ISSN 1004-4140
CN 11-3017/P
Founded in 1992, bimonthly
Institute of Geophysics, China Earthquake Administration
Nuctech Company Limited



Articles Online First have been peer-reviewed and accepted, which are not yet assigned to volumes /issues, but are citable by Digital Object Identifier (DOI).
Display Method:
Research on the Application Value of 18F-FDG PET/CT Comwith Neuronal Antibody Detection in the Diagnosis and Treatment of PNS Patients
YUAN Leilei, CHEN Qian, QIAO Zhen, LI Xiaotong, FAN Di, ZHANG Wei, AI Lin
 doi: 10.15953/j.ctta.2022.070
Abstract(82) HTML PDF(12)
Objective: To explore the whole body 18F-FDGClinical value of PET/CT combined neuroantibody detection in the diagnosis and treatment of paraneoplastic neurological syndromes (PNS). Methods: Clinical, laboratory and imaging data of 56 hospitalized patients with suspected PNS who underwent systemic 18F-FDG PET/CT and neuropathic tumor antibody detection were retrospectively collected and followed up. ROC curve was performed to compare the diagnostic efficacy of PET/CT, neuronal antibody and the combined detection results of them. Results: Among the 56 patients with suspected PNS, there were 20 patients with malignant tumor, including 19 cases complicated with PNS and 1 patient with spinal cord metastasis which also lead to neurological symptoms. 18F-FDG PET/CT imaging indicated tumor or possible tumor in 23 cases, of which 20 cases were true positive, 3 cases were false positive (the follow-up results were reflux esophagitis, reactive bone changes, inflammatory lesions in neck), and the remaining 33 cases were true negative. The sensitivity, specificity and accuracy of PET/CT were 100%, 91.7% and 94.6%, respectively. There were 33 cases with positive neuroantibody, including 8 cases of tumor with PNS (3 cases with anti-Amphiphysin antibody encephalitis, 2 cases with anti-GABAb antibody encephalitis, 1 case with anti-Yo antibody encephalitis, and 2 cases with anti-Hu antibody encephalitis). There were 25 cases without tumor (10 cases with LGI1 antibody encephalitis, 3 cases with anti-amphiphysin antibody encephalitis, 1 case with anti-Hu antibody encephalitis, 3 cases with anti-GABAb antibody encephalitis, 3 cases with anti-Yo antibody encephalitis and 1 case with Anti-caspr2, 1 case with GAD65, 1 case with NMDA, 1case with PNMA and 1 case with SOX1 antibody (1 case each) 23 cases were negative (12 cases with tumor). The sensitivity, specificity and accuracy of neuronal antibody test were 40.0%, 30.6% and 33.9%, respectively. The sensitivity, specificity and accuracy of the combined detection were 100.0%, 33.3%, 57.1%, 50%, 94.4%, 78.6%, respectively. ROC analysis showed that AUC were 0.958 (P=0.000<0.05; 95% CI 0.904~1.000), 0.353 (P=0.070>0.05; 95% CI 0.199~0.506), 0.667 (P=0.040<0.05; 95% CI 0.528~0.806) and 0.672 (P=0.034<0.05; 95% CI 0.514~0.830). Conclusion: Whole body 18F-FDG PET/CT can be the first choice for noninvasive tumor screening in patients with suspected PNS.
