio. Significance. This investigation demonstrates the utility of 2D antiscatter grids and grid-based scatter sampling in suppressing MV cross-scatter. Our approach successfully minimized the effects of MV cross-scatter in concurrent kV CBCT imaging and high dose MV treatment delivery scenarios. Hence, robust MV cross-scatter suppression is potentially feasible without MV beam delivery interruption or compromising kV image acquisition rate.

A Contouring Strategy and Reference Atlases for the Full Abdominopelvic Bowel Bag on Treatment Planning and Cone Beam Computed Tomography Images

PurposeTo establish a practical contouring strategy with reference atlases for the abdominopelvic bowel bag on treatment planning computed tomography (TPCT) and cone beam computed tomography (CBCT) images. Methods and MaterialsA scoping literature review was done to evaluate the existing definitions and contouring guidelines for bowel bag and small bowel planning-at-risk volume–like structures. A comprehensive definition was proposed for the abdominopelvic bowel bag that expanded the Radiation Therapy Oncology Group Pelvic Normal Tissue Consensus definition. Seven patients with TPCT and first-treatment-day CBCT images were selected from an institutional database to represent a range of normal anatomy and CBCT image quality. The TPCT and CBCT images were contoured using the proposed definition. During contouring, the Radiation Therapy Oncology Group definition's list of inclusion and exclusion structures was expanded. For areas with limited visibility of the bowel bag on either TPCT or CBCT, a set of operational definitions was developed based on consistently visible reference structures. ResultsA literature review showed that previously existing bowel bag definitions predominantly focused on the pelvic region and did not provide a complete and practical description of the full abdominopelvic contour relative to structures consistently visible in all radiation therapy images. The proposed contouring strategy had 4 components: a definition, a list of inclusion and exclusion structures, 15 tabulated operational definitions, and a set of atlases. The bowel bag was defined as the peritoneal cavity and retroperitoneal duodenum and ascending and descending colon, as visualized at the time of image acquisition. The operational definitions formalized the location of the peritoneal fascial planes through a simple look-up table. The proposed contouring strategy and reference atlases were successfully used on both TPCT and CBCT images. ConclusionsThis study produced a practical contouring strategy and reference atlases to enable reproducible delineation of the full bowel bag on TPCT and CBCT images. The strategy is a necessary first step toward consensus contouring with reduced observer variability, which is a prerequisite for evaluation of cumulative dose and its correlation with toxic effects, adaptive planning strategies, and automated contouring potential.

Deep learning methods for enhancing cone‐beam CT image quality toward adaptive radiation therapy: A systematic review

The use of deep learning (DL) to improve cone‐beam CT (CBCT) image quality has gained popularity as computational resources and algorithmic sophistication have advanced in tandem. CBCT imaging has the potential to facilitate online adaptive radiation therapy (ART) by utilizing up‐to‐date patient anatomy to modify treatment parameters before irradiation. Poor CBCT image quality has been an impediment to realizing ART due to the increased scatter conditions inherent to cone‐beam acquisitions. Given the recent interest in DL applications in radiation oncology, and specifically DL for CBCT correction, we provide a systematic theoretical and literature review for future stakeholders. The review encompasses DL approaches for synthetic CT generation, as well as projection domain methods employed in the CBCT correction literature. We review trends pertaining to publications from January 2018 to April 2022 and condense their major findings—with emphasis on study design and DL techniques. Clinically relevant endpoints relating to image quality and dosimetric accuracy are summarized, highlighting gaps in the literature. Finally, we make recommendations for both clinicians and DL practitioners based on literature trends and the current DL state‐of‐the‐art methods utilized in radiation oncology.

CDFRegNet: A cross-domain fusion registration network for CT-to-CBCT image registration

Background and ObjectiveComputer tomography (CT) to cone-beam computed tomography (CBCT) image registration plays an important role in radiotherapy treatment placement, dose verification, and anatomic changes monitoring during radiotherapy. However, fast and accurate CT-to-CBCT image registration is still very challenging due to the intensity differences, the poor image quality of CBCT images, and inconsistent structure information. MethodsTo address these problems, a novel unsupervised network named cross-domain fusion registration network (CDFRegNet) is proposed. First, a novel edge-guided attention module (EGAM) is designed, aiming at capturing edge information based on the gradient prior images and guiding the network to model the spatial correspondence between two image domains. Moreover, a novel cross-domain attention module (CDAM) is proposed to improve the network's ability to guide the network to effectively map and fuse the domain-specific features. ResultsExtensive experiments on a real clinical dataset were carried out, and the experimental results verify that the proposed CDFRegNet can register CT to CBCT images effectively and obtain the best performance, while compared with other representative methods, with a mean DSC of 80.01±7.16%, a mean TRE of 2.27±0.62 mm, and a mean MHD of 1.50±0.32 mm. The ablation experiments also proved that our EGAM and CDAM can further improve the accuracy of the registration network and they can generalize well to other registration networks. ConclusionThis paper proposed a novel CT-to-CBCT registration method based on EGAM and CDAM, which has the potential to improve the accuracy of multi-domain image registration.

