Deep Learning Enabled Deblurring of Computed Tomography Images of Porous Media

2021 ◽  
Author(s):  
Khalid Labib Alsamadony ◽  
Ertugrul Umut Yildirim ◽  
Guenther Glatz ◽  
Umair bin Waheed ◽  
Sherif M. Hanafy
Keyword(s):  
Computed Tomography ◽  
Image Quality ◽  
Transfer Learning ◽  
Temporal Resolution ◽  
Exposure Time ◽  
Ct Images ◽  
Loss Functions ◽  
High Quality ◽  
X Ray ◽  
Training Images

Abstract Computed tomography (CT) is an important tool to characterize rock samples allowing quantification of physical properties in 3D and 4D. The accuracy of a property delineated from CT data is strongly correlated with the CT image quality. In general, high-quality, lower noise CT Images mandate greater exposure times. With increasing exposure time, however, more wear is put on the X-Ray tube and longer cooldown periods are required, inevitably limiting the temporal resolution of the particular phenomena under investigation. In this work, we propose a deep convolutional neural network (DCNN) based approach to improve the quality of images collected during reduced exposure time scans. First, we convolve long exposure time images from medical CT scanner with a blur kernel to mimic the degradation caused because of reduced exposure time scanning. Subsequently, utilizing the high- and low-quality scan stacks, we train a DCNN. The trained network enables us to restore any low-quality scan for which high-quality reference is not available. Furthermore, we investigate several factors affecting the DCNN performance such as the number of training images, transfer learning strategies, and loss functions. The results indicate that the number of training images is an important factor since the predictive capability of the DCNN improves as the number of training images increases. We illustrate, however, that the requirement for a large training dataset can be reduced by exploiting transfer learning. In addition, training the DCNN on mean squared error (MSE) as a loss function outperforms both mean absolute error (MAE) and Peak signal-to-noise ratio (PSNR) loss functions with respect to image quality metrics. The presented approach enables the prediction of high-quality images from low exposure CT images. Consequently, this allows for continued scanning without the need for X-Ray tube to cool down, thereby maximizing the temporal resolution. This is of particular value for any core flood experiment seeking to capture the underlying dynamics.

scholarly journals Improving grazing-incidence small-angle X-ray scattering–computed tomography images by total variation minimization

2020 ◽  
Vol 53 (1) ◽  
pp. 140-147
Author(s):  
Hiroki Ogawa ◽  
Shunsuke Ono ◽  
Yukihiro Nishikawa ◽  
Akihiko Fujiwara ◽  
Taizo Kabe ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Image Quality ◽  
Total Variation ◽  
Small Angle ◽  
Grazing Incidence ◽  
Ct Images ◽  
Measurement Time ◽  
X Ray ◽  
X Ray Scattering ◽  
Ray Scattering

Grazing-incidence small-angle X-ray scattering (GISAXS) coupled with computed tomography (CT) has enabled the visualization of the spatial distribution of nanostructures in thin films. 2D GISAXS images are obtained by scanning along the direction perpendicular to the X-ray beam at each rotation angle. Because the intensities at the q positions contain nanostructural information, the reconstructed CT images individually represent the spatial distributions of this information (e.g. size, shape, surface, characteristic length). These images are reconstructed from the intensities acquired at angular intervals over 180°, but the total measurement time is prolonged. This increase in the radiation dosage can cause damage to the sample. One way to reduce the overall measurement time is to perform a scanning GISAXS measurement along the direction perpendicular to the X-ray beam with a limited interval angle. Using filtered back-projection (FBP), CT images are reconstructed from sinograms with limited interval angles from 3 to 48° (FBP-CT images). However, these images are blurred and have a low image quality. In this study, to optimize the CT image quality, total variation (TV) regularization is introduced to minimize sinogram image noise and artifacts. It is proposed that the TV method can be applied to downsampling of sinograms in order to improve the CT images in comparison with the FBP-CT images.

Detection of COVID-19 Using Deep Transfer Learning-Based Approach from X-Ray and Computed Tomography(CT) Images

2021 ◽  
pp. 303-313
Author(s):  
Kumar Kalpadiptya Roy ◽  
Ipsita Mazumder ◽  
Arijit Das ◽  
Subhram Das
Keyword(s):  
Computed Tomography ◽  
Transfer Learning ◽  
Ct Images ◽  
X Ray

Image Quality Analysis of X-ray Grating Interferometer Using Integrating-bucket Method

2020 ◽  
Vol 64 (2) ◽  
pp. 20503-1-20503-5
Author(s):  
Faiz Wali ◽  
Shenghao Wang ◽  
Ji Li ◽  
Jianheng Huang ◽  
Yaohu Lei ◽  
...  
Keyword(s):  
Image Quality ◽  
Phase Contrast ◽  
Quality Analysis ◽  
Phase Contrast Imaging ◽  
Phase Image ◽  
Contrast Imaging ◽  
High Quality ◽  
Image Quality Analysis ◽  
X Ray ◽  
Grating Interferometer

