Regularization strategies in statistical image reconstruction of low-dose x-ray CT: A review

When are Diffusion Priors Helpful in Sparse Reconstruction? A Study with Sparse-view CT

Diffusion models demonstrate state-of-the-art performance on image generation, and are gaining traction for sparse medical image reconstruction tasks. However, compared to classical reconstruction algorithms relying on simple analytical priors, diffusion models have the dangerous property of producing realistic looking results even when incorrect, particularly with few observations. We investigate the utility of diffusion models as priors for image reconstruction by varying the number of observations and comparing their performance to classical priors (sparse and Tikhonov regularization) using pixel-based, structural, and downstream metrics. We make comparisons on low-dose chest wall computed tomography (CT) for fat mass quantification. First, we find that classical priors are superior to diffusion priors when the number of projections is "sufficient". Second, we find that diffusion priors can capture a large amount of detail with very few observations, significantly outperforming classical priors. However, they fall short of capturing all details, even with many observations. Finally, we find that the performance of diffusion priors plateau after extremely few (≈10-15) projections. Ultimately, our work highlights potential issues with diffusion-based sparse reconstruction and underscores the importance of further investigation, particularly in high-stakes clinical settings.

Data-driven and model-guided iterative reconstruction framework for simultaneous sparse-view and ring artifact reduction in synchrotron X-ray microtomography

Synchrotron X-ray microtomography (S-µCT) is a highly valuable technique for investigating organ function and pathologies. However, its application is often limited by high radiation doses and the occurrence of ring artifacts. While S-µCT utilizing sparse-view projections can effectively decrease radiation doses, the reconstructed images frequently exhibit severe streaking artifacts, which are exacerbated by ring artifacts, ultimately compromising reconstruction accuracy, image quality, and resolution. Previous research has primarily focused on either sparse-view CT reconstruction or ring artifact reduction, leaving the issue of simultaneous sparse-view and ring artifact reduction under-explored. In this study, we propose a data-driven and model-guided iterative reconstruction framework for S-µCT to address this issue. Specifically, this framework integrates a data prior derived from a score-based generative model to tackle the streaking artifacts introduced by sparse-view projections, along with a model prior obtained from a regularization model to suppress ring artifacts. To assess the effectiveness and capabilities of the proposed framework, simulations using foam phantoms and real S-µCT experiments involving rat liver samples were conducted. The results demonstrated that the proposed framework effectively reduces both streaking and ring artifacts, yielding high-quality S-µCT images with significant reconstruction accuracy and improved image resolution. These findings suggest that the proposed framework holds considerable promise for expanding the application of S-µCT in the future.

Comparison of iterative reconstruction implementations for multislice helical CT

The most mature image reconstruction algorithms in multislice helical computed tomography are based on analytical and iterative methods. Over the past decades, several methods have been developed for iterative reconstructions that improve image quality by reducing noise and artifacts. In the regularization step of iterative reconstruction, noise can be significantly reduced, thereby making low-dose CT. The quality of the reconstructed image can be further improved by using model-based reconstructions. In these reconstructions, the main focus is on modeling the data acquisition process, including the behavior of the photon beams, the geometry of the system, etc. In this article, we propose two model-based reconstruction algorithms using a virtual detector for multislice helical CT. The aim of this study is to compare the effect of using a virtual detector on image quality for the two proposed algorithms with a model-based iterative reconstruction using the original detector model. Since the algorithms are implemented using multiple GPUs, the merging of separately reconstructed volumes can significantly affect image quality. This issue is often referred to as the "long object" problem, for which we also present a solution that plays an important role in the proposed reconstruction processes. The algorithms were evaluated using mathematical and physical phantoms, as well as patient cases. The SSIM, MS-SSIM and L1 metrics were utilized to evaluate the image quality of the mathematical phantom case. To demonstrate the effectiveness of the algorithms, we used the CatPhan 600 phantom. Additionally, anonymized patient scans were used to showcase the improvements in image quality on real scan data.

Machine Learned Texture Prior From Full-Dose CT Database via Multi-Modality Feature Selection for Bayesian Reconstruction of Low-Dose CT

In our earlier study, we proposed a regional Markov random field type tissue-specific texture prior from previous full-dose computed tomography (FdCT) scan for current low-dose CT (LdCT) imaging, which showed clinical benefits through task-based evaluation. Nevertheless, two assumptions were made for early study. One assumption is that the center pixel has a linear relationship with its nearby neighbors and the other is previous FdCT scans of the same subject are available. To eliminate the two assumptions, we proposed a database assisted end-to-end LdCT reconstruction framework which includes a deep learning texture prior model and a multi-modality feature based candidate selection model. A convolutional neural network-based texture prior is proposed to eliminate the linear relationship assumption. And for scenarios in which the concerned subject has no previous FdCT scans, we propose to select one proper prior candidate from the FdCT database using multi-modality features. Features from three modalities are used including the subjects' physiological factors, the CT scan protocol, and a novel feature named Lung Mark which is deliberately proposed to reflect the z-axial property of human anatomy. Moreover, a majority vote strategy is designed to overcome the noise effect from LdCT scans. Experimental results showed the effectiveness of Lung Mark. The selection model has accuracy of 84% testing on 1,470 images from 49 subjects. The learned texture prior from FdCT database provided reconstruction comparable to the subjects having corresponding FdCT. This study demonstrated the feasibility of bringing clinically relevant textures from available FdCT database to perform Bayesian reconstruction of any current LdCT scan.