A Motion Artifact Correction Algorithm for Cone-Beam CT in Patients with Hepatic Malignancies Treated with Transarterial Chemoembolization

PurposeTo evaluate the effect of a motion artifact correction algorithm (MACA) on cone-beam computed tomography (CT) during transarterial chemoembolization for hepatic malignancies. Materials and MethodsFrom June 2020 to March 2021, 42 patients with mild-to-severe motion artifacts detected using single cone-beam CT scans were evaluated retrospectively. The image quality of native and motion-corrected data was compared. The maximum intensity, sharpness, and full width at half maximum (FWHM) of 5 segmental hepatic arteries were quantitatively measured. The overall quality of maximum intensity projection (MIP) images, conspicuity of tumor-supplying arteries, and need for selective angiography to ascertain the vascular anatomy were qualitatively evaluated by multiple readers. Paired t and Wilcoxon signed rank tests were used to compare the parameters. ResultsThe mean maximum intensity and sharpness increased from 2,792.01 HU ± 451.36 to 3,148.40 HU ± 594.46 and from 0.31 ± 0.02/mm to 0.34 ± 0.02/mm, respectively, using the MACA (both P < .001). The MACA decreased the mean FWHM from 2.02 mm ± 0.27 to 1.78 mm ± 0.26 (P < .001). The overall quality of the MIP images and the conspicuity of the tumor-supplying artery were enhanced from 2.5 to 3.0 points and from 3.0 to 4.0 points, respectively (both P < .001). Selective angiography was expected to be omitted in 7 cases (16.7%, 7/42) after using the MACA. ConclusionsThe MACA significantly improved both quantitative and qualitative image quality of cone-beam CT in selected patients with motion artifacts during transarterial chemoembolization for hepatic malignancies.

Improving cone-beam CT quality using a cycle-residual connection with a dilated convolution-consistent generative adversarial network

Objective.Cone-Beam CT (CBCT) often results in severe image artifacts and inaccurate HU values, meaning poor quality CBCT images cannot be directly applied to dose calculation in radiotherapy. To overcome this, we propose a cycle-residual connection with a dilated convolution-consistent generative adversarial network (Cycle-RCDC-GAN). Approach. The cycle-consistent generative adversarial network (Cycle-GAN) was modified using a dilated convolution with different expansion rates to extract richer semantic features from input images. Thirty pelvic patients were used to investigate the effect of synthetic CT (sCT) from CBCT, and 55 head and neck patients were used to explore the generalizability of the model. Three generalizability experiments were performed and compared: the pelvis trained model was applied to the head and neck; the head and neck trained model was applied to the pelvis, and the two datasets were trained together. Main results. The mean absolute error (MAE), the root mean square error (RMSE), peak signal to noise ratio (PSNR), the structural similarity index (SSIM), and spatial nonuniformity (SNU) assessed the quality of the sCT generated from CBCT. Compared with CBCT images, the MAE improved from 28.81 to 18.48, RMSE from 85.66 to 69.50, SNU from 0.34 to 0.30, and PSNR from 31.61 to 33.07, while SSIM improved from 0.981 to 0.989. The sCT objective indicators of Cycle-RCDC-GAN were better than Cycle-GAN’s. The objective metrics for generalizability were also better than Cycle-GAN’s. Significance. Cycle-RCDC-GAN enhances CBCT image quality and has better generalizability than Cycle-GAN, which further promotes the application of CBCT in radiotherapy.

