Associating Peritoneal Metastasis With T2-Weighted MRI Images in Epithelial Ovarian Cancer Using Deep Learning and Radiomics: A Multicenter Study.

BACKGROUND: The preoperative diagnosis of peritoneal metastasis (PM) in epithelial ovarian cancer (EOC) is challenging and can impact clinical decision-making. PURPOSE: To investigate the performance of T2 -weighted (T2W) MRI-based deep learning (DL) and radiomics methods for PM evaluation in EOC patients. STUDY TYPE: Retrospective. POPULATION: Four hundred seventy-nine patients from five centers, including one training set (N = 297 [mean, 54.87 years]), one internal validation set (N = 75 [mean, 56.67 years]), and two external validation sets (N = 53 [mean, 55.58 years] and N = 54 [mean, 58.22 years]). FIELD STRENGTH/SEQUENCE: 1.5 or 3 T/fat-suppression T2W fast or turbo spin-echo sequence. ASSESSMENT: ResNet-50 was used as the architecture of DL. The largest orthogonal slices of the tumor area, radiomics features, and clinical characteristics were used to construct the DL, radiomics, and clinical models, respectively. The three models were combined using decision-level fusion to create an ensemble model. Diagnostic performances of radiologists and radiology residents with and without model assistance were evaluated. STATISTICAL TESTS: Receiver operating characteristic analysis was used to assess the performances of models. The McNemar test was used to compare sensitivity and specificity. A two-tailed P < 0.05 was considered significant. RESULTS: The ensemble model had the best AUCs, outperforming the DL model (0.844 vs. 0.743, internal validation set; 0.859 vs. 0.737, external validation set I) and clinical model (0.872 vs. 0.730, external validation set II). After model assistance, all readers had significantly improved sensitivity, especially for those with less experience (junior radiologist1, from 0.639 to 0.820; junior radiologist2, from 0.689 to 0.803; resident1, from 0.623 to 0.803; resident2, from 0.541 to 0.738). One resident also had significantly improved specificity (from 0.633 to 0.789). DATA CONCLUSIONS: T2W MRI-based DL and radiomics approaches have the potential to preoperatively predict PM in EOC patients and assist in clinical decision-making. EVIDENCE LEVEL: 4 TECHNICAL EFFICACY: Stage 2.

Numerical Simulation Investigation on Natural Gas Leakage under Different Kitchen Layouts in China.

The safety of an open kitchen is a controversial topic in China. In this study, natural gas leakage and ventilation processes under an open kitchen layout and closed kitchen layout are simulated by CFD. The evolution of a hazardous gas cloud and the triggering behaviors of alarms are analyzed and discussed. For closing all windows in the leakage process, the state of the partition door is a major factor. A closed kitchen layout with a closing partition door performs best in confining a hazardous gas cloud. At this point, it is difficult for a hazardous gas cloud to appear in the living area. With the partition door open, the gas cloud develops rapidly. For opening windows in the leakage process, a large scale hazardous gas cloud is not able to form under all layouts. For alarm-triggering behaviors, a closed kitchen layout when closing the partition door also performs best. When opening the partition door, alarms cannot trigger stably under all layouts. For the ventilation process, hazardous gas cloud dissipation under an open kitchen layout is slightly faster than the closed kitchen layout. Under a weak convection effect, there is a transition stage which delays the time point for exhausting leaked gas. Based on the analysis, some recommendations for accident prevention are proposed. Residents should try to use closed kitchens and close partition doors when not cooking. If open kitchens are used, multiple alarms with lower trigger values should be installed. It is better to choose a ceiling type for gas alarms. The windows of the house are recommended to select two layers type. Higher layers can open during the ventilation process to accelerate the exhaust of leaked gas. These recommendations provide a reference for preventing fires and explosions.

Numerical Simulation Analysis of the Hydrogen-Blended Natural Gas Leakage and Ventilation Processes in a Domestic House.