Evaluation of Invasion of Pulmonary Subsolid Nodules by Artificial Intelligence Volumetric Density Method
WANG Jingchen, CHAI Jun
 doi: 10.15953/j.ctta.2022.099
Abstract(40) HTML PDF(3)
Objective: To explore the value of artificial intelligence (AI) volumetric density method in determining the invasion of pulmonary hyposolid nodules (SSNs). Methods: 108 SSNs and pathological results of 106 patients were reviewed, divided into glandular prodromal lesions group and adenocarcinoma group. Pulmonary nodule AI software was used to measure and compare the CT quantitative parameters of the two groups, including maximum CT value, minimum CT value, average CT value, kurtosis, skewness, Perc.25%, Perc.50%, Perc.75%, Perc.90%, nodule volume and mean nodule diameter. Receiver operating characteristic curve (ROC) was obtained by Medcalc software to evaluate the sensitivity, specificity, positive predictive value and negative predictive value for the diagnosis of SSNs infiltration, and their diagnostic performance was evaluated by logistic regression analysis. Results: There were significant differences in most CT quantitative parameters of SSNs. The highest diagnostic efficiency was Perc.25% and AUC was 0.797. Perc.50% and CT mean, AUC were 0.787. Logistic regression analysis showed that Perc.25% with the highest diagnostic efficiency was combined with Perc.50% and CT mean value, respectively. The model with Perc.25% and CT mean value had the highest diagnostic efficiency, and the combined diagnostic model had higher diagnostic efficiency than Perc.25% and CT mean value alone. According to Medcalc software, SSNs with Perc.25% ≥−578 HU and mean CT value ≥ −468 HU are more likely to be adenocarcinoma group. In this study, Perc.25% was combined with the mean diameter of nodules, and a very valuable combined diagnostic model II was obtained to judge the infiltration of SSNs. Conclusion: AI volume density method has a high diagnostic value for SSNs invasion. The combination of Perc.25% and mean CT value can accurately judge the invasion than the use of average CT value alone, thus providing a quantitative basis for the clinical management of SSNs.
Evaluation of Diagnostic Value of Pulmonary Nodules Based on Two AI Softwares
CHEN Xinhua, HUANG Xiaoqi, LI Jianlong, GUO Youmin
 doi: 10.15953/j.ctta.2022.087
Abstract(32) HTML PDF(9)
Objective: To explore the clinical value of two kinds of AI detection software in ≥ 5 mm pulmonary nodules. Methods: A total of 92 patients with pulmonary nodules (483 nodules) were selected from June 2021 to October 2021 in the affiliated Hospital of Yan'an University. The nodules detected by AI software were evaluated and the number and type of nodules were recorded by two senior radiologists. Manual film reading was evaluated by two senior imaging doctors, which was used as the gold standard for nodule recognition. The detection rate, false positive rate and false negative rate of the two software’s were calculated, and the nodule detection value of the two AI software was evaluated. Chi-square test and Fisher precision test are used to compare the differences between different software and gold standard. Finally, the diagnostic value of the combination of two kinds of AI software for pulmonary nodules was calculated. Results: The detection rates of software A and software B nodules were 92.1% and 87.0% respectively. The coincidence degree between software A and manual reading was general (Kappa=0.213), while that between software B and manual reading was weak (Kappa=0.150). There was significant difference in the detection of solid nodules and calcified nodules between software A and manual reading, and between software B and pure ground glass nodules. The detection rate of nodules combined with two kinds of AI software was 97.1%. Compared with manual reading, there was no significant difference in the detection of nodule types. The combination of the two AI software’s had a good agreement with manual reading (Kappa=0.439). Conclusion: Two kinds of AI software association improve the ability of nodule detection and classification analysis. The method of joint diagnosis is recommended for clinical use, and it also provides evidence for further improving the homogenization management of AI data sets.
Application of Comprehensive Geophysical Prospecting Method in Goaf Detection
TANG Su, WU Yinting, XING Hao, WEI Yongshan, WANG Xuan
 doi: 10.15953/j.ctta.2022.176
Abstract(62) HTML PDF(5)
To accurately explore distribution of a historical small coal mine goaf in Hami, Xinjiang, integrated application of microtremor and transient electromagnetic is used for goaf exploration researching. This work use different physical properties data into lithology, use microtremor exploration to determine the goaf position and use transient electromagnetic exploration identify goaf water. Two methods can complement each other, verify each other, and effectively reduce the wrong results of single geophysical method. Which also provides a reference for the selection and application of geophysical exploration means in small mine goaf exploration.
