Traditional Computed Tomography Research Articles

Reservoir Computing Based on Oxygen-Vacancy-Mediated X-ray Optical Synaptic Device for Medical CT Bone Diagnosis.

Recognition and judgment of X-ray computed tomography (CT) images play a crucial role in medical diagnosis and disease prevention. However, the storage and calculation of the X-ray imaging system applied in the traditional CT diagnosis is separate, and the pathological judgment is based on doctors' experience, which will affect the timeliness and accuracy of decision-making. In this paper, a simple-structured reservoir computing network (RC) is proposed based on Ga2O3 X-ray optical synaptic devices to recognize medical skeletal CT images with high accuracy. Through oxygen vacancy engineering, Ga2O3 X-ray optical synaptic devices with adjustable photocurrent gain and a persistent photoconductivity effect were obtained. By using the Ga2O3 X-ray optical synaptic device as a reservoir, we constructed an RC network for medical skeletal CT diagnosis and verified its image recognition capability using the MNIST data set with an accuracy of 78.08%. In the elbow skeletal CT image recognition task, the recognition rate is as high as 100%. This work constructs a simple-structured RC network for X-ray image recognition, which is of great significance in applications in medical fields.

Assessing Tissue Differentiation Capabilities in X-ray Imaging Through Cadmium Zinc Telluride (CZT) Photon Counting Detectors.

In recent years, the use of photon-counting detection technology has resulted in significant X-ray imaging research. This advanced technology holds considerable promise for enhancing Computed Tomography (CT) scanners, offering a potential solution to the limitations inherent in traditional CT detectors. The ongoing studies are dedicated to assessing the effectiveness and sensitivity of semiconductor detector materials in photon counting detectors, particularly in their ability to detect soft tissue contrasts. This study aimed to characterize the performance of the Cadmium Zinc Telluride (CZT) photon counting detector in identifying various tissues with a focus on assessing its sensitivity in distinguishing tissue types in X-ray imaging. An optimal frame rate per second (FPS) of the CZT detector was evaluated by setting the X-ray tube voltage and current at 25 keV and 35 keV and 0.5 mA, 1.0 mA respectively. By keeping the optimum FPS fixed, the detector energy thresholds were set in small steps from 15 keV to 35 keV and the tube currents were set for X-ray tubes in ranges of 0.1 mA to 1.0 mA to evaluate the relationship between voltage and current of the X-ray source and counts per second (CPS). The specimens, including fat, liver, muscles, paraffin wax, and contrast media were placed on a stair-step chamber constructed from Plexi-glass, spanning six distinct thickness levels. We investigated X-ray transmission across these tissue samples at varying thicknesses, exploring five energy thresholds: 21 keV, 25 keV, 29 keV, 31 keV, and 45 keV to evaluate their impact on the CPS. This study found that a frame rate of 12 frames per second optimally aligns with the spectral response of the X-ray source. Furthermore, a linear correlation is present between the CPS and the X-ray tube current. The transmission of X-rays through a sample depends on its thickness, a phenomenon observed across different energy thresholds in this study. The detectors exhibit high sensitivity and linearity, ensuring accurate and reliable measurements in diverse imaging scenarios, making them invaluable tools for precise data acquisition in both research and clinical applications.

Precision Identification of Locally Advanced Rectal Cancer in Denoised CT Scans Using EfficientNet and Voting System Algorithms.

