Low-dose CT Images Research Articles

A residual multi-scale feature extraction network with hybrid loss for low-dose computed tomography image denoising

A residual multi-scale feature extraction network with hybrid loss for low-dose computed tomography image denoising

Assessment of skeletal muscle using deep learning on low-dose CT images.

The visceral fat area obtained by computed tomography (CT) at the navel level is clinically used as an indicator of visceral fat obesity in Japan. Analysis of skeletal muscle mass using CT images at the navel level may potentially support concurrent assessment of sarcopenia and sarcopenic obesity. The purpose of this study was to assess the performance of deep learning models (DLMs) for skeletal muscle mass measurement using low-dose abdominal CT. The primary dataset used in this study included 11,494 low-dose abdominal CT images at navel level acquired in 7,370 subjects for metabolic syndrome screening. The publicly available Cancer Imaging Archive (TCIA) dataset, including 5,801 abdominal CT images, was used as a complementary dataset. For abdominal CT image segmentation, we used the SegU-net DLM with different filter size and hierarchical depth. The segmentation accuracy was assessed by measuring the dice similarity coefficient (DSC), cross-sectional area (CSA) error, and Bland-Altman plots. The proposed DLM achieved a DSC of 0.992 ± 0.012, a CSA error of 0.41 ± 1.89%, and a Bland-Altman percent difference of -0.1 ± 3.8%. The proposed DLM was able to automatically segment skeletal muscle mass measurements from low-dose abdominal CT with high accuracy.

Low-dose CT image quality evaluation method based on radiomics and deep residual network with attention mechanism

Low-dose computed tomography (CT) produces less radiation but may degrade image quality. Research on automatic evaluation and classification of low-dose CT image quality is necessary, which is helpful to clinicians. A low-dose CT image quality evaluation method is proposed based on radiomics and deep learning. After pre-training on the ImageNet dataset, the trained deep learning model is then transferred to the target dataset of low-dose CT image. Region of interest is obtained through a deep residual network with attention mechanism. Pyradiomics is used to extract radiomics features of ROI, and T-test and LASSO algorithm are used for feature screening. Finally, support vector machine, random forest and logistic regression models are adopted to classify and predict the quality of low-dose CT image based on the selected radiomics features. Experimental results show that the trained deep residual network model with attention module can achieve 92% recognition accuracy of region of interest, and the evaluation accuracy of the three machine learning models achieve 86.6% (support vector machine), 84.5% (random forest) and 61.7% (logistic regression), respectively. Transfer learning can effectively reduce the required number of samples for the training of deep learning model and avoid overfitting. After the introduction of attention module in deep residual network model, accuracy rises faster, and the final accuracy is slightly higher than that of deep residual network model without attention module, but the improvement is slightly lower than expected. Radiomics features of region of interest reflect the quality differences of low-dose CT images which can be employed for the quality evaluation of low-dose CT image.

Characterization of hypermetabolic lymph nodes after SARS-CoV-2 vaccination using PET-CT derived node-RADS, in patients with melanoma

This study aimed to evaluate the diagnostic accuracy of Node Reporting and Data System (Node-RADS) in discriminating between normal, reactive, and metastatic axillary LNs in patients with melanoma who underwent SARS-CoV-2 vaccination. Patients with proven melanoma who underwent a 2-[18F]-fluoro-2-deoxy-D-glucose positron emission tomography/computed tomography (2-[18F]-FDG PET/CT) between February and April 2021 were included in this retrospective study. Primary melanoma site, vaccination status, injection site, and 2-[18F]-FDG PET/CT were used to classify axillary LNs into normal, inflammatory, and metastatic (combined classification). An adapted Node-RADS classification (A-Node-RADS) was generated based on LN anatomical characteristics on low-dose CT images and compared to the combined classification. 108 patients were included in the study (54 vaccinated). HALNs were detected in 42 patients (32.8%), of whom 97.6% were vaccinated. 172 LNs were classified as normal, 30 as inflammatory, and 14 as metastatic using the combined classification. 152, 22, 29, 12, and 1 LNs were classified A-Node-RADS 1, 2, 3, 4, and 5, respectively. Hence, 174, 29, and 13 LNs were deemed benign, equivocal, and metastatic. The concordance between the classifications was very good (Cohen’s k: 0.91, CI 0.86–0.95; p-value < 0.0001). A-Node-RADS can assist the classification of axillary LNs in melanoma patients who underwent 2-[18F]-FDG PET/CT and SARS-CoV-2 vaccination.

