Search

Background: The detection of amyloid-β (Aβ) deposition in the brain provides crucial evidence in the clinical diagnosis of Alzheimer's disease (AD). However, the current positron emission tomography (PET)-based brain Aβ examination suffers from the problems of coarse visual inspection (in many cases, with 2-class stratification) and high scanning cost.Objective: 1) To characterize the non-binary Aβ deposition levels in the AD continuum based on clustering of PET data, and 2) to explore the feasibility of predicting individual Aβ deposition grades with non-invasive functional magnetic resonance imaging (fMRI).Methods: 1) Individual whole-brain Aβ-PET images from the OASIS-3 dataset (N = 258) were grouped into three clusters (grades) with t-SNE and k-means. The demographical data as well as global and regional standard uptake value ratios (SUVRs) were compared among the three clusters with Chi-square tests or ANOVA tests. 2) From resting-state fMRI, both conventional functional connectivity (FC) and high-order FC networks were constructed and the topological architectures of the two networks were jointly learned with graph convolutional networks (GCNs) to predict the Aβ-PET grades for each individual.Results: We found three clearly separated clusters, indicating three Aβ-PET grades. There were significant differences in gender, age, cognitive ability, APOE type, as well as global and regional SUVRs among the three grades we found. The prediction of Aβ-PET grades with GCNs on FC for the 258 samples in the AD continuum reached a satisfactory averaged accuracy (78.8%) in the two-class classification tasks.Conclusion: The results demonstrated the feasibility of using deep learning on a non-invasive brain functional imaging technique to approximate PET-based Aβ deposition grading. [ABSTRACT FROM AUTHOR]

[Display omitted] • Propose deep learning-based detection method of time-delayed water quality indicator. • Contribution of each input indicator is quantified based on SHAP value. • NH 3 -H, DO, turbidity contribute significantly to detecting time-delayed WQIs. • Sensitivity analysis provides insight for higher accuracy and cost performance. Rapid and accurate detection of time-delayed water quality indicators (WQIs) is the key to achieving fast feedback regulation of wastewater treatment plants (WWTPs) that enables its energy-efficient operation and high tolerance towards shock sewage loads. However, advanced oxidation methods are costly, and data-driven modeling methods based on traditional machine learning algorithms for detecting time-delayed WQIs have limited detection accuracy. This work develops deep learning models based on long short-term memory (LSTM) neural networks to detect time-delayed WQIs in WWTPs intake accurately. The lack of interpretability of the deep learning models hampers the optimization of the developed LSTM models in applications. Therefore, a global sensitivity analysis (GSA) based on Shapley additive explanations (SHAP) is performed to quantify the contribution of the input indicators to detection results of the developed LSTM models. The direct contributions provide the basis for optimizing the input indicators to achieve more cost-effective modeling detection. In the case study, the developed LSTM models achieved good accuracy (R2 of 0.9141, 0.9239, and 0.9040, respectively) in detecting chemical oxygen demand, total nitrogen, and total phosphorus in the influent of a WWTP, outperforming the four types of baseline models. According to the SHAP values, the contributions of dissolved oxygen, turbidity, and ammonia nitrogen to the above detection targets are always in the top third of all input indicators, which are more outstanding than meteorological indicators. Removing the indicator with the smallest SHAP value reduces the build and run costs of the models with minimal loss of detection accuracy. Combining deep learning and GSA to detect WWTPs influent is a novel and effective attempt. This attempt provides a more sustainable solution for rapid and accurate detection of time-delayed WQIs, which drives WWTPs' operation in an intelligent, clean, and safe direction. [ABSTRACT FROM AUTHOR]

• A model combining two-stage decomposition and LSTM for TP detection in sewer networks. • Two-stage decomposition reduces the instability and complexity of TP time series. • Mode decomposition procedure setting significant impact modeling efficiency, accuracy. • Proposed hybrid model achieves excellent accuracy of average R2 = 0.946. • Proposed hybrid model has the advantage of being energy-efficient and cost-effective. Timely and accurately grasping total phosphorus (TP) concentration in sewer networks is crucial for urban phosphorus flow management and shock load early warning of sewage treatment facilities. Modeling-based methods are cleaner and more energy-saving than traditional methods requiring digestion procedures. However, the TP time series' strong nonlinearity and complexity result in unsatisfactory accuracy in these methods. This work proposes a hybrid model named CEEMDAN-SE-VMD-LSTM that combines complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), sample entropy (SE), variational mode decomposition (VMD), and long short-term memory (LSTM) neural network to improve the accuracies of modeling-based methods. In proposed hybrid model, the two-stage decomposition procedure can decompose the TP time series into several lower-complexity modes, and the powerful nonlinear mapping ability of the LSTM neural network enables accurate prediction of these modes. In case study, the proposed hybrid model achieves excellent detection accuracy with an average R2 of 0.9460 ± 0.0243. Compared with the hybrid models formed by combining other decomposition procedures (i.e., CEEMDAN, VMD, singular spectrum analysis (SSA), CEEMDAN-SE-SSA) and LSTM neural network, the proposed hybrid model has the highest detection accuracy (1.36–3.94 % higher Nash-Sutcliffe efficiency, 1.28–4.58 % higher R2, 12.14–24.86 % lower RMSE). The strategy of setting the mode decomposition procedure based on a comprehensive analysis of the decomposition algorithms and the obtained modes ensures high detection accuracy of the proposed hybrid model while avoiding costly computational burdens. This work is enlightening for improving the accuracy and modeling efficiency of soft detection methods, which are expected to reduce energy consumption and pollution caused by water quality detection. [ABSTRACT FROM AUTHOR]

