Computed Tomography Noise Research Articles

Simulated deep CT characterization of liver metastases with high-resolution filtered back projection reconstruction

Early diagnosis and accurate prognosis of colorectal cancer is critical for determining optimal treatment plans and maximizing patient outcomes, especially as the disease progresses into liver metastases. Computed tomography (CT) is a frontline tool for this task; however, the preservation of predictive radiomic features is highly dependent on the scanning protocol and reconstruction algorithm. We hypothesized that image reconstruction with a high-frequency kernel could result in a better characterization of liver metastases features via deep neural networks. This kernel produces images that appear noisier but preserve more sinogram information. A simulation pipeline was developed to study the effects of imaging parameters on the ability to characterize the features of liver metastases. This pipeline utilizes a fractal approach to generate a diverse population of shapes representing virtual metastases, and then it superimposes them on a realistic CT liver region to perform a virtual CT scan using CatSim. Datasets of 10,000 liver metastases were generated, scanned, and reconstructed using either standard or high-frequency kernels. These data were used to train and validate deep neural networks to recover crafted metastases characteristics, such as internal heterogeneity, edge sharpness, and edge fractal dimension. In the absence of noise, models scored, on average, 12.2% (α=0.012\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\alpha =0.012$$\\end{document}) and 7.5% (α=0.049)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\alpha =0.049)$$\\end{document} lower squared error for characterizing edge sharpness and fractal dimension, respectively, when using high-frequency reconstructions compared to standard. However, the differences in performance were statistically insignificant when a typical level of CT noise was simulated in the clinical scan. Our results suggest that high-frequency reconstruction kernels can better preserve information for downstream artificial intelligence-based radiomic characterization, provided that noise is limited. Future work should investigate the information-preserving kernels in datasets with clinical labels.

Wasserstein-GAN-based noise reduction for preserving anatomical structures in low-dose CT

The purpose of low-dose computed tomography (LDCT) is to reduce patient's radiation dose. However, LDCT imaging often suffers from low image quality due to computed tomography (CT) noise patterns. While conventional methods such as iterative reconstruction and image processing have been studied to address this issue, their clinical application remains limited due to the degradation of critical anatomical structures. Recently, deep-learning-based noise reduction approaches have shown promising results in improving the image quality of LDCT. Particularly, generative adversarial network (GAN)-based noise reduction methods have produced outstanding results, approaching the quality of normal-dose CT (NDCT) images. In this study, we propose a hybrid framework based on Wasserstein-GAN with the non-subsampled contourlet transform for enhanced noise reduction in LDCT. The aim is to meticulously address CT noise patterns and generate CT images similar to NDCT. We implemented the proposed algorithm and conducted an experiment using clinical abdominal CT datasets to demonstrate its feasibility. According to our results, the proposed method effectively reduced noise components in LDCT images while preserving the anatomical structures of NDCT. This was confirmed through both qualitative and quantitative evaluations. The implications of our findings for clinical practice are significant, as the proposed method can improve the accuracy of LDCT and reduce radiation dose to patients, ultimately enhancing patient safety.

Improving the quality of computed tomography brain images in the presence of cochlear implant induced metal artefacts through the additional use of tissue mimicking materials alongside metal artefact reduction software

IntroductionMetal artefact reduction software (MAR) can be used to improve Computed Tomography (CT) image quality in the presence of implanted metalwork; however, this software is not effective for superficial metallic structures such as cochlear implants (CI).This study aimed to investigate whether the effectiveness of MAR software could be improved for brain scans with CI present through the use of tissue mimicking materials (TMM) placed exteriorly to the implant. MethodsIn this two-part study, a CI was positioned on the surface of water and anthropomorphic phantoms and imaged using a helical CT brain protocol. Three TMM, Superflab, Sure Thermal heat packs, and Bart's Bolus, were utilised and images were acquired to assess the resulting artefact reduction in terms of CT numbers, noise and artefact index (Aind). Changes in CTDIvol were assessed for the anthropomorphic phantom scans. ResultsIn the water phantom, statistically significant reductions in CT number (p = 0.038) and noise (p = 0.033) were observed for Superflab, whilst the heat packs produced similar significant reductions in CT number (p < 0.001) and noise (p = 0.001) for the anthropomorphic phantom images. Aind values were significantly reduced through the use of Superflab (p = 0.009) and the heat packs (p < 0.001). No significant effects were observed for Bart's Bolus. CTDIvol increases of generally less than 5% were observed for scans with TMM in place. ConclusionThe additional use of TMM alongside MAR software yielded statistically significant reductions in CI induced metal artefacts on both water and anthropomorphic phantom scans with minimal dose increases. Implications for practiceThe extent of metal artefacts in clinical head scans with CI in place could be significantly reduced through combined use of TMM and MAR software, consequently providing greater diagnostic confidence in the images.

