Search

Accurate prediction of PM2.5 concentration is important for pollution control, public health, and ecological protection. However, due to the nonlinear nature of PM2.5 data, the accuracy of existing methods suffers and performs poorly in both short-term and long-term predictions. In this study, a deep learning hybrid prediction model based on clustering and quadratic decomposition is proposed. The model utilizes the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) to decompose the PM2.5 sequences into multiple intrinsic modal function components (IMFs), and clusters and re-fuses the subsequences with similar complexity by permutation entropy (PE) and K-means clustering. For the fused high-frequency sequences, a secondary decomposition is performed using the whale optimization algorithm (WOA) optimized variational modal decomposition (VMD). Finally, the nonlinear and temporal features are captured for prediction using the long- and short-term memory neural network (LSTM). Experiments show that this proposed model exhibits good stability and generalization ability. It does not only make accurate predictions in the short term, but also captures the trends in the long-term prediction. There is a significant performance improvement over the baseline models. Further comparisons with existing models outperform the current state-of-the-art models. [ABSTRACT FROM AUTHOR]

PM2.5 pollution poses an important threat to the atmospheric environment and human health. To precisely forecast PM2.5 concentration, this study presents an innovative combined model: EMD-SE-GWO-VMD-ZCR-CNN-LSTM. First, empirical mode decomposition (EMD) is used to decompose PM2.5, and sample entropy (SE) is used to assess the subsequence complexity. Secondly, the hyperparameters of variational mode decomposition (VMD) are optimized by Gray Wolf Optimization (GWO) algorithm, and the complex subsequences are decomposed twice. Next, the sequences are divided into high-frequency and low-frequency parts by using the zero crossing rate (ZCR); the high-frequency sequences are predicted by a convolutional neural network (CNN), and the low-frequency sequences are predicted by a long short-term memory network (LSTM). Finally, the predicted values of the high-frequency and low-frequency sequences are reconstructed to obtain the final results. The experiment was conducted based on the data of 1009A, 1010A, and 1011A from three air quality monitoring stations in the Beijing area. The results indicate that the R2 value of the designed model increased by 2.63%, 0.59%, and 1.88% on average in the three air quality monitoring stations, respectively, compared with the other single model and the mixed model, which verified the significant advantages of the proposed model. [ABSTRACT FROM AUTHOR]

The aim was to estimate the association of egg consumption with dyslipidemia and its subtypes in Chinese adults. The cross-sectional health survey was conducted in Hanzhong city of Shaanxi province in 2010. Consumption of eggs (alone and mixed in foods) and nutrients were assessed using a semi-quantitative 81-item food frequency questionnaire (FFQ). Serum lipids and other physical measurements were measured by trained medical staff. Odds ratios (ORs) and 95% confidence intervals (CIs) were estimated from multivariable logistic regressions. Restricted cubic spline was used to visualize the possible non-linear association. A total of 2241 participants (mean age, 48.0 years; 34.5% men) were included. The age- and gender- standardization dyslipidemia prevalence was 21.1% in men and 18.4% in women, respectively. Approximately 42.1% of individuals never or rarely consumed egg, while 9.4% consumed ≥2 times/week. The mean egg intake was 21.5 g/d in men and 17.4 in women. In women, after adjustment for demographics and lifestyle factors, the ORs (95%CI) for dyslipidemia and hypertriglyceridemia were 0.57(0.32–1.05) and 0.38 (0.17–0.85) for individuals consuming egg ≥2 times/week, compared with never or rarely consumers. When further adjustment of BMI and dietary factors, the inverse relation persisted for hypertriglyceridemia but attenuated to margin of significance for dyslipidemia (P-trend = 0.050). Increment of 25 g/d was associated with 15% (OR: 0.85, (0.71–1.00)) lower risk of dyslipidemia. Results from restricted cubic splines showed a substantial reduction of the risk within lower range of consumption, which reached the minimum risk around 20–50 g/d (P-nonlinearity = 0.069). No associations were observed in men. Egg consumption was associated with decreased risk of dyslipidemia among Chinese women but not men. Additionally, the beneficial effect tended to be stronger at 20–50 g egg per day. Further study is warranted to validate these findings and clarify the causality. [ABSTRACT FROM AUTHOR]

