Physics-informed model-based generative neural network for synthesizing scanner- and algorithm-specific low-dose CT exams.

Deep learning based denoising techniques have effectively reduced noise in low-dose computed tomography (LDCT) images. However, the need for pairs of LDCT and normal-dose CT (NDCT) training images is the main reason that makes the use of deep learning networks difficult. Therefore, we propose a method of generating synthesized LDCT images by adding synthesized quantum noise to the NDCT images. For this, we estimate the quantum noise power spectrum (NPS) of patient CT images by modeling anatomical NPS and quantum NPS, respectively. We use the quantum NPS as a noise filter and generate synthesized quantum noise by filtering random Gaussian noise. Using the proposed method, LDCT denoiser can be trained without using pairs of LDCT and NDCT images, and the trained denoiser effectively reduces noise in LDCT images. The proposed method facilitates data acquisition for training deep learning networks and reduces the risk of radiation exposure to patients.

Computer algorithms that simulate lower-doses computed tomography (CT) images from clinical-dose images are widely available. However, most operate in the projection domain and assume access to the reconstruction method. Access to commercial reconstruction methods may often not be available in medical research, making image-domain noise simulation methods useful. However, the introduction of non-linear reconstruction methods, such as iterative and deep learning-based reconstruction, makes noise insertion in the image domain intractable, as it is not possible to determine the noise textures analytically. To develop a deep learning-based image-domain method to generate low-dose CT images from clinical-dose CT (CDCT) images for non-linear reconstruction methods. We propose a fully image domain-based method, utilizing a series of three convolutional neural networks (CNNs), which, respectively, denoise CDCT images, predict the standard deviation map of the low-dose image, and generate the noise power spectra (NPS) of local patches throughout the low-dose image. All three models have U-net-based architectures and are partly or fully three-dimensional. As a use case for this study and with no loss of generality, we use paired low-dose and clinical-dose brain CT scans. A dataset of paired scans was retrospectively obtained. All images were acquired with a wide-area detector clinical system and reconstructed using its standard clinical iterative algorithm. Each pair was registered using rigid registration to correct for motion between acquisitions. The data was randomly partitioned into training ( samples), validation ( samples), and test ( samples) sets. The performance of each of these three CNNs was validated separately. For the denoising CNN, the local standard deviation decrease, and bias were determined. For the standard deviation map CNN, the real and estimated standard deviations were compared locally. Finally, for the NPS CNN, the NPS of the synthetic and real low-dose noise were compared inside and outside the skull. Two proof-of-concept denoising studies were performed to determine if the performance of a CNN- or a gradient-based denoising filter on the synthetic low-dose data versus real data differed. The denoising network had a median decrease in noise in the cerebrospinal fluid by a factor of and introduced a median bias of HU. The network for standard deviation map estimation had a median error of HU. The noise power spectrum estimation network was able to capture the anisotropic and shift-variant nature of the noise structure by showing good agreement between the synthetic and real low-dose noise and their corresponding power spectra. The two proof of concept denoising studies showed only minimal difference in standard deviation improvement ratio between the synthetic and real low-dose CT images with the median difference between the two being 0.0 and +0.05 for the CNN- and gradient-based filter, respectively. The proposed method demonstrated good performance in generating synthetic low-dose brain CT scans without access to the projection data or to the reconstruction method. This method can generate multiple low-dose image realizations from one clinical-dose image, so it is useful for validation, optimization, and repeatability studies of image-processing algorithms.

