Journal of X-Ray Science and Technology - Volume 28, issue 4

Authors: Li, Zicong | Zeng, Bingliang | Lei, Pinggui | Liu, Jiaqi | Fan, Bing | Shen, Qinglin | Pang, Peipei | Xu, Rongchun

Article Type: Research Article

Abstract: BACKGROUND: Pneumonia caused by COVID-19 shares overlapping imaging manifestations with other types of pneumonia. How to objectively and quantitatively differentiate pneumonia patients with and without COVID-19 virus remains clinical challenge. OBJECTIVE: To formulate standardized scoring criteria and an objective quantization standard to guide decision making in detection and diagnosis of COVID-19 virus induced pneumonia in clinical practice. METHODS: A retrospective dataset includes computed tomography (CT) images acquired from 43 pneumonia patients with COVID-19 virus detected by reverse transcription-polymerase chain reaction (RT-PCR) tests and 49 pneumonia patients without COVID-19 virus. All patients were treated during the same …time period in two hospitals. Key indicators of differential diagnosis were identified in relevant literature and the scores were quantified namely, patients with more than 8 points were identified as high risk, those with 6–8 points as moderate risk, and those with fewer than 6 points as low risk for COVID-19 virus. In the study, 3 radiologists determined the scores for all patients. Diagnostic sensitivity and specificity were subsequently calculated. RESULTS: A total of 61 patients were determined as high risk, among which 42 were COVID-19 positive by RT-PCR tests. Next, 9 were identified as moderate risk, one of whom was COVID-19 positive. Last, 22 were classified into the low-risk group, all of them are COVID-19 negative. Based on these results, the sensitivity of detection COVID-19 positive cases between the high-risk group and the non-high-risk group was 0.98 with 95% confidence interval [0.88, 1.00], and the specificity was 0.61 [0.46, 0.75]. The detection sensitivity between the moderate-/high-risk group and the low-risk group was 1.00 [0.92, 1.00], and the specificity was 0.45 [0.31, 0.60]. CONCLUSION: The proposed quantitative scoring criteria showed high sensitivity and moderate specificity in detecting COVID-19 using CT images, which indicates that these criteria may be beneficial for screening in real-world practice and helpful for long-term disease control. Show more

Keywords: COVID-19, Coronavirus, Pneumonia, X-ray computed tomography

DOI: 10.3233/XST-200689

Citation: Journal of X-Ray Science and Technology, vol. 28, no. 4, pp. 583-589, 2020

Authors: Mastouri, Rekka | Khlifa, Nawres | Neji, Henda | Hantous-Zannad, Saoussen

Article Type: Research Article

Abstract: BACKGROUND: Lung cancer is the most common cancer in the world. Computed tomography (CT) is the standard medical imaging modality for early lung nodule detection and diagnosis that improves patient’s survival rate. Recently, deep learning algorithms, especially convolutional neural networks (CNNs), have become a preferred methodology for developing computer-aided detection and diagnosis (CAD) schemes of lung CT images. OBJECTIVE: Several CNN-based research projects have been initiated to design robust and efficient CAD schemes for the detection and classification of lung nodules. This paper reviews the recent works in this area and gives an insight into technical progress. …METHODS: First, a brief overview of CNN models and their basic structures is presented in this investigation. Then, we provide an analytic comparison of the existing approaches to discover recent trend and upcoming challenges. We also introduce an objective description of both handcrafted and deep learning features, as well as the types of nodules, the medical imaging modalities, the widely used databases, and related works in the last three years. The articles presented in this work were selected from various databases. About 57% of reviewed articles published in the last year. RESULTS: Our analysis reveals that several methods achieved promising performance with high sensitivity rates ranging from 66% to 100% under the false-positive rates ranging from 1 to 15 per CT scan. It can be noted that CNN models have contributed to the accurate detection and early diagnosis of lung nodules. CONCLUSIONS: From the critical discussion and an outline for prospective directions, this survey provide researchers valuable information to master the deep learning concepts and to deepen their knowledge of the trend and latest techniques in developing CAD schemes of lung CT images. Show more

Keywords: CAD of CT images, deep learning, lung cancer screening, lung nodule detection, lung nodule classification

DOI: 10.3233/XST-200660

Citation: Journal of X-Ray Science and Technology, vol. 28, no. 4, pp. 591-617, 2020

Authors: Shi, Liu | Liu, Baodong | Yu, Hengyong | Wei, Cunfeng | Wei, Long | Zeng, Li | Wang, Ge

Article Type: Research Article

Abstract: Computed tomography (CT) has been widely applied in medical diagnosis, nondestructive evaluation, homeland security, and other science and engineering applications. Image reconstruction is one of the core CT imaging technologies. In this review paper, we systematically reviewed the currently publicly available CT image reconstruction open source toolkits in the aspects of their environments, object models, imaging geometries, and algorithms. In addition to analytic and iterative algorithms, deep learning reconstruction networks and open codes are also reviewed as the third category of reconstruction algorithms. This systematic summary of the publicly available software platforms will help facilitate CT research and development.