Dual-domain attention-guided convolutional neural network for low-dose cone-beam computed tomography reconstruction

Excessive ionizing radiation in cone-beam computed tomography (CBCT) causes damage to patients, whereas a low radiation dose degrades the imaging quality. To improve the quality of low-dose CBCT images, deep-learning-based methods are developed and have obtained good performance. However, most previous studies only process the reconstructed CT images, and cannot recover the structures already lost in the reconstruction process. In this paper, a dual-domain attention-guided network framework (Dual-AGNet) is developed to process images in both projection and reconstruction domains. Spatial attention modules are included in the AGNet to effectively and adaptively compensate the intra- and inter-images information in both domains. Moreover, a joint loss function is developed to circumvent the structures loss and over-smoothness in CT images. Our method is evaluated and compared with the state-of-the-art methods on a simulated and a real low-dose CBCT datasets of walnuts. Our Dual-AGNet obtains significantly better performance than the state-of-the-art methods; on the simulated and real datasets, it decreases the root mean square error by approximately 11% and 19%, increases the peak signal-to-noise ratio by approximately 5% and 7%, and increases the structural similarity by approximately 5% and 2%, respectively. In qualitative evaluation, our Dual-AGNet not only suppresses the noise, but also provides realistic CT images with many delicate structures. In addition, the developed Dual-AGNet can be integrated into the existing CBCT system to promote the development of low-dose CBCT imaging. Testing code is available at https://github.com/LianyingChao/Dual-AGNet.

Reference-free learning-based similarity metric for motion compensation in cone-beam CT

Purpose. Patient motion artifacts present a prevalent challenge to image quality in interventional cone-beam CT (CBCT). We propose a novel reference-free similarity metric (DL-VIF) that leverages the capability of deep convolutional neural networks (CNN) to learn features associated with motion artifacts within realistic anatomical features. DL-VIF aims to address shortcomings of conventional metrics of motion-induced image quality degradation that favor characteristics associated with motion-free images, such as sharpness or piecewise constancy, but lack any awareness of the underlying anatomy, potentially promoting images depicting unrealistic image content. DL-VIF was integrated in an autofocus motion compensation framework to test its performance for motion estimation in interventional CBCT. Methods. DL-VIF is a reference-free surrogate for the previously reported visual image fidelity (VIF) metric, computed against a motion-free reference, generated using a CNN trained using simulated motion-corrupted and motion-free CBCT data. Relatively shallow (2-ResBlock) and deep (3-Resblock) CNN architectures were trained and tested to assess sensitivity to motion artifacts and generalizability to unseen anatomy and motion patterns. DL-VIF was integrated into an autofocus framework for rigid motion compensation in head/brain CBCT and assessed in simulation and cadaver studies in comparison to a conventional gradient entropy metric. Results. The 2-ResBlock architecture better reflected motion severity and extrapolated to unseen data, whereas 3-ResBlock was found more susceptible to overfitting, limiting its generalizability to unseen scenarios. DL-VIF outperformed gradient entropy in simulation studies yielding average multi-resolution structural similarity index (SSIM) improvement over uncompensated image of 0.068 and 0.034, respectively, referenced to motion-free images. DL-VIF was also more robust in motion compensation, evidenced by reduced variance in SSIM for various motion patterns (σ DL-VIF = 0.008 versus σ gradient entropy = 0.019). Similarly, in cadaver studies, DL-VIF demonstrated superior motion compensation compared to gradient entropy (an average SSIM improvement of 0.043 (5%) versus little improvement and even degradation in SSIM, respectively) and visually improved image quality even in severely motion-corrupted images.Conclusion: The studies demonstrated the feasibility of building reference-free similarity metrics for quantification of motion-induced image quality degradation and distortion of anatomical structures in CBCT. DL-VIF provides a reliable surrogate for motion severity, penalizes unrealistic distortions, and presents a valuable new objective function for autofocus motion compensation in CBCT.

A two-step method to improve image quality of CBCT with phantom-based supervised and patient-based unsupervised learning strategies

