Journal of X-Ray Science and Technology - Volume 29, issue 2

Authors: Sethy, Prabira Kumar | Behera, Santi Kumari | Anitha, Komma | Pandey, Chanki | Khan, M.R.

Article Type: Research Article

Abstract: The objective of this study is to conduct a critical analysis to investigate and compare a group of computer aid screening methods of COVID-19 using chest X-ray images and computed tomography (CT) images. The computer aid screening method includes deep feature extraction, transfer learning, and machine learning image classification approach. The deep feature extraction and transfer learning method considered 13 pre-trained CNN models. The machine learning approach includes three sets of handcrafted features and three classifiers. The pre-trained CNN models include AlexNet, GoogleNet, VGG16, VGG19, Densenet201, Resnet18, Resnet50, Resnet101, Inceptionv3, Inceptionresnetv2, Xception, MobileNetv2 and ShuffleNet. The handcrafted features are GLCM, …LBP & HOG, and machine learning based classifiers are KNN, SVM & Naive Bayes. In addition, the different paradigms of classifiers are also analyzed. Overall, the comparative analysis is carried out in 65 classification models, i.e., 13 in deep feature extraction, 13 in transfer learning, and 39 in the machine learning approaches. Finally, all classification models perform better when applying to the chest X-ray image set as comparing to the use of CT scan image set. Among 65 classification models, the VGG19 with SVM achieved the highest accuracy of 99.81%when applying to the chest X-ray images. In conclusion, the findings of this analysis study are beneficial for the researchers who are working towards designing computer aid tools for screening COVID-19 infection diseases. Show more

Keywords: Computer-aided screening, coronavirus, X-Ray, CT images, machine learning, transfer learning, deep learning

DOI: 10.3233/XST-200784

Citation: Journal of X-Ray Science and Technology, vol. 29, no. 2, pp. 197-210, 2021

Authors: Abbas, M.A. | Alqahtani, M.S. | Alkulib, A.J. | Almohiy, H.M. | Alshehri, R.F. | Alamri, E.A. | Alamri, A.A.

Article Type: Research Article

Abstract: BACKGROUND: Recent occurrence of the 2019 coronavirus disease (COVID-19) outbreak, caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has highlighted the need for fast, accurate, and simple strategies to identify cases on a large scale. OBJECTIVE: This study aims to develop and test an accurate detection and severity classification methodology that may help medical professionals and non-radiologists recognize the behavior and propagation mechanisms of the virus by viewing computed tomography (CT) images of the lungs with implicit materials. METHODS: In this study, the process of detecting the virus began with the deployment of a …virtual material inside CT images of the lungs of 128 patients. Virtual material is a hypothetical material that can penetrate the healthy regions in the image by performing sequential numerical measurements to interpret images with high data accuracy. The proposed method also provides a segmented image of only the healthy parts of the lung. RESULTS: The resulting segmented images, which represent healthy parts of the lung, are classified into six levels of severity. These levels are classified according to physical symptoms. The results of the proposed methodology are compared with those of the radiologists’ reports. This comparison revealed that the gold-standard reports correlated with the results of the proposed methodology with a high accuracy rate of 93%. CONCLUSION: The study results indicate the possibility of relying on the proposed methodology for discovering the effects of the SARS-CoV-2 virus in the lungs through CT imaging analysis with limited dependency on radiologists. Show more

Keywords: COVID-19, viral detection, image processing, computed tomography, implicit materials

DOI: 10.3233/XST-200794

Citation: Journal of X-Ray Science and Technology, vol. 29, no. 2, pp. 211-228, 2021

Authors: Rezaeijo, Seyed Masoud | Abedi-Firouzjah, Razzagh | Ghorvei, Mohammadreza | Sarnameh, Samad