The blending of hydrogen in natural gas may have effects on the safety of its usage in a domestic house. In this work, the leakage accident of hydrogen-blended natural gas (HBNG) in the kitchen of a domestic house is analyzed by CFD with a hydrogen blending ratio (HBR) ≤ 30%. The whole process is divided into the gas accumulation process and the ventilation process. In the initial leakage stage, the influence of heights and the HBR on the gas distribution is analyzed. HBNG concentration increases with increasing height. Based on the exit Froude number, the formation of a gas cloud in the kitchen is significantly influenced by the initial momentum and buoyancy, while it is more driven by the concentration gradient beyond the kitchen. In contrast to height, the variation of HBR on the HBNG distribution is not significant. In the ventilation process, the evolution of the hazardous gas cloud volume is analyzed. With windows and doors closed, the hazardous gas cloud fills the house in approximately 3600 s after the leakage occurs. When windows and doors are open for ventilation, the volume of the hazardous gas cloud first declines rapidly and then slowly. The reasons for the variation rate of hazardous gas cloud volume are analyzed according to ventilation conditions. The difference during the decline stage for different HBRs is analyzed according to the gas layering properties. Under a lack of convection condition, the ventilation process finally reaches a stagnant stage. In addition, another ventilation process has been investigated after extending the gas accumulation time. After extending the gas accumulation time, the effect of different HBRs on the ventilation process remains the same as before. However, it postpones the time point to enter the stagnation stage. As gas accumulation time extends from 3600 to 5400 and 7200 s, the ventilation time into the stagnation stage increases from about 4800 to 5400 and 6000 s, respectively. This study has implications for the establishment of a risk assessment system based on hazardous gas cloud volume.

CAPNet: Context attention pyramid network for computer-aided detection of microcalcification clusters in digital breast tomosynthesis.

BACKGROUND AND OBJECTIVE: Computer-aided detection (CADe) of microcalcification clusters (MCs) in digital breast tomosynthesis (DBT) is crucial in the early diagnosis of breast cancer. Although convolutional neural network (CNN)-based detection models have achieved excellent performance in medical lesion detection, they are subject to some limitations in MC detection: 1) Most existing models employ the feature pyramid network (FPN) for multi-scale object detection; however, the rough feature sharing between adjacent layers in the FPN may limit the detection ability for small and low-contrast MCs; and 2) the MCs region only accounts for a small part of the annotation box, so the features extracted indiscriminately within the whole box may easily be affected by the background. In this paper, we develop a novel CNN-based CADe method to alleviate the impacts of the above limitations for the accurate and rapid detection of MCs in DBT. METHODS: The proposed method has two parts: a novel context attention pyramid network (CAPNet) for intra-layer MC detection in two-dimensional (2D) slices and a three-dimensional (3D) aggregation procedure for aggregating 2D intra-layer MCs into a 3D result according to their connectivity in 3D space. The proposed CAPNet is based on an anchor-free and one-stage detection architecture and contains a context feature selection fusion (CFSF) module and a microcalcification response (MCR) branch. The CFSF module can efficiently enrich shallow layers' features by the complementary selection of local context features, aiming to reduce the missed detection of small and low-contrast MCs. The MCR branch is a one-layer branch parallel to the classification branch, which can alleviate the influence of the background region within the annotation box on feature extraction and enhance the ability of the model to distinguish MCs from normal breast tissue. RESULTS: We performed a comparison experiment on an in-house clinical dataset with 648 DBT volumes, and the proposed method achieved impressive performance with a sensitivity of 91.56 % at 1 false positive per DBT volume (FPs/volume) and 93.51 % at 2 FPs/volume, outperforming other representative detection models. CONCLUSIONS: The experimental results indicate that the proposed method is effective in the detection of MCs in DBT. This method can provide objective, accurate, and quick diagnostic suggestions for radiologists, presenting potential clinical value for early breast cancer screening.

Right upper lobe segmentectomy and subsegmentectomy guided by classification pattern of peripheral segmental veins.