Backgrounds: Abdominal obesity has been linked to risk of mortality, but whether and how trajectory of waist circumstance (WC) underpins this association remains unclear. The study aimed to identify long-term WC change trajectories and examine their association and joint effect with body mass index (BMI) on mortality among Chinese older adults.Methods: This present study included participants 60 years of age or older from China Health and Nutrition Survey (CHNS) from 1991 to 2015. The duration of follow-up was defined as period from the first to latest visit date attended with information on mortality, end of follow-up, or loss to follow-up (censoring). Latent class trajectory analysis (LCTA) was used to assess the changes of WC trajectories overtime. Cox proportional hazard models were used to assess hazard ratios (HRs) and corresponding 95% confidence internal (CIs) for mortality.Results: A total of 2601 participants with 8700 visits were included, and 562 mortality (21.6%) occurred during a median follow-up of 8.7 years. Using a group-based modeling approach, four distinct trajectories of WC change among Chinese older adults were identified as loss (13.5%), stable (46.8%), moderate gain (31.2%) and substantial gain (8.5%). With WC stable group as reference, the multivariable adjusted HRs for mortality were 1.34(95%CI:1.01-1.78) in loss group, 1.13(0.91-1.41) in moderate gain and 1.54(1.12-2.12) in substantial gain group. Compared with participants with normal BMI at baseline and maintained WC stable, the risk of mortality generally increased for all WC change group in initial overweight/obesity individuals, and the highest risk were observed for WC loss and stable pattern (HR:2.43, 95%CI: 1.41-4.19; HR:1.67 (1.07-2.60)).Conclusions: In older Chinese, both long-term WC loss and substantial gain conferred excess risk for mortality. The baseline BMI might modify the effect as overweight individuals had a greater risk imposed by WC loss than those in normal weight. Maintaining stable WC and normal weight might be necessary to reduce the risk of mortality. [ABSTRACT FROM AUTHOR]

Objectives: This study assessed the associations between long-term trajectories of percentage of energy from fat (PEF) and obesity among Chinese adults. Methods: Longitudinal data collected by the China Health and Nutrition Survey from 1991 to 2015 were analyzed. A body mass index ≥28.0 was defined as general obesity. Participants' baseline PEF levels were categorized as lower than the recommendation of the Chinese Dietary Guideline (<20%), meeting the recommendation (20–30%), and higher than the recommendation (>30%). Patterns of PEF trajectories were identified by latent class trajectory analysis for overall participants and participants in different baseline PEF groups, respectively. Cox proportional hazards regression models with shared frailty were used to estimate associations between PEF and obesity. Results: Data on 13,025 participants with 72,191 visits were analyzed. Four patterns of PEF trajectory were identified for overall participants and participants in three different baseline PEF groups, respectively. Among overall participants, compared with "Baseline Low then Increase Pattern" (from 12% to 20%), participants with "Baseline Normal-Low then Increase-to-High Pattern" (from 20% to 32%) had a higher hazard of obesity (hazard ratio (HR) and 95% confident interval (CI) at 1.18 (1.01–1.37)). Compared with the "Stable Pattern" group (stable at around 18% and 22%, respectively), participants with "Sudden-Increase Pattern" (from 18% to 30%) in the baseline group whose PEF levels were lower than the recommendation and those with "Sudden-Increase then Decrease Pattern" (rapidly increased from 25% to 40%, and then decreased) in the baseline group who met the recommendation had higher hazards of obesity (HRs and 95% CIs being 1.65 (1.13–2.41) and 1.59 (1.03–2.46), respectively). Conclusions: Adults with a trajectory that involved a sudden increase to a high-level PEF had a higher risk of general obesity. People should avoid increasing PEF suddenly. [ABSTRACT FROM AUTHOR]

Purpose: To characterize the quality of life and mental health status of patients with uveitis and investigate predictors of psychological problems.Methods: A total of 245 patients and 105 controls were enrolled in this cross-sectional study. Quality of life, psychological status, socio-demographic and clinical data were obtained from questionnaires and medical records. Multivariate regression analyses and Receiver Operating Characteristic (ROC) were applied to obtain the model predicting psychological problems of patients.Results: Of 245 patients, 16.7% and 26.5% (P< .0001) screened positive for anxiety and depression, respectively. The model predicting anxiety was comprised of low annual household income and poor self-reported visual function (P= .029,P< .0001, respectively), with an AUC of ROC of 0.744. The model predicting depression was comprised of poor self-reported visual function and ocular complications (P< .0001, P= .012, respectively), with an AUC of 0.78.Conclusions: Economic hardship, ocular complications, and poor self-reported visual function are predictors of mental problems in patients with uveitis. [ABSTRACT FROM AUTHOR]