Noise is inherent to low-dose CT acquisition. We propose to train a convolutional neural network (CNN) jointly with an adversarial CNN to estimate routine-dose CT images from low-dose CT images and hence reduce noise. A generator CNN was trained to transform low-dose CT images into routine-dose CT images using voxelwise loss minimization. An adversarial discriminator CNN was simultaneously trained to distinguish the output of the generator from routine-dose CT images. The performance of this discriminator was used as an adversarial loss for the generator. Experiments were performed using CT images of an anthropomorphic phantom containing calcium inserts, as well as patient non-contrast-enhanced cardiac CT images. The phantom and patients were scanned at 20% and 100% routine clinical dose. Three training strategies were compared: the first used only voxelwise loss, the second combined voxelwise loss and adversarial loss, and the third used only adversarial loss. The results showed that training with only voxelwise loss resulted in the highest peak signal-to-noise ratio with respect to reference routine-dose images. However, CNNs trained with adversarial loss captured image statistics of routine-dose images better. Noise reduction improved quantification of low-density calcified inserts in phantom CT images and allowed coronary calcium scoring in low-dose patient CT images with high noise levels. Testing took less than 10 s per CT volume. CNN-based low-dose CT noise reduction in the image domain is feasible. Training with an adversarial network improves the CNNs ability to generate images with an appearance similar to that of reference routine-dose CT images.

Coronary artery calcification (CAC) measurement from low-dose CT images can be used to assess the risk of coronary artery disease. A fully automatic algorithm to detect and measure CAC from low-dose non-contrast, non-ECG-gated chest CT scans is presented. Based on the automatically detected CAC, the Agatston score (AS), mass score and volume score were computed. These were compared with scores obtained manually from standard-dose ECG-gated scans and low-dose un-gated scans of the same patient. The automatic algorithm segments the heart region based on other pre-segmented organs to provide a coronary region mask. The mitral valve and aortic valve calcification is identified and excluded. All remaining voxels greater than 180HU within the mask region are considered as CAC candidates. The heart segmentation algorithm was evaluated on 400 non-contrast cases with both low-dose and regular dose CT scans. By visual inspection, 371 (92.8%) of the segmentations were acceptable. The automated CAC detection algorithm was evaluated on 41 low-dose non-contrast CT scans. Manual markings were performed on both low-dose and standard-dose scans for these cases. Using linear regression, the correlation of the automatic AS with the standard-dose manual scores was 0.86; with the low-dose manual scores the correlation was 0.91. Standard risk categories were also computed. The automated method risk category agreed with manual markings of gated scans for 24 cases while 15 cases were 1 category off. For low-dose scans, the automatic method agreed with 33 cases while 7 cases were 1 category off.

Low-dose computed tomography (LDCT) is a diagnostic imaging technique designed to minimize radiation exposure to the patient. However, this reduction in radiation may compromise computed tomography (CT) image quality, adversely impacting clinical diagnoses. Various advanced LDCT methods have emerged to mitigate this challenge, relying on well-matched LDCT and normal-dose CT (NDCT) image pairs for training. Nevertheless, these methods often face difficulties in distinguishing image details from nonuniformly distributed noise, limiting their denoising efficacy. Additionally, acquiring suitably paired datasets in the medical domain poses challenges, further constraining their applicability. Hence, the objective of this study was to develop an innovative denoising framework for LDCT images employing unpaired data. In this paper, we propose a LDCT denoising network (DNCNN) that alleviates the need for aligning LDCT and NDCT images. Our approach employs generative adversarial networks (GANs) to learn and model the noise present in LDCT images, establishing a mapping from the pseudo-LDCT to the actual NDCT domain without the need for paired CT images. Within the domain of weakly supervised methods, our proposed model exhibited superior objective metrics on the simulated dataset when compared to CycleGAN and selective kernel-based cycle-consistent GAN (SKFCycleGAN): the peak signal-to-noise ratio (PSNR) was 43.9441, the structural similarity index measure (SSIM) was 0.9660, and the visual information fidelity (VIF) was 0.7707. In the clinical dataset, we conducted a visual effect analysis by observing various tissues through different observation windows. Our proposed method achieved a no-reference structural sharpness (NRSS) value of 0.6171, which was closest to that of the NDCT images (NRSS =0.6049), demonstrating its superiority over other denoising techniques in preserving details, maintaining structural integrity, and enhancing edge contrast. Through extensive experiments on both simulated and clinical datasets, we demonstrated the superior efficacy of our proposed method in terms of denoising quality and quantity. Our method exhibits superiority over both supervised techniques, including block-matching and 3D filtering (BM3D), residual encoder-decoder convolutional neural network (RED-CNN), and Wasserstein generative adversarial network-VGG (WGAN-VGG), and over weakly supervised approaches, including CycleGAN and SKFCycleGAN.