Objective. In this study, we aimed to develop deep learning framework to improve cone-beam computed tomography (CBCT) image quality for adaptive radiation therapy (ART) applications. Approach. Paired CBCT and planning CT images of 2 pelvic phantoms and 91 patients (15 patients for testing) diagnosed with prostate cancer were included in this study. First, well-matched images of rigid phantoms were used to train a U-net, which is the supervised learning strategy to reduce serious artifacts. Second, the phantom-trained U-net generated intermediate CT images from the patient CBCT images. Finally, a cycle-consistent generative adversarial network (CycleGAN) was trained with intermediate CT images and deformed planning CT images, which is the unsupervised learning strategy to learn the style of the patient images for further improvement. When testing or applying the trained model on patient CBCT images, the intermediate CT images were generated from the original CBCT image by U-net, and then the synthetic CT images were generated by the generator of CycleGAN with intermediate CT images as input. The performance was compared with conventional methods (U-net/CycleGAN alone trained with patient images) on the test set. Results. The proposed two-step method effectively improved the CBCT image quality to the level of CT scans. It outperformed conventional methods for region-of-interest contouring and HU calibration, which are important to ART applications. Compared with the U-net alone, it maintained the structure of CBCT. Compared with CycleGAN alone, our method improved the accuracy of CT number and effectively reduced the artifacts, making it more helpful for identifying the clinical target volume. Significance. This novel two-step method improves CBCT image quality by combining phantom-based supervised and patient-based unsupervised learning strategies. It has immense potential to be integrated into the ART workflow to improve radiotherapy accuracy.

AI Denoising Significantly Enhances Image Quality and Diagnostic Confidence in Interventional Cone-Beam Computed Tomography.

(1) To investigate whether interventional cone-beam computed tomography (cbCT) could benefit from AI denoising, particularly with respect to patient body mass index (BMI); (2) From 1 January 2016 to 1 January 2022, 100 patients with liver-directed interventions and peri-procedural cbCT were included. The unenhanced mask run and the contrast-enhanced fill run of the cbCT were reconstructed using weighted filtered back projection. Additionally, each dataset was post-processed using a novel denoising software solution. Place-consistent regions of interest measured signal-to-noise ratio (SNR) per dataset. Corrected mixed-effects analysis with BMI subgroup analyses compared objective image quality. Multiple linear regression measured the contribution of “Radiation Dose”, “Body-Mass-Index”, and “Mode” to SNR. Two radiologists independently rated diagnostic confidence. Inter-rater agreement was measured using Spearman correlation (r); (3) SNR was significantly higher in the denoised datasets than in the regular datasets (p < 0.001). Furthermore, BMI subgroup analysis showed significant SNR deteriorations in the regular datasets for higher patient BMI (p < 0.001), but stable results for denoising (p > 0.999). In regression, only denoising contributed positively towards SNR (0.6191; 95%CI 0.6096 to 0.6286; p < 0.001). The denoised datasets received overall significantly higher diagnostic confidence grades (p = 0.010), with good inter-rater agreement (r ≥ 0.795, p < 0.001). In a subgroup analysis, diagnostic confidence deteriorated significantly for higher patient BMI (p < 0.001) in the regular datasets but was stable in the denoised datasets (p ≥ 0.103).; (4) AI denoising can significantly enhance image quality in interventional cone-beam CT and effectively mitigate diagnostic confidence deterioration for rising patient BMI.

Diagnostic usefulness of cone-beam computed tomography versus multi-detector computed tomography for sinonasal structure evaluation.

ObjectiveCone‐beam computed tomography (CBCT) is a promising imaging modality for sinonasal evaluation, with advantages of relatively low radiation dose, low cost, and quick outpatient imaging. Our study aimed to compare the diagnostic performance and image quality of CBCT with those of multi‐detector computed tomography (MDCT) with different slice thickness.MethodsWe retrospectively reviewed 60 consecutive patients who had undergone both CBCT and MDCT. MDCT images was reconstructed with 1 and 3 mm slice thickness. The quantitative image quality parameters (image noise, signal‐to‐noise ratio [SNR], and contrast‐to noise ratio [CNR] were calculated and compared between the two imaging modalities. Two observers (ENT surgeon and neuroradiologist) evaluated the presence of seven sinonasal anatomic variations in each patient and interobserver agreements were analyzed. The diagnostic performance of CBCT (0.3 mm) and MDCT (3 mm) was assessed and compared with that of high resolution MDCT (1 mm), which is considered as the gold standard.ResultsThe image noise was significantly higher and SNR and CNR values were lower in the CBCT (0.3 mm) group than in the MDCT groups (1 and 3 mm). The diagnostic performance of CBCT (0.3 mm) was similar to that of MDCT (1 mm) and superior to that of MDCT (3 mm). The highest interobserver agreement was for high resolution MDCT (1 mm), followed by CBCT (0.3 mm), and MDCT (3 mm).ConclusionConsidering its low radiation dose, low cost, and ease of clinical access, CBCT may be a useful imaging modality for as first line sinonasal evaluation and repeated follow up.Study design: Retrospective study in a tertiary referral university center.Level of evidence: NA.