Noise aware content-noise complementary GAN with local and global discrimination for low-dose CT denoising

In response to rising concerns over radiation exposure in computed tomography (CT) imaging, effective denoising methods for low-dose CT (LDCT) images are crucial. In recent years, the use of deep learning techniques especially generative adversarial networks (GANs) significantly enhanced the efficiency of LDCT denoising methods, surpassing traditional methods. However, GAN-based denoising methods often face challenges in preserving structural consistency and fine details. This study introduces a novel GAN framework with three accretions to enhance the effectiveness of LDCT denoising. Firstly, our generator is designed to leverage a complementary learning scheme between image noise and image content via two distinct paths. One path focuses on exploring the anatomical information of the image, while the second path is dedicated to learning the noise pattern. This complementary learning scheme provides stable noise cancellation while preserving the maximum structural information of the image. Subsequently, we propose a novel noise-conscious mean absolute error loss to address the challenge posed by the non-stationary characteristic of CT noise. In contrast to conventional MAE loss, this loss attentively prioritizes the different parts of the image based on the local distribution of noise in that region. We also incorporate a gradient-domain loss into the loss function, which inspires the generator to preserve precise image details through explicit guidance. Finally, this study adopted a U-Net-based design for the discriminator to better regularize the model by discriminating between the clean image and the denoised image at both global and local levels. The merit of this discriminator is that it can better adapt to the non-stationary environment of GAN training and guide the generator to produce denoised images that are locally and globally consistent. Our thorough experiments using abdominal CT and lung CT datasets demonstrate the superior performance of our approach compared to state-of-the-art methods.

Virtual Monoenergetic Images Facilitate Better Identification of the Arc of Riolan During Splenic Flexure Takedown.

This study aimed to investigate whether virtual monoenergetic images (VMIs) can aid radiologists and surgeons in better identifying the arc of Riolan (AOR) and to determine the optimal kilo electron volt (keV) level. Thirty-three patients were included. Conventional images (CIs) and VMI (40-100 keV) were reconstructed using arterial phase spectral-based images. The computed tomography (CT) attenuation and noise of the AOR, the CT attenuation of the erector spinal muscle, and the background noise on VMI and CI were measured, respectively. The signal-to-noise ratio, contrast-to-noise ratio (CNR), and signal intensity ratio were calculated. The image quality of the AOR was evaluated according to a 4-point Likert grade. The CT attenuation, noise, CNR, and signal intensity ratio of the AOR were significantly higher in VMI at 40 and 50 keV compared with CI ( P < 0.001); VMI at 40 keV was significantly higher than 50 keV ( P < 0.05). No significant difference in signal-to-noise ratio, background noise, and CT attenuation of the spinal erector muscle was observed between VMI and CI ( P > 0.05). virtual monoenergetic image at 40 keV produced the best subjective scores. Virtual monoenergetic image at 40 keV makes it easier to observe the AOR with optimized subjective and objective image quality. This may prompt radiologists and surgeons to actively search for it and encourage surgeons to preserve it during splenic flexure takedown.

Direct estimation of the noise power spectrum from patient data to generate synthesized CT noise for denoising network training.