Background: Studies have analyzed the simplified branching pattern of peripheral segmental veins and developed a standardized approach for intersegmental vein identification in the right upper lobe (RUL). However, the identification approach of intersubsegmental veins has not been reported. This study aimed to supplement the identification approach of intersubsegmental veins and the classification pattern of peripheral segmental veins by using three-dimensional computed tomography bronchography and angiography (3D-CTBA). Materials and methods: A total of 600 patients with ground glass opacity (GGO) who had undergone 3D-CTBA preoperatively at Hebei General Hospital between September 2020 and September 2022 were used for the retrospective study. We reviewed the anatomical variations of RUL veins in these patients using 3D-CTBA images. Results: According to the anatomical position, the peripheral segmental veins structures of RUL were classified into five categories: "Iab type of anterior with central vein" (256/600, 42.7%), "Ib type of anterior with central vein" (166/600, 27.7%), "Central vein type" (38/600, 6.3%), "Anterior vein type" (81/600, 13.5%), "Right top pulmonary vein type" (57/600, 9.5%). The approach for intersegmental vein and intersubsegmental veins identification was divided into five types: anterior approach, posterobronchial approach, central vein approach, V2t approach, and intermediate bronchus posterior surface approach. Conclusions: The classification pattern of peripheral segmental veins should find wide application. Further, approaches identifying intersegmental veins and intersubsegmental veins may help thoracic surgeons perform safe and accurate RUL segmentectomy.

Predicting pleural invasion of invasive lung adenocarcinoma in the adjacent pleura by imaging histology.

The aim of the present study was to develop a non-invasive method based on histological imaging and clinical features for predicting the preoperative status of visceral pleural invasion (VPI) in patients with lung adenocarcinoma (LUAD) located near the pleura. VPI is associated with a worse prognosis of LUAD; therefore, early and accurate detection is critical for effective treatment planning. A total of 112 patients with preoperative computed tomography presentation of adjacent pleura and postoperative pathological findings confirmed as invasive LUAD were retrospectively enrolled. Clinical and histological imaging features were combined to develop a preoperative VPI prediction model and validate the model's efficacy. Finally, a nomogram for predicting LUAD was established and validated using a logistic regression algorithm. Both the clinical signature and radiomics signature (Rad signature) exhibited a perfect fit in the training cohort. The clinical signature was overfitted in the testing cohort, whereas the Rad signature showed a good fit. To combine clinical and radiomics signatures for optimal performance, a nomogram was created using the logistic regression algorithm. The results indicated that this approach had the highest predictive performance, with an area under the curve of 0.957 for the clinical signature and 0.900 for the Rad signature. In conclusion, histological imaging and clinical features can be combined in columnar maps to predict the preoperative VPI status of patients with adjacent pleural infiltrative lung carcinoma.

Deep Learning Radiomics Nomogram Based on Magnetic Resonance Imaging for Differentiating Type I/II Epithelial Ovarian Cancer.

RATIONALE AND OBJECTIVES: To develop and validate a T2-weighted magnetic resonance imaging (MRI)-based deep learning radiomics nomogram (DLRN) to differentiate between type I and type II epithelial ovarian cancer (EOC). MATERIALS AND METHODS: This multicenter study incorporated 437 patients from five centers, divided into training (n = 271), internal validation (n = 68), and external validation (n = 98) sets. The deep learning (DL) model was constructed using the largest orthogonal slices of the tumor area. The extracted radiomics features were employed in building the radiomics model. The clinical model was developed based on clinical characteristics. A DLRN was built by integrating the DL signature, radiomics signature, and independent clinical predictors. Model performances were evaluated through receiver operating characteristic (ROC) analysis, Brier score, calibration curve, and decision curve analysis (DCA). The areas under the ROC curve (AUCs) were compared using the DeLong test. A two-tailed P < 0.05 was considered significantly different. RESULTS: The DLRN exhibited satisfactory discrimination between type I and type II EOC with the AUC of 0.888 (95% confidence interval [CI] 0.810, 0.966) and 0.866 (95% CI 0.786, 0.946) in the internal and external validation sets, respectively. These AUCs significantly exceeded those of the clinical model (P = 0.013 and 0.043, in the internal and external validation sets, respectively). The DLRN demonstrated optimal classification accuracy and clinical application value, according to Brier scores, calibration curves, and DCA. CONCLUSION: A T2-weighted MRI-based DLRN showed promising potential in differentiating between type I and type II EOC, which could offer assistance in clinical decision-making.