The valuable structure features in full-dose computed tomography (FdCT) scans can be exploited as prior knowledge for low-dose CT (LdCT) imaging. However, lacking the capability to represent local characteristics of interested structures of the LdCT image adaptively may result in poor preservation of details/textures in LdCT image. This paper aims to explore a novel prior knowledge retrieval and representation paradigm, called adaptive prior features assisted restoration algorithm, for the purpose of better restoration of the low-dose lung CT images by capturing local features from FdCT scans adaptively. The innovation lies in the construction of an offline training database and the online patch-search scheme integrated with the principal components analysis (PCA). Specifically, the offline training database is composed of 3-D patch samples extracted from existing full-dose lung scans. For online patch-search, 3-D patches with structure similar to the noisy target patch are first selected from the database as the training samples. Then, PCA is applied on the training samples to retrieve their local prior principal features adaptively. By employing the principal features to decompose the noisy target patch and using an adaptive coefficient shrinkage technique for inverse transformation, the noise of the target patch can be efficiently removed and the detailed texture can be well preserved. The effectiveness of the proposed algorithm was validated by CT scans of patients with lung cancer. The results show that it can achieve a noticeable gain over some state-of-the-art methods in terms of noise suppression and details/textures preservation.

PurposeRepeated computed tomography (CT) scans are prescribed for some clinical applications such as lung nodule surveillance. Several studies have demonstrated that incorporating a high-quality prior image into the reconstruction of subsequent low-dose CT (LDCT) acquisitions can either improve image quality or reduce data fidelity requirements. Our proposed previous normal-dose image induced nonlocal means (ndiNLM) regularization method for LDCT is an example of such a method. However, one major concern with prior image based methods is that they might produce false information when the prior image and the current LDCT image show different structures (for example, if a lung nodule emerges, grows, shrinks or disappears over time). This study aims to assess the performance of the ndiNLM regularization method in situations with change in anatomy.MethodWe incorporated the ndiNLM regularization into the statistical image reconstruction (SIR) framework for reconstruction of subsequent LDCT images. Because of its patch-based search mechanism, a rough registration between the prior image and the current LDCT image is adequate for the SIR-ndiNLM method. We assessed the performance of the SIR-ndiNLM method in lung nodule surveillance for two different scenarios: (1) the nodule was not found in a baseline exam but appears in a follow-up LDCT scan; (2) the nodule was present in a baseline exam but disappears in a follow-up LDCT scan. We further investigated the effect of nodule size on the performance of the SIR-ndiNLM method.ResultsWe found that a relatively large search-window (e.g., 33×33) should be used for the SIR-ndiNLM method to account for misalignment between the prior image and the current LDCT image, and to ensure that enough similar patches can be found in the prior image. With proper selection of other parameters, experimental results with two patient datasets demonstrated that the SIR-ndiNLM method did not miss true nodules nor introduce false nodules in the lung nodule surveillance scenarios described above. We also found that the SIR-ndiNLM reconstruction shows improved image quality when the prior image is similar to the current LDCT image in anatomy. These gains in image quality might appear small upon visual inspection, but they can be detected using quantitative measures. Finally, the SIR-ndiNLM method also performed well in ultra-low-dose conditions and with different nodule sizes.ConclusionsThis study assessed the performance of the SIR-ndiNLM method in situations in which the prior image and the current LDCT image show substantial anatomical differences, specifically, changes in lung nodules. The experimental results demonstrate that the SIR-ndiNLM method does not introduce false lung nodules nor miss true nodules, which relieves the concern that this method might produce false information. However, there is insufficient evidence that these findings will hold true for all kinds of anatomical changes.