Evidence was limited on trajectory of body mass index (BMI) through adulthood and its association with hypertension. We aimed to evaluate their association by sex in large-scale study. Data were obtained from the China Health and Nutrition Survey (CHNS) from 1991 to 2015. Latent class trajectory analysis (LCTA) was used to capture BMI change trajectories. Hazard risks (HRs) were estimated from Cox proportion hazard regression. Among 14,262 participants (mean age, 38.8; 47.8% men), 5138 hypertension occurred (2687 men and 2451 women) occurred during a mean follow-up 9.6 years. Four body mass trajectory groups were identified as BMI loss, stable, moderate and substantial gain. Appropriately half of participants (48.0%) followed 1 of the 2 BMI gain trajectories, where BMI increased at least 3 kg/m2 overtime. Compared with participants with stable BMI, those gaining BMI substantially had higher risk of hypertension by 65% (HR 1.65, 95% CI 1.45–1.86) in male and 83% (HR 1.83, 95% CI 1.58–2.12) in female. The HRs in BMI loss patterns were 0.74 (0.62–0.89) in men and 0.87 (0.75–1.00) in women. Our findings imply that majority of Chinese adults transited up to a higher BMI level during follow-up. Avoiding excessive weight gain and maintaining stable weight might be important for hypertension prevention. [ABSTRACT FROM AUTHOR]

Objective: Egg is the vital source for many nutrients. The present study aimed to evaluate how egg consumption contributes to body fat mass and fat distribution in general Chinese adults.Methods: In a representative study of Chinese adults, 2241 rural adults aged 18-80 years completed physical measurements and semi quantitative food frequency questionnaire (FFQ). Body fat percentage (BF%) was estimated through validated Chinese-specific equations including age, body mass index (BMI), waist circumference (WC), gender and their interaction.Results: The prevalence of excessive body fat and central obesity in males were 44.2% and 31.3%, lower than that in females (60.7% and 36.1%, p < 0.05). 37.3% of lean or normal-weight individuals by BMI experienced excessive body fat given their BF% and 43.5% were central obesity according to WC. In females, egg consumption was associated with better body fat mass and its distribution. Compared with non-consumers, individuals consumed egg >50 g/d had a 34% lower risk of central obesity (OR:0.66, 0.39-0.99) and 38% lower risk of excessive body fat (OR:0.62, 0.37-1.00). There were significant dose-response relationships between egg consumption and overweight with excessive body fat/central obesity (p for linear trend <0.05). However, these associations were not significant in males.Conclusion: Our findings underscore the necessity for future public health guidelines to assess body fat mass and distribution simultaneously in Chinese adults. Egg consumption might be beneficial for weight regulation, especially for overweight with excessive body fat/central obesity in females. Further research is warrant to establish the causal inference. [ABSTRACT FROM AUTHOR]

Human key-point detection is a challenging research field in computer vision. Convolutional neural models limit the number of parameters and mine the local structure, and have made great progress in significant target detection and key-point detection. However, the features extracted by shallow layers mainly contain a lack of semantic information, while the features extracted by deep layers contain rich semantic information but a lack of spatial information that results in information imbalance and feature extraction imbalance. With the complexity of the network structure and the increasing amount of computation, the balance between the time of communication and the time of calculation highlights the importance. Based on the improvement of hardware equipment, network operation time is greatly improved by optimizing the network structure and data operation methods. However, as the network structure becomes deeper and deeper, the communication consumption between networks also increases, and network computing capacity is optimized. In addition, communication overhead is also the focus of recent attention. We propose a novel network structure PGNet, which contains three parts: pipeline guidance strategy (PGS); Cross-Distance-IoU Loss (CIoU); and Cascaded Fusion Feature Model (CFFM). [ABSTRACT FROM AUTHOR]