Prognostic significance of Memorial Sloan-Kettering prognostic score in small-cell lung cancer patients.

Aim: To assess the potential factors with predictive value for survival in small-cell lung cancer (SCLC) patients and to develop a nomogram prediction model. Patients & methods: We retrospectively screened and analyzed patients with pathologically confirmed SCLC from April 2015 to December 2021. Results: A total of 167 patients with SCLC were included. According to the Memorial Sloan-Kettering prognostic score (MPS), patients were divided into three groups: group 0 (n = 65), group 1 (n = 69) and group 2 (n = 33). The multivariate analysis showed that MPS was an independent prognostic factor for progression-free and overall survival in SCLC patients (p < 0.05). The nomogram showed that MPS was the most influential factor for overall survival. Conclusion: MPS is an independent prognostic factor for overall and progression-free survival in SCLC patients and performed better than other indicators used in this study.

Analysis of variations in the bronchovascular pattern of the lingular segment to explore the correlations between the lingular segment artery and left superior division veins.

Background: With the development of anatomical segmentectomy, the thoracic surgeons must master the anatomical variations of the pulmonary bronchi and vessels. However, there are only a few reports showing anatomic variations of the lingular segment (LS) using three-dimensional computed tomography bronchography and angiography (3D-CTBA). Thus, the present study aimed to analyze the bronchovascular patterns of the LS and explore the correlation between the lingular segment artery (LSA) and left superior division veins (LSDV). Materials and methods: The 3D-CTBA data of the left upper lobe (LUL) were collected from patients who had undergone lobectomy or segmentectomy at Hebei General Hospital between October 2020 and October 2022. We reviewed the clinical characteristics and variations in bronchi and pulmonary vessels and grouped them according to different classifications. Results: Among all 540 cases of 3D-CTBA, the branching patterns of LSA included 369 (68.3%) cases with the interlobar origin, 126 (23.3%) cases with the interlobar and mediastinal origin, and 45 (8.3%) cases with the mediastinal origin. The branching pattern of LSDV could be classified into three forms: Semi-central vein type (345/540, 63.9%), Non-central vein type (76/540, 14.1%), and Central vein type (119/540, 22.0%). There were 51 cases (9.4%) with Non-central vein type, 50 cases (9.3%) with Central vein type, 268 cases (49.6%) with Semi-central vein type in the interlobar type, and 7 cases (1.3%) with Non-central vein type, 9 cases (1.7%) with Central vein type, 29 cases (5.4%) with Semi-central vein type in the mediastinal type. Moreover, the Non-central vein type, the Central vein type, and the Semi-central vein type accounted for 18 (3.3%), 60 (11.1%), and 48 (8.9%) in the interlobar and mediastinal type. Combinations of the branching patterns of the LSA and LSDV were significantly dependent (p < 0.005). The combinations of the interlobar and mediastinal type with the Central vein type, and the interlobar type and the mediastinal type with the Semi-central vein type were frequently observed. Conclusions: This study found the relationship between the LSA and LSDV and clarified the bifurcation patterns of the bronchovascular in the LS. Our data can be used by thoracic surgeons to perform safe and precise LS segmentectomy.

Automatic Classification of Mass Shape and Margin on Mammography with Artificial Intelligence: Deep CNN Versus Radiomics.

The purpose of this study is to test the feasibility for deep CNN-based artificial intelligence methods for automatic classification of the mass margin and shape, while radiomic feature-based machine learning methods were also implemented in this study as baseline and for comparison study. In this retrospective study, 596 patients with breast mass that underwent mammography from 4 hospitals were enrolled from January 2012 to October 2019. Margin and shape of each mass were annotated according to BI-RADS by 2 experienced radiologists. Deep CNN-based AI was implemented for the classification task based on Resnet50. Balanced sampler and CBAM were also used to improve the performance of the Deep CNNs. As comparison, image texture features were extracted and then dimensionality reduction methods (such as PCA, ICA) and classical classifiers (such as SVM, DT, KNN) were used for classification task. Based on Python programming software, accuracy (ACC) was used to evaluate the performance of the model, and the model with the highest ACC value was selected. Deep CNN based on Resnet50 with balanced sampler and CBAM achieved the best performance for both margin and shape classification, with ACC of 0.838 and 0.874, respectively. For the radiomics-based machine learning, the best performance for margin was achieved as 0.676 by the combination of FA + RF, while the best performance for shape was 0.802 by the combination of PCA + MLP. The feasibility for automatic classification with coarse labeling of the mass shape and margin was testified with the deep CNN-based AI methods, while radiomic feature-based machine learning methods achieved inferior classification results.