A Comprehensive Review of GAN-Based Denoising Models for Low-Dose Computed Tomography Images

Computed Tomography (CT) offers great visualization of the intricate internal body structures. To protect a patient from the potential radiation-related health risks, the acquisition of CT images should adhere to the “as low as reasonably allowed” (ALARA) standard. However, the acquired Low-dose CT (LDCT) images are inadvertently corrupted by artifacts and noise during the processes of acquisition, storage, and transmission, degrading the visual quality of the image and also causing the loss of image features and relevant information. Most recently, generative adversarial network (GAN) models based on deep learning (DL) have demonstrated ground-breaking performance to minimize image noise while maintaining high image quality. These models’ ability to adapt to uncertain noise distributions and representation-learning ability makes them highly desirable for the denoising of CT images. The state-of-the-art GANs used for LDCT image denoising have been comprehensively reviewed in this research paper. The aim of this paper is to highlight the potential of DL-based GAN for CT dose optimization and present future scope of research in the domain of LDCT image denoising.

Adherence to Annual Lung Cancer Screening in a Centralized Academic Program

BackgroundAdherence to lung cancer screening (LCS) protocols is critical for achieving mortality reductions. However, adherence rates, particularly for recommended annual screening among patients with low-risk findings, are often sub-optimal. We evaluated annual LCS adherence for patients with low-risk findings participating in a centralized screening program at a tertiary academic center. Patients and MethodsWe conducted a retrospective, observational cohort study of a centralized lung cancer screening program launched in July 2018. We performed electronic medical review of 337 patients who underwent low-dose CT (LDCT) screening before February 1, 2021 (to ensure ≥ 15 months follow up) and had a low-risk Lung-RADS score of 1 or 2. Captured data included patient characteristics (smoking history, Fagerstrom score, environmental exposures, lung cancer risk score), LDCT imaging dates, and Lung-RADS results. The primary outcome measure was adherence to annual screening. We used multivariable logistic regression models to identify factors associated with adherence. ResultsOverall, 337 patients had an initial Lung-RADS result of 1 (n = 189) or 2 (n = 148). Among this cohort, 139 (73.5%) of Lung-RADS 1 and 111 (75.0%) of Lung-RADS 2 patients completed the annual repeat LDCT within 15 months, respectively. The only patient characteristic associated with adherence was having Medicaid coverage; compared to having private insurance, Medicaid patients were less adherent (adjusted OR = 0.37, 95% CI = 0.15-0.92). No other patient characteristic was associated with adherence. ConclusionOur centralized screening program achieved a high initial annual adherence rate. Although LCS has first-dollar insurance coverage, other socioeconomic concerns may present barriers to annual screening for Medicaid recipients.

Feasibility analysis of low-dose CT with asynchronous quantitative computed tomography to assess vBMD

BackgroundTo explore the feasibility of low-dose computed tomography (LDCT) with asynchronous quantitative computed tomography (asynchronous QCT) for assessing the volumetric bone mineral density (vBMD).Methods416 women patients, categorized into 4 groups, were included and underwent chest CT examinations combined with asynchronous QCT, and CT scanning dose protocols (LDCT or CDCT) were self-determined by the participants. Radiation dose estimations were retrieved from patient protocols, including volume CT dose index (CTDIvol) and dose-length-product (DLP), and then calculated effective dose (ED). Delimiting ED by 1.0 mSv, chest CT examinations were categorized into 2 groups, LDCT group and CDCT group. vBMD of T12-L2 was obtained by transferring the LDCT and CDCT images to the QCT workstation, without extra radiation.ResultsThere was no difference of vBMD among 4 age groups in LDCT group (P = 0.965), and no difference in CDCT group (P = 0.988). In LDCT group and CDCT group, vBMD was not correlated to mAs, CTDIvol and DLP (P > 0.05), respectively. Between LDCT group and CDCT group, there was no difference of vBMD (P ≥ 0.480), while differences of mAs, CTDIvol and DLP.ConclusionThere was no difference of vBMD between LDCT group and CDCT group and vBMD was not correlated to mAs. While screening for diseases such as lung cancer and mediastinal lesions, LDCT combined with asynchronous QCT can be also used to assess vBMD simultaneously with no extra imaging equipment, patient visit time, radiation dose and no additional economic cost.