Local advanced rectal cancer (LARC) poses significant treatment challenges due to its location and high recurrence rates. Accurate early detection is vital for treatment planning. With magnetic resonance imaging (MRI) being resource-intensive, this study explores using artificial intelligence (AI) to interpret computed tomography (CT) scans as an alternative, providing a quicker, more accessible diagnostic tool for LARC. In this retrospective study, CT images of 1070 T3-4 rectal cancer patients from 2010 to 2022 were analyzed. AI models, trained on 739 cases, were validated using two test sets of 134 and 197 cases. By utilizing techniques such as nonlocal mean filtering, dynamic histogram equalization, and the EfficientNetB0 algorithm, we identified images featuring characteristics of a positive circumferential resection margin (CRM) for the diagnosis of locally advanced rectal cancer (LARC). Importantly, this study employs an innovative approach by using both hard and soft voting systems in the second stage to ascertain the LARC status of cases, thus emphasizing the novelty of the soft voting system for improved case identification accuracy. The local recurrence rates and overall survival of the cases predicted by our model were assessed to underscore its clinical value. The AI model exhibited high accuracy in identifying CRM-positive images, achieving an area under the curve (AUC) of 0.89 in the first test set and 0.86 in the second. In a patient-based analysis, the model reached AUCs of 0.84 and 0.79 using a hard voting system. Employing a soft voting system, the model attained AUCs of 0.93 and 0.88, respectively. Notably, AI-identified LARC cases exhibited a significantly higher five-year local recurrence rate and displayed a trend towards increased mortality across various thresholds. Furthermore, the model's capability to predict adverse clinical outcomes was superior to those of traditional assessments. AI can precisely identify CRM-positive LARC cases from CT images, signaling an increased local recurrence and mortality rate. Our study presents a swifter and more reliable method for detecting LARC compared to traditional CT or MRI techniques.

CT-based radiomics for differentiating peripherally located pulmonary sclerosing pneumocytoma from carcinoid.

Pulmonary sclerosing pneumocytoma (PSP) and pulmonary carcinoid (PC) are difficult to distinguish based on conventional imaging examinations. In recent years, radiomics has been used to discriminate benign from malignant pulmonary lesions. However, the value of radiomics based on computed tomography (CT) images to differentiate PSP from PC has not been well explored. We aimed to investigate the feasibility of radiomics in the differentiation between PSP and PC. Fifty-three PSP and fifty-five PC were retrospectively enrolled and then were randomly divided into the training and test sets. Univariate and multivariable logistic analyses were carried to select clinical predictor related to differential diagnosis of PSP and PC. A total of 1316 radiomics features were extracted from the unenhanced CT (UECT) and contrast-enhanced CT (CECT) images, respectively. The minimum redundancy maximum relevance and the least absolute shrinkage and selection operator were used to select the most significant radiomics features to construct radiomics models. The clinical predictor and radiomics features were integrated to develop combined models. Two senior radiologists independently categorized each patient into PSP or PC group based on traditional CT method. The performances of clinical, radiomics, and combined models in differentiating PSP from PC were investigated by the receiver operating characteristic (ROC) curve. The diagnostic performance was also compared between the combined models and radiologists. In regard to differentiating PSP from PC, the area under the curves (AUCs) of the clinical, radiomics, and combined models were 0.87, 0.96, and 0.99 in the training set UECT, and were 0.87, 0.97, and 0.98 in the training set CECT, respectively. The AUCs of the clinical, radiomics, and combined models were 0.84, 0.92, and 0.97 in the test set UECT, and were 0.84, 0.93, and 0.98 in the test set CECT, respectively. In regard to the differentiation between PSP and PC, the combined model was comparable to the radiomics model, but outperformed the clinical model and the two radiologists, whether in the test set UECT or CECT. Radiomics approaches show promise in distinguishing between PSP and PC. Moreover, the integration of clinical predictor (gender) has the potential to enhance the diagnostic performance even further.

Image reconstruction based on nonlinear diffusion model for limited-angle computed tomography

The problem of limited-angle computed tomography (CT) imaging reconstruction has a wide range of practical applications. Due to various factors such as high x-ray absorption, structural characteristics of the scanned object, and equipment limitations, it is often impractical to obtain a complete angular scan, resulting in limited-angle scan data. In this paper, we propose an iterative image reconstruction algorithm for limited-angle CT. The algorithm carries out a traditional CT reconstruction and a nonlinear diffusion process alternatively. Specifically, a subtle partial differential equation is constructed to guide the nonlinear diffusion process to eliminate limited-angle artifacts in the reconstructed image. Numerical experiments on both analytic data and real data validate the efficacy of the proposed nonlinear diffusion reconstruction algorithm. Furthermore, a linear diffusion reconstruction algorithm which combines a traditional CT reconstruction algorithm and a linear diffusion process is also presented in the paper.