Multimodal cross enhanced fusion network for diagnosis of Alzheimer's disease and subjective memory complaints.

Deep learning methods using multimodal imagings have been proposed for the diagnosis of Alzheimer's disease (AD) and its early stages (SMC, subjective memory complaints), which may help to slow the progression of the disease through early intervention. However, current fusion methods for multimodal imagings are generally coarse and may lead to suboptimal results through the use of shared extractors or simple downscaling stitching. Another issue with diagnosing brain diseases is that they often affect multiple areas of the brain, making it important to consider potential connections throughout the brain. However, traditional convolutional neural networks (CNNs) may struggle with this issue due to their limited local receptive fields. To address this, many researchers have turned to transformer networks, which can provide global information about the brain but can be computationally intensive and perform poorly on small datasets. In this work, we propose a novel lightweight network called MENet that adaptively recalibrates the multiscale long-range receptive field to localize discriminative brain regions in a computationally efficient manner. Based on this, the network extracts the intensity and location responses between structural magnetic resonance imagings (sMRI) and 18-Fluoro-Deoxy-Glucose Positron Emission computed Tomography (FDG-PET) as an enhancement fusion for AD and SMC diagnosis. Our method is evaluated on the publicly available ADNI datasets and achieves 97.67% accuracy in AD diagnosis tasks and 81.63% accuracy in SMC diagnosis tasks using sMRI and FDG-PET. These results achieve state-of-the-art (SOTA) performance in both tasks. To the best of our knowledge, this is one of the first deep learning research methods for SMC diagnosis with FDG-PET.

What is the correlation between the defective and splitting posterior segmental bronchus and recurrent artery crossing intersegmental planes in the right upper lobe?

Background: With the prevalence of three-dimensional computed tomography bronchography and angiography (3D-CTBA) and the development of anatomical segmentectomy, studies have confirmed the increased incidence of anomalous veins in patients with tracheobronchial abnormalities. Nevertheless, the characteristic anatomical correlation between bronchus and artery variation remains unknown. Thus, we conducted a retrospective study to investigate recurrent artery crossing intersegmental planes and their associated pulmonary anatomical features by analyzing the incidence and types of the right upper lobe (RUL) bronchus and the artery composition of the posterior segment. Materials and Methods: A total of 600 patients with ground-glass opacity who had undergone 3D-CTBA preoperatively at Hebei General Hospital between September 2020 and September 2022 were included. We reviewed the anatomical variations of the RUL bronchus and artery in these patients using 3D-CTBA images. Results: Among all 600 cases, the defective and splitting B2 contained four types of the RUL bronchial structure: B1 + BX2a, B2b, B3 (11/600, 1.8%); B1, B2a, BX2b + B3 (3/600, 0.5%); B1 + BX2a, B3 + BX2b (18/600, 3%); B1, B2a, B2b, B3 type (29/600, 4.8%). The incidence of recurrent artery crossing intersegmental planes was 12.7% (70/600). The incidence of recurrent artery crossing intersegmental planes with and without the defective and splitting B2 was 26.2% (16/61) and 10.0% (54/539), respectively (p < 0.005). Conclusions: In patients with defective and splitting B2, the incidence of recurrent artery crossing intersegmental planes was increased. Our study provides certain references that surgeons can use to plan and perform RUL segmentectomy.

SAH-NET: Structure-Aware Hierarchical Network for Clustered Microcalcification Classification in Digital Breast Tomosynthesis.