Comparison of endoscopic versus CT assessment of stone-free status after percutaneous nephrolithotomy (PCNL)

This study is aimed to determine whether postoperative low dose computed tomography (LDCT) imaging is necessary after percutaneous nephrolithotomy (PCNL), or the surgeon's intraoperative assessment of residual fragments (RF) is sufficient and avoidance of postoperative imaging with reduction of radiation exposure can be achieved. Data of all 610 patients who underwent PCNL in prone position in our institution from February 2009 to September 2020 was collected. Parameters such as age, gender, BMI, ASA-Classification, stone related parameters and the surgeon’s assessment of stone-free status were analyzed. The LDCT performed postoperatively was compared to the intraoperative assessment of the surgeon regarding RF. The mean age of patients was 52.82 years; the mean BMI was 28.18 kg/m2. In 418 cases, the surgeon made a clear statement about the presence of RF and postoperative LDCT was carried out. The discrepancy between the two methods (surgeon´s assessment vs. LDCT) was significant at p < 0.0001. The sensitivity, specificity, positive and negative predictive value of the surgeon when assessing RF were 24.05%, 99.45%, 98.28% and 50%. Stone free rate (SFR) after primary PCNL was 45.57%. The overall SFR at discharge was 96.23%. Although the surgeon´s assessment of RF was reliable, postoperative LDCT imaging should still be performed if endoscopic stone clearance is suspected due to the high false negative rate and the low negative predictive value. The optimal timing of postoperative imaging following PCNL remains unclear.

Long-term outcomes of lung cancer screening in males and females

BackgroundThis study explored female and male overall mortality and lung cancer (LC) survival in two LC screening (LCS) populations, focusing on the predictive value of coronary artery calcification (CAC) at baseline low-dose computed tomography (LDCT). MethodsThis retrospective study analysed data of 6495 heavy smokers enrolled in the MILD and BioMILD LCS trials between 2005 and 2016. The primary objective of the study was to assess sex differences in all-cause mortality and LC survival. CAC scores were automatically calculated on LDCT images by a validated artificial intelligence (AI) software. Sex differences in 12-year cause-specific mortality rates were stratified by age, pack-years and CAC score. ResultsThe study included 2368 females and 4127 males. The 12-year all-cause mortality rates were 4.1 % in females and 7.7 % in males (p < 0.0001), and median CAC score was 8.7 vs. 41 respectively (p < 0.0001). All-cause mortality increased with rising CAC scores (log-rank test, p < 0.0001) for both sexes. Although LC incidence was not different between the two sexes, females had lower rates of 12-year LC mortality (1.0 % vs. 1.9 %, p = 0.0052), and better LC survival from diagnosis (72.3 % vs. 51.7 %; p = 0.0005), with a similar proportion of stage I (58.1 % vs. 51.2 %, p = 0.2782). ConclusionsOur findings demonstrate that female LCS participants had lower rates of all-cause mortality at 12 years and better LC survival than their male counterparts, with similar LC incidence rates and stage at diagnosis. The lower CAC burden observed in women at all ages might contribute to explain their lower rates of all-cause mortality and better LC survival.