Keywords: CT, image reconstruction, open source, toolkits, algorithm, analytic, iterative, deep learning

DOI: 10.3233/XST-200666

Citation: Journal of X-Ray Science and Technology, vol. 28, no. 4, pp. 619-639, 2020

Authors: Farhood, Bagher | Mohammadi ASL, Kamal | Sarvizadeh, Mostafa | Aliasgharzadeh, Akbar

Article Type: Research Article

Abstract: OBJECTIVE: Several physical factors such as dose rate and photon energy may change response and sensitivity of polymer gel dosimeters. This study aims to evaluate the R 2 -dose response and sensitivity dependence of PASSAG-U gel dosimeters with 3% and 5% urea on dose rate and photon energy. MATERIALS AND METHODS: The PASSAG-U gel dosimeters were prepared under normal atmospheric conditions. The obtained gel dosimeters were irradiated to different dose rates (100, 200, and 300 cGy/min) and photon energies (6 and 15 MV). Finally, responses (R 2 ) of the PASSAG-U gel dosimeters with 3% and 5% urea were analyzed …by MRI technique at 1, 10, 14 days after the irradiation process. RESULTS: The findings showed that the R 2 -dose responses of PASSAG-U gel dosimeters with 3% and 5% urea do not vary under the differently evaluated dose rates and photon energies. The R 2 -dose sensitivity of PASSAG-U polymer gel dosimeter with 3% urea does not change under the differently evaluated dose rates and photon energies, but it changes for PASSAG-U polymer gel dosimeter with 5% urea. The dose resolution values ranged from 0.20 to 0.86 Gy and from 0.27 to 2.20 Gy for the PASSAG-U gel dosimeter with 3% and 5% urea for the different dose rates and photon energies, respectively. Furthermore, it was revealed that the R 2 -dose response and sensitivity dependence of PASSAG-U gel dosimeters with 3% and 5% urea on dose rate and photon energy can vary over post irradiation time. CONCLUSIONS: The study results demonstrated that dosimetric characteristics (dependence of dose rate and photon energy, and dose resolution) of PASSAG-U gel dosimeter with 3% were better than those of PASSAG-U gel dosimeter with 5% urea. Show more

Keywords: Polymer gel dosimetry, PASSAG-U, urea, photon energy, dose rate

DOI: 10.3233/XST-190625

Citation: Journal of X-Ray Science and Technology, vol. 28, no. 4, pp. 641-658, 2020

Authors: AlKubeyyer, Atheer | Ben Ismail, Mohamed Maher | Bchir, Ouiem | Alkubeyyer, Metab

Article Type: Research Article

Abstract: Meningioma is among the most common primary tumors of the brain. The firmness of Meningioma is a critical factor that influences operative strategy and patient counseling. Conventional methods to predict the tumor firmness rely on the correlation between the consistency of Meningioma and their preoperative MRI findings such as the signal intensity ratio between the tumor and the normal grey matter of the brain. Machine learning techniques have not been investigated yet to address the Meningioma firmness detection problem. The main purpose of this research is to couple supervised learning algorithms with typical descriptors for developing a computer-aided detection (CAD) …of the Meningioma tumor firmness in MRI images. Specifically, Local Binary Patterns (LBP), Gray Level Co-occurrence Matrix (GLCM) and Discrete Wavelet Transform (DWT) are extracted from real labeled MRI-T2 weighted images and fed into classifiers, namely support vector machine (SVM) and k-nearest neighbor (KNN) algorithm to learn association between the visual properties of the region of interest and the pre-defined firm and soft classes. The learned model is then used to classify unlabeled MRI-T2 weighted images. This paper represents a baseline comparison of different features used in CAD system that intends to accurately recognize the Meningioma tumor firmness. The proposed system was implemented and assessed using a clinical dataset. Using LBP feature yielded the best performance with 95% of F-score, 87% of balanced accuracy and 0.87 of the area under ROC curve (AUC) when coupled with KNN classifier, respectively. Show more