Developing a deep-learning network for denoising low-dose CT (LDCT) images necessitates paired computed tomography (CT) images acquired at different dose levels. However, it is challenging to obtain these images from the samepatient. In this study, we introduce a novel approach to generate CT images at different doselevels. Our method involves the direct estimation of the quantum noise power spectrum (NPS) from patient CT images without the need for prior information. By modeling the anatomical NPS using a power-law function and estimating the quantum NPS from the measured NPS after removing the anatomical NPS, we create synthesized quantum noise by applying the estimated quantum NPS as a filter to random noise. By adding synthesized noise to CT images, synthesized CT images can be generated as if these are obtained at a lower dose. This leads to the generation of paired images at different dose levels for training denoisingnetworks. The proposed method accurately estimates the reference quantum NPS. The denoising network trained with paired data generated using synthesized quantum noise achieves denoising performance comparable to networks trained using Mayo Clinic data, as justified by the mean-squared-error (MSE), structural similarity index (SSIM)and peak signal-to-noise ratio (PSNR)scores. This approach offers a promising solution for LDCT image denoising network development without the need for multiple scans of the same patient at differentdoses.

A comparison of image quality of cerium oxide nanoparticles and iodine contrast agents in computed tomography scan

Computed tomography (CT) scan, with iodine-based contrast, produces good image quality by improving the visualisation of relatively low-contrast internal body structures. However, the impact of using iodinated difference should be considered in patients susceptible to contrast allergy and renal impairment. Therefore, alternative contrast materials, such as cerium oxide nanoparticles (CeO2 NPs), must be used, with biocompatible properties and strong X-ray attenuation capabilities. This study compared the CT scan image quality of CeO2 NPs and iodinated contrast agents. This experimental study started by preparing a suspension of CeO2 NPs and iodine in aquabidest at a concentration of 500 ppm. The suspension was scanned using a CT scan with a helical scanning method. The exposure coefficient parameters were set for the tube voltage of 80 kV, Field of View of 28 cm, slice thickness of 5 mm, and tube current time of 150 mAs, 200 mAs, and 250 mAs. Then, CT images in DICOM data format were processed using MicroDICOM Viewer software. The quality of the CT scan images was analysed based on the CT number value, noise level, and contrast resolution. The images of CeO2 nanoparticles have higher CT values, lower noise levels, and better contrast resolution than those of iodine contrast agents. The results show that the CT image results of CeO2 NPs have better quality than those of iodine-containing contrast agents.

Technical Note: Assessment of Pulse Pileup on Single-Energy and Multienergy Images From a Clinical Photon-Counting Detector Computed Tomography.

Pulse pileup effects occur when pulses occur so close together that they fall on top of one another, resulting in count loss and errors in energy thresholding. To date, there has been little work systematically detailing the quantitative effects of pulse pileup on material decomposition accuracy for photon-counting detector (PCD) computed tomography (CT). Our aim in this work was to quantify the effects of pulse pileup on single-energy and multienergy CT images, including low-energy bin (BL), high-energy bin (BH), iodine map, and virtual noncontrast images from a commercial PCD-CT. Scans of a 20-cm diameter multienergy CT phantom with 10 solid inserts were acquired at a fixed tube potential as the tube current was varied across the available range. Four types of images (BL, BH, iodine map, and virtual noncontrast) were reconstructed using an iterative reconstruction algorithm at strength 2, a quantitative reconstruction kernel (Qr40), 2-/1-mm slice thickness/increment, and a 210-mm field-of-view. The mean and standard deviation of CT numbers were recorded and the ratios of CT number between BL and BH images were calculated and plotted, along with noise versus tube current and noise × versus tube current. As tube current was increased, the range of variations in CT numbers was less than 13.4 HU for all inserts and image types evaluated. Noise × versus tube current showed a small positive slope equal to a noise increase from 100 mA of 10% at 500 mA and 15% at 900 mA compared with what would be expected if the slope was zero. Minimal impact on single-energy and multienergy CT numbers and noise performance was observed for the evaluated clinical PCD-CT system.

Discrimination tasks in simulated low-dose CT noise.