Article Type: Research Article

Abstract: BACKGROUND AND OBJECTIVE: Radiomics has been widely used in quantitative analysis of medical images for disease diagnosis and prognosis assessment. The objective of this study is to test a machine-learning (ML) method based on radiomics features extracted from chest CT images for screening COVID-19 cases. METHODS: The study is carried out on two groups of patients, including 138 patients with confirmed and 140 patients with suspected COVID-19. We focus on distinguishing pneumonia caused by COVID-19 from the suspected cases by segmentation of whole lung volume and extraction of 86 radiomics features. Followed by feature extraction, nine feature-selection procedures …are used to identify valuable features. Then, ten ML classifiers are applied to classify and predict COVID-19 cases. Each ML models is trained and tested using a ten-fold cross-validation method. The predictive performance of each ML model is evaluated using the area under the curve (AUC) and accuracy. RESULTS: The range of accuracy and AUC is from 0.32 (recursive feature elimination [RFE]+Multinomial Naive Bayes [MNB] classifier) to 0.984 (RFE+bagging [BAG], RFE+decision tree [DT] classifiers) and 0.27 (mutual information [MI]+MNB classifier) to 0.997 (RFE+k-nearest neighborhood [KNN] classifier), respectively. There is no direct correlation among the number of the selected features, accuracy, and AUC, however, with changes in the number of the selected features, the accuracy and AUC values will change. Feature selection procedure RFE+BAG classifier and RFE+DT classifier achieve the highest prediction accuracy (accuracy: 0.984), followed by MI+Gaussian Naive Bayes (GNB) and logistic regression (LGR)+DT classifiers (accuracy: 0.976). RFE+KNN classifier as a feature selection procedure achieve the highest AUC (AUC: 0.997), followed by RFE+BAG classifier (AUC: 0.991) and RFE+gradient boosting decision tree (GBDT) classifier (AUC: 0.99). CONCLUSION: This study demonstrates that the ML model based on RFE+KNN classifier achieves the highest performance to differentiate patients with a confirmed infection caused by COVID-19 from the suspected cases. Show more

Keywords: COVID-19, machine-learning, radiomics, chest CT images

DOI: 10.3233/XST-200831

Citation: Journal of X-Ray Science and Technology, vol. 29, no. 2, pp. 229-243, 2021

Authors: Tang, Hui | Lin, Yu Bing | Sun, Guo Yan | Bao, Xu Dong

Article Type: Research Article

Abstract: OBJECTIVE: To reduce secondary artifactes generated by the current interpolation-based metal artifact reduction (MAR) methods, this study proposes and tests a new Poisson fusion sinogram based metal artifact reduction (FS-MAR) method. METHODS: The proposed FS-MAR method consists of (1) generating the prior image, (2) forward projecting this prior image and applying the Poisson blending technique to seamlessly replace the metal-affected sinogram of the original projection in the metal projection region (MPR) by the prior image projection to get the corrected metal-free sinogram, and (3) performing the filtered back projection (FBP) on the corrected sinogram and filling the metal …image back to the metal-free corrected image to get the final artifact reduced image. Simulated images are calculated by taking clinical metal-free CT images as phantoms and inserting metals during the simulated projection process to get the corresponding metal-affected images by the FBP. After the simulated images are processed by the proposed MAR method, two metrics structural similarity index (SSIM) and peak signal-to-noise ratio (PSNR) are used to evaluate image quality. Finally, visual evaluation is also performed using several real clinical metal-affected images obtained from the Revision Radiology group. RESULTS: In two testing samples, using FS-MAR method yields the highest SSIM and PSNR of 0.8912 and 30.6693, respectively. Visual evaluation results on both simulated and clinical images also show that using FS-MAR method generates less image artifacts than using the interpolation-based algorithm. CONCLUSIONS: This study demonstrated that with the same prior image, applying the proposed Poisson FS-MAR method can achieve the higher image quality than using the interpolation-based algorithm. Show more

Keywords: Computed tomography, metal artifact reduction, image fusion, Poisson blending

DOI: 10.3233/XST-200799

Citation: Journal of X-Ray Science and Technology, vol. 29, no. 2, pp. 245-257, 2021

Authors: Manerikar, Ankit | Li, Fangda | Kak, Avinash C.

