RESUMO
Traffic Prediction based on graph structures is a challenging task given that road networks are typically complex structures and the data to be analyzed contains variable temporal features. Further, the quality of the spatial feature extraction is highly dependent on the weight settings of the graph structures. In the transportation field, the weights of these graph structures are currently calculated based on factors like the distance between roads. However, these methods do not take into account the characteristics of the road itself or the correlations between different traffic flows. Existing approaches usually pay more attention to local spatial dependencies extraction while global spatial dependencies are ignored. Another major problem is how to extract sufficient information at limited depth of graph structures. To address these challenges, we propose a Random Graph Diffusion Attention Network (RGDAN) for traffic prediction. RGDAN comprises a graph diffusion attention module and a temporal attention module. The graph diffusion attention module can adjust its weights by learning from data like a CNN to capture more realistic spatial dependencies. The temporal attention module captures the temporal correlations. Experiments on three large-scale public datasets demonstrate that RGDAN produces predictions with 2%-5% more precision than state-of-the-art methods.
Assuntos
DifusãoRESUMO
Introduction: Papillary thyroid cancer (PTC) is one of the most common malignancies involving the endocrine system. Aim: To explore the clinical value of ultrasound-based radiomics for predicting the recurrence of PTC after complete endoscopic resection. Material and methods: The general data of 361 PTC patients were collected. They were randomly assigned to the modeling group (n = 253) and the validation group (n = 108) according to the ratio of 7 : 3. In the modeling group, the PyRadiomics package was applied to extract radiomic features from preoperative ultrasound images, and least absolute shrinkage and selection operator (LASSO) was used to screen and to construct a radiomics score (Rad-score). Independent prognostic predictors were identified using the Cox proportional hazards model, and a nomogram prediction model was constructed by R software. Results: Using the LASSO regression model, 7 radiomic features were screened and then the Rad-score was calculated. Based on the Rad-score, modeling and validation groups were divided into low-, medium- and high-risk groups, and the 10-year recurrence-free survival rates were 94.7% vs. 95.9%, 83.6% vs. 80.0%, and 50.0% vs. 66.6%, respectively (p < 0.001). Multivariate analysis revealed that age, lymph node metastasis and Rad-score were independent predictors for recurrence-free survival (p < 0.05). Conclusions: The ultrasound-based radiomics score can effectively predict the postoperative recurrence-free survival in patients with PTC. The nomogram prediction model is superior to the AJCC staging system in terms of predictive accuracy and consistency.
RESUMO
OBJECTIVE: To investigate the aortic elastic properties and its clinical significance in intracranial aneurysms (IAs). METHODS: One hundred and seven IAs patients (57 with hypertension) and 108 healthy subjects were recruited. The internal aortic diameters in systole and diastole were measured by the M-mode echocardiography, the aortic elasticity indexes were calculated and compared. RESULTS: The aortic distensibility (DIS) was lower and the aortic stiffness index (SI) was higher in IAs patients than those in controls (both P <0.001). DIS was lower and SI was higher in IAs patients with hypertension (IAs-HP) than those in IAs with no hypertension (P <0.001). Similar results were obtained when the aortic elasticity index were adjusted for body surface area and body mass index. CONCLUSION: Abnormal aortic elasticity is a common finding in IAs patients and hypertension is closely related to the severity of aortic elasticity.