This study reports the results of a set of discrimination experiments using simulated images that represent the appearance of subtle lesions in low-dose computed tomography (CT) of the lungs. Noise in these images has a characteristic ramp-spectrum before apodization by noise control filters. We consider three specific diagnostic features that determine whether a lesion is considered malignant or benign, two system-resolution levels, and four apodization levels for a total of 24 experimental conditions. The goal of the investigation is to better understand how well human observers perform subtle discrimination tasks like these, and the mechanisms of that performance. We use a forced-choice psychophysical paradigm to estimate observer efficiency and classification images. These measures quantify how effectively subjects can read the images, and how they use images to perform discrimination tasks across the different imaging conditions. The simulated CT images used as stimuli in the psychophysical experiments are generated from high-resolution objects passed through a modulation transfer function (MTF) before down-sampling to the image-pixel grid. Acquisition noise is then added with a ramp noise-power spectrum (NPS), with subsequent smoothing through apodization filters. The features considered are lesion size, indistinct lesion boundary, and a nonuniform lesion interior. System resolution is implemented by an MTF with resolution (10% max.) of 0.47 or 0.58 cyc/mm. Apodization is implemented by a Shepp-Logan filter (Sinc profile) with various cutoffs. Six medically naïve subjects participated in the psychophysical studies, entailing training and testing components for each condition. Training consisted of staircase procedures to find the 80% correct threshold for each subject, and testing involved 2000 psychophysical trials at the threshold value for each subject. Human-observer performance is compared to the Ideal Observer to generate estimates of task efficiency. The significance of imaging factors is assessed using ANOVA. Classification images are used to estimate the linear template weights used by subjects to perform these tasks. Classification-image spectra are used to analyze subject weights in the spatial-frequency domain. Overall, average observer efficiency is relatively low in these experiments (10%-40%) relative to detection and localization studies reported previously. We find significant effects for feature type and apodization level on observer efficiency. Somewhat surprisingly, system resolution is not a significant factor. Efficiency effects of the different features appear to be well explained by the profile of the linear templates in the classification images. Increasingly strong apodization is found to both increase the classification-image weights and to increase the mean-frequency of the classification-image spectra. A secondary analysis of "Unapodized" classification images shows that this is largely due to observers undoing (inverting) the effects of apodization filters. These studies demonstrate that human observers can be relatively inefficient at feature-discrimination tasks in ramp-spectrum noise. Observers appear to be adapting to frequency suppression implemented in apodization filters, but there are residual effects that are not explained by spatial weighting patterns. The studies also suggest that the mechanisms for improving performance through the application of noise-control filters may require further investigation.

The value of using a deep learning image reconstruction algorithm of thinner slice thickness to balance the image noise and spatial resolution in low-dose abdominal CT.

Traditional reconstruction techniques have certain limitations in balancing image quality and reducing radiation dose. The deep learning image reconstruction (DLIR) algorithm opens the door to a new era of medical image reconstruction. The purpose of the study was to evaluate the DLIR images at 1.25 mm thickness in balancing image noise and spatial resolution in low-dose abdominal computed tomography (CT) in comparison with the conventional adaptive statistical iterative reconstruction-V at 40% strength (ASIR-V40%) at 5 and 1.25 mm. This retrospective study included 89 patients who underwent low-dose abdominal CT. Five sets of images were generated using ASIR-V40% at a 5 mm slice thickness and 1.25 mm (high-resolution) with DLIR at 1.25 mm using 3 strengths: low (DLIR-L), medium (DLIR-M), and high (DLIR-H). Qualitative evaluation was performed for image noise, artifacts, and visualization of small structures, while quantitative evaluation was performed for standard deviation (SD), signal-to-noise ratio (SNR), and spatial resolution (defined as the edge rising slope). At 1.25 mm, DLIR-M and DLIR-H images had significantly lower noise (SD in fat: 14.29±3.37 and 9.65±3.44 HU, respectively), higher SNR for liver (3.70±0.78 and 5.64±1.20, respectively), and higher overall image quality (4.30±0.44 and 4.67±0.40, respectively) than did the respective values in ASIR-V40% images (20.60±4.04 HU, 2.60±0.63, and 3.77±0.43; all P values <0.05). Compared with the 5 mm ASIR-V40% images, the 1.25 mm DLIR-H images had lower noise (SD: 9.65±3.44 vs. 13.63±10.03 HU), higher SNR (5.64±1.20 vs. 4.69±1.28), and higher overall image quality scores (4.67±0.40 vs. 3.94±0.46) (all P values <0.001). In addition, DLIR-L, DLIR-M, and DLIR-H images had a significantly higher spatial resolution in terms of edge rising slope (59.66±21.46, 58.52±17.48, and 59.26±13.33, respectively, vs. 33.79±9.23) and significantly higher image quality scores in the visualization of fine structures (4.43±0.50, 4.41±0.49, and 4.38±0.49, respectively vs. 2.62±0.49) than did the 5 mm ASIR-V40 images. The 1.25 mm DLIR-M and DLIR-H images had significantly reduced image noise and improved SNR and overall image quality compared to the 1.25 mm ASIR-V40% images, and they had significantly improved the spatial resolution and visualization of fine structures compared to the 5 mm ASIR-V40% images. DLIR-H images had further reduced image noise compared with the 5 mm ASIR-V40% images, and DLIR-H was the most effective technique at balancing the image noise and spatial resolution in low-dose abdominal CT.