ATNet: A Defect Detection Framework for X-ray Images of DIP Chip Lead Bonding

In order to improve the production quality and qualification rate of chips, X-ray nondestructive imaging technology has been widely used in the detection of chip defects, which represents an important part of the quality inspection of products after packaging. However, the current traditional defect detection algorithm cannot meet the demands of high accuracy, fast speed, and real-time chip defect detection in industrial production. Therefore, this paper proposes a new multi-scale feature fusion module (ATSPPF) based on convolutional neural networks, which can more fully extract semantic information at different scales. In addition, based on this module, we design a deep learning model (ATNet) for detecting lead defects in chips. The experimental results show that at 8.2 giga floating point operations (GFLOPs) and 146 frames per second (FPS), mAP_0.5 and mAP_0.5-0.95 can achieve an average accuracy of 99.4% and 69.3%, respectively, while the detection speed is faster than the baseline yolov5s by nearly 50%.

Quantitative Analysis of Image Quality in Low-Dose Computed Tomography Imaging for COVID-19 Patients

Background

Computed tomography (CT) scan is one of the main tools to diagnose and grade COVID-19 progression. To avoid the side effects of CT imaging, low-dose CT imaging is of crucial importance to reduce population absorbed dose. However, this approach introduces considerable noise levels in CT images.

Methods

In this light, we set out to simulate four reduced dose levels (60% dose, 40% dose, 20% dose, and 10% dose) of standard CT imaging using Beer-Lambert's law across 49 patients infected with COVID-19. Then, three denoising filters, namely Gaussian, bilateral, and median, were applied to the different low-dose CT images, the quality of which was assessed prior to and after the application of the various filters via calculation of peak signal-to-noise ratio, root mean square error (RMSE), structural similarity index measure, and relative CT-value bias, separately for the lung tissue and whole body.

Results

The quantitative evaluation indicated that 10%-dose CT images have inferior quality (with RMSE = 322.1 ± 104.0 HU and bias = 11.44% ± 4.49% in the lung) even after the application of the denoising filters. The bilateral filter exhibited superior performance to suppress the noise and recover the underlying signals in low-dose CT images compared to the other denoising techniques. The bilateral filter led to RMSE and bias of 100.21 ± 16.47 HU and - 0.21% ± 1.20%, respectively, in the lung regions for 20%-dose CT images compared to the Gaussian filter with RMSE = 103.46 ± 15.70 HU and bias = 1.02% ± 1.68% and median filter with RMSE = 129.60 ± 18.09 HU and bias = -6.15% ± 2.24%.

Conclusions

The 20%-dose CT imaging followed by the bilateral filtering introduced a reasonable compromise between image quality and patient dose reduction.

Physics-informed sinogram completion for metal artifact reduction in CT imaging

Objective.Metal artifacts in the computed tomography (CT) imaging are unavoidably adverse to the clinical diagnosis and treatment outcomes. Most metal artifact reduction (MAR) methods easily result in the over-smoothing problem and loss of structure details near the metal implants, especially for these metal implants with irregular elongated shapes. To address this problem, we present the physics-informed sinogram completion (PISC) method for MAR in CT imaging, to reduce metal artifacts and recover more structural textures.Approach.Specifically, the original uncorrected sinogram is firstly completed by the normalized linear interpolation algorithm to reduce metal artifacts. Simultaneously, the uncorrected sinogram is also corrected based on the beam-hardening correction physical model, to recover the latent structure information in metal trajectory region by leveraging the attenuation characteristics of different materials. Both corrected sinograms are fused with the pixel-wise adaptive weights, which are manually designed according to the shape and material information of metal implants. To furtherly reduce artifacts and improve the CT image quality, a post-processing frequency split algorithm is adopted to yield the final corrected CT image after reconstructing the fused sinogram.Main results.We qualitatively and quantitatively evaluated the presented PISC method on two simulated datasets and three real datasets. All results demonstrate that the presented PISC method can effectively correct the metal implants with various shapes and materials, in terms of artifact suppression and structure preservation.Significance.We proposed a sinogram-domain MAR method to compensate for the over-smoothing problem existing in most MAR methods by taking advantage of the physical prior knowledge, which has the potential to improve the performance of the deep learning based MAR approaches.

Comparison of supervised-learning approaches for designing a channelized observer for image quality assessment in CT

BACKGROUND

The current paradigm for evaluating computed tomography (CT) system performance relies on a task-based approach. As the Hotelling observer (HO) provides an upper bound of observer performances in specific signal detection tasks, the literature advocates HO use for optimization purposes. However, computing the HO requires calculating the inverse of the image covariance matrix, which is often intractable in medical applications. As an alternative, dimensionality reduction has been extensively investigated to extract the task-relevant features from the raw images. This can be achieved by using channels, which yields the channelized-HO (CHO). The channels are only considered efficient when the channelized observer (CO) can approximate its unconstrained counterpart. Previous work has demonstrated that supervised learning-based methods can usually benefit CO design, either for generating efficient channels using partial least squares (PLS) or for replacing the Hotelling detector with machine-learning (ML) methods.

PURPOSE

Here we investigated the efficiency of a supervised ML-algorithm used to design a CO for predicting the performance of unconstrained HO. The ML-algorithm was applied either (1) in the estimator for dimensionality reduction, or (2) in the detector function.