Computed tomography-based radiomics improves non-invasive diagnosis of Pneumocystis jirovecii pneumonia in non-HIV patients: a retrospective study

BackgroundPneumocystis jirovecii pneumonia (PCP) could be fatal to patients without human immunodeficiency virus (HIV) infection. Current diagnostic methods are either invasive or inaccurate. We aimed to establish an accurate and non-invasive radiomics-based way to identify the risk of PCP infection in non-HIV patients with computed tomography (CT) manifestation of pneumonia.MethodsThis is a retrospective study including non-HIV patients hospitalized for suspected PCP from January 2010 to December 2022 in one hospital. The patients were randomized in a 7:3 ratio into training and validation cohorts. Computed tomography (CT)-based radiomics features were extracted automatically and used to construct a radiomics model. A diagnostic model with traditional clinical and CT features was also built. The area under the curve (AUC) were calculated and used to evaluate the diagnostic performance of the models. The combination of the radiomics features and serum β-D-glucan levels was also evaluated for PCP diagnosis.ResultsA total of 140 patients (PCP: N = 61, non-PCP: N = 79) were randomized into training (N = 97) and validation (N = 43) cohorts. The radiomics model consisting of nine radiomic features performed significantly better (AUC = 0.954; 95% CI: 0.898-1.000) than the traditional model consisting of serum β-D-glucan levels (AUC = 0.752; 95% CI: 0.597–0.908) in identifying PCP (P = 0.002). The combination of radiomics features and serum β-D-glucan levels showed an accuracy of 95.8% for identifying PCP infection (positive predictive value: 95.7%, negative predictive value: 95.8%).ConclusionsRadiomics showed good diagnostic performance in differentiating PCP from other types of pneumonia in non-HIV patients. A combined diagnostic method including radiomics and serum β-D-glucan has the potential to provide an accurate and non-invasive way to identify the risk of PCP infection in non-HIV patients with CT manifestation of pneumonia.Trial registrationClinicalTrials.gov (NCT05701631).

SOUL-Net: A Sparse and Low-Rank Unrolling Network for Spectral CT Image Reconstruction.

Spectral computed tomography (CT) is an emerging technology, that generates a multienergy attenuation map for the interior of an object and extends the traditional image volume into a 4-D form. Compared with traditional CT based on energy-integrating detectors, spectral CT can make full use of spectral information, resulting in high resolution and providing accurate material quantification. Numerous model-based iterative reconstruction methods have been proposed for spectral CT reconstruction. However, these methods usually suffer from difficulties such as laborious parameter selection and expensive computational costs. In addition, due to the image similarity of different energy bins, spectral CT usually implies a strong low-rank prior, which has been widely adopted in current iterative reconstruction models. Singular value thresholding (SVT) is an effective algorithm to solve the low-rank constrained model. However, the SVT method requires a manual selection of thresholds, which may lead to suboptimal results. To relieve these problems, in this article, we propose a sparse and low-rank unrolling network (SOUL-Net) for spectral CT image reconstruction, that learns the parameters and thresholds in a data-driven manner. Furthermore, a Taylor expansion-based neural network backpropagation method is introduced to improve the numerical stability. The qualitative and quantitative results demonstrate that the proposed method outperforms several representative state-of-the-art algorithms in terms of detail preservation and artifact reduction.

Assessing bone mineral density in non-weight bearing regions of the body: a texture analysis approach using abdomen and pelvis computed tomography Hounsfield units-a cross-sectional study.

Highlighting a gap in comprehending bone microarchitecture's intricacies using dual-energy X-ray absorptiometry (DXA), this study aims to bridge this chasm by analyzing texture in non-weight bearing regions on axial computed tomography (CT) scans. Our goal is to enrich osteoporosis patient management by enhancing bone quality and microarchitecture insights. Conducted at Busan Medical Center from March 1, 2013, to August 30, 2022, 1,320 cases (782 patients) were screened. After applying exclusion criteria, 458 samples (296 patients) underwent bone mineral density (BMD) assessment with both CT and DXA. Regions of interest (ROIs) included spine pedicle's maximum trabecular area, sacrum Zone 1, superior/inferior pubic ramus, and femur's greater/lesser trochanters. Texture features (n=45) were extracted from ROIs using gray-level co-occurrence matrices. A regression model predicted BMD, spotlighting the top five influential texture features. Correlation coefficients ranged from 0.709 (lowest for total femur BMD) to 0.804 (highest for femur intertrochanter BMD). Mean squared error (MSE) values were also provided for lumbar and femur BMD/bone mineral content (BMC) metrics. The most influential texture features included contrast_32, correlation_32_v, and three other metrics. By melding traditional DXA and CT texture analysis, our approach presents a comprehensive bone health perspective, potentially revolutionizing osteoporosis diagnostics.