Evaluation of silicone rubber-lead shield's effectiveness in protecting the breast during thoracic CT

Radiation of thoracic computed tomography (CT) involves the breast although it is not considered an organ of interest. According to the International Commission on Radiological Protection (ICRP) No. 103, the breast is an organ with a high level of sensitivity when interacting with x-rays, increasing the potential risk of breast cancer. Therefore, the radiation dose must be optimized while maintaining image quality. The dose optimization can be accomplished using a radiation shield. This study aims to determine the effect of silicone rubber (SR)-lead (Pb) in various thicknesses as an alternative protective material limiting dose and preserving the image quality of the breast in thoracic CT. SR-Pb was made from SR and Pb by a simple method. The SR-Pb had thicknesses of 3, 6, 9, and 12 mm. The breast dose was measured using a CT dose profiler on the surface of the breast phantom. The CT number and the noise level of the resulting image were determined quantitatively. The dose without the radiation shield was 5.4 mGy. The doses measured using shielding with thicknesses of 3, 6, 9, and 12 mm were 5.2, 4.5, 4.3, and 3.3 mGy, respectively. Radiation shielding with a thickness of 12 mm reduced breast surface dose by up to 38%. The CT numbers and noise levels for the left and right breast phantom images were almost the same as those ​​without radiation shields indicating there were only slight artifacts in the image. Therefore, SR-Pb is considered a good shielding material which can be pplied in a clinical setting by placing it directly on the breast surface for dose optimization.

Noise Suppression of Computed Tomography (CT) Images Using Residual Encoder-Decoder Convolutional Neural Network (RED-CNN)

In this study, an in-house residual encoder-decoder convolutional neural network (RED-CNN)-based algorithm was composed and trained using images of cylindrical polymethyl-methacrylate (PMMA) phantom with a diameter of 26 cm at different simulated noise levels. The model was tested on 21 × 26 cm elliptical PMMA computed tomography (CT) phantom images with simulated noise to evaluate its denoising capability using signal to noise ratio (SNR), comparative peak signal-to-noise ratio (cPSNR), structural similarity (SSIM) index, modulation transfer function frequencies (MTF 10 %) and noise power spectra (NPS) values as parameters. Evaluation of a possible decrease of image quality was also performed by testing the model using homogenous water phantom and wire phantom images acquired using different mAs values. Results show that the model was able to consistently increase SNR, cPSNR, SSIM values, and decrease the integral noise power spectra (NPS). However, the noise level on either training or testing data affects the model’s final denoising performance. The lower noise level on testing data images tends to result in over-smoothed images, as indicated by the shift of the NPS curves. In contrast, higher simulated noise level tends to result in less satisfactory denoising performance, as indicated by lower SNR, cPSNR, and SSIM values. Meanwhile, the higher noise level on training data images tends to produce denoised images with reduced sharpness, as indicated by the decrease of the MTF 10 % values. Further studies are required to better understand the character of RED-CNN for CT noise suppression regarding the optimum parameters for best results.