QS-ADN: quasi-supervised artifact disentanglement network for low-dose CT image denoising by local similarity among unpaired data

Deep learning has been successfully applied to low-dose CT (LDCT) image denoising for reducing potential radiation risk. However, the widely reported supervised LDCT denoising networks require a training set of paired images, which is expensive to obtain and cannot be perfectly simulated. Unsupervised learning utilizes unpaired data and is highly desirable for LDCT denoising. As an example, an artifact disentanglement network (ADN) relies on unpaired images and obviates the need for supervision but the results of artifact reduction are not as good as those through supervised learning. An important observation is that there is often hidden similarity among unpaired data that can be utilized. This paper introduces a new learning mode, called quasi-supervised learning, to empower ADN for LDCT image denoising. For every LDCT image, the best matched image is first found from an unpaired normal-dose CT (NDCT) dataset. Then, the matched pairs and the corresponding matching degree as prior information are used to construct and train our ADN-type network for LDCT denoising. The proposed method is different from (but compatible with) supervised and semi-supervised learning modes and can be easily implemented by modifying existing networks. The experimental results show that the method is competitive with state-of-the-art methods in terms of noise suppression and contextual fidelity. The code and working dataset are publicly available at https://github.com/ruanyuhui/ADN-QSDL.git.

Structure-preserving low-dose computed tomography image denoising using a deep residual adaptive global context attention network.

Low-dose computed tomography (LDCT) scans can effectively reduce the radiation damage to patients, but this is highly detrimental to CT image quality. Deep convolutional neural networks (CNNs) have shown their potential in improving LDCT image quality. However, the conventional CNN-based approaches rely fundamentally on the convolution operations, which are ineffective for modeling the correlations among nonlocal similar structures and the regionally distinct statistical properties in CT images. This modeling deficiency hampers the denoising performance for CT images derived in this manner. In this paper, we propose an adaptive global context (AGC) modeling scheme to describe the nonlocal correlations and the regionally distinct statistics in CT images with negligible computation load. We further propose an AGC-based long-short residual encoder-decoder (AGC-LSRED) network for efficient LDCT image noise artifact-suppression tasks. Specifically, stacks of residual AGC attention blocks (RAGCBs) with long and short skip connections are constructed in the AGC-LSRED network, which allows valuable structural and positional information to be bypassed through these identity-based skip connections and thus eases the training of the deep denoising network. For training the AGC-LSRED network, we propose a compound loss that combines the L1 loss, adversarial loss, and self-supervised multi-scale perceptual loss. Quantitative and qualitative experimental studies were performed to verify and validate the effectiveness of the proposed method. The simulation experiments demonstrated the proposed method exhibits the best result in terms of noise suppression [root-mean-square error (RMSE) =9.02; peak signal-to-noise ratio (PSNR) =33.17] and fine structure preservation [structural similarity index (SSIM) =0.925] compared with other competitive CNN-based methods. The experiments on real data illustrated that the proposed method has advantages over other methods in terms of radiologists' subjective assessment scores (averaged scores =4.34). With the use of the AGC modeling scheme to characterize the structural information in CT images and of residual AGC-attention blocks with long and short skip connections to ease the network training, the proposed AGC-LSRED method achieves satisfactory results in preserving fine anatomical structures and suppressing noise in LDCT images.

Transformer With Double Enhancement for Low-Dose CT Denoising.

Increasingly serious health problems have made the usage of computed tomography surge. Therefore, algorithms for processing CT images are becoming more and more abundant. These algorithms can lessen the harm of cumulative radiation in CT technology for the patient while eliminating the noise of image caused by dose reduction. However, the mainstream CNN-based algorithms are inefficient when dealing with features in broad regions. Inspired by the large receptive field of transformer framework, this paper designs an end-to-end low-dose CT (LDCT) denoising network based on the transformer. The overall framework of network contains a main branch and dual side branches. Specifically, the overlapping-free window-based self-attention transformer block is adopted on the main branch to realize image denoising. On the dual side branches, we propose double enhancement module to enrich edge, texture, and context information of LDCT images. Meanwhile, the receptive field of network is further enlarged after processing, which is helpful for building model's long-range dependencies. The outputs of the side branches are concatenated to the middle layer of the network for enhancing information and generating high-quality CT images. In addition, to better train the network and achieve the expected results, we introduce a compound loss function including mean squared error (MSE), multi-scale perceptual (MSP), and Sobel-L1 (SL) to make the denoised image closer to the targeted norm-dose CT (NDCT) image. Lastly, we conducted experiments on two clinical datasets including abdomen, head, and chest LDCT images with 25%, 25%, and 10% of the full dose, respectively. The experimental results demonstrated that the proposed DEformer achieved better denoising performance than the existing algorithms.