Iterative Image Reconstruction Methodology in Optical CT Radiochromic Gel Dosimetry

Modern advancements in radiation therapy require paralleled advancements in the dosimetric tools used to verify dose distributions. Optical computed tomography (CT) imaged radiochromic gel dosimeters provide comprehensive, tissue equivalent, 3D dosimetric information with high spatial resolution and low imaging times. Traditional CT image reconstruction methods (filtered backprojection) do not account for light refraction within the optical CT system reducing the image quality. Iterative reconstruction methods make use of a system matrix that describes this light refraction thus, improving the reconstructed image quality. However, use of iterative reconstruction methods is not widespread, largely due to the impractical storage size of the required system matrix. Furthermore, current iterative reconstruction methods do not address the issue of image degradation due to a single detector element collecting light from multiple raypaths. For optical CT radiochromic gel dosimetry to be used effectively as a radiation therapy treatment plan verification tool, the system must be both practical and accurate. Thus, this work has two main objectives: (i) reduce the size of system matrices by means of polar coordinate discretization in lieu of the traditional Cartesian coordinate discretization, and (ii) reduce image degradation due to multiple raypaths by a novel approach to populating the system matrix that accounts for multiple raypaths.

Application of spectral CT in diagnosis, classification and prognostic monitoring of gastrointestinal cancers: progress, limitations and prospects

Gastrointestinal (GI) cancer is the leading cause of cancer-related deaths worldwide. Computed tomography (CT) is an important auxiliary tool for the diagnosis, evaluation, and prognosis prediction of gastrointestinal tumors. Spectral CT is another major CT revolution after spiral CT and multidetector CT. Compared to traditional CT which only provides single-parameter anatomical diagnostic mode imaging, spectral CT can achieve multi-parameter imaging and provide a wealth of image information to optimize disease diagnosis. In recent years, with the rapid development and application of spectral CT, more and more studies on the application of spectral CT in the characterization of GI tumors have been published. For this review, we obtained a substantial volume of literature, focusing on spectral CT imaging of gastrointestinal cancers, including esophageal, stomach, colorectal, liver, and pancreatic cancers. We found that spectral CT can not only accurately stage gastrointestinal tumors before operation but also distinguish benign and malignant GI tumors with improved image quality, and effectively evaluate the therapeutic response and prognosis of the lesions. In addition, this paper also discusses the limitations and prospects of using spectral CT in GI cancer diagnosis and treatment.

Diagnostic Accuracy for Acute Rib Fractures: A Cross-sectional Study Utilizing Automatic Rib Unfolding and 3D Volume-Rendered Reformation

The aim of this study was to compare the use of computed tomography (CT) with automatic rib unfolding and three-dimensional (3D) volume-rendered imaging in the detection and characterization of rib fractures and flail chest. A total of 130 patients with blunt chest trauma underwent whole-body CT, and five independent readers assessed the presence and characterization of rib fractures using traditional CT images, automatic rib unfolding, and 3D volume-rendered images in separate readout sessions at least 2weeks apart. A gold standard was established by consensus among the readers based on the combined analysis of conventional and reformatted images. Automatic rib unfolding significantly reduced mean reading time by 47.5%-74.9% (P<0.0001) while maintaining a comparable diagnostic performance for rib fractures (positive predictive value [PPV] of 82.1%-93.5%, negative predictive value [NPV] of 96.8%-98.2%, and 69.4%-94.2% and 96.9%-99.1% for conventional axial images and 70.4%-85.1% and 95.2%-96.6% for 3D images) and better interobserver agreement (kappa of 0.74-0.87). For flail chest, automatic rib unfolding showed a PPV of 85.7%-100%, NPV of 90.4%-99.0%, and 80.0%-100% and 89.7%-100% for conventional axial images and 76.9%-100% and 89.0%-92.1% for 3D images. Automatic rib unfolding demonstrated equivalent diagnostic performance to conventional images in detecting acute rib fractures and flail chest, with good interobserver agreement and time-saving benefits.