DFSNE-Net: Deviant feature sensitive noise estimate network for low-dose CT denoising

Background:Computed tomography (CT) has radiation problems, and high-quality CT scanning is usually accompanied by high doses of radiation that can be harmful to humans, so low-dose CT denoising has received extensive academic attention. Method:In this paper, firstly, the concept of deviant features is proposed which lead to the hypothesis of correlation between noise and deviant features. Secondly, in order to estimate the noise in CT, a new method of deviant feature perception and downsampling is proposed. Specifically, the deviant feature perception module based on the multi-scale convolutional cooperative (MSC-DFPM) and Filtering module based on the self-information space attention(SISA-FM) are proposed, and construct the deviant feature sensitive noise estimate network (DFSNE-Net), then a balanced loss function and training strategy adapted to the DFSNE-Net are proposed. Finally, noise distribution normalization based on skewness and kurtosis(SK-NDN) and low credible noise suppression based on confidence interval(CI-LCNS) as the noise optimization methods are proposed to optimize the estimated noise of DFSNE-Net, which is applied to the denoising task of CT. Conclusions:Experiments and results demonstrate that our proposed denoising method in this paper obtains a better denoising effect than the current state-of-the-art denoising methods in different evaluation indicator, which proves the hypothesis that noise is strongly correlated with deviant features. It is also proved that the denoising method proposed in this paper can achieve the denoising task for different doses of CT.

Relative stopping power resolution in time-of-flight proton CT

Objective. Proton computed tomography (CT) is similar to x-ray CT but relies on protons rather than photons to form an image. In its most common operation mode, the measured quantity is the amount of energy that a proton has lost while traversing the imaged object from which a relative stopping power map can be obtained via tomographic reconstruction. To this end, a calorimeter which measures the energy deposited by protons downstream of the scanned object has been studied or implemented as energy detector in several proton CT prototypes. An alternative method is to measure the proton’s residual velocity and thus its kinetic energy via the time of flight (TOF) between at least two sensor planes. In this work, we study the RSP resolution, seen as image noise, which can be expected from TOF proton CT systems. Approach. We rely on physics models on the one hand and statistical models of the relevant uncertainties on the other to derive closed form expressions for the noise in projection images. The TOF measurement error scales with the distance between the TOF sensor planes and is reported as velocity error in ps/m. We use variance reconstruction to obtain noise maps of a water cylinder phantom given the scanner characteristics and additionally reconstruct noise maps for a calorimeter-based proton CT system as reference. We use Monte Carlo simulations to verify our model and to estimate the noise due to multiple Coulomb scattering inside the object. We also provide a comparison of TOF helium and proton CT. Main results. We find that TOF proton CT with 30 ps m−1 velocity error reaches similar image noise as a calorimeter-based proton CT system with 1% energy error (1 sigma error). A TOF proton CT system with a 50 ps m−1 velocity error produces slightly less noise than a 2% calorimeter system. Noise in a reconstructed TOF proton CT image is spatially inhomogeneous with a marked increase towards the object periphery. Our modelled noise was consistent with Monte Carlo simulated images. TOF helium CT offers lower RSP noise at equal fluence, but is less advantageous at equal imaging dose. Significance. This systematic study of image noise in TOF proton CT can serve as a guide for future developments of this alternative solution for estimating the residual energy of protons and helium ions after the scanned object.

Impact of Denoising on Deep-Learning-Based Automatic Segmentation Framework for Breast Cancer Radiotherapy Planning.