Benign and malignant classification of clustered microcalcifications (MCs) in digital breast tomosynthesis (DBT) is an essential task in computer-aided diagnosis. However, due to the anisotropic resolution of DBT, three-dimensional (3-D) convolutional neural network (CNN)-based methods cannot extract hierarchical features efficiently. Moreover, the sparse distribution of MC points in the cluster makes it difficult for the CNN to extract discriminative structural information for classification. To comprehensively address these challenges, we propose a novel structure-aware hierarchical network (SAH-Net) for benign and malignant classification of clustered MC in a DBT volume. Specifically, the two-dimensional (2-D) group convolution is used to extract intraslice features. The one-to-one correspondence between group convolutions and slices ensures the independence of hierarchical feature extraction. Then, a partial deformable Transformer-based 3-D structural feature learning module is proposed to capture the long-range dependency between MC points in the cluster. We evaluate the proposed method on an in-house dataset with 495 clustered MCs collected from 462 DBT images. Experimental results confirm the validity of our proposed modules. The results also show that the proposed SAH-Net outperforms several other representative methods on this topic, and achieves the best classification result, with an area under the receiver operation curve (AUC) of 86.87%. The implementation of the proposed model is available at https://github.com/sunhaotian130911/SAHNet.

T2-weighted MRI-based radiomics for discriminating between benign and borderline epithelial ovarian tumors: a multicenter study.

BACKGROUND: Preoperative differentiation between benign and borderline epithelial ovarian tumors (EOTs) is challenging and can significantly impact clinical decision making. The purpose was to investigate whether radiomics based on T2-weighted MRI can discriminate between benign and borderline EOTs preoperatively. METHODS: A total of 417 patients (309, 78, and 30 samples in the training and internal and external validation sets) with pathologically proven benign and borderline EOTs were included in this multicenter study. In total, 1130 radiomics features were extracted from manually delineated tumor volumes of interest on images. The following three different models were constructed and evaluated: radiomics features only (radiomics model); clinical and radiological characteristics only (clinic-radiological model); and a combination of them all (combined model). The diagnostic performances of models were assessed using receiver operating characteristic (ROC) analysis, and area under the ROC curves (AUCs) were compared using the DeLong test. RESULTS: The best machine learning algorithm to distinguish borderline from benign EOTs was the logistic regression. The combined model achieved the best performance in discriminating between benign and borderline EOTs, with an AUC of 0.86 ± 0.07. The radiomics model showed a moderate AUC of 0.82 ± 0.07, outperforming the clinic-radiological model (AUC of 0.79 ± 0.06). In the external validation set, the combined model performed significantly better than the clinic-radiological model (AUCs of 0.86 vs. 0.63, p = 0.021 [DeLong test]). CONCLUSIONS: Radiomics, based on T2-weighted MRI, can provide critical diagnostic information for discriminating between benign and borderline EOTs, thus having the potential to aid personalized treatment options.

Naples Prognostic Score is an Independent Prognostic Factor in Patients with Small Cell Lung Cancer and Nomogram Predictive Model Established.

Background: The routine clinical nutritional and inflammatory indicators such as serum albumin, total cholesterol and lymphocytes have been widely investigated in the prognosis of small cell lung cancer (SCLC). The Naples prognostic score (NPS), based on nutritional and inflammatory status, has been identified as a prognostic impactor in several malignancies. However, the prognostic role of NPS in SCLC has not been elucidated. This study aims to evaluate the prognostic effect of NPS in SCLC patients. Patients and Methods: Patients with SCLC were recruited at Hebei General Hospital between April 2015 and August 2021. Pretreatment clinical and laboratory data were obtained. Participants were assigned into three groups according to NPS (group 0: NPS=0, group 1: NPS=1 or 2, group 2: NPS=3 or 4). Kaplan-Meier and Cox regression analysis were performed to assess the prognostic significance of NPS. The RMS package in R software was used to draw the nomogram predictive model. Results: A total of 128 patients were enrolled. The median progression-free survival (PFS) and overall survival (OS) was 7.2 and 12.3 months, respectively. The median PFS and OS was 12.3 vs 19.8 months, 7.6 vs 14.1 months and 6.0 vs 8.45 months for the three groups respectively. There were significant differences in both OS and FPS among the three groups. Survival analysis showed that NPS was significantly correlated with both OS and PFS (P<0.05). Lower NPS is associated with longer OS and PFS. Multivariate analysis showed that NPS has an independent prognostic impact on OS (P<0.05). The nomogram predictive model showed that NPS has good predictive power for survival rates. Conclusion: NPS is an independent prognostic factor for OS in SCLC patients. Low NPS may predict longer OS. Therefore NPS plays a vital role in the nomogram predictive model of survival rates in SCLC patients.