GA-Net: A geographical attention neural network for the segmentation of body torso tissue composition

PurposeBody composition analysis (BCA) of the body torso plays a vital role in the study of physical health and pathology and provides biomarkers that facilitate the diagnosis and treatment of many diseases, such as type 2 diabetes mellitus, cardiovascular disease, obstructive sleep apnea, and osteoarthritis. In this work, we propose a body composition tissue segmentation method that can automatically delineate those key tissues, including subcutaneous adipose tissue, skeleton, skeletal muscle tissue, and visceral adipose tissue, on positron emission tomography/computed tomography scans of the body torso. MethodsTo provide appropriate and precise semantic and spatial information that is strongly related to body composition tissues for the deep neural network, first we introduce a new concept of the body area and integrate it into our proposed segmentation network called Geographical Attention Network (GA-Net). The body areas are defined following anatomical principles such that the whole body torso region is partitioned into three non-overlapping body areas. Each body composition tissue of interest is fully contained in exactly one specific minimal body area. Secondly, the proposed GA-Net has a novel dual-decoder schema that is composed of a tissue decoder and an area decoder. The tissue decoder segments the body composition tissues, while the area decoder segments the body areas as an auxiliary task. The features of body areas and body composition tissues are fused through a soft attention mechanism to gain geographical attention relevant to the body tissues. Thirdly, we propose a body composition tissue annotation approach that takes the body area labels as the region of interest, which significantly improves the reproducibility, precision, and efficiency of delineating body composition tissues. ResultsOur evaluations on 50 low-dose unenhanced CT images indicate that GA-Net outperforms other architectures statistically significantly based on the Dice metric. GA-Net also shows improvements for the 95% Hausdorff Distance metric in most comparisons. Notably, GA-Net exhibits more sensitivity to subtle boundary information and produces more reliable and robust predictions for such structures, which are the most challenging parts to manually mend in practice, with potentially significant time-savings in the post hoc correction of these subtle boundary placement errors. Due to the prior knowledge provided from body areas, GA-Net achieves competitive performance with less training data. Our extension of the dual-decoder schema to TransUNet and 3D U-Net demonstrates that the new schema significantly improves the performance of these classical neural networks as well. Heatmaps obtained from attention gate layers further illustrate the geographical guidance function of body areas for identifying body tissues. Conclusions(i) Prior anatomic knowledge supplied in the form of appropriately designed anatomic container objects significantly improves the segmentation of bodily tissues. (ii) Of particular note are the improvements achieved in the delineation of subtle boundary features which otherwise would take much effort for manual correction. (iii) The method can be easily extended to existing networks to improve their accuracy for this application.

A constructive non-local means algorithm for low-dose computed tomography denoising with morphological residual processing.

Low-dose computed tomography (LDCT) has attracted significant attention in the domain of medical imaging due to the inherent risks of normal-dose computed tomography (NDCT) based X-ray radiations to patients. However, reducing radiation dose in CT imaging produces noise and artifacts that degrade image quality and subsequently hinders medical disease diagnostic performance. In order to address these problems, this research article presents a competent low-dose computed tomography image denoising algorithm based on a constructive non-local means algorithm with morphological residual processing to achieve the task of removing noise from the LDCT images. We propose an innovative constructive non-local image filtering algorithm by means of applications in low-dose computed tomography technology. The nonlocal mean filter that was recently proposed was modified to construct our denoising algorithm. It constructs the discrete property of neighboring filtering to enable rapid vectorized and parallel implantation in contemporary shared memory computer platforms while simultaneously decreases computing complexity. Subsequently, the proposed method performs faster computation compared to a non-vectorized and serial implementation in terms of speed and scales linearly with image dimension. In addition, the morphological residual processing is employed for the purpose of edge-preserving image processing. It combines linear lowpass filtering with a nonlinear technique that enables the extraction of meaningful regions where edges could be preserved while removing residual artifacts from the images. Experimental results demonstrate that the proposed algorithm preserves more textural and structural features while reducing noise, enhances edges and significantly improves image quality more effectively. The proposed research article obtains better results both qualitatively and quantitively when compared to other comparative algorithms on publicly accessible datasets.