Experimental investigation on interfacial defect detection for SCCS with conventional and novel contact NDT techniques

In order to ensure the mechanical performance and structural safety of steel–concrete composite structures (SCCS), advanced nondestructive testing (NDT) technique for bonding status at steel–concrete interfaces needs to be developed. In this study, the feasibility of multichannel analysis of surface waves (MASW)-based interfacial debonding detection is validated using the contact sensors array. Herein, the multichannel sensor arrays are composed of piezoceramic lead zirconate titanate patches and high-frequency acceleration meters, respectively. For comparison, the impact-response (IR) method and impact-transmission (IT) method are performed utilizing a force hammer and high-frequency acceleration meters, and corresponding damage imaging algorithms are developed. The applicability of ultrasonic computed tomography (CT) scanning test, scanning impact echo (IE) method, ultrasonic tomography (UT) technique on interfacial debonding detection is further discussed in depth. Research findings indicate that the developed contact MASW measurement can fully capture the variation of dispersion characteristics of surface waves induced by debonding defects in SCCS. The developed IR and IT methods are suitable for detecting interfacial debondings in different dimensions. Besides, the damage nephogram resolution of IR is higher than that of the IT method. In addition, the practicability of traditional ultrasonic CT scanning tests, scanning IE method, and UT techniques using commercial equipment is investigated. Experimental observations show that classical NDT testing techniques are incapable of effectively identifying the existence of interfacial damage and are unsuitable for NDT tests on SCCS. The research findings in this study clearly exhibit the precision and limitation of various contact NDT techniques and lay a solid foundation for interfacial debonding detection in practical SCCS.

Spectral Computed Tomography-Guided Radiotherapy of Osteosarcoma utilizing BiOI Nanosheets.

As an aggressive malignant bone tumor, osteosarcoma (OS) is usually found in children and adolescents. Computed tomography (CT) is an important tool for the clinical evaluation of osteosarcoma, but limits to low diagnostic specificity due to single parameters of traditional CT and modest signal-to-noise ratio of clinical iodinated contrast agents. As one kind of spectral CT, dual-energy CT (DECT), with the advantage of a provision of multi-parameter information, makes it possible to acquire the best signal-to-noise ratio image, accurate detection, as well as imaging-guided therapy of bone tumors. Hereby, we synthesized BiOI nanosheets (BiOI NSs) as a DECT contrast agent with superior imaging capability compared to iodine agents for clinical detection of OS. Meanwhile, the synthesized BiOI NSs with great biocompatibility is able to achieve effective radiotherapy (RT) by enhancing X-ray dose deposition at the tumor site, leading to DNA damage, which in turn inhibits tumor growth. This study offers a promising new avenue for DECT imaging-guided treatment of OS. STATEMENT OF SIGNIFICANCE: Osteosarcoma (OS) is a common primary malignant bone tumor. Traditional surgical procedures and conventional CT scans are often used for the treatment and monitoring of OS, but the effects are generally unsatisfactory. In this work, BiOI nanosheets (NSs) was reported for dual-energy CT (DECT) imaging-guided OS radiotherapy. The powerful and constant X-ray absorption of BiOI NSs at any energy guarantees excellent enhanced DECT imaging performance, allowing detailed visualization of OS through images with a better signal-to-noise ratio and guiding radiotherapy process. The deposition of X-rays could be greatly enhanced by Bi atoms to induce serious DNA damage in radiotherapy. Taken together, the BiOI NSs for DECT-guided radiotherapy will greatly improve the current treatment status of OS.

Impact of Nasal Cavity CT Opacification Upon Sinonasal Quality of Life.