A deep unsupervised learning framework for the 4D CBCT artifact correction

Objective. Four-dimensional cone-beam computed tomography (4D CBCT) has unique advantages in moving target localization, tracking and therapeutic dose accumulation in adaptive radiotherapy. However, the severe fringe artifacts and noise degradation caused by 4D CBCT reconstruction restrict its clinical application. We propose a novel deep unsupervised learning model to generate the high-quality 4D CBCT from the poor-quality 4D CBCT. Approach. The proposed model uses a contrastive loss function to preserve the anatomical structure in the corrected image. To preserve the relationship between the input and output image, we use a multilayer, patch-based method rather than operate on entire images. Furthermore, we draw negatives from within the input 4D CBCT rather than from the rest of the dataset. Main results. The results showed that the streak and motion artifacts were significantly suppressed. The spatial resolution of the pulmonary vessels and microstructure were also improved. To demonstrate the results in the different directions, we make the animation to show the different views of the predicted correction image in the supplementary animation. Significance. The proposed method can be integrated into any 4D CBCT reconstruction method and maybe a practical way to enhance the image quality of the 4D CBCT.

Investigation of image quality of MV and kV CBCT with low-Z beams and high DQE detector.

To investigate cone beam computed tomography (CBCT) image quality using novel combinations of kilovoltage (kV) and megavoltage (MV) beams and detectormaterials. MV and kV CBCT imaging was simulated using the Fastcat hybrid Monte Carlo application. CBCT imaging with various beam energies was investigated: 2.5 and 6 MV photon beams generated with carbon, aluminum, and tungsten targets and a 120 kVp x-ray tube beam based off of a Varian Truebeam on-board imager (OBI). Cadmium tungstate (CWO), gadolinium oxysulfide (GOS), and cesium iodide (CsI) detectors with identical pixel pitch of 0.784 mm were evaluated. Modulation transfer functions (MTF) for all detector/beam combinations were calculated. MV and kV CBCT images for each detector/beam combination of a contrast phantom containing inserts with rib and spongiosa bone, lung, and adipose tissues were simulated with an imaging dose of 7 mGy. Contrast to noise ratio (CNR) of all inserts were compared for all detector/beam combinations. CBCT images of an anthropomorphic head phantom with silver amalgam fillings were also generated. The CWO/120 kVp beam combination resulted in the highest MTF at low frequencies and the CsI detector showed the highest MTF for all other beams and at high frequencies. The CWO/120 kVp beam combination showed the highest CNR for all tissues. The unoptimized CWO/2.5 MV carbon target beam showed the highest CNR of the MV beam/detector combinations with CNR 4% and 17% worse than the optimized Truebeam CsI 120 kVp setup with a bowtie filter and antiscatter grid. Additionally, the CWO 2.5 MV setup showed qualitative reduction of metal artifacts surrounding silver amalgam fillings in an anthropomorphic headphantom. This finding makes a compelling case that further optimization of this CWO carbon target setup could produce CBCT images with similar CNR to current OBI CBCT for equivalent dose with added resilience to metalartifacts.

Impact of metal artefacts on subjective perception of image quality of 13 CBCT devices

The overall objective of this study was to assess how metal artefacts impact image quality of 13 CBCT devices. As a secondary objective, the influence of scanning protocols and field of view on CBCT image quality with and without metal artefacts was also assessed. CBCT images were acquired of a dry human skull phantom considering three clinical simulated conditions: one without metal and two with metallic materials (metallic pin and implant). An industrial micro-CT was used as a reference to register the CBCT images. Afterwards, four observers evaluated 306 representative image slices from 13 devices, ranking them from best to worst. Furthermore, within each device, medium FOV and small FOV standard images were compared. General linear mixed models were used to assess subjective perception of examiners on overall image quality in the absence and presence of metal-related artefacts (p < 0.05). Image quality perception significantly differed amongst CBCT devices (p < 0.05). Some devices performed significantly better, independently of scanning protocol and clinical condition. In the presence of metal artefacts, medium FOV standard scanning protocols scored significantly better, while in the absence of metal, small FOV standard yielded the highest performance. Subjective image quality differs significantly amongst CBCT devices and scanning protocols. Metal-related artefacts may highly impact image quality, with a significant device-dependent variability and only few scanners being more robust against metal artefacts. Often, metal artefact expression may be somewhat reduced by proper protocol selection. Metallic objects may severely impact image quality in several CBCT devices.