Chest computed tomography (CT) plays a vital role in the early diagnosis, treatment, and follow-up of COVID-19 pneumonia during the pandemic. However, this raises concerns about excessive exposure to ionizing radiation. This study aimed to survey radiation doses in low-dose chest CT (LDCT) and ultra-low-dose chest CT (ULD) protocols used for imaging COVID-19 pneumonia relative to standard CT (STD) protocols so that the best possible practice and dose reduction techniques could be recommended. A total of 564 articles were identified by searching major scientific databases, including ISI Web of Science, Scopus, and PubMed. After evaluating the content and applying the inclusion criteria to technical factors and radiation dose metrics relevant to the LDCT protocols used for imaging COVID-19 patients, data from ten articles were extracted and analyzed. Technique factors that affect the application of LDCT and ULD are discussed, including tube current (mA), peak tube voltage (kVp), pitch factor, and iterative reconstruction (IR) algorithms. The CTDIvol values for the STD, LDCT, and ULD chest CT protocols ranged from 2.79–13.2 mGy, 0.90–4.40 mGy, and 0.20–0.28 mGy, respectively. The effective dose (ED) values for STD, LDCT, and ULD chest CT protocols ranged from 1.66–6.60 mSv, 0.50–0.80 mGy, and 0.39–0.64 mSv, respectively. Compared with the standard (STD), LDCT reduced the dose reduction by a factor of 2–4, whereas ULD reduced the dose reduction by a factor of 8–13. These dose reductions were achieved by applying scan parameters and techniques such as iterative reconstructions, ultra-long pitches, and fast spectral shaping with a tin filter. Using LDCT, the cumulative radiation dose of serial CT examinations during the acute period of COVID-19 may have been inferior or equivalent to that of conventional CT.

Deep learning has recently attracted widespread interest as a means of reducing noise in low-dose CT (LDCT) images. Deep convolutional neural networks (CNNs) are typically trained to transfer high-quality image features of normal-dose CT (NDCT) images to LDCT images. However, existing deep learning approaches for denoising LDCT images often overlook the statistical property of CT images. In this paper, we propose an approach to statistical image restoration for LDCT using deep learning. We introduce a loss function to incorporate the noise property in image domain derived from the noise statistics in sinogram domain. In order to capture the spatially-varying statistics of CT images, we increase the receptive fields of the neural network to cover full-size CT slices. In addition, the proposed network utilizes z-directional correlation by taking multiple consecutive CT slices as input. For performance evaluation, the proposed networks are trained and validated with a public dataset consisting of LDCT-NDCT image pairs. We also perform a retrospective study by testing the networks with clinical LDCT images. The experimental results show that the denoising networks successfully reduce the noise level and restore the image details without adding artifacts. This study demonstrates that the statistical deep learning approach can restore the image quality of LDCT without loss of anatomical information.

The resurgence of deep neural networks has created an alternative pathway for low-dose computed tomography denoising by learning a nonlinear transformation function between low-dose CT (LDCT) and normal-dose CT (NDCT) image pairs. However, those paired LDCT and NDCT images are rarely available in the clinical environment, making deep neural network deployment infeasible. This study proposes a novel method for self-supervised low-dose CT denoising to alleviate the requirement of paired LDCT and NDCT images. Specifically, we have trained an invertible neural network to minimize the pixel-based mean square distance between a noisy slice and the average of its two immediate adjacent noisy slices. We have shown the aforementioned is similar to training a neural network to minimize the distance between clean NDCT and noisy LDCT image pairs. Again, during the reverse mapping of the invertible network, the output image is mapped to the original input image, similar to cycle consistency loss. Finally, the trained invertible network’s forward mapping is used for denoising LDCT images. Extensive experiments on two publicly available datasets showed that our method performs favourably against other existing unsupervised methods.