Hybrid Multiple-Organ Segmentation Method Using Multiple U-Nets in PET/CT Images

PET/CT can scan low-dose computed tomography (LDCT) images with morphological information and PET images with functional information. Because the whole body is targeted for imaging, PET/CT examinations are important in cancer diagnosis. However, the several images obtained by PET/CT place a heavy burden on radiologists during diagnosis. Thus, the development of computer-aided diagnosis (CAD) and technologies assisting in diagnosis has been requested. However, because FDG accumulation in PET images differs for each organ, recognizing organ regions is essential for developing lesion detection and analysis algorithms for PET/CT images. Therefore, we developed a method for automatically extracting organ regions from PET/CT images using U-Net or DenseUNet, which are deep-learning-based segmentation networks. The proposed method is a hybrid approach combining morphological and functional information obtained from LDCT and PET images. Moreover, pre-training using ImageNet and RadImageNet was performed and compared. The best extraction accuracy was obtained by pre-training ImageNet with Dice indices of 94.1, 93.9, 91.3, and 75.1% for the liver, kidney, spleen, and pancreas, respectively. This method obtained better extraction accuracy for low-quality PET/CT images than did existing studies on PET/CT images and was comparable to existing studies on diagnostic contrast-enhanced CT images using the hybrid method and pre-training.

Optimization of Median Modified Wiener Filter for Improving Lung Segmentation Performance in Low-Dose Computed Tomography Images

In low-dose computed tomography (LDCT), lung segmentation effectively improves the accuracy of lung cancer diagnosis. However, excessive noise is inevitable in LDCT, which can decrease lung segmentation accuracy. To address this problem, it is necessary to derive an optimized kernel size when using the median modified Wiener filter (MMWF) for noise reduction. Incorrect application of the kernel size can result in inadequate noise removal or blurring, degrading segmentation accuracy. Therefore, various kernel sizes of the MMWF were applied in this study, followed by region-growing-based segmentation and quantitative evaluation. In addition to evaluating the segmentation performance, we conducted a similarity assessment. Our results indicate that the greatest improvement in segmentation performance and similarity was at a kernel size 5 × 5. Compared with the noisy image, the accuracy, F1-score, intersection over union, root mean square error, and peak signal-to-noise ratio using the optimized MMWF were improved by factors of 1.38, 33.20, 64.86, 7.82, and 1.30 times, respectively. In conclusion, we have demonstrated that by applying the MMWF with an appropriate kernel size, the optimization of noise and blur reduction can enhance segmentation performance.

Spatiotemporal denoising of low-dose cardiac CT image sequences using RecycleGAN.

Electrocardiogram (ECG)-gated multi-phase computed tomography angiography (MP-CTA) is frequently used for diagnosis of coronary artery disease. Radiation dose may become a potential concern as the scan needs to cover a wide range of cardiac phases during a heart cycle. A common method to reduce radiation is to limit the full-dose acquisition to a predefined range of phases while reducing the radiation dose for the rest. Our goal in this study is to develop a spatiotemporal deep learning method to enhance the quality of low-dose CTA images at phases acquired at reduced radiation dose. Recently, we demonstrated that a deep learning method, Cycle-Consistent generative adversarial networks (CycleGAN), could effectively denoise low-dose CT images through spatial image translation without labeled image pairs in both low-dose and full-dose image domains. As CycleGAN does not utilize the temporal information in its denoising mechanism, we propose to use RecycleGAN, which could translate a series of images ordered in time from the low-dose domain to the full-dose domain through an additional recurrent network. To evaluate RecycleGAN, we use the XCAT phantom program, a highly realistic simulation tool based on real patient data, to generate MP-CTA image sequences for 18 patients (14 for training, 2 for validation and 2 for test). Our simulation results show that RecycleGAN can achieve better denoising performance than CycleGAN based on both visual inspection and quantitative metrics. We further demonstrate the superior denoising performance of RecycleGAN using clinical MP-CTA images from 50 patients.