Article Type: Research Article

Abstract: BACKGROUND: Materials characterization made possible by dual energy CT (DECT) scanners is expected to considerably improve automatic detection of hazardous objects in checked and carry-on luggage at our airports. Training a computer to identify the hazardous items from DECT scans however implies training on a baggage dataset that can represent all the possible ways a threat item can packed inside a bag. Practically, however, generating such data is made challenging by the logistics (and the permissions) related to the handling of the hazardous materials. OBJECTIVE: The objective of this study is to present a software simulation pipeline that …eliminates the need for a human to handle dangerous materials and that allows for virtually unlimited variability in the placement of such materials in a bag alongside benign materials. METHODS: In this paper, we present our DEBISim software pipeline that carries out an end-to-end simulation of a DECT scanner for virtual bags. The key highlights of DEBISim are: (i) A 3D user-interactive graphics editor for constructing a virtual 3D bag with manual placement of different types of objects in it; (ii) An automated virtual bag generation algorithm for creating randomized baggage datasets; (iii) An ability to spawn deformable sheets and liquid-filled containers in a virtual bag to represent plasticized and liquid explosives; and (iv) A GPU-based X-ray forward modelling block for spiral cone-beam scanners used in checked baggage screening. RESULTS: We have tested our simulator using two standard CT phantoms: the American College of Radiology (ACR) phantom and the NIST security screening phantom as well as on a set of reference materials representing commonly encountered items in checked baggage. For these phantoms, we have assessed the quality of the simulator by comparing the simulated data reconstructions with real CT scans of the same phantoms. The comparison shows that the material-specific properties as well as the CT artifacts in the scans generated by DEBISim are close to those produced by an actual scanner. CONCLUSION: DEBISim is an end-to-end simulation framework for rapidly generating X-ray baggage data for dual energy cone-beam scanners. Show more

Keywords: X-ray based threat detection, baggage security screening, dual energy CT

DOI: 10.3233/XST-200808

Citation: Journal of X-Ray Science and Technology, vol. 29, no. 2, pp. 259-285, 2021

Authors: Shen, Wenting | Chen, Yang | Qian, Wen | Liu, Wangyan | Zhu, Yinsu | Xu, Yi | Zhu, Xiaomei

Article Type: Research Article

Abstract: BACKGROUND: Accuracy of CT-derived fractional flow reserve depends on good image quality. Thus, improving image quality during coronary CT angiography (CCTA) is important. OBJECTIVE: To investigate impact of respiratory motion artifact on coronary image quality focusing on vessel diameter and territory during one beat CCTA by a 256-row detector. METHODS: We retrospectively reviewed patients who underwent CCTA under free-breathing (n  = 100) and breath-holding (n  = 100), respectively. Coronary image quality is defined as 4-1 from excellent to poor (non-diagnostic) and respiratory motion artifact severity is also scored on a 4-point scale from no artifact to severe artifact. …Coronary image quality and respiratory motion artifact severity of all images were evaluated by two radiologists independently. RESULTS: Compared with free-breathing group, the image qualities are significantly higher in per-segment, per-vessel and per-patient levels (P  < 0.001) and proportion of segments with excellent image quality also improves significantly (73.6% vs 60.1%, P  < 0.001) in breath-holding group. The image quality improvement occurs in medium-sized coronary arterial segments. Coronary image quality improves with respiratory motion artifacts decreasing in both groups, respectively. CONCLUSION: During one heartbeat CCTA, breath-holding is still recommended to improve coronary image quality due to improvement of the image quality in the medium-sized coronary arteries. Show more

Keywords: Motion artifacts, computed tomography angiography, coronary angiography, breath holding

DOI: 10.3233/XST-200812

Citation: Journal of X-Ray Science and Technology, vol. 29, no. 2, pp. 287-296, 2021