METHODS

A channelized support vector machine (CSVM) was employed and compared against the CHO in terms of ability to predict HO performances. Both the CSVM and the CHO were estimated with channels derived from the singular value decomposition (SVD) of the system operator, principal component analysis (PCA), and PLS. The huge variety of regularization strategies proposed by CT system vendors for statistical image reconstruction (SIR) make the generalization capability of an observer a key point to consider upfront of implementation in clinical practice. To evaluate the generalization properties of the observers, we adopted a 2-step testing process: (1) achieved with the same regularization strategy (as in the training phase) and (2) performed using different reconstruction properties. We generated simulated- signal-known-exactly/background-known-exactly (SKE/BKE) tasks in which different noise structures were generated using Markov random field (MRF) regularizations using either a Green or a quadratic, function.

RESULTS

The CSVM outperformed the CHO for all types of channels and regularization strategies. Furthermore, even though both COs generalized well to images reconstructed with the same regularization strategy as the images considered in the training phase, the CHO failed to generalize to images reconstructed differently whereas the CSVM managed to successfully generalize. Lastly, the proposed CSVM observer used with PCA channels outperformed the CHO with PLS channels while using a smaller training data set.

CONCLUSION

These results argue for introducing the supervised-learning paradigm in the detector function rather than in the operator of the channels when designing a CO to provide an accurate estimate of HO performance. The CSVM with PCA channels proposed here could be used as a surrogate for HO in image quality assessment.

Deep learning-based noise reduction for coronary CT angiography: using four-dimensional noise-reduction images as the ground truth

Background

To assess low-contrast areas such as plaque and coronary artery stenosis, coronary computed tomography angiography (CCTA) needs to provide images with lower noise without increasing radiation doses.

Purpose

To develop a deep learning-based noise-reduction method for CCTA using four-dimensional noise reduction (4DNR) as the ground truth for supervised learning.

Material and Methods

\We retrospectively collected 100 retrospective ECG-gated CCTAs. We created 4DNR images using non-rigid registration and weighted averaging three timeline CCTA volumetric data with intervals of 50 ms in the mid-diastolic phase. Our method set the original reconstructed image as the input and the 4DNR as the target image and obtained the noise-reduced image via residual learning. We evaluated the objective image quality of the original and deep learning-based noise-reduction (DLNR) images based on the image noise of the aorta and the contrast-to-noise ratio (CNR) of the coronary arteries. Further, a board-certified radiologist evaluated the blurring of several heart structures using a 5-point Likert scale subjectively and assigned a coronary artery disease reporting and data system (CAD-RADS) category independently.

Results

DLNR CCTAs showed 64.5% lower image noise ( P < 0.001) and achieved a 2.9 times higher CNR of coronary arteries than that in original images, without significant blurring in subjective comparison ( P > 0.1). The intra-observer agreement of CAD-RADS in the DLNR image was excellent (0.87, 95% confidence interval = 0.77–0.99) with original CCTAs.

Conclusion

Our DLNR method supervised by 4DNR significantly reduced the image noise of CCTAs without affecting the assessment of coronary stenosis.

Noise Characteristics Modeled Unsupervised Network for Robust CT Image Reconstruction

Deep learning (DL)-based methods show great potential in computed tomography (CT) imaging field. The DL-based reconstruction methods are usually evaluated on the training and testing datasets which are obtained from the same distribution, i.e., the same CT scan protocol (i.e., the region setting, kVp, mAs, etc.). In this work, we focus on analyzing the robustness of the DL-based methods against protocol-specific distribution shifts (i.e., the training and testing datasets are from different region settings, different kVp settings, or different mAs settings, respectively). The results show that the DL-based reconstruction methods are sensitive to the protocol-specific perturbations which can be attributed to the noise distribution shift between the training and testing datasets. Based on these findings, we presented a low-dose CT reconstruction method using an unsupervised strategy with the consideration of noise distribution to address the issue of protocol-specific perturbations. Specifically, unpaired sinogram data is enrolled into the network training, which represents unique information for specific imaging protocol, and a Gaussian mixture model (GMM) is introduced to characterize the noise distribution in CT images. It can be termed as GMM based unsupervised CT reconstruction network (GMM-unNet) method. Moreover, an expectation-maximization algorithm is designed to optimize the presented GMM-unNet method. Extensive experiments are performed on three datasets from different scan protocols, which demonstrate that the presented GMM-unNet method outperforms the competing methods both qualitatively and quantitatively.

Real-World and Regulatory Perspectives of Artificial Intelligence in Cardiovascular Imaging

Recent progress in digital health data recording, advances in computing power, and methodological approaches that extract information from data as artificial intelligence are expected to have a disruptive impact on technology in medicine. One of the potential benefits is the ability to extract new and essential insights from the vast amount of data generated during health care delivery every day. Cardiovascular imaging is boosted by new intelligent automatic methods to manage, process, segment, and analyze petabytes of image data exceeding historical manual capacities. Algorithms that learn from data raise new challenges for regulatory bodies. Partially autonomous behavior and adaptive modifications and a lack of transparency in deriving evidence from complex data pose considerable problems. Controlling new technologies requires new controlling techniques and ongoing regulatory research. All stakeholders must participate in the quest to find a fair balance between innovation and regulation. The regulatory approach to artificial intelligence must be risk-based and resilient. A focus on unknown emerging risks demands continuous surveillance and clinical evaluation during the total product life cycle. Since learning algorithms are data-driven, high-quality data is fundamental for good machine learning practice. Mining, processing, validation, governance, and data control must account for bias, error, inappropriate use, drifts, and shifts, particularly in real-world data. Regulators worldwide are tackling twenty-first century challenges raised by "learning" medical devices. Ethical concerns and regulatory approaches are presented. The paper concludes with a discussion on the future of responsible artificial intelligence.

