A Dual Robust Graph Neural Network Against Graph Adversarial Attacks.

Graph Neural Networks (GNNs) have gained widespread usage and achieved remarkable success in various real-world applications. Nevertheless, recent studies reveal the vulnerability of GNNs to graph adversarial attacks that fool them by modifying graph structure. This vulnerability undermines the robustness of GNNs and poses significant security and privacy risks across various applications. Hence, it is crucial to develop robust GNN models that can effectively defend against such attacks. One simple approach is to remodel the graph. However, most existing methods cannot fully preserve the similarity relationship among the original nodes while learning the node representation required for reweighting the edges. Furthermore, they lack supervision information regarding adversarial perturbations, hampering their ability to recognize adversarial edges. To address these limitations, we propose a novel Dual Robust Graph Neural Network (DualRGNN) against graph adversarial attacks. DualRGNN first incorporates a node-similarity-preserving graph refining (SPGR) module to prune and refine the graph based on the learned node representations, which contain the original nodes' similarity relationships, weakening the poisoning of graph adversarial attacks on graph data. DualRGNN then employs an adversarial-supervised graph attention (ASGAT) network to enhance the model's capability in identifying adversarial edges by treating these edges as supervised signals. Through extensive experiments conducted on four benchmark datasets, DualRGNN has demonstrated remarkable robustness against various graph adversarial attacks.

The Application of Computer Technology to Clinical Practice Guideline Implementation: A Scoping Review.

Implementation of clinical practice guidelines (CPG) is a complex and challenging task. Computer technology, including artificial intelligence (AI), has been explored to promote the CPG implementation. This study has reviewed the main domains where computer technology and AI has been applied to CPG implementation. PubMed, Embase, Web of science, the Cochrane Library, China National Knowledge Infrastructure database, WanFang DATA, VIP database, and China Biology Medicine disc database were searched from inception to December 2021. Studies involving the utilization of computer technology and AI to promote the implementation of CPGs were eligible for review. A total of 10429 published articles were identified, 117 met the inclusion criteria. 21 (17.9%) focused on the utilization of AI techniques to classify or extract the relative content of CPGs, such as recommendation sentence, condition-action sentences. 47 (40.2%) focused on the utilization of computer technology to represent guideline knowledge to make it understandable by computer. 15 (12.8%) focused on the utilization of AI techniques to verify the relative content of CPGs, such as conciliation of multiple single-disease guidelines for comorbid patients. 34 (29.1%) focused on the utilization of AI techniques to integrate guideline knowledge into different resources, such as clinical decision support systems. We conclude that the application of computer technology and AI to CPG implementation mainly concentrated on the guideline content classification and extraction, guideline knowledge representation, guideline knowledge verification, and guideline knowledge integration. The AI methods used for guideline content classification and extraction were pattern-based algorithm and machine learning. In guideline knowledge representation, guideline knowledge verification, and guideline knowledge integration, computer techniques of knowledge representation were the most used.

Effects of ambient air pollution on cause-specific hospitalizations in Wuhan during 2017-2019.

BACKGROUND: Few studies have focused on the associations between air pollutants and multiple organ system diseases in the entire hospitalized population. The present study aims to explore the short-term effects of six routinely monitored air pollutants on the broad causes of hospital admissions and estimate the resulting hospital admission burdens. METHODS: Daily hospital admission records from 2017 to 2019 were obtained from the Wuhan Information center of Health and Family Planning. Generalized additive models (GAMs) were employed to evaluate the effects of air pollutants on the percent increase in the cause-specific daily number of hospital admissions. Increased hospital admission numbers, days, and expenses were also estimated. RESULTS: A total of 2636,026 hospital admissions were identified. We found that both PM2.5 and PM10 increased the risk of hospital admissions for most disease categories. Short-term exposure to PM2.5 was positively associated with hospitalizations of several rarely studied disease categories, such as diseases of the eye and adnexa (2.83%, 95%CI: 0.96-4.73%, P < 0.01) and diseases of the musculoskeletal system and connective tissue (2.17%, 95% CI: 0.88-3.47%, P < 0.001). NO2 was observed to have a robust effect on diseases of the respiratory system (1.36%, 95%CI: 0.74-1.98%, P < 0.001). CO was significantly associated with hospital admissions for six disease categories. Furthermore, each 10-µg/m3 increase in PM2.5 was associated with an annual increase of 13,444 hospital admissions (95% CI: 6239-20,649), 124,344 admission days (95% CI: 57,705-190,983), and 166-million-yuan admission expenses (95% CI: 77-255). CONCLUSION: Our study suggested that particulate matter (PM) had a short-term effect on hospital admissions of most major disease categories and resulted in a considerable hospital admission burden. In addition, the health effects of NO2 and CO emissions require more attention in megacities.