Abstract Grating-based x-ray phase-contrast imaging has the potential to enhance image quality and provide inner structure details non-destructively. In this work, using grating-based x-ray phase-contrast imaging system and employing integrating-bucket method, the quantitative expressions of signal-to-noise ratios due to photon statistics and mechanical error are analyzed in detail. Photon statistical noise and mechanical error are the main sources affecting the image noise in x-ray grating interferometry. Integrating-bucket method is a new phase extraction method translated to x-ray grating interferometry; hence, its image quality analysis would be of great importance to get high-quality phase image. The authors’ conclusions provide an alternate method to get high-quality refraction signal using grating interferometer, and hence increases applicability of grating interferometry in preclinical and clinical usage.

scholarly journals Vulnerability in Deep Transfer Learning Models to Adversarial Fast Gradient Sign Attack for COVID-19 Prediction from Chest Radiography Images

2021 ◽  
Vol 11 (9) ◽  
pp. 4233
Author(s):  
Biprodip Pal ◽  
Debashis Gupta ◽  
Md. Rashed-Al-Mahfuz ◽  
Salem A. Alyami ◽  
Mohammad Ali Moni
Keyword(s):  
Deep Learning ◽  
Transfer Learning ◽  
Chest Radiography ◽  
High Sensitivity ◽  
Ct Images ◽  
Classification Performance ◽  
Learning Models ◽  
X Ray ◽  
Image Capturing ◽  
Fast Gradient

The COVID-19 pandemic requires the rapid isolation of infected patients. Thus, high-sensitivity radiology images could be a key technique to diagnose patients besides the polymerase chain reaction approach. Deep learning algorithms are proposed in several studies to detect COVID-19 symptoms due to the success in chest radiography image classification, cost efficiency, lack of expert radiologists, and the need for faster processing in the pandemic area. Most of the promising algorithms proposed in different studies are based on pre-trained deep learning models. Such open-source models and lack of variation in the radiology image-capturing environment make the diagnosis system vulnerable to adversarial attacks such as fast gradient sign method (FGSM) attack. This study therefore explored the potential vulnerability of pre-trained convolutional neural network algorithms to the FGSM attack in terms of two frequently used models, VGG16 and Inception-v3. Firstly, we developed two transfer learning models for X-ray and CT image-based COVID-19 classification and analyzed the performance extensively in terms of accuracy, precision, recall, and AUC. Secondly, our study illustrates that misclassification can occur with a very minor perturbation magnitude, such as 0.009 and 0.003 for the FGSM attack in these models for X-ray and CT images, respectively, without any effect on the visual perceptibility of the perturbation. In addition, we demonstrated that successful FGSM attack can decrease the classification performance to 16.67% and 55.56% for X-ray images, as well as 36% and 40% in the case of CT images for VGG16 and Inception-v3, respectively, without any human-recognizable perturbation effects in the adversarial images. Finally, we analyzed that correct class probability of any test image which is supposed to be 1, can drop for both considered models and with increased perturbation; it can drop to 0.24 and 0.17 for the VGG16 model in cases of X-ray and CT images, respectively. Thus, despite the need for data sharing and automated diagnosis, practical deployment of such program requires more robustness.

scholarly journals Multi-Cycle-Consistent Adversarial Networks for Edge Denoising of Computed Tomography Images

2021 ◽  
Vol 17 (4) ◽  
pp. 1-16
Author(s):  
Xiaowe Xu ◽  
Jiawei Zhang ◽  
Jinglan Liu ◽  
Yukun Ding ◽  
Tianchen Wang ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Image Denoising ◽  
Ct Images ◽  
High Dose ◽  
Ct Image ◽  
Paired Data ◽  
High Quality ◽  
Computed Tomography Images ◽  
Significant Difference ◽  
Global Cycle

As one of the most commonly ordered imaging tests, the computed tomography (CT) scan comes with inevitable radiation exposure that increases cancer risk to patients. However, CT image quality is directly related to radiation dose, and thus it is desirable to obtain high-quality CT images with as little dose as possible. CT image denoising tries to obtain high-dose-like high-quality CT images (domain Y ) from low dose low-quality CT images (domain X ), which can be treated as an image-to-image translation task where the goal is to learn the transform between a source domain X (noisy images) and a target domain Y (clean images). Recently, the cycle-consistent adversarial denoising network (CCADN) has achieved state-of-the-art results by enforcing cycle-consistent loss without the need of paired training data, since the paired data is hard to collect due to patients’ interests and cardiac motion. However, out of concerns on patients’ privacy and data security, protocols typically require clinics to perform medical image processing tasks including CT image denoising locally (i.e., edge denoising). Therefore, the network models need to achieve high performance under various computation resource constraints including memory and performance. Our detailed analysis of CCADN raises a number of interesting questions that point to potential ways to further improve its performance using the same or even fewer computation resources. For example, if the noise is large leading to a significant difference between domain X and domain Y , can we bridge X and Y with a intermediate domain Z such that both the denoising process between X and Z and that between Z and Y are easier to learn? As such intermediate domains lead to multiple cycles, how do we best enforce cycle- consistency? Driven by these questions, we propose a multi-cycle-consistent adversarial network (MCCAN) that builds intermediate domains and enforces both local and global cycle-consistency for edge denoising of CT images. The global cycle-consistency couples all generators together to model the whole denoising process, whereas the local cycle-consistency imposes effective supervision on the process between adjacent domains. Experiments show that both local and global cycle-consistency are important for the success of MCCAN, which outperforms CCADN in terms of denoising quality with slightly less computation resource consumption.