Research Progress of Scattering Artifact Correction in Medical Cone-beam Computed Tomography Imaging Based on Deep Learning
ZHANG Wenjun, HUANG Gang, DING Haining, XU Hongchun
 doi: 10.15953/j.ctta.2022.131
Abstract(56) HTML PDF(14)
In the medical computed tomography imaging system, Compton scattered photons and Rayleigh scattered photons generated by the interaction between X-ray and objects have serious impact on image quality, especially in cone beam computed tomography and multi-layer detector system compared to fan beam computed tomography system. At present, there have been many scattering artifacts correction methods, which are summarized into three categories: hardware correction methods, software correction methods and hybrid software and hardware correction methods. However, with the advances in computing power of computers and the development of deep learning in medical image processing, some new methods of scattering artifacts correction have appeared in recent years. This paper first introduces the traditional correction methods. Then, the method of scattering artifacts correction based on deep learning is introduced in detail, which is divided into image domain learning method and projection domain learning method. Different deep-learning neural networks in this method are introduced in detail too. Finally, the application prospects of the deep learning method in multi-source computed tomography imaging scattering artifacts are prospected.
CT and MRI Findings of Multiple Infarcted Regenerative Nodules in Liver Cirrhosis after Variceal Hemorrhage
XU Xiaoli, ZHANG Tao, ZHANG Xueqin
 doi: 10.15953/j.ctta.2022.063
Abstract(18) HTML PDF(3)
Objective: To investigate the CT and MRI features of multiple infarcted regenerative nodules in cirrhosis after variceal hemorrhage. Methods: A total of 21 patients, including 13 males and 8 females, who were diagnosed with multiple infarction regenerating nodules in cirrhosis after variceal hemorrhage, were included in this study. All patients were examined by 3.0TMR scanner or 256 slice spiral CT, the enhancement pattern, signal intensity, shape, number, size, edge, location and distribution of lesions were analyzed. Results: In CT or MRI imaging, 3 patients had 10 or less lesions, and 19 patients had more than 10 lesions. The diameter of liver lesions was 3~26mm.Most lesions are round nodules, most of the lesions were clustered and distributed. The lesions were mainly distributed in the subcapsular region of liver. After dynamic enhancement of CT and MRI, most nodules have no obvious enhancement, and a few can have marginal enhancement. On T1WI images, all lesions showed equal signal or slightly low signal. On T2WI images, most lesions are high signals with clear boundary. During CT and MRI follow-up, the lesions disappeared in 13 patients and shrank or significantly reduced in 8 patients. Conclusions: CT and MRI can show the imaging features of multiple infarcted regenerative nodules in liver cirrhosis after variceal hemorrhage, which can be differentiated from liver malignancy by image follow-up, clinical history, and tumor indicators.
Display Method:
2022, 31(6).  
Low-dose CT Image Reconstruction Method Based on CNN and Transformer Coupling Network
QIAO Yiyu, QIAO Zhiwei
2022, 31(6): 697-707.   doi: 10.15953/j.ctta.2022.114
Abstract(334) HTML PDF(38)
Under the condition that the number of projection angles is constant, reducing the radiation dose under each angle is an effective way to realize low-dose CT. However, the reconstructed images obtained through this method can be very noisy. At present, the deep learning image denoising method represented by convolutional neural networks (CNN) has become a classical method for low-dose CT image denoising. Inspired by the good performance of transformer in computer vision tasks, this paper proposes a CNN transformer coupling network (CTC) to further improve the performance of CT image denoising. CTC network makes comprehensive use of local information association ability of CNN and global information capture ability of transformer, constructs eight core network blocks composed of CNN components and an improved transformer component, which are interconnected based on residual connection mechanism and information reuse mechanism. Compared with the existing four denoising networks, CTC network demonstrate better denoising ability and can realize high-precision low-dose CT image reconstruction.
Low-dose CT Reconstruction Based on Deep Energy Models
ZHU Yuanzheng, LV Qiwen, GUAN Yu, LIU Qiegen
2022, 31(6): 709-720.   doi: 10.15953/j.ctta.2021.077
Abstract(647) HTML PDF(70)
Reducing the dose of computed tomography (CT) is essential for reducing the radiation risk in clinical applications. With the rapid development and wide application of deep learning, it has brought new directions for the development of low-dose CT imaging algorithms. Unlike most existing prior-driven algorithms that benefit from manually designed prior functions or supervised learning schemes, in this paper, we use an energy-based deep model to learn the prior knowledge of normal-dose CT, and then in the iterative reconstruction phase, we integrate data consistency as a conditional item into the iterative generation model of low-dose CT, and realize the low-dose CT reconstruction through the prior experience of iterative updating training of Langevin dynamics. The experimental results show that the proposed method hold excellent noise reduction and detail retention capabilities.
