The reliable and automatic segmentation of pulmonary lobes in computed tomography scans is an important pre-condition for the diagnosis, assessment, and treatment of lung diseases. However, due to the incomplete lobar structures and morphological changes caused by diseases, the lobe segmentation still encounters great challenges. Recently, convolution neural network has exerted a tremendous impact on medical image analysis. Nevertheless, the basic convolution operations mainly obtain local features that are insufficient for accurate lobe segmentation. The idea that the global features are equally crucial especially when lesions appear is considered. Here, a dual-attention V-network named DAV-Net for pulmonary lobe segmentation is proposed. First, a novel dual-attention module to capture global contextual information and model the semantic dependencies in spatial and channel dimensions is introduced. Second, a progressive output scheme is used to avoid the vanishing gradient phenomenon and obtain relatively effective features in hidden layers. Finally, an improved combo loss is devised to address input and output lobe imbalance problem during training and inference. In the evaluation using the LUNA16 dataset and our in-house dataset, the proposed DAV-Net obtains Dice similarity coefficients of 0.947 and 0.934, respectively; these values are superior to those obtained by existing methods.This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
BackgroundSchool bullying among adolescents has been a worldwide public health issue. It has been observed that adolescents who are exposed to violent video games (VVGs) are often more aggressive. However, research on the association between violent video game exposure (VVGE) and different types of school bullying is limited in the Chinese context.ObjectiveThe purpose of this study was to explore whether VVGE is linked to school bullying behaviors among Chinese adolescents and to examine the relationship between different levels of violent game exposure and four sub-types (physical, verbal, relational, and cyber) of school bullying involvement.MethodsThis was a cross-sectional study of 1,992 Chinese students (55.02% boys and 44.98% girls) with the average age of 15.84 ± 1.62 years. Sub-types of school bullying victimization and perpetration, Internet addiction, and VVGE were measured by using a self-administrated questionnaire. The association was examined by multiple logistic regression analysis, adjusting for covariates.ResultsPhysical, verbal, relational, and cyber school bullying victimization were reported by 18.12, 60.34, 11.75, and 12.05% of the adolescents, and physical, verbal, relational, and cyber school bullying perpetration were reported by 16.62, 54.62, 21.49, and 8.23% of them. Of the students, 1,398 (70.18%) were normal Internet users, 514 (25.80%) showed moderate Internet addictive behaviors, and 31 (1.56%) of the students showed severe Internet addictive behaviors. The prevalence of no VVGE, low-level VVGE, medium-level VVGE, and high-level of VVGE were 27.70, 24.10, 24.20, and 24.00%, respectively. The risk of physical victimization and physical perpetration significantly increased with the increasing degree of violent video game exposure (P for trend < 0.001), with the highest adjusted odds ratios (ORs) of 2.251 (95% CI 1.501–3.375) and 2.554 (95% CI 1.685–3.870), when comparing high-level VVGE with no VVGE.ConclusionThese findings highlight the specific association between different sub-types of school bullying involvement and violent video game exposure. Physical school bullying prevention and intervention programs should be conducted after adolescents are exposed to violent video games.
Objectives: The New Chinese Diabetes Risk Score (NCDRS) is a noninvasive tool to assess the risk of type 2 diabetes mellitus (T2DM) in the Chinese population. Our study aimed to evaluate the performance of the NCDRS in predicting T2DM risk with a large cohort.Methods: The NCDRS was calculated, and participants were categorized into groups by optimal cutoff or quartiles. Hazard ratios (HRs) and 95% confidential intervals (CIs) in Cox proportional hazards models were used to estimate the association between the baseline NCDRS and the risk of T2DM. The performance of the NCDRS was assessed by the area under the curve (AUC).Results: The T2DM risk was significantly increased in participants with NCDRS ≥25 (HR = 2.12, 95% CI 1.88–2.39) compared with NCDRS <25 after adjusting for potential confounders. T2DM risk also showed a significant increasing trend from the lowest to the highest quartile of NCDRS. The AUC was 0.777 (95% CI 0.640–0.786) with a cutoff of 25.50.Conclusion: The NCDRS had a significant positive association with T2DM risk, and the NCDRS is valid for T2DM screening in China.
Smoking is prohibited in many places. In order to detect smoking behaviour and make accurate judgments in a timely manner, machine vision technology is used to carry out research on the detection of small targets of cigarette sticks to improve the problem of difficult and inaccurate recognition of small targets in current detection algorithms and to improve the accuracy of cigarette sticks being recognized. In this study, we propose an improved method to detect smoking behaviour of Yolox based on the attention mechanism, in view of the fact that cigarette sticks occupy a very small part of the image and the presence of hand occlusion. The Attention mechanism module is added to the feature extraction network to focus on local information. Meanwhile, the use of the deep network is increased by adding a scale to focus attention within the target region; and the loss function is optimally replaced and the GIoU loss function is chosen to solve the drawback problem of IoU; experimental results show that the improved Yolox-c algorithm mAP reaches 95.96%, compared with Yolox and Yolov5 mAP, which are improved by 5.87% and 8.85%, respectively.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.