Using 100- instead of 120-kVp computed tomography to diagnose pulmonary embolism almost halves the radiation dose with preserved diagnostic quality

2010 ◽  
Vol 51 (3) ◽  
pp. 260-270 ◽  
Author(s):  
Peter Björkdahl ◽  
Ulf Nyman
Keyword(s):  
Computed Tomography ◽  
Pulmonary Embolism ◽  
Image Quality ◽  
Radiation Dose ◽  
Effective Dose ◽  
Ct Number ◽  
Subjective Image Quality ◽  
Diagnostic Quality ◽  
X Ray ◽  
Tube Potential

Background: Concern has been raised regarding the mounting collective radiation doses from computed tomography (CT), increasing the risk of radiation-induced cancers in exposed populations. Purpose: To compare radiation dose and image quality in a chest phantom and in patients for the diagnosis of pulmonary embolism (PE) at 100 and 120 peak kilovoltage (kVp) using 16-multichannel detector computed tomography (MDCT). Material and Methods: A 20-ml syringe containing 12 mg I/ml was scanned in a chest phantom at 100/120 kVp and 25 milliampere seconds (mAs). Consecutive patients underwent 100 kVp ( n = 50) and 120 kVp ( n = 50) 16-MDCT using a “quality reference” effective mAs of 100, 300 mg I/kg, and a 12-s injection duration. Attenuation (CT number), image noise (1 standard deviation), and contrast-to-noise ratio (CNR; fresh clot = 70 HU) of the contrast medium syringe and pulmonary arteries were evaluated on 3-mm-thick slices. Subjective image quality was assessed. Computed tomography dose index (CTDIvol) and dose–length product (DLP) were presented by the CT software, and effective dose was estimated. Results: Mean values in the chest phantom and patients changed as follows when X-ray tube potential decreased from 120 to 100 kVp: attenuation +23% and +40%, noise +38% and +48%, CNR −6% and 0%, and CTDIvol −38% and −40%, respectively. Mean DLP and effective dose in the patients decreased by 42% and 45%, respectively. Subjective image quality was excellent or adequate in 49/48 patients at 100/120 kVp. No patient with a negative CT had any thromboembolism diagnosed during 3-month follow-up. Conclusion: By reducing X-ray tube potential from 120 to 100 kVp, while keeping all other scanning parameters unchanged, the radiation dose to the patient may be almost halved without deterioration of diagnostic quality, which may be of particular benefit in young individuals.

scholarly journals A novel range telescope concept for proton CT

2022 ◽  
Author(s):  
Marc Granado-González ◽  
César Jesús-Valls ◽  
Thorsten Lux ◽  
Tony Price ◽  
Federico Sánchez
Keyword(s):  
Computed Tomography ◽  
Proton Beam ◽  
Energy Resolution ◽  
Plastic Scintillator ◽  
Proton Beam Therapy ◽  
Stopping Power ◽  
High Quality ◽  
X Ray ◽  
Proton Computed Tomography ◽  
Clinical Conditions

Abstract Proton beam therapy can potentially offer improved treatment for cancers of the head and neck and in paediatric patients. There has been asharp uptake of proton beam therapy in recent years as improved delivery techniques and patient benefits are observed. However, treatments are currently planned using conventional x-ray CT images due to the absence of devices able to perform high quality proton computed tomography(pCT) under realistic clinical conditions. A new plastic-scintillator-based range telescope concept, named ASTRA, is proposed here to measure the proton’s energy loss in a pCT system. Simulations conducted using GEANT4 yield an expected energy resolution of 0.7%. If calorimetric information is used the energy resolution could be further improved to about 0.5%. In addition, the ability of ASTRA to track multiple protons simultaneously is presented. Due to its fast components, ASTRA is expected to reach unprecedented data collection rates, similar to 10^8 protons/s.The performance of ASTRA has also been tested by simulating the imaging of phantoms. The results show excellent image contrast and relative stopping power reconstruction.

Sign in / Sign up

Export Citation Format

Share Document