Practical simulations of low-dose CT images have a possibility of being helpful means for optimization of the CT exposure dose. Because current methods reported by several researchers are limited to specific vendor platforms and generally rely on raw sinogram data that are difficult to access, we have developed a new computerized scheme for producing simulated low-dose CT images from real high-dose images without use of raw sinogram data or of a particular phantom. Our computerized scheme for low-dose CT simulation was based on the addition of a simulated noise image to a real high-dose CT image reconstructed by the filtered back-projection algorithm. First, a sinogram was generated from the forward projection of a high-dose CT image. Then, an additional noise sinogram resulting from use of a reduced exposure dose was estimated from a predetermined noise model. Finally, a noise CT image was reconstructed with a predetermined filter and was added to the real high-dose CT image to create a simulated low-dose CT image. The noise power spectrum and modulation transfer function of the simulated low-dose images were very close to those of the real low-dose images. In order to confirm the feasibility of our method, we applied this method to clinical cases which were examined with the high dose initially and then followed with a low-dose CT. In conclusion, our proposed method could simulate the low-dose CT images from their real high-dose images with sufficient accuracy and could be used for determining the optimal dose setting for various clinical CT examinations.

To evaluate commercial deep learning-based software for fully automated coronary artery calcium (CAC) scoring on non-electrocardiogram (ECG)-gated low-dose CT (LDCT) with different slice thicknesses compared with manual ECG-gated calcium-scoring CT (CSCT). This retrospective study included 567 patients who underwent both LDCT and CSCT. All LDCT images were reconstructed with a 2.5-mm slice thickness (LDCT2.5-mm), and 453 LDCT scans were reconstructed with a 1.0-mm slice thickness (LDCT1.0-mm). Automated CAC scoring was performed on CSCT (CSCTauto), LDCT1.0-mm, and LDCT2.5-mm images. The reliability of CSCTauto, LDCT1.0-mm, and LDCT2.5-mm was compared with manual CSCT scoring (CSCTmanual) using intraclass correlation coefficients (ICCs) and Bland-Altman analysis. Agreement, in CAC severity category, was analyzed using weighted kappa statistics. Diagnostic performance at various Agatston score cutoffs was also calculated. CSCTauto, LDCT1.0-mm, and LDCT2.5-mm demonstrated excellent agreement with CSCTmanual (ICC [95% confidence interval, CI]: 1.000 [1.000, 1.000], 0.937 [0.917, 0.952], and 0.955 [0.946, 0.963], respectively). The mean difference with 95% limits of agreement was lower with LDCT1.0-mm than with LDCT2.5-mm (19.94 [95% CI, -244.0, 283.9] vs. 45.26 [-248.2, 338.7]). Regarding CAC severity, LDCT1.0-mm achieved almost perfect agreement, and LDCT2.5-mm achieved substantial agreement (kappa [95% CI]: 0.809 [0.776, 0.838], 0.776 [0.740, 0.809], respectively). Diagnostic performance for detecting Agatston score ≥ 400 was also higher with LDCT1.0-mm than with LDCT2.5-mm (F1 score, 0.929 vs. 0.855). Fully automated CAC-scoring software with both CSCT and LDCT yielded excellent reliability and agreement with CSCTmanual. LDCT1.0-mm yielded more accurate Agatston scoring than LDCT2.5-mm using fully automated commercial software. • Total Agatston scores and all vessels of CSCTauto, LDCT1.0-mm, and LDCT2.5-mm demonstrated excellent agreement with CSCTmanual (all ICC > 0.85). • The diagnostic performance for detecting all Agatston score cutoffs was better with LDCT1.0-mm than with LDCT2.5-mm. • This automated software yielded a lower degree of underestimation compared with methods described in previous studies, and the degree of underestimation was lower with LDCT1.0-mm than with LDCT2.5-mm.