Fast Model Predictive Control of PEM Fuel Cell System Using the L1 Norm

This work describes the development of a fast Model Predictive Control (MPC) algorithm for a Proton Exchange Membrane (PEM) fuel cell. The MPC cost-function used considers the sum of absolute values of predicted control errors (the L1 norm). Unlike previous approaches to nonlinear MPC-L1, in which quite complicated neural approximators have been used, two analytical approximators of the absolute value function are utilised. An advanced trajectory linearisation is performed on-line. As a result, an easy-to-solve quadratic optimisation task is derived. All implementation details of the discussed algorithm are detailed for two considered approximators. Furthermore, the algorithm is thoroughly compared with the classical MPC-L2 method in which the sum of squared predicted control errors is minimised. A multi-criteria control quality assessment is performed as the MPC-L1 and MPC-L2 algorithms are compared using four control quality indicators. It is shown that the presented MPC-L1 scheme gives better results for the PEM.

Artificial Intelligence in Radiation Therapy

Artificial intelligence (AI) has great potential to transform the clinical workflow of radiotherapy. Since the introduction of deep neural networks, many AI-based methods have been proposed to address challenges in different aspects of radiotherapy. Commercial vendors have started to release AI-based tools that can be readily integrated to the established clinical workflow. To show the recent progress in AI-aided radiotherapy, we have reviewed AI-based studies in five major aspects of radiotherapy including image reconstruction, image registration, image segmentation, image synthesis, and automatic treatment planning. In each section, we summarized and categorized the recently published methods, followed by a discussion of the challenges, concerns, and future development. Given the rapid development of AI-aided radiotherapy, the efficiency and effectiveness of radiotherapy in the future could be substantially improved through intelligent automation of various aspects of radiotherapy.

Low-dose cone-beam computed tomography reconstruction through a fast three-dimensional compressed sensing method based on the three-dimensional pseudo-polar fourier transform

Background:

Reconstruction of high quality two dimensional images from fan beam computed tomography (CT) with a limited number of projections is already feasible through Fourier based iterative reconstruction method. However, this article is focused on a more complicated reconstruction of three dimensional (3D) images in a sparse view cone beam computed tomography (CBCT) by utilizing Compressive Sensing (CS) based on 3D pseudo polar Fourier transform (PPFT).

Method:

In comparison with the prevalent Cartesian grid, PPFT re gridding is potent to remove rebinning and interpolation errors. Furthermore, using PPFT based radon transform as the measurement matrix, reduced the computational complexity.

Results:

In order to show the computational efficiency of the proposed method, we compare it with an algebraic reconstruction technique and a CS type algorithm. We observed convergence in <20 iterations in our algorithm while others would need at least 50 iterations for reconstructing a qualified phantom image. Furthermore, using a fast composite splitting algorithm solver in each iteration makes it a fast CBCT reconstruction algorithm. The algorithm will minimize a linear combination of three terms corresponding to a least square data fitting, Hessian (HS) Penalty and l1 norm wavelet regularization. We named it PP-based compressed sensing-HS-W. In the reconstruction range of 120 projections around the 360° rotation, the image quality is visually similar to reconstructed images by Feldkamp-Davis-Kress algorithm using 720 projections. This represents a high dose reduction.

Conclusion:

The main achievements of this work are to reduce the radiation dose without degrading the image quality. Its ability in removing the staircase effect, preserving edges and regions with smooth intensity transition, and producing high-resolution, low-noise reconstruction results in low-dose level are also shown.

Self-Bayesian aberration removal via constraints for ultracold atom microscopy

High-resolution imaging of ultracold atoms typically requires custom high numerical aperture (NA) optics, as is the case for quantum gas microscopy. These high NA objectives involve many optical elements, each of which contributes to loss and light scattering, making them unsuitable for quantum backaction limited "weak" measurements. We employ a low-cost high NA aspheric lens as an objective for a practical and economical-although aberrated-high-resolution microscope to image ⁸⁷Rb Bose-Einstein condensates. Here, we present a methodology for digitally eliminating the resulting aberrations that is applicable to a wide range of imaging strategies and requires no additional hardware. We recover nearly the full NA of our objective, thereby demonstrating a simple and powerful digital aberration correction method for achieving optimal microscopy of quantum objects. This reconstruction relies on a high-quality measure of our imaging system's even-order aberrations from density-density correlations measured with differing degrees of defocus. We demonstrate our aberration compensation technique using phase-contrast imaging, a dispersive imaging technique directly applicable to quantum backaction limited measurements. Furthermore, we show that our digital correction technique reduces the contribution of photon shot noise to density-density correlation measurements which would otherwise contaminate the desired quantum projection noise signal in weak measurements.

Core Imaging Library - Part I: a versatile Python framework for tomographic imaging

We present the Core Imaging Library (CIL), an open-source Python framework for tomographic imaging with particular emphasis on reconstruction of challenging datasets. Conventional filtered back-projection reconstruction tends to be insufficient for highly noisy, incomplete, non-standard or multi-channel data arising for example in dynamic, spectral and in situ tomography. CIL provides an extensive modular optimization framework for prototyping reconstruction methods including sparsity and total variation regularization, as well as tools for loading, preprocessing and visualizing tomographic data. The capabilities of CIL are demonstrated on a synchrotron example dataset and three challenging cases spanning golden-ratio neutron tomography, cone-beam X-ray laminography and positron emission tomography. This article is part of the theme issue 'Synergistic tomographic image reconstruction: part 2'.

