Search

Inferential modeling has been of significance for modern manufacturing in estimating the quality-related process variables. As an effective inferential model, probabilistic slow feature analysis (PSFA) has gained attention in regression tasks to interpret dynamic properties with a slowness preference. However, PSFA is often challenged by the nonlinear sequential data due to its linear state-space structure. In this study, a new nonlinear extension of PSFA is proposed under the deep learning framework to enhance the dynamic feature extraction with limited labels, incorporating variational inference and Monte Carlo inference to derive the objective function. The proposed model considers the relevance of inputs with outputs as the input weights to upgrade prediction performance. The proposed model is verified through an industrial hydrocracking process to predict diesel yield with missing labels ranged from 0% to 50%, and the root mean squared error is reduced by at least 8.78% compared to PSFA.

Lately, variational autoencoder (VAE) has arisen as one of the most prevalent methods for unsupervised learning of intricate distributions. Despite being successful in deep feature extraction and uncertain data modelling, it still suffers from the instability and reconstruction error due to random sampling in the latent subspace representation of original input space. In this paper, to deal with those limitations, constrained VAE (CVAE) is proposed by utilizing input sample information. Enthused by parallel interaction mechanism between the ventral and dorsal stream of human brain in object recognition, parallel interaction spatial-temporal CVAE (PIST-CVAE) is proposed to extract spatial and temporal features from input samples. Lower dimensional nonlinear features extracted from PIST-CVAE are used to build the soft sensor. The effectiveness of CVAE and PIST-CVAE is demonstrated in an industrial case study, a polyester polymerization process.

Aiming to addressing the nonrepetitive uncertainties of multiagent systems, this work proposes a discrete-time-distributed adaptive iterative learning control (DDAILC) scheme for an output consensus problem, where two fundamental requirements in the traditional distributed iterative learning control (ILC) methods, i.e., the identical initial states and the repetitive desired trajectories, are removed. Furthermore, the algorithm design and analysis are directly aimed at discrete-time nonlinear multiagent systems, rather than continuous-time ones, to meet the needs of practical implementations. The iteration-varying trajectory of the virtual leader is included in the learning control protocol for a compensation. The adaptive parameter-updating law works along the iteration dimension by using a general consensus error that contains the output data of adjacent agents. To ensure the estimation of the control gain to be nonzero, a semisaturator is utilized in the parameter-updating law. The convergence of the output consensus is shown rigorously. Both numerical and practical examples are used to test the theoretical results. Moreover, the DDAILC efficiently improves performance of the building heating, ventilation, and air conditioning (HVAC) system by utilizing both the distributed topology and the repetitive dynamic characteristic.

Modern industrial plants generally consist of multiple manufacturing units, and the local correlation within each unit can be used to effectively alleviate the effect of spurious correlation and meticulously reflect the operation status of the process system. Therefore, the local correlation, which is called spatial information here, should also be taken into consideration when developing the monitoring model. In this study, a cascaded monitoring network (MoniNet) method is proposed to develop the monitoring model with concurrent analytics of temporal and spatial information. By implementing convolutional operation to each variable, the temporal information that reveals dynamic correlation of process data and spatial information that reflects local characteristics within individual operation unit can be extracted simultaneously. For each convolutional feature, a submodel is developed and then all the submodels are integrated to generate a final monitoring model. Based on the developed model, the operation status of the newly collected sample can be identified by comparing the calculated statistics with their corresponding control limits. Similar to the convolutional neural network (CNN), the MoniNet can also expand its receptive field and capture deeper information by adding more convolutional layers. Besides, the filter selection and submodel development in MoniNet can be replaced to generalize the proposed network to many existing monitoring strategies. The performance of the proposed method is validated using two real industrial processes. The illustration results show that the proposed method can effectively detect process anomalies by concurrent analytics of temporal and spatial information.

Objectives The aim of this study was to establish a time-resolved fluorescent immunoassay (TRFIA) for the detection of serum galactose agglutinin 3 (Gal-3) and apply this method to evaluate the clinical significance of serum Gal-3 in predicting Idiopathic Membranous Nephropathy (IMN) progression. Methods The Gal-3-TRFIA was established using the double antibody sandwich method, with the capture antibodies coated on a 96-well microplate and the detection antibodies chelated to Europium (III) (Eu3+). Serum Gal-3 was detected in 81 patients with IMN and 123 healthy controls to further evaluate the value of the Gal-3 in staging of IMN. Results The sensitivity of the Gal-3-TRFIA assay was 0.85 ng/mL, and the detection range was 0.85–1000 ng/mL. The Gal-3 intra-batch and inter-batch coefficients of variation were 3.45% and 5.12%, respectively. The correlation coefficient (R) between the Gal-3-TRFIA assay and commercially available enzyme-linked immunosorbent assay kits was 0.83. The serum Gal-3 concentration was higher in patients with IMN (65.57 ± 55.90 ng/mL) compared to healthy controls (16.29 ± 9.91 ng/mL, P 3+) chelates for the detection of Gal-3 concentrations in serum. The Gal-3-TRFIA also detected the severity of the IMN course.