Authors: Kwak, Jong Hyeok | Lee, Chi Hyung | Kim, Gyeong Rip | Lee, Sang Weon | Kim, Young Ha | Song, Geun Sung | Son, Dong Wuk | Sung, Hynu Chul | Kwak, Jin Sung | Sung, Soon Ki

Article Type: Research Article

Abstract: OBJECTIVE: In this study, we present an appropriate angle of incidence to reduce the distortions in images of L4 and L5 during a general anteroposterior radiograph examination. METHOD: We selected 170 patients who had normal radiological findings among those who underwent anteroposterior and lateral examination for lumbar vertebrae. An optimum angle of incidence wa suggested through the statistical analysis by measuring the lumbar lordosis angle and the intervertebral disc angle in these 170 patients. RESULT: We suggested the incident angle (10.28°) of L4 and the incident angle (23.49°) of L5. We compared the distorted area ratios …when the incident angle was 0°, 10°, and 23.5° using the ATOM® phantom. The ratio for the L4 decreased from 14.90% to 12.11% and that of the L5 decreased from 15.25% to 13.72% after applying the angle of incidence. We determined the incident angle (9.34°) of L4 and (21.26°) of L5 below 30° of LLA. Thus, we determined the incident angle (11.21°) of L4 and (25.73°) of L5 above 30° of LLA. CONCLUSION: When you apply the optimum angle of incidence, the distortion of image was minimized and an image between the joints adjacent to the anteroposterior vertebral image with an accurate structure was obtained. As a result, we were able to improve the quality of the image and enhance diagnostic information. Show more

Keywords: Lumbar vertebrae, lordotic angle, intervertebral angle, incident angle, anteroposterior, radiography

DOI: 10.3233/XST-200786

Citation: Journal of X-Ray Science and Technology, vol. 29, no. 2, pp. 297-306, 2021

Authors: Kang, Zhen | Xu, Anhui | Wang, Liang

Article Type: Research Article

Abstract: BACKGROUND: Since Gleason score (GS) 4 + 3 prostate cancer (PCa) has a worse prognosis than GS 3 + 4 PCa, differentiating these two types of PCa is of clinical significance. OBJECTIVE: To assess the predictive roles of using T2WI and ADC-derived image texture parameters in differentiating GS 3 + 4 from GS 4 + 3 PCa. METHODS: Forty-eight PCa patients of GS 3 + 4 and 37 patients of GS 4 + 3 are retrieved and randomly divided into training (60%) and testing (40%) sets. Axial image showing the maximum tumor size is selected in the T2WI and ADC maps for further image texture feature analysis. Three …hundred texture features are computed from each region of interest (ROI) using MaZda software. Feature reduction is implemented to obtain 30 optimal features, which are then used to generate the most discriminative features (MDF). Receiver operating characteristic (ROC) curve analysis is performed on MDF values in the training sets to achieve cutoff values for determining the correct rates of discrimination between two Gleason patterns in the testing sets. RESULTS: ROC analysis on T2WI and ADC-derived MDF values in the training set (n  = 51) results in a mean area under the curve (AUC) of 0.953±0.025 (with sensitivity 0.9274±0.0615 and specificity 0.897±0.069), and 0.985±0.013 (with sensitivity 0.9636±0.0446 and specificity 0.9726±0.0258), respectively. Using the corresponding MDF cutoffs, 95.3% (ranges from 76.5% to 100%) and 94.1% (ranged from 76.5% to 100%) of test cases (n  = 34) are correctly discriminated using T2WI and ADC-derived MDF values, respectively. CONCLUSIONS: The study demonstrates that using T2WI and ADC-derived image texture parameters has a potential predictive role in differentiating GS 3 + 4 and GS 4 + 3 PCa. Show more

Keywords: Classification of prostate cancer, magnetic resonance imaging, gleason score, image texture analysis

DOI: 10.3233/XST-200785

Citation: Journal of X-Ray Science and Technology, vol. 29, no. 2, pp. 307-315, 2021