Core Imaging Library - Part I: a versatile Python framework for tomographic imaging

We present the Core Imaging Library (CIL), an open-source Python framework for tomographic imaging with particular emphasis on reconstruction of challenging datasets. Conventional filtered back-projection reconstruction tends to be insufficient for highly noisy, incomplete, non-standard or multi-channel data arising for example in dynamic, spectral and in situ tomography. CIL provides an extensive modular optimization framework for prototyping reconstruction methods including sparsity and total variation regularization, as well as tools for loading, preprocessing and visualizing tomographic data. The capabilities of CIL are demonstrated on a synchrotron example dataset and three challenging cases spanning golden-ratio neutron tomography, cone-beam X-ray laminography and positron emission tomography. This article is part of the theme issue 'Synergistic tomographic image reconstruction: part 2'.

Musiré: multimodal simulation and reconstruction framework for the radiological imaging sciences

A software-based workflow is proposed for managing the execution of simulation and image reconstruction for SPECT, PET, CBCT, MRI, BLI and FMI packages in single and multimodal biomedical imaging applications. The workflow is composed of a Bash script, the purpose of which is to provide an interface to the user, and to organize data flow between dedicated programs for simulation and reconstruction. The currently incorporated simulation programs comprise GATE for Monte Carlo simulation of SPECT, PET and CBCT, SpinScenario for simulating MRI, and Lipros for Monte Carlo simulation of BLI and FMI. Currently incorporated image reconstruction programs include CASToR for SPECT and PET as well as RTK for CBCT. MetaImage (mhd) standard is used for voxelized phantom and image data format. Meshlab project (mlp) containers incorporating polygon meshes and point clouds defined by the Stanford triangle format (ply) are employed to represent anatomical structures for optical simulation, and to represent tumour cell inserts. A number of auxiliary programs have been developed for data transformation and adaptive parameter assignment. The software workflow uses fully automatic distribution to, and consolidation from, any number of Linux workstations and CPU cores. Example data are presented for clinical SPECT, PET and MRI systems using the Mida head phantom and for preclinical X-ray, PET and BLI systems employing the Digimouse phantom. The presented method unifies and simplifies multimodal simulation setup and image reconstruction management and might be of value for synergistic image research. This article is part of the theme issue 'Synergistic tomographic image reconstruction: part 2'.

Dual-path deep learning reconstruction framework for propagation-based X-ray phase-contrast computed tomography with sparse-view projections

Propagation-based X-ray phase-contrast computed tomography (PB-PCCT) can serve as an effective tool for studying organ function and pathologies. However, it usually suffers from a high radiation dose due to the long scan time. To alleviate this problem, we propose a deep learning reconstruction framework for PB-PCCT with sparse-view projections. The framework consists of dual-path deep neural networks, where the edge detection, edge guidance, and artifact removal models are incorporated into two subnetworks. It is worth noting that the framework has the ability to achieve excellent performance by exploiting the data-based knowledge of the sample material characteristics and the model-based knowledge of PB-PCCT. To evaluate the effectiveness and capability of the proposed framework, simulations and real experiments were performed. The results demonstrated that the proposed framework could significantly suppress streaking artifacts and produce high-contrast and high-resolution computed tomography images.

[Artificial intelligence in image evaluation and diagnosis]

Die Radiologie ist von stetem Wandel geprägt und definiert sich über den technologischen Fortschritt. Künstliche Intelligenz (KI) wird die praktische Tätigkeit in der Kinder- und Jugendradiologie künftig in allen Belangen verändern. Bildakquisition, Befunderkennung und -segmentierung sowie die Erkennung von Gewebeeigenschaften und deren Kombination mit Big Data werden die Haupteinsatzgebiete in der Radiologie sein. Höhere Effektivität, Beschleunigung von Untersuchung und Befundung sowie Kosteneinsparung sind mit der Anwendung von KI verbundene Erwartungshaltungen. Ein verbessertes Patientenmanagement, Arbeitserleichterungen für medizinisch-technische Radiologieassistenten und Kinder- und Jugendradiologen sowie schnellere Untersuchungs- und Befundzeiten markieren die Meilensteine der KI-Entwicklung in der Radiologie. Von der Terminkommunikation und Gerätesteuerung bis zu Therapieempfehlung und -monitoring wird der Alltag durch Elemente der KI verändert. Kinder- und Jugendradiologen müssen daher grundlegend über KI informiert sein und mit Datenwissenschaftlern bei der Etablierung und Anwendung von KI-Elementen zusammenarbeiten.