The high concentrations of heavy metals in municipal industrial sewer networks will seriously impact the microorganisms of the activated sludge in the wastewater treatment plant (WWTP), thus deteriorating the effluent quality and destroying the stability of sewage treatment. Therefore, timely prediction and early warning of heavy metal concentrations in industrial sewer networks is crucial. However, due to the complex sources of heavy metals in industrial sewer networks, traditional physical modeling and linear methods cannot establish an accurate prediction model. Herein, we developed a Gated Recurrent Unit (GRU) neural network model based on a deep learning algorithm for predicting the concentrations of heavy metals in industrial sewer networks. To train the GRU model, we used low-cost and easy-to-obtain urban multi-source data, including socio-environmental indicator data, air environmental indicator data, water quantity indicator data, and easily measurable water quality indicator data. The model was applied to predict the concentrations of heavy metals (Cu, Zn, Ni, and Cr) in the sewer networks of an industrial area in southern China. The results are compared with the commonly used Artificial Neural Network (ANN) model. In this study, it was shown that the GRU had better prediction performance for Cu, Zn, Ni, and Cr concentrations, with the average R2 significantly increased by 12.35%, 11.94%, 9.21%, and 8.13%, respectively, compared to ANN predictions. The sensitivity analysis based on Shapley (SHAP) values revealed that conductivity (σ), temperature (T), pH, and sewage flow (Flow) contributed significantly to the prediction results of the model. Furthermore, the three input variables including air pressure (AP), land area (A), and population (Pop.) were removed without affecting the prediction performance of the model, which maximized the modeling efficiency and reduced the operational cost. This study provides an economical and feasible technical method for early warning of abnormal heavy metal concentrations in urban industrial sewer networks. [Display omitted] • This study reveals the variation of Cu, Zn, Ni, and Cr concentrations in industrial sewer networks. • Deep learning method based on multi-source data can predict heavy metal concentrations in sewers. • The Gated Recurrent Unit model has superior prediction performance with more than 93% accuracy. • The Shapely value method identifies the key indicators for predicting heavy metals. [ABSTRACT FROM AUTHOR]

• Data-driven method based on deep learning can predict FOG content in sewers. • The FOG prediction model based on GRU has good generalization ability. • GRU network structure adjustment achieves the optimal model performance. • The key indicators are identified based on GSA for rapid FOG content monitoring. The content of fat, oil and grease (FOG) in the sewer network sediments is the key indicator for diagnosing sewer blockage and overflow. However, the traditional FOG detection is time-consuming and costly, and the establishment of mathematical models based on statistical methods to predict the content of FOG fail to provide satisfactory accuracy. Herein, a deep learning algorithm used a data-driven FOG content prediction model is proposed to achieve a more accurate prediction of FOG content. Meanwhile, global sensitivity analysis (GSA) is exploited to evaluate the contribution of input indicators to the output indicator (FOG) in the model, so that some input indicators that have less impact on the prediction performance can be screened out, the best combination of input indicators can be determined, and the operation cost of the model can be reduced. To evaluate the effectiveness of the proposed model, a case study was conducted in a city in southern China. The experimental results indicate that the prediction model obtains good FOG estimations and performs well from a single site to multiple sites with a mean R2 of 0.922, showing a good generalization performance. Through GSA, the key input indicators in the model were identified as pH, water temperature (T), relative humidity (RH), sewage flow (Flow), drinking water supply (DWS), velocity (V) and conductivity (σ), and the input indicators such as air pressure (AP), population (Pop.), and liquid level (LV) can be reduced without affecting the prediction accuracy of the model. [Display omitted] [ABSTRACT FROM AUTHOR]

Point source pollution in urban drainage networks, which is difficult to monitor and control, has been regarded as an intractable problem. To solve the problem, key water quality indicators must be tracked in the evaluation and prediction of sewer water quality. However, some of these important chemical indicators (e.g. biological oxygen demand (BOD 5), chemical oxygen demand (COD), ammonia nitrogen (NH 4 +-N), total nitrogen (TN), and total phosphorus (TP)) require a great deal of time and effort to measure, which will adversely affect the prediction in a sewage network. Existing statistical methods and machine learning algorithms cannot effectively solve the detection time problem or provide limited accuracy. Moreover, the lack of various factors taken into account in these methods results in unsatisfactory predictive performance. Few studies consider the impact of urban multi-source data on water quality prediction of sewer networks while developing statistical methods or machine learning algorithms. To address this problem, we propose a deep learning approach based on multi-source data fusion. This approach takes into account the following indicators to comprehensively analyze and predict drainage water quality: environmental indicators (such as area and diameter); social indicators (such as population); water quantity indicators (such as drinking water supply, sewage flow, water velocity, and liquid level); and easily monitored water quality criteria indicators (such as pH, temperature, and conductivity). To test the effectiveness of this method, we conducted a case study in a city in southern China. By comparing this new method with the linear method (multiple linear regression, MLR) and traditional learning algorithm (multilayer perception, MLP), it is found that the deep learning algorithm—which includes recurrent neural network (RNN), long-short term memory (LSTM), and gated recurrent unit (GRU)—has good predictive performance, in which GRU shows superior ability in predicting the chemical index of water quality and the learning curve is faster. The results showed that the GRU achieved 0.82%–5.07% higher R2 than RNN and LSTM, 9.13%–15.03% higher R2 than traditional machine learning algorithms, and 37.26%–43.38% higher R2 than linear methods. [Display omitted] • Deep learning model based on multi-source data can predict sewer water quality. • RNN-based models have better predictive performance than MLR and MLP models. • In deep learning models, the GRU has fast learning curve and simple architecture. • GRU model prediction performance outperformed RNN and LSTM model. [ABSTRACT FROM AUTHOR]