Current methods of quantifying inflammation in chronic rhinosinusitis (CRS) on computed tomography (CT) imaging focus on opacification of the paranasal sinuses and show limited correlation with patient-reported outcome measures. This study aimed to determine if quantifying CT opacification of the nasal cavity correlated with Sino-Nasal Outcomes Test scores (SNOT-22). Thirty patients with CRS were enrolled. Lund-Mackay and SNOT-22 scores were measured. Nasal cavity regions of interest (ROIs) were measured by 2 independent raters using ImageJ at 3 points on coronal CT scans: anteriorly at the lacrimal duct, at the approximate midpoint demarcated by the posterior aspect of the globe, and posteriorly at the transition from the hard to soft palate. Superior and inferior regions were defined based upon the root of the inferior turbinate. Percent opacification was calculated for each ROI. Analyses were conducted bilaterally and for the side with greater opacification (worse side). Interrater reliability was strong for all ROIs. Lund-Mackay scores correlated with nasal blockage only (r = .495, P = .01) and did not correlate with nasal cavity ROI opacification. Inferior nasal cavity opacification for worse-sided anterior ROI and middle ROI correlated with SNOT-22 scores for nasal blockage (anterior r = .41, P = .03) (middle r = .42, P = .023) and runny nose (anterior r = .44, P = .02) (middle r = .38, P = .04). Posterior ROIs did not correlate with SNOT-22. Traditional CT scoring of sinus opacification does not correlate well with nasal cavity opacification or SNOT-22. Inferior nasal cavity inflammation provides unique correlations with SNOT-22 nasal questions and may guide targeted interventions in these regions.

Perfusion parameters of triphasic computed tomography hold preoperative prediction value for microvascular invasion in hepatocellular carcinoma

The purpose of this study was to evaluate perfusion parameters of triphasic computed tomography (CT) scans in predicting microvascular invasion (MVI) in hepatocellular carcinoma (HCC). All patients were pathologically diagnosed as HCC and underwent triple-phase enhanced CT imaging, which was used to calculate the blood perfusion parameters of hepatic arterial supply perfusion (HAP), portal vein blood supply perfusion (PVP), hepatic artery perfusion Index (HPI), and arterial enhancement fraction (AEF). Receiver operating characteristic (ROC) curve was used to evaluate the performance. The mean values of PVP(Min), AEF(Min), the difference in PVP, HPI and AEF related parameters, the relative PVP(Min) and AEF(Min) in MVI negative group were significantly higher than those in MVI positive group, while for the difference in HPI(Max), the relative HPI(Max) and AEF(Max), the value of MVI positive group significantly higher than that of negative group. The combination of PVP, HPI and AEF had the highest diagnostic efficacy. The two parameters related to HPI had the highest sensitivity, while the combination of PVP related parameters had higher specificity. A combination of perfusion parameters in patients with HCC derived from traditional triphasic CT scans can be used as a preoperative biomarker for predicting MVI.

Automatic Segmentation Using Deep Convolutional Neural Networks for Tumor CT Images

Tumor computed tomography (CT) image segmentation can provide a basis for the determination of tumor location and type. Therefore, it is of great significance to study the automatic segmentation method of tumor CT images. To address the problem of poor segmentation effect of traditional automatic tumor CT images segmentation methods, we propose an automatic segmentation method for tumor CT images using deep convolutional neural networks (DCNNs). First, the CT tumor image is simplified. According to the features of the target region and the background region, the distribution features of the tumor region in the CT images are obtained by convolution calculation, and the feature extraction is completed by feature fusion. Second, based on the feature extraction results, a deep supervised network is constructed to determine the image depth, which lays a solid foundation for accurate segmentation of tumor regions. Finally, DCNN was used to construct automatic segmentation for tumor CT images, which achieves the automatic segmentation of tumor CT images by mode calculation. The results show that the segmented tumor region is close to the actual region and the maximum pixel loss coefficient is 0.07, the maximum segmentation sensitivity is 7865[Formula: see text]kbps/s, the pixel segmentation specific coefficient and the segmentation edge distance are kept at a low level, which has a certain application value in the field of tumor CT images.

A Multienergy Computed Tomography Method Based on a Blind Decomposition Model for Multivoltage X-Ray Transmission Images

Multienergy computed tomography (CT) can be implemented based on traditional CT systems to reduce the hardware cost. However, prior knowledge of the X-ray spectra or material composition and multiple scans are required in most such implementation methods, which imposes an extra burden. To avoid this, a method based on a blind decomposition model of multivoltage X-ray transmission images is proposed in this paper. The model considers X-ray scattering as an important factor resulting in errors of theoretical and measured polychromatic X-ray transmission. Since X-ray scatter has low-frequency characteristics, it is estimated via local averaging of the errors and subtracted from the errors. The solution algorithm for the model is derived with the Karush–Kuhn–Tucker conditions. In addition, image segmentation of the projection or reconstructed image is not needed. Samples representing different imaging situations are used in the experiments. Beam hardening artifacts are obviously reduced in the reconstructed images obtained with the proposed method. The relative errors of the X-ray energies of different component materials are calculated for the reconstructed images. For the first sample, the errors are less than 1% in the reconstructed images obtained with the proposed method and are seen to be the best among all compared methods. For the second sample, the errors are less than 10% in most reconstructed images obtained with the proposed method, whereas they are larger than 10% in most reconstructed images obtained with the compared methods. These results demonstrate the effectiveness of the proposed method.