In industrial processes, soft sensor techniques are used to predict the hard-to-measure quality variables under the classic supervised learning paradigm. However, the data challenges, i.e., the widespread noises and inadequate labeled samples, usually make the data-driven sensors weak for application. In this paper, a simple but effective regularization method termed as adversarial smoothing regularization (ASR) is proposed, which measures the local smoothness of prediction around each input sample. By minimizing the divergence between the predictions for noisy and clean inputs, the proposed ASR regulates the learning model to be robust against local perturbation in a semi-supervised manner. Theoretical analysis explains the smoothing capability of ASR, and shows that it is helpful for the improvement of generalization performance. We also design a tri-regression framework to further use the information of unlabeled samples with pseudo labels and present the adversarial smoothing tri-regression (ASTR) model for soft sensor. Based on two industrial processes, comprehensive soft sensor experiments and noise tests are performed to show the robust semi-supervised learning capability of the proposed ASR and ASTR.

Converting CO2 into chemicals with electricity generated by renewable energy is a promising way to achieve the goal of carbon neutrality. Carbon-based materials have the advantages of low cost, wide sources and environmental friendliness. In this work, we prepared a series of boron-doped covalent triazine frameworks and found that boron doping can significantly improve the CO selectivity up to 91.2% in the CO2 electroreduction reactions(CO2RR). The effect of different doping ratios on the activity by adjusting the proportion of doped atoms was systematically investigated. This work proves that the doping modification of non-metallic materials is a very effective way to improve their activity, and also lays a foundation for the study of other element doping in the coming future.

Background We evaluated the use of thyroid stimulating immunoglobulin (TSI) assay to optimize the detection strategy for Graves’ disease. Methods Five hundred and forty-four well characterized serum samples from the Clinical Laboratory of Shanghai Tongren Hospital were collected from August 2019 to April 2020. The serum samples were obtained from 52 untreated GD patients, 155 treated GD patients, 83 patients with Hashimoto's thyroiditis, 70 patients with thyroid nodules, 83 patients with thyroid cancer, and 101 healthy subjects. All samples were evaluated by both TSI assay and TSH receptor autoantibodies (TRAb) assay. Moreover, 23 patients without a distinct thyroid disease diagnosis at the first visit were monitored for 6 months to make a final diagnosis. Results The clinical sensitivity of the TSI and TRAb assays was 98.10% and 94.20% respectively, while the clinical specificity was 92.30% and 96.70% respectively. ROC plot analysis based on sera of UT-GD (newly diagnosed GD patients) revealed an area under the curve (AUC) of 0.974 for the TSI assay. The best cutoff value was 0.58 IU/l (98.0% of sensitivity, 92.8% of specificity). The AUC for the TRAb assay was 0.961. Furthermore, combining TSI and TRAb results, the area under the curve was 0.981. In a pilot study of 23 patients with an uncertain initial diagnosis, the follow-up results showed the clinical diagnosis of 22 out of 23 cases were resolved in agreement with the results obtained by the TSI assay, and one case matched the result obtained by TRAb assay. Conclusion The TSI assay presents very promising analytical characteristics and could be adopted in clinical practice to improve GD diagnosis. The TSI assay might be better than TRAb assay in initial differential diagnosis of GD from other thyroid diseases.

Lingli Chen,1,&ast; Jianfeng Hong,2,&ast; Renjing Hu,3,&ast; Xiaomei Yu,1 Xindong Chen,1 Shaoxiong Zheng,1 Yuan Qin,1 Xiumei Zhou,1 Yigang Wang,1 Liping Zheng,2 Hongming Fang,2 Pengfei Liu,4 Biao Huang1 1Department of Immunoassay Laboratory, College of Life Sciences and Medicine, Zhejiang Sci-Tech University, Hangzhou, People’s Republic of China; 2Department of Laboratory, Affiliated Xiaoshan Hospital, Hangzhou Normal University, Hangzhou, People’s Republic of China; 3Department of Laboratory, The Affiliated Wuxi No.2 People’