Unsupervised learning-based dual-domain method for low-dose CT denoising

Objective. Low-dose CT (LDCT) is an important research topic in the field of CT imaging because of its ability to reduce radiation damage in clinical diagnosis. In recent years, deep learning techniques have been widely applied in LDCT imaging and a large number of denoising methods have been proposed. However, One major challenge of supervised deep learning-based methods is the exactly geometric pairing of datasets with different doses. Therefore, the aim of this study is to develop an unsupervised learning-based LDCT imaging method to address the aforementioned challenges. Approach. In this paper, we propose an unsupervised learning-based dual-domain method for LDCT denoising, which consists of two stages: the first stage is projection domain denoising, in which the unsupervised learning method Noise2Self is applied to denoise the projection data with statistically independent and zero-mean noise. The second stage is an iterative enhancement approach, which combines the prior information obtained from the generative model with an iterative reconstruction algorithm to enhance the details of the reconstructed image. Main results. Experimental results show that our proposed method outperforms the comparison method in terms of denoising effect. Particularly, in terms of SSIM, the denoised results obtained using our method achieve the highest SSIM. Significance. In conclusion, our unsupervised learning-based method can be a promising alternative to the traditional supervised methods for LDCT imaging, especially when the availability of the labeled datasets is limited.

Vector Quantized Convolutional Autoencoder for Low-Dose CT Image Reconstruction with Perceptual and Bias Reducing Loss

In medical imaging, Computed Tomography (CT) is one of the most often utilized imaging modalities for diagnosing different disorders. Deep learning has become significant in the field of medical imaging, specifically investigated for low-dose CT. In recently available CT Scanners, Low Dose CT (LDCT) reconstruction is presented with a post-processing approach, which uses deep learning-based methods to reduce the noise level. Applying low radiation dosage can decrease damage to patients but the projected image is corrupted with noise due to lower intensity and fewer angle measurements, resulting in excessive noise in the reconstructed CT image. Thus, the presence of noise and artifacts in LDCT images limits their potential use. Here, a vector quantized convolutional encoder network is proposed for the image reconstruction task. The network is trained using the LoDoPaB-CT Dataset and tested on chest CT images. The qualitative and quantitative results produced better quality results when compared with the recent state-of-art deep learning-based methods. The quality of the results is optimized with perceptual and Bias-Reducing loss functions.

Low-Dose CT Image Synthesis for Domain Adaptation Imaging Using a Generative Adversarial Network With Noise Encoding Transfer Learning.

Deep learning (DL) based image processing methods have been successfully applied to low-dose x-ray images based on the assumption that the feature distribution of the training data is consistent with that of the test data. However, low-dose computed tomography (LDCT) images from different commercial scanners may contain different amounts and types of image noise, violating this assumption. Moreover, in the application of DL based image processing methods to LDCT, the feature distributions of LDCT images from simulation and clinical CT examination can be quite different. Therefore, the network models trained with simulated image data or LDCT images from one specific scanner may not work well for another CT scanner and image processing task. To solve such domain adaptation problem, in this study, a novel generative adversarial network (GAN) with noise encoding transfer learning (NETL), or GAN-NETL, is proposed to generate a paired dataset with a different noise style. Specifically, we proposed a method to perform noise encoding operator and incorporate it into the generator to extract a noise style. Meanwhile, with a transfer learning (TL) approach, the image noise encoding operator transformed the noise type of the source domain to that of the target domain for realistic noise generation. One public and two private datasets are used to evaluate the proposed method. Experiment results demonstrated the feasibility and effectiveness of our proposed GAN-NETL model in LDCT image synthesis. In addition, we conduct additional image denoising study using the synthesized clinical LDCT data, which verified the merit of the proposed synthesis in improving the performance of the DL based LDCT processing method.