Association of NO2 with daily hospital admissions for mental disorders: Investigation of the modification effects of green spaces and long-term NO2 exposure.

Air pollution is a risk factor for increased hospital admissions due to mental disorders, while green spaces have been linked with better mental health. We linked daily hospital admission records from Wuhan's 74 municipal hospitals from 2017 to 2019 with modeled annual average NO2 concentrations and added data on the residential surrounding green spaces with 250 m and 500 m buffers based on the normalized difference vegetation index (NDVI) using a land use regression model (LUR). The conditional logistic regression model was used to estimate the acute effect of short-term NO2 exposure, and stratification analyses were applied to explore the modification effect of long-term NO2 exposure and green spaces by estimating the odds ratios in the single- and dual-environmental factor groups. A total of 42,705 hospital admissions for mental disorders were identified. Short-term exposure to NO2 was associated with an increased risk of hospital admission for mental disorders. A 10 µg/m3 increase in NO2 (lag01 day) was associated with an increase in hospital admissions of 2.86% (95% CI, 2.05-3.68) for the total mental disorders. Compared with patients in the "low-NDVI/low-NO2" group (ER = 2.27%, 95% CI, 0.27-4.31), patients in the "high-NDVI/low-NO2" group (ER = 1.93%, -0.10-3.99) showed a lower and insignificant increase in hospitalizations for the total mental disorders, while greenness had a slight moderating effect in the high-level long-term NO2 exposure areas. This study suggested that green spaces may moderate the acute effect of NO2 exposure for mental disorder hospitalizations, especially in low-level long-term NO2 exposure areas.

PM2.5-Associated Hospitalization Risk of Cardiovascular Diseases in Wuhan: Cases Alleviated by Residential Greenness.

PM2.5, a type of particulate matter with an aerodynamic diameter of less than 2.5 µm, is associated with the occurrence of cardiovascular diseases (CVDs), while greenness seems to be associated with better cardiovascular health. We identified 499,336 CVD cases in Wuhan's 74 municipal hospitals between 2017 and 2019. A high-resolution PM2.5 model and a normalized difference vegetation index (NDVI) map were established to estimate individual exposures. The time-stratified case-crossover design and conditional logistic regression models were applied to explore the associations between PM2.5 and CVDs under different levels of environmental factors. Greenness could alleviate PM2.5-induced hospitalization risks of cardiovascular diseases. Compared with patients in the low-greenness group (ER = 0.99%; 95% CI: 0.71%, 1.28%), patients in the high-greenness group (ER = 0.45%; 95% CI: 0.13%, 0.77%) showed a lower increase in total CVD hospitalizations. After dividing the greenness into quartiles and adding long-term PM2.5 exposure as a control factor, no significant PM2.5-associated hospitalization risks of CVD were identified in the greenest areas (quartile 4), whether the long-term PM2.5 exposure level was high or low. Intriguingly, in the least green areas (quartile 1), the PM2.5-induced excess risk of CVD hospitalization was 0.58% (95% CI: 0.04%, 1.11%) in the long-term high-level PM2.5 exposure group, and increased to 1.61% (95% CI: 0.95%, 2.27%) in the long-term low-level PM2.5 exposure group. In the subgroup analysis, males and participants aged 55-64 years showed more significant increases in the PM2.5-induced risk of contracting CVDs with a reduction in greenness and fine particle exposure conditions. High residential greenness can greatly alleviate the PM2.5-induced risk of cardiovascular admission. Living in the areas with long-term low-level PM2.5 may make people more sensitive to short-term increases in PM2.5, leading to CVD hospitalization.

Rhesus monkey model of liver disease reflecting clinical disease progression and hepatic gene expression analysis.

Alcoholic liver disease (ALD) is a significant public health issue with heavy medical and economic burdens. The aetiology of ALD is not yet completely understood. The development of drugs and therapies for ALD is hampered by a lack of suitable animal models that replicate both the histological and metabolic features of human ALD. Here, we characterize a rhesus monkey model of alcohol-induced liver steatosis and hepatic fibrosis that is compatible with the clinical progression of the biochemistry and pathology in humans with ALD. Microarray analysis of hepatic gene expression was conducted to identify potential molecular signatures of ALD progression. The up-regulation of expression of hepatic genes related to liver steatosis (CPT1A, FASN, LEPR, RXRA, IGFBP1, PPARGC1A and SLC2A4) was detected in our rhesus model, as was the down-regulation of such genes (CYP7A1, HMGCR, GCK and PNPLA3) and the up-regulation of expression of hepatic genes related to liver cancer (E2F1, OPCML, FZD7, IGFBP1 and LEF1). Our results demonstrate that this ALD model reflects the clinical disease progression and hepatic gene expression observed in humans. These findings will be useful for increasing the understanding of ALD pathogenesis and will benefit the development of new therapeutic procedures and pharmacological reagents for treating ALD.