A multienergy computed tomography method without image segmentation or prior knowledge of X-ray spectra or materials

Many methods have been proposed for multienergy computed tomography (CT) imaging based on traditional CT systems. Usually, either prior knowledge of the X-ray spectra distribution or materials or the segmentation of the projection or reconstructed image is needed. To avoid these requirements, a multienergy CT method is proposed in this paper. A CT image can be seen as a linear combination of energy-dependent components and spatially dependent components. The latter components are the base images, while the former components are the coefficients. A blind decomposition model is constructed to decompose the multivoltage projections to obtain the base images and the energies. Multienergy CT images are computationally synthesized with the base images and the energies. Multivoltage projections can be acquired based on one scan with stepped voltages. X-ray scattering is considered an important factor in imaging errors and appears as a low-frequency signal. The variance is used to describe the low-frequency features and is minimized as the optimized objective function of the decomposition model. The solution of the model uses Karush–Kuhn–Tucker (KKT) conditions. In the experiments, the images reconstructed with the proposed method exhibit weak beam-hardening artifacts. Additionally, the X-ray energies of the different materials represented have small relative errors. Therefore, the reconstructed images have narrow energy intervals. This shows the effectiveness of the proposed method.

CT image super-resolution reconstruction based on global hybrid attention

Computer tomography (CT) has played an essential role in the field of medical diagnosis, but the blurry edges and unclear textures in traditional CT images usually interfere the subsequent judgement from radiologists or clinicians. Deep learning based image super-resolution methods have been applied for CT image restoration recently. However, different levels of information of CT image details are mixed and difficult to be mapped from deep features by traditional convolution operations. Moreover, features representing regions of interest (ROIs) in CT images are treated equally as those for background, resulting in low concentration of meaningful features and high redundancy of computation. To tackle these issues, a CT image super-resolution network is proposed based on hybrid attention mechanism and global feature fusion, which consists of the following three parts: 1) stacked Swin Transformer blocks are used as the backbone to extract initial features from the degraded CT image; 2) a multi-branch hierarchical self-attention module (MHSM) is proposed to adaptively map multi-level features representing different levels of image information from the initial features and establish the relationship between these features through a self-attention mechanism, where three branches apply different strategies of integrating convolution, down-sampling and up-sampling operations according to three different scale factors; 3) a multidimensional local topological feature enhancement module (MLTEM) is proposed and plugged into the end of the backbone to refine features in the channel and spatial dimension simultaneously, so that the features representing ROIs could be enhanced while meaningless ones eliminated. Experimental results demonstrate that our method outperform the state-of-the-art super-resolution methods on restoring CT images with respect to peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) indices.

Statistical Interior Tomography via L1 Norm Dictionary Learning without Assuming an Object Support.

Interior tomography of X-ray computed tomography (CT) has many advantages, such as a lower radiation dose and lower detector hardware cost compared to traditional CT. However, this imaging technique only uses the projection data passing through the region of interest (ROI) for imaging; accordingly, the projection data are truncated at both ends of the detector, so the traditional analytical reconstruction algorithm cannot satisfy the demand of clinical diagnosis. To solve the above limitations, in this paper we propose a high-quality statistical iterative reconstruction algorithm that uses the zeroth-order image moment as novel prior knowledge; the zeroth-order image moment can be estimated in the projection domain using the Helgason–Ludwig consistency condition. Then, the L1norm of sparse representation, in terms of dictionary learning, and the zeroth-order image moment constraints are incorporated into the statistical iterative reconstruction framework to construct an objective function. Finally, the objective function is minimized using an alternating minimization iterative algorithm. The chest CT image simulated and CT real data experimental results demonstrate that the proposed approach can remove shift artifacts effectively and has superior performance in removing noise and persevering fine structures than the total variation (TV)-based approach.