In computed tomography (CT)-imaging an optimal compromise between the radiation burden and the image quality for the imaging task is needed. Lower-dose CT is desirable, however, lowering the dose results in a lower signal-to-noise ratio and therefore in a reduced image quality. In this research, we aim to develop a tool to simulate lower-dose scans from an existing standard-dose scan. The main application of this tool is to determine the lowest possible radiation dose that still produces sufficient clinical information. The x-ray tube current reduction is modeled by estimating the noise equivalent number of photons in the high exposure scan and applying a thinning technique to reduce that number. The proposed method accounts for the bowtie filter, for the electronic system noise, for the noise correlation between neighboring detector elements, for the beam hardening effect, and for the non-linear smoothing filter in very low-dose scans. Several phantom studies with different acquisition protocols were performed to evaluate the accuracy of the proposed framework. The results demonstrate a close agreement between the noise magnitude and texture of the measured and the simulated lower-dose scans. For instance, the standard deviation of noise in the simulation of lower-dose scans with 90% tube current reduction matches the reconstructions from the real scans with less than 1% and 3% error for sequential and helical scans, respectively. The noise texture was also assessed by analyzing the noise power spectrum of the simulated lower-dose images which matches those from the real scans. Furthermore, the relation between the measured and predicted noise in projection domain is very close to the line of identity which confirms the accuracy of the model.

Deep learning-based image analysis of pretreatment FDG-PET/CT predicts CAR-T cell treatment outcome at month-12 for patients with Relapsed/Refractory large B-cell lymphomas

Photon counting detector CT (PCD-CT) has the potential to sort detected x-ray photons into more than two energy bins to perform multi-material decomposition using K-edge materials without additional mathematical constraints/assumptions. The availability of multiple energy thresholds in PCD-CT opens the possibility for unique clinical applications such as simultaneous multi-contrast-agent imaging. To evaluate the technical performance of a PCD-CT system with four energy thresholds for conducting multi-material decomposition of iodine, gadolinium, and water after a comprehensive system calibration procedure. Experiments were performed on a commercial PCD-CT scanner (NAEOTOM Alpha, Siemens Healthineers) using a research mode with four energy thresholds. Material calibration was performed by scanning contrast materials with known concentration placed in various sizes of cylindrical water-equivalent phantoms at various tube currents and positions. From these data, per-voxel contrast models were derived for each material by multiple linear regressions up to second order, where the predictor variables were combinations of tube current, phantom diameter, and distance to isocenter. After calibration, small (20 cm diameter) and large (30cm×40cm) uniform multi-energy CT phantoms and an anthropomorphic thorax phantom (31cm×39cm), each containing bone, soft tissue, and several quantitative inserts of iodine (0.2-5.0mg/ml), gadolinium (0.3-5.0mg/ml), and a mixture of both (5.0mg/ml each) were scanned using the four-energy threshold mode at 140 kV with energy thresholds set at 20, 52, 75, and 82 keV. Images for each energy threshold were reconstructed with a filtered back-projection algorithm and a quantitative kernel (Qr44). Subsequent image-based material decomposition was performed on individual voxels of the energy threshold images using the models derived from the calibration process to generate water, iodine, and gadolinium basis images. Quantitative material concentrations and image noise magnitudes, and noise power spectra (NPS) were measured. Iodine and gadolinium concentrations above 1.25 mg/ml were visible in the material basis images in all phantoms. Averaged quantitative accuracy for iodine and gadolinium in the small, large, and thorax phantoms were 0.39±0.30 mg/ml, 0.96±0.53 mg/ml, and 0.53±0.35 mg/ml, respectively. In the water basis images, accuracy was within 24 HU in all phantoms. Noise magnitudes in iodine and gadolinium basis images were 2.91 and 3.52 times that of the lowest energy threshold image for the small phantom, and were 6.92 and 7.17 times that of the lowest threshold image for the large phantom. Noise texture did not change between the energy threshold and material basis images for either size phantom as measured by the NPS. After a comprehensive system calibration, a commercial PCD-CT with four energy thresholds performed multiple material decomposition of iodine, gadolinium, and water for concentrations above 1.25 mg/ml in human-sized phantoms. Noise amplification and quantitative accuracy from the decomposition greatly depended on object size.