Authors: Chang, Chia-Hao | Ni, Yu-Ching | Tseng, Sheng-Pin

Article Type: Research Article

Abstract: The study aims to develop a rational polynomial approximation method for improving the accuracy of the effective atomic number calculation with a dual-energy X-ray imaging system. This method is based on a multi-materials calibration model with iterative optimization, which can improve the calculation accuracy of the effective atomic number by adding a rational term without increasing the computation time. The performance of the proposed rational polynomial approximation method is demonstrated and validated by both simulated and experimental studies. The twelve reference materials are used to establish the effective atomic number calibration model, and the value of the effective atomic numbers …are between 5.444 and 22. For the accuracy of the effective atomic number calculation, the relative differences between calculated and experimental values are less than 8.5%for all sample cases in this study. The average calculation accuracy of the method proposed in this study can be improved by about 40%compared with the conventional polynomial approximation method. Additionally, experimental quality assurance phantom imaging result indicates that the proposed method is compliant with the international baggage inspection standards for detecting the explosives. Moreover, the experimental imaging results reveal that the difference of color between explosives and the surrounding materials is in significant contrast for the dual-energy image with the proposed method. Show more

Keywords: Effective atomic numbers, rational polynomial approximation, dual-energy X-ray, baggage inspection

DOI: 10.3233/XST-200790

Citation: Journal of X-Ray Science and Technology, vol. 29, no. 2, pp. 317-330, 2021

Authors: Hussein, Khalid I. | Alqahtani, Mohammed S. | Grelowska, Iwona | Reben, Manuela | Afifi, Hesham | Zahran, Heba | Yaha, I. S. | Yousef, El Sayed

Article Type: Research Article

Abstract: BACKGROUND: Metal oxide glass composites have attracted huge interest as promising shielding materials to replace toxic, heavy, and costly conventional shielding materials. OBJECTIVE: In this work, we evaluate shielding effectiveness of four novel tellurite-based glasses samples doped with oxide metals (namely, A, B, C, and D, which are 75TeO2 - 10P2 O5 - 10ZnO- 5PbF2 - 0.24Er2 O3 ; 70TeO2 - 10P2 O5 - 10ZnO- 5PbF2 -5MgO- 0.24Er2 O3 ; 70TeO2 - 10P2 O5 - 10ZnO- 5PbF2 - 5BaO- 0.24Er2 O3 ; and 7 0TeO2 - 10P2 O5 -10ZnO- 5PbF2 - 5SrO; respectively) by assessing them through …a wide range of ionizing radiation energies (0.015–15 MeV). METHODS: The radiation-shielding parameters including mass attenuation coefficient (MAC), linear attenuation coefficient (LAC), half-value layer (HVL), mean free path, (MFP), effective atomic number (Zeff ), effective electron number (Neff ), and the transmission factor are computed in the selected range of ionizing radiation energies. Furthermore, the proposed samples were compared with the most common shielding glass materials. The optical parameters viz oscillator, dispersion energy, nonlinear refractive indices, molar, and electronic polarizability of these transparent glasses are reported at different wavelengths. RESULTS: The results show that the proposed samples have considerable effectiveness as transparent shielding glass materials at various ionizing radiation energies. They can be employed for effective radiation-protection outcomes. Sample C demonstrated slightly better shielding properties than the other samples with differences of 1.33%, 4.6%, and 4.2% for samples A, B, and D, respectively. A similar trend is observed regarding the mass attenuation coefficients. Nevertheless, sample B shows better optical properties than the other prepared glass samples. CONCLUSIONS: Our findings indicate that the proposed novel glass samples have good shielding properties and optical characteristics, which can pave the way for their utilization as transparent radiation-shielding materials in medical and industrial applications. Show more

Keywords: Optical parameters, glasses, half-value layer, mass attenuation coefficient, mean free path, radiation protection

DOI: 10.3233/XST-200780

Citation: Journal of X-Ray Science and Technology, vol. 29, no. 2, pp. 331-345, 2021