Simple SummaryWe investigated the contouring data of organs at risk from 40 patients with breast cancer who underwent radiotherapy. The performance of denoising-based auto-segmentation was compared with manual segmentation and conventional deep-learning-based auto-segmentation without denoising. Denoising-based auto-segmentation achieved superior segmentation accuracy on the liver compared with AccuContourTM-based auto-segmentation. This denoising-based auto-segmentation method could provide more precise contour delineation of the liver and reduce the clinical workload.Objective: This study aimed to investigate the segmentation accuracy of organs at risk (OARs) when denoised computed tomography (CT) images are used as input data for a deep-learning-based auto-segmentation framework. Methods: We used non-contrast enhanced planning CT scans from 40 patients with breast cancer. The heart, lungs, esophagus, spinal cord, and liver were manually delineated by two experienced radiation oncologists in a double-blind manner. The denoised CT images were used as input data for the AccuContourTM segmentation software to increase the signal difference between structures of interest and unwanted noise in non-contrast CT. The accuracy of the segmentation was assessed using the Dice similarity coefficient (DSC), and the results were compared with those of conventional deep-learning-based auto-segmentation without denoising. Results: The average DSC outcomes were higher than 0.80 for all OARs except for the esophagus. AccuContourTM-based and denoising-based auto-segmentation demonstrated comparable performance for the lungs and spinal cord but showed limited performance for the esophagus. Denoising-based auto-segmentation for the liver was minimal but had statistically significantly better DSC than AccuContourTM-based auto-segmentation (p < 0.05). Conclusions: Denoising-based auto-segmentation demonstrated satisfactory performance in automatic liver segmentation from non-contrast enhanced CT scans. Further external validation studies with larger cohorts are needed to verify the usefulness of denoising-based auto-segmentation.

QAMaster: A new software framework for phantom-based computed tomography quality assurance.

The regular evaluation of imaging performance of computed tomography (CT) scanners is essential for CT quality assurance. For automation of this process, the software QAMaster was developed at the Universitätsklinikum Erlangen, which provides based on CT scans of the CatPhan® 504 (The Phantom Laboratory, Salem, USA) automated image quality analysis and documentation by evaluating CT number accuracy, spatial linearity, uniformity, contrast‐noise‐ratio, spatial resolution, noise, and slice thickness. Dose assessment is supported by calculations of the weighted computed tomography dose index (CTDI w ) and weighted cone beam dose index (CBDI w ). QAMaster was tested with CatPhan® 504 scans and compared to manual evaluations of these scans, whereby high consistency of the respective results was observed. The CT numbers, spatial linearity, uniformity, contrast‐noise‐ratio, noise, and slice thickness deviated by only (0.13 ± 0.25) HU, (0.02 ± 0.05) mm, (−0.01 ± 0.03)%, 0.8 ± 1.8, (0.131 ± 0.05) HU, and (0.004 ± 0.005) mm between both evaluations, respectively. The QAMaster results for spatial resolution did not differ significantly (p = 0.34) from the CatPhan® 504 based manual resolution assessment. Dose computations were fully consistent between QAMaster and manual calculations. Thus, QAMaster proved to be a comprehensive and functional software for performing an automated CT quality assurance routine. QAMaster will be open‐source after its release.

Comparison of image quality and lesion diagnosis in abdominopelvic unenhanced CT between reduced-dose CT using deep learning post-processing and standard-dose CT using iterative reconstruction: A prospective study

PurposeTo compare image quality and lesion diagnosis between reduced-dose abdominopelvic unenhanced computed tomography (CT) using deep learning (DL) post-processing and standard-dose CT using iterative reconstruction (IR). MethodTotally 251 patients underwent two consecutive abdominopelvic unenhanced CT scans of the same range, including standard and reduced doses, respectively. In group A, standard-dose data were reconstructed by (blend 30 %) IR. In group B, reduced-dose data were reconstructed by filtered back projection reconstruction to obtain group B1 images, and post-processed using the DL algorithm (NeuAI denosing, Neusoft medical, Shenyang, China) with 50 % and 100 % weights to obtain group B2 and B3 images, respectively. Then, CT values of the liver, the second lumbar vertebral centrum, the erector spinae and abdominal subcutaneous fat were measured. CT values, noise levels, signal-to-noise ratios (SNRs), contrast-to-noise ratios (CNRs), radiation doses and subjective scores of image quality were compared. Subjective evaluations of low-density liver lesions were compared by diagnostic results from enhanced CT or Magnetic Resonance Imaging. ResultsGroups B3 and B1 showed the lowest and highest noise levels, respectively (P < 0.001). The SNR and CNR in group B3 were highest (P < 0.001). The radiation dose in group B was reduced by 71.5 % on average compared to group A. Subjective scores in groups A and B2 were highest (P < 0.001). Diagnostic sensitivity and confidence for liver metastases in groups A and B2 were highest (P < 0.001). ConclusionsReduced-dose abdominopelvic unenhanced CT combined with DL post-processing could ensure image quality and satisfy diagnostic needs.