Prior-image-based CT reconstruction using attenuation-mismatched priors

Prior-image-based reconstruction (PIBR) methods are powerful tools for reducing radiation doses and improving the image quality of low-dose computed tomography (CT). Apart from anatomical changes, prior and current images can also have different attenuations because they originated from different scanners or from the same scanner but with different x-ray beam qualities (e.g., kVp settings, beam filters) during data acquisition. In such scenarios, with attenuation-mismatched priors, PIBR is challenging. In this work, we investigate a specific PIBR method, called statistical image reconstruction, using normal-dose image-induced nonlocal means regularization (SIR-ndiNLM), to address PIBR with such attenuation-mismatched priors and achieve quantitative low-dose CT imaging. We propose two corrective schemes for the original SIR-ndiNLM method, (1) a global histogram-matching approach and (2) a local attenuation correction approach, to account for the attenuation differences between the prior and current images in PIBR. We validate the efficacy of the proposed schemes using images acquired from dual-energy CT scanners to simulate attenuation mismatches. Meanwhile, we utilize different CT slices to simulate anatomical mismatches or changes between the prior and the current low-dose image. We observe that the original SIR-ndiNLM introduces artifacts to the reconstruction when an attenuation-mismatched prior is used. Furthermore, we find that a larger attenuation mismatch between the prior and current images results in more severe artifacts in the SIR-ndiNLM reconstruction. Our two proposed corrective schemes enable SIR-ndiNLM to effectively handle the attenuation mismatch and anatomical changes between the two images and successfully eliminate the artifacts. We demonstrate that the proposed techniques permit SIR-ndiNLM to leverage the attenuation-mismatched prior and achieve quantitative low-dose CT reconstruction from both low-flux and sparse-view data acquisitions. This work permits robust and reliable PIBR for CT data acquired using different beam settings.

Improving Image Quality and Reducing Radiation Dose for Pediatric CT by Using Deep Learning Reconstruction

Background CT deep learning reconstruction (DLR) algorithms have been developed to remove image noise. How the DLR affects image quality and radiation dose reduction has yet to be fully investigated. Purpose To investigate a DLR algorithm's dose reduction and image quality improvement for pediatric CT. Materials and Methods DLR was compared with filtered back projection (FBP), statistical-based iterative reconstruction (SBIR), and model-based iterative reconstruction (MBIR) in a retrospective study by using data from CT examinations of pediatric patients (February to December 2018). A comparison of object detectability for 15 objects (diameter, 0.5-10 mm) at four contrast difference levels (50, 150, 250, and 350 HU) was performed by using a non-prewhitening-matched mathematical observer model with eye filter (d'_NPWE), task transfer function, and noise power spectrum analysis. Object detectability was assessed by using area under the curve analysis. Three pediatric radiologists performed an observer study to assess anatomic structures with low object-to-background signal and contrast to noise in the azygos vein, right hepatic vein, common bile duct, and superior mesenteric artery. Observers rated from 1 to 10 (worst to best) for edge definition, quantum noise level, and object conspicuity. Analysis of variance and Tukey honest significant difference post hoc tests were used to analyze differences between reconstruction algorithms. Results Images from 19 patients (mean age, 11 years ± 5 [standard deviation]; 10 female patients) were evaluated. Compared with FBP, SBIR, and MBIR, DLR demonstrated improved object detectability by 51% (16.5 of 10.9), 18% (16.5 of 13.9), and 11% (16.5 of 14.8), respectively. DLR reduced image noise without noise texture effects seen with MBIR. Radiologist ratings were 7 ± 1 (DLR), 6.2 ± 1 (MBIR), 6.2 ± 1 (SBIR), and 4.6 ± 1 (FBP); two-way analysis of variance showed a difference on the basis of reconstruction type (P < .001). Radiologists consistently preferred DLR images (intraclass correlation coefficient, 0.89; 95% CI: 0.83, 0.93). DLR demonstrated 52% (1 of 2.1) greater dose reduction than SBIR. Conclusion The DLR algorithm improved image quality and dose reduction without sacrificing noise texture and spatial resolution. © RSNA, 2020 Online supplemental material is available for this article.

Scalable Double Regularization for 3D Nano-CT Reconstruction

Nano-CT (computerized tomography) has emerged as a non-destructive high-resolution cross-sectional imaging technique to effectively study the sub-µm pore structure of shale, which is of fundamental importance to the evaluation and development of shale oil and gas. Nano-CT poses unique challenges to the inverse problem of reconstructing the 3D structure due to the lower signal-to-noise ratio (than Micro-CT) at the nano-scale, increased sensitivity to the misaligned geometry caused by the movement of object manipulator, limited sample size, and a larger volume of data at higher resolution. We propose a scalable double regularization (SDR) method to utilize the entire dataset for simultaneous 3D structural reconstruction across slices through total variation regularization within slices and L ₁ regularization between adjacent slices. SDR allows information borrowing both within and between slices, contrasting with the traditional methods that usually build on slice by slice reconstruction. We develop a scalable and memory-efficient algorithm by exploiting the systematic sparsity and consistent geometry induced by such Nano-CT data. We illustrate the proposed method using synthetic data and two Nano-CT imaging datasets of Jiulaodong (JLD) shale and Longmaxi (LMX) shale acquired in the Sichuan Basin. These numerical experiments show that the proposed method substantially outperforms selected alternatives both visually and quantitatively.

Spatial Distribution of Noise Reduction in Four Iterative Reconstruction Algorithms in CT—A Technical Evaluation

Iterative reconstruction (IR) is a computed tomgraphy (CT) reconstruction algorithm aiming at improving image quality by reducing noise in the image. During this process, IR also changes the noise properties in the images. To assess how IR algorithms from four vendors affect the noise properties in CT images, an anthropomorphic phantom was scanned and images reconstructed with filtered back projection (FBP), and a medium and high level of IR. Each image acquisition was performed 30 times at the same slice position, to create noise maps showing the inter-image pixel standard deviation through the 30 images. We observed that IR changed the noise properties in the CT images by reducing noise more in homogeneous areas than at anatomical edges between structures of different densities. This difference increased with increasing IR level, and with increasing difference in density between two adjacent structures. Each vendor’s IR algorithm showed slightly different noise reduction properties in how much noise was reduced at different positions in the phantom. Users need to be aware of these differences when working with optimization of protocols using IR across scanners from different vendors.