Low-dose CT Denoising Based on Subspace Projection and Edge Enhancement
WEI Yili, YANG Ziyuan, XIA Wenjun, WANG Tao, ZHANG Yi
2022, 31(6): 721-729.   doi: 10.15953/j.ctta.2022.108
Abstract(152) HTML PDF(38)
Low-dose computed tomography (CT) is a relatively safe method for disease screening. But low-dose CT images often contain severe noise and artifacts, which seriously affect the subsequent diagnosis. To solve this problem, this paper proposes a subspace projection and edge enhancement network (SPEENet). SPEENet hold an architecture of autoencoder, including two main modules: dual stream encoder and decoder. The dual stream encoder can be divided into two parts: noise image coding stream and edge information coding stream. The noise image coding stream removes the noise and artifacts in low-dose CT images by using the image features extracted from the low-dose CT images. The edge information coding stream mainly focuses on the edge information of low-dose CT images and fully utilize the edge information to preserve the structures. In order to make full use of the encoder features, this paper introduces the noise basis projection module to establish a basis based on the features of encoder and decoder, and uses this basis to project the features extracted by the encoder into the corresponding subspace to obtain better feature representation. In this paper, experiments are conducted on the public database to verify the effectiveness. The experimental results show that SPEENet can achieve better denoising performance than other low-dose CT denoising networks.
Nuclear TV Multi-channel Image Reconstruction Algorithm Based on Chambolle-pock Framework
MA Jingyi, QIAO Zhiwei
2022, 31(6): 731-747.   doi: 10.15953/j.ctta.2022.111
Abstract(243) HTML PDF(48)
Total variation (TV) minimum algorithm is an image reconstruction algorithm based on compressed sensing theory, which can realize the reconstruction of images with high accuracy from sparse projection or noisy projection data and has been widely used in computed tomography (CT), magnetic resonance imaging (MRI) and electronic paramagnetic resonance imaging (EPRI). Energy spectrum CT, T1 or T2 weighted MRI and EPRI both belong to multi-channel imaging. The channel-by-channel TV algorithm can achieve high-precision image reconstruction, but it ignores the similarity among the images of each channel while Nuclear TV algorithm is a TV algorithm that considers the image similarity among channels, and can realize high-precision image reconstruction. For multi-channel image reconstruction, taking CT reconstruction as an example, this paper proposes a nuclear TV multi-channel image reconstruction algorithm based on the framework of Chambolle-pock algorithm. Through the reconstruction experiments of simulated phantom and real CT image phantom, the accuracy of the algorithm is verified, the convergence of the algorithm is analyzed, the influence of algorithm parameters on the convergence rate is explored, and the sparse reconstruction ability and noisy projection reconstruction ability of the algorithm are evaluated. The experimental results show that the proposed algorithm can achieve higher reconstruction accuracy than the channel-by-channel TV algorithm. Nuclear TV algorithm is a high-precision multi-channel image reconstruction algorithm, which can be applied to multi-channel reconstruction of various imaging modes.
Research on Image Analysis Method of Spectral CT Based on Principal Component Analysis
DI Yunxia, KONG Huihua, NIU Xiaowei
2022, 31(6): 749-760.   doi: 10.15953/j.ctta.2022.077
Abstract(255) HTML PDF(32)
Spectral computed tomography (CT) based on photon counting detector can simultaneously collect projection data of multiple spectral channels and obtain absorption characteristics of material within corresponding energy ranges, so it can be effectively applied to material identification and material decomposition. Principal component analysis is an excellent multivariate analysis technique, which can be applied to process multi-energy spectral CT data. In this paper, principal component analysis was performed on spectral CT data in projection domain and image domain respectively, and the analysis results were compared systematically. Meanwhile, in order to reduce the influence of noise and improve the color characterization performance of spectral CT images, the method of combining double domain filtering with pixel value square was proposed to denoise the noisy principal component images, and then the selected principal component images were mapped to RGB color channels. The experimental results demonstrate that the principal component analysis can obtain clear CT images and identify the different components of the substance, whether in the projection domain or the image domain. However, compared with the principal component analysis method in the image domain, principal component analysis in the projection domain can retain more details of the substance and acquire clearer color CT images.