Keywords: Meningioma, tumor firmness, visual features, supervised learning

DOI: 10.3233/XST-200644

Citation: Journal of X-Ray Science and Technology, vol. 28, no. 4, pp. 659-682, 2020

Authors: Yin, Jin | Qiu, Jia-Jun | Qian, Wei | Ji, Lin | Yang, Dan | Jiang, Jing-Wen | Wang, Jun-Ren | Lan, Lan

Article Type: Research Article

Abstract: BACKGROUND: In regular examinations, it may be difficult to visually identify benign and malignant liver tumors based on plain computed tomography (CT) images. RCAD (radiomics-based computer-aided diagnosis) has proven to be helpful and provide interpretability in clinical use. OBJECTIVE: This work aims to develop a CT-based radiomics signature and investigate its correlation with malignant/benign liver tumors. METHODS: We retrospectively analyzed 168 patients of hepatocellular carcinoma (malignant) and 117 patients of hepatic hemangioma (benign). Texture features were extracted from plain CT images and used as candidate features. A radiomics signature was developed from the candidate features. We …performed logistic regression analysis and used a multiple-regression coefficient (termed as R ) to assess the correlation between the developed radiomics signature and malignant/benign liver tumors. Finally, we built a logistic regression model to classify benign and malignant liver tumors. RESULTS: Thirteen features were chosen from 1223 candidate features to constitute the radiomics signature. The logistic regression analysis achieved an R  = 0.6745, which was much larger than R α  = 0.3703 (the critical value of R at significant level α  = 0.001). The logistic regression model achieved an average AUC of 0.87. CONCLUSIONS: The developed radiomics signature was statistically significantly correlated with malignant/benign liver tumors (p  < 0.001). It has potential to help enhance physicians’ diagnostic abilities and play an important role in RCADs. Show more

Keywords: Radiomics signature, texture analysis, liver tumor, logistic regression model, classification between malignant and benign tumors

DOI: 10.3233/XST-200675

Citation: Journal of X-Ray Science and Technology, vol. 28, no. 4, pp. 683-694, 2020

Authors: Anam, Choirul | Adhianto, Dwi | Sutanto, Heri | Adi, Kusworo | Ali, Mohd Hanafi | Rae, William Ian Duncombe | Fujibuchi, Toshioh | Dougherty, Geoff

Article Type: Research Article

Abstract: The objective of this study is to determine X-ray dose distribution and the correlation between central, peripheral and weighted-centre peripheral doses for various phantom sizes and tube voltages in computed tomography (CT). We used phantoms developed in-house, with various water-equivalent diameters (Dw) from 8.5 up to 42.1 cm. The phantoms have one hole in the centre and four holes at the periphery. By using these five holes, it is possible to measure the size-specific central dose (Ds,c), peripheral dose (Ds,p), and weighted dose (Ds,w).The phantoms are scanned using a CT scanner (Siemens Somatom Definition AS), with the tube voltage varied from …80 up to 140 kVps. The doses are measured using a pencil ionization chamber (Ray safe X2 CT Sensor) in every hole for all phantoms. The relationships between Ds,c, Ds,p, and Ds,w, and the water-equivalent diameter are established. The size-conversion factors are calculated. Comparisons between Ds,c, Ds,p, and Ds,ware also established. We observe that the dose is relatively homogeneous over the phantom for water-equivalent diameters of 12–14 cm. For water-equivalent diameters less than 12 cm, the dose in the centre is higher than at the periphery, whereas for water-equivalent diameters greater than 14 cm, the dose at the centre is lower than that at the periphery. We also find that the distribution of the doses is influenced by the tube voltage. These dose distributions may be useful for calculating organ doses for specific patients using their CT images in future clinical practice. Show more

Keywords: Dose distribution, size-specific dose estimate (SSDE), central dose, peripheral dose, weighted dose

DOI: 10.3233/XST-200667

Citation: Journal of X-Ray Science and Technology, vol. 28, no. 4, pp. 695-708, 2020