BMNet: A New Region-Based Metric Learning Method for Early Alzheimer's Disease Identification With FDG-PET Images.

18F-fluorodeoxyglucose (FDG)-positron emission tomography (PET) reveals altered brain metabolism in individuals with mild cognitive impairment (MCI) and Alzheimer's disease (AD). Some biomarkers derived from FDG-PET by computer-aided-diagnosis (CAD) technologies have been proved that they can accurately diagnosis normal control (NC), MCI, and AD. However, existing FDG-PET-based researches are still insufficient for the identification of early MCI (EMCI) and late MCI (LMCI). Compared with methods based other modalities, current methods with FDG-PET are also inadequate in using the inter-region-based features for the diagnosis of early AD. Moreover, considering the variability in different individuals, some hard samples which are very similar with both two classes limit the classification performance. To tackle these problems, in this paper, we propose a novel bilinear pooling and metric learning network (BMNet), which can extract the inter-region representation features and distinguish hard samples by constructing the embedding space. To validate the proposed method, we collect 898 FDG-PET images from Alzheimer's disease neuroimaging initiative (ADNI) including 263 normal control (NC) patients, 290 EMCI patients, 147 LMCI patients, and 198 AD patients. Following the common preprocessing steps, 90 features are extracted from each FDG-PET image according to the automatic anatomical landmark (AAL) template and then sent into the proposed network. Extensive fivefold cross-validation experiments are performed for multiple two-class classifications. Experiments show that most metrics are improved after adding the bilinear pooling module and metric losses to the Baseline model respectively. Specifically, in the classification task between EMCI and LMCI, the specificity improves 6.38% after adding the triple metric loss, and the negative predictive value (NPV) improves 3.45% after using the bilinear pooling module. In addition, the accuracy of classification between EMCI and LMCI achieves 79.64% using imbalanced FDG-PET images, which illustrates that the proposed method yields a state-of-the-art result of the classification accuracy between EMCI and LMCI based on PET images.

Analysis and Prediction of Methane Invasion Distance Considering Real Ground Boundary.

Natural gas has become a global energy consumption hotspot because of its large reserves and clean combustion. Due to soil corrosion, construction damage, and natural disasters, leakage accidents of buried natural gas pipelines often occur. In this paper, the steady simulation method was used to study the methane invasion limit state (MILS) and the methane invasion limit distance (MILD) under the conditions of hardened surface ground (HSG), unhardened surface ground (UHSG), and semihardened surface ground (SHSG), and the transient simulation of methane invasion distance (MID) under the condition of HSG with the largest MILD was carried out. The results showed that regardless of ground conditions, with the increase of leakage time, the diffusion range of methane in soil will not increase all the time, and there was a limit state (MILS). The distribution range and concentration of methane in the soil under HSG condition were the largest, followed by the SHSG condition, and the UHSG condition was the smallest. When the ground condition changed from UHSG to HSG, the MILD increased from 3.41 to 9.32 m. The HSG condition will increase the MILD and the range of dangerous areas. The buried depth of the pipeline had a serious impact on the MILD. When the buried depth of the pipeline increased from 0.3 to 1.5 m, the MILD increased from 1.75 to 3.49 m under the condition of UHSG and exceeded 10 m under the condition of HSG. The average error of the MID prediction model was 2.37% under the condition of HSG, which can accurately predict the leakage of buried pipeline. The MID provides a reference for the layout of urban underground gas leakage monitoring points. The MILD can provide guidance for the safe distance between natural gas pipeline and structures in the design code of natural gas pipeline.