Characterization of tissue-specific pre-log Bayesian CT reconstruction by texture-dose relationship

PURPOSE

Tissue textures have been recognized as biomarkers for various clinical tasks. In computed tomography (CT) image reconstruction, it is important but challenging to preserve the texture when lowering x-ray exposure from full- toward low-/ultra-low dose level. Therefore, this paper aims to explore the texture-dose relationship within one tissue-specific pre-log Bayesian CT reconstruction algorithm.

METHODS

To enhance the texture in ultra-low dose CT (ULdCT) reconstruction, this paper presents a Bayesian type algorithm. A shifted Poisson model is adapted to describe the statistical properties of pre-log data, and a tissue-specific Markov random field prior (MRFt) is used to incorporate tissue texture from previous full-dose CT, thus called SP-MRFt algorithm. Utilizing the SP-MRFt algorithm, we investigated tissue texture degradation as a function of x-ray dose levels from full dose (100 mAs/120 kVp) to ultralow dose (1 mAs/120 kVp) by using quantitative texture-based evaluation metrics.

RESULTS

Experimental results show the SP-MRFt algorithm outperforms conventional filtered back projection (FBP) and post-log domain penalized weighted least square MRFt (PWLS-MRFt) in terms of noise suppression and texture preservation. Comparable results are also obtained with shifted Poisson model with 7 × 7 Huber MRF weights (SP-Huber7). The investigation on texture-dose relationship shows that the quantified texture measures drop monotonically as dose level decreases, and interestingly a turning point is observed on the texture-dose response curve.

CONCLUSIONS

This important observation implies that there exists a minimum dose level, at which a given CT scanner (hardware configuration and image reconstruction software) can achieve without compromising clinical tasks. Moreover, the experiment results show that the variance of electronic noise has higher impact than the mean to the texture-dose relationship.

Model-Based Iterative Reconstruction for Propagation-Based Phase-Contrast X-Ray CT including Models for the Source and the Detector

Propagation-based phase-contrast X-ray computed tomography is a valuable tool for high-resolution visualization of biological samples, giving distinct improvements in terms of contrast and dose requirements compared to conventional attenuation-based computed tomography. Due to its ease of implementation and advances in laboratory X-ray sources, this imaging technique is increasingly being transferred from synchrotron facilities to laboratory environments. This however poses additional challenges, such as the limited spatial coherence and flux of laboratory sources, resulting in worse resolution and higher noise levels. Here we extend a previously developed iterative reconstruction algorithm for this imaging technique to include models for the reduced spatial coherence and the signal spreading of efficient scintillator-based detectors directly into the physical forward model. Furthermore, we employ a noise model which accounts for the full covariance statistics of the image formation process. In addition, we extend the model describing the interference effects such that it now matches the formalism of the widely-used single-material phase-retrieval algorithm, which is based on the sample-homogeneity assumption. We perform a simulation study as well as an experimental study at a laboratory inverse Compton source and compare our approach to the conventional analytical approaches. We find that the modeling of the source and the detector inside the physical forward model can tremendously improve the resolution at matched noise levels and that the modeling of the covariance statistics reduces overshoots (i.e. incorrect increase / decrease in sample properties) at the sample edges significantly.

Ten-year prospective evaluation of whole-body cancer screening with multiple modalities including [18F]fluorodeoxyglucose positron emission tomography in a healthy population

PURPOSE

To prospectively evaluate the value of whole-body cancer screening with multiple modalities including FDG-PET in a healthy population.

METHODS

The study was conducted in 1197 healthy individuals aged ≥ 35 years at enrollment between August 2003 and July 2004. All participants were scheduled to receive annual whole-body cancer screening five times (screening period) with subsequent long-term follow-up (follow-up period). The endpoints of the study were definitive cancer diagnosis, cancer-related death, and all-cause death.

RESULTS

The follow-up rate was 99.8% for the screening period and 96.2% for the follow-up period. Forty-five cancers were confirmed during the screening period (August 2003 to July 2009), and 37 of the 45 were detected by the screening. Fourteen of the 45 were PET positive. Sixteen, 5, 4, 9 and 11 cancers were confirmed after the first, the second, the third, the fourth, and the fifth (took 2 years) screening, respectively. Eight participants died, of whom five died of cancer. The rate of cancer incidence (per 100,000) of 628.7 (95% confidence interval [CI] 445.0-812.4) was significantly high, and the rates of cancer mortality and all-cause mortality of 69.9 (95% CI 8.6-131.1) and 111.8 (95% CI 34.3-189.2), respectively, were significantly low, compared with the corresponding rates of 379.3, 138.2 and 354.2, respectively, in the age-rank- and sex-matched general population. During the follow-up period (August 2009 to July 2013), 37 cancers were confirmed and 30 of the 37 were detected. Seven participants died, of whom three died of cancer. The rate of cancer incidence was 809.6 (95% CI 548.7-1070.5). The rates of cancer mortality and all-cause mortality of 65.6 (95% CI 0-139.9) and 153.2 (95% CI 39.7-266.6), respectively, were significantly low compared with 190.1 and 462.3, respectively, in the general population.

CONCLUSION

Cancer detection by PET alone was limited. While the high cancer incidence was attributed to the extensive screening, the low cancer and all-cause mortality may indicate the potential value of this type of cancer screening. Cancer incidence increases with aging and it has been shown that continuous screening may reduce the risk caused by the cancer progression.