Authors: Yang, Tiejun | Zhou, Yudan | Li, Lei | Zhu, Chunhua

Article Type: Research Article

Abstract: BACKGROUND: Brain tumor segmentation plays an important role in assisting diagnosis of disease, treatment plan planning, and surgical navigation. OBJECTIVE: This study aims to improve the accuracy of tumor boundary segmentation using the multi-scale U-Net network. METHODS: In this study, a novel U-Net with dilated convolution (DCU-Net) structure is proposed for brain tumor segmentation based on the classic U-Net structure. First, the MR brain tumor images are pre-processed to alleviate the class imbalance problem by reducing the input of the background pixels. Then, the multi-scale spatial pyramid pooling is used to replace the max pooling at …the end of the down-sampling path. It can expand the feature receptive field while maintaining image resolution. Finally, a dilated convolution residual block is combined to improve the skip connections in the training networks to improve the network’s ability to recognize the tumor details. RESULTS: The proposed model has been evaluated using the Brain Tumor Segmentation (BRATS) 2018 Challenge training dataset and achieved the dice similarity coefficients (DSC) score of 0.91, 0.78 and 0.83 for whole tumor, core tumor and enhancing tumor segmentation, respectively. CONCLUSIONS: The experiment results indicate that the proposed model yields a promising performance in automated brain tumor segmentation. Show more

Keywords: Brain tumor segmentation, dilated convolution, multi-scale spatial pyramid pooling, DCU-Net, U-Net

DOI: 10.3233/XST-200650

Citation: Journal of X-Ray Science and Technology, vol. 28, no. 4, pp. 709-726, 2020

Authors: Hou, Yubao

Article Type: Research Article

Abstract: The automatic classification of breast cancer pathological images has important clinical application value. However, to develop the classification algorithm using the artificially extracted image features faces several challenges including the requirement of professional domain knowledge to extract and compute highiquality image features, which are often time-consuming, laborious, and difficult. For overcoming these challenges, this study developed and applied an improved deep convolutional neural network model to perform automatic classification of breast cancer using pathological images. Specifically, in this study, data enhancement and migration learning methods are used to effectively avoid the overfitting problems with deep learning models when they are …limited by training image sample size. Experimental results show that a 91% recognition rate or accuracy when applying this improved deep learning model to a publicly available dataset of BreaKHis. Comparing with other previously used models, the new model yields good robustness and generalization. Show more

Keywords: Breast cancer histopathological image classification, deep leaning, convolutional neural network, transfer learning, data augmentation, open dataset of BreaKHis

DOI: 10.3233/XST-200658

Citation: Journal of X-Ray Science and Technology, vol. 28, no. 4, pp. 727-738, 2020

Authors: Hsiao, Chia-Chi | Chen, Po-Chou | Kuo, Pei-Chi | Ho, Chih-Hao | Jao, Jo-Chi

Article Type: Research Article

Abstract: BACKGROUND: Although computed tomography (CT) is a powerful diagnostic imaging modality for diagnosing vascular diseases, it is some what risky to human health due to the high radiation dosage. Thus, CT vendors have developed low dose computed tomography (LDCT) aiming to solve this problem. Nowadays, LDCT has gradually become a main stream of CT examination. OBJECTIVE: This study aimed to assess the feasibility of LDCTAin an animal model and compare the imaging features and doses in two clinical scanners. METHODS: Twenty-two New Zealand rabbit head and neck CTA images pre- and post-contrast agent injection were performed …using256-sliceand 64-slice CT scanners. The tube voltages used in the 256-slice and the 64-slice CTA were 70 kVp and 80 kVp, respectively. Quantitative images indices and radiation doses obtained from CTA in these two scanners were compared. RESULTS: More neck arterial vessels could be visualized in multi-planar reconstruction (MPR) CTA on the 256-slice CT scanner than on the 64-slice CT scanner. After contrast agent injection, all observed neck arterial vessels had higher CT numbers in 256-slice CTA than in 64-slice CTA. There was no significant difference in contrast-to-noise (CNR) of CTA images between these two scanners. CT dose index (CTDI) and dose length product (DLP) for the 256-slice CTA were lower than those for the 64-slice CTA. CONCLUSIONS: Low dose CTA of rabbits with 70 or 80 kVp is feasible in a 256-slice or a 64-slice CT scanner. The radiation dose from the 256-slice CTA was much lower than that from the 64-slice CTA with comparable SNR and CNR. The technique can be further applied in longitudinal monitoring of an animal stroke model in the future. Show more

Keywords: Low dose computed tomography angiography, 256-slice CTA, 64-slice CTA, New Zealand rabbit model, CT number, signal-to-noise ratio, contrast-to-noise ratio

DOI: 10.3233/XST-200669

Citation: Journal of X-Ray Science and Technology, vol. 28, no. 4, pp. 739-750, 2020