Evaluation of radiation treatment planning by computed tomography metal artifact reduction algorithm

Introduction: The objective of this study was to evaluate the usefulness of Computed Tomography(CT) images acquired through repeated subtraction reconstruction algorithms to reduce metal artifacts in CT Treatment Planning System(TPS).Materials and methods: Origin images of Gammex phantom and Rando phantom and non orthopedic metal artifact reduction(O-MAR) images were obtained after high density implantation. O-MAR applied images were also obtained. For evaluation of images, regions of interest(ROI) were set at five tissue rods and three points directly affected by artifacts in Gammex phantom. CT number and noise were compared and analyzed. To investigate dose change according to CT number change of the surrounding artifact, a three virtual cylindrical target volume was formed on Rando phantom CT images. The average dose was then compared and analyzed.Results: CT number difference according to the application of O-MAR showed significant difference among lung and bone rod and 3 ROI directly affected. Noise difference according to O-MAR application was significantly different in rod except for bone rod. In the treatment plan using Rando phantom, non-O-MAR and O-MAR images showed -4.3 ~ 1.9% and -0.4 ~ 2.3% dose differences, respectively.Conclusion: Applying a metal artifact reduction reconstruction algorithm can reduce image distortion due to high density implantation, improve image quality, and correct CT number.

Half2Half: deep neural network based CT image denoising without independent reference data

Reducing radiation dose of x-ray computed tomography (CT) and thereby decreasing the potential risk to patients are desirable in CT imaging. Deep neural network (DNN) has been proposed to reduce noise in low-dose CT (LdCT) images and showed promising results. However, most existing DNN-based methods require training a neural network using high-quality CT images as the reference. Lack of high-quality reference data has therefore been the bottleneck in the current DNN-based methods. Recently, a noise-to-noise (Noise2Noise) training method was proposed to train a denoising neural network with only noisy images. It has also been applied to LdCT data in both the count domain and image domain. However, the method still requires a separately acquired independent noisy reference image for supervising the training procedure. To address this limitation, we propose a novel method to generate both training inputs and training labels from the existing CT scans, which does not require any additional high-dose CT images or repeated scans. Therefore, existing large noisy dataset can be fully exploited for training a denoising neural network. Our experimental results show that the trained networks can reduce noise in existing CT image and hence improve the image quality for clinical diagnosis.

Performance Assessment of Texture Reproduction in High-Resolution CT.

Assessment of computed tomography (CT) images can be complex due to a number of dependencies that affect system performance. In particular, it is well-known that noise in CT is object-dependent. Such object-dependence can be more pronounced and extend to resolution and image textures with the increasing adoption of model-based reconstruction and processing with machine learning methods. Moreover, such processing is often inherently nonlinear complicating assessments with simple measures of spatial resolution, etc. Similarly, recent advances in CT system design have attempted to improve fine resolution details - e.g., with newer detectors, smaller focal spots, etc. Recognizing these trends, there is a greater need for imaging assessment that are considering specific features of interest that can be placed within an anthropomorphic phantom for realistic emulation and evaluation. In this work, we devise a methodology for 3D-printing phantom inserts using procedural texture generation for evaluation of performance of high-resolution CT systems. Accurate representations of texture have previously been a hindrance to adoption of processing methods like model-based reconstruction, and texture serves as an important diagnostic feature (e.g. heterogeneity of lesions is a marker for malignancy). We consider the ability of different systems to reproduce various textures (as a function of the intrinsic feature sizes of the texture), comparing microCT, cone-beam CT, and diagnostic CT using normal- and high-resolution modes. We expect that this general methodology will provide a pathway for repeatable and robust assessments of different imaging systems and processing methods.