Improving Low-Dose Pediatric Abdominal CT by Using Convolutional Neural Networks

Purpose

To evaluate the efficacy of convolutional neural networks (CNNs) to improve the image quality of low-dose pediatric abdominal CT images.

Materials and Methods

Images from 11 pediatric abdominal CT examinations acquired between June and July 2018 were reconstructed with filtered back projection (FBP) and an iterative reconstruction (IR) algorithm. A residual CNN was trained using the FBP image as the input and the difference between FBP and IR as the target such that the network was able to predict the residual image and simulate the IR. CNN-based postprocessing was applied to 20 low-dose pediatric image datasets acquired between December 2016 and December 2017 on a scanner limited to reconstructing FBP images. The FBP and CNN images were evaluated based on objective image noise and subjective image review by two pediatric radiologists. For each of five features, readers rated images on a five-point Likert scale and also indicated their preferred series. Readers also indicated their "overall preference" for CNN versus FBP. Preference and Likert scores were analyzed for individual and combined readers. Interreader agreement was assessed.

Results

The CT number remained unchanged between FBP and CNN images. Image noise was reduced by 31% for CNN images (P < .001). CNN was preferred for overall image quality for individual and combined readers. For combined Likert scores, at least one of the two score types (Likert or binary preference) indicated a significant favoring of CNN over FBP for low contrast, image noise, artifacts, and high contrast, whereas the reverse was true for spatial resolution.

Conclusion

FBP images can be improved in image space by a well-trained CNN, which may afford a reduction in dose or improvement in image quality on scanners limited to FBP reconstruction.© RSNA, 2019.

Generative Noise Reduction in Dental Cone-Beam CT by a Selective Anatomy Analytic Iteration Reconstruction Algorithm

Dental cone-beam computed tomography (CBCT) is a powerful tool in clinical treatment planning, especially in a digital dentistry platform. Currently, the “as low as diagnostically acceptable” (ALADA) principle and diagnostic ability are a trade-off in most of the 3D integrated applications, especially in the low radio-opaque densified tissue structure. The CBCT benefits in comprehensive diagnosis and its treatment prognosis for post-operation predictability are clinically known in modern dentistry. In this paper, we propose a new algorithm called the selective anatomy analytic iteration reconstruction (SA2IR) algorithm for the sparse-projection set. The algorithm was simulated on a phantom structure analogous to a patient’s head for geometric similarity. The proposed algorithm is projection-based. Interpolated set enrichment and trio-subset enhancement were used to reduce the generative noise and maintain the scan’s clinical diagnostic ability. The results show that proposed method was highly applicable in medico-dental imaging diagnostics fusion for the computer-aided treatment planning, because it had significant generative noise reduction and lowered computational cost when compared to the other common contemporary algorithms for sparse projection, which generate a low-dosed CBCT reconstruction.

Blind deconvolution in model-based iterative reconstruction for CT using a normalized sparsity measure

Model-based iterative reconstruction techniques for CT that include a description of the noise statistics and a physical forward model of the image formation process have proven to increase image quality for many applications. Specifically, including models of the system blur into the physical forward model and thus implicitly performing a deconvolution of the projections during tomographic reconstruction, could demonstrate distinct improvements, especially in terms of resolution. However, the results strongly rely on an exact characterization of all components contributing to the system blur. Such characterizations can be laborious and even a slight mismatch can diminish image quality significantly. Therefore, we introduce a novel objective function, which enables us to jointly estimate system blur parameters during tomographic reconstruction. Conventional objective functions are biased in terms of blur and can yield lowest cost to blurred reconstructions with low noise levels. A key feature of our objective function is a new normalized sparsity measure for CT based on total-variation regularization, constructed to be less biased in terms of blur. We outline a solving strategy for jointly recovering low-dimensional blur parameters during tomographic reconstruction. We perform an extensive simulation study, evaluating the performance of the regularization and the dependency of the different parts of the objective function on the blur parameters. Scenarios with different regularization strengths and system blurs are investigated, demonstrating that we can recover the blur parameter used for the simulations. The proposed strategy is validated and the dependency of the objective function with the number of iterations is analyzed. Finally, our approach is experimentally validated on test-bench data of a human wrist phantom, where the estimated blur parameter coincides well with visual inspection. Our findings are not restricted to attenuation-based CT and may facilitate the recovery of more complex imaging model parameters.

Statistical CT reconstruction using region-aware texture preserving regularization learning from prior normal-dose CT image

In some clinical applications, prior normal-dose CT (NdCT) images are available, and the valuable textures and structure features in them may be used to promote follow-up low-dose CT (LdCT) reconstruction. This study aims to learn texture information from the NdCT images and leverage it for follow-up LdCT image reconstruction to preserve textures and structure features. Specifically, the proposed reconstruction method first learns the texture information from those patches with similar structures in NdCT image, and the similar patches can be clustered by searching context features efficiently from the surroundings of the current patch. Then it utilizes redundant texture information from the similar patches as a priori knowledge to describe specific regions in the LdCT image. The advanced region-aware texture preserving prior is termed as 'RATP'. The main advantage of the PATP prior is that it can properly learn the texture features from available NdCT images and adaptively characterize the region-specific structures in the LdCT image. The experiments using patient data were performed to evaluate the performance of the proposed method. The proposed RATP method demonstrated superior performance in LdCT imaging compared to the filtered back projection (FBP) and statistical iterative reconstruction (SIR) methods using Gaussian regularization, Huber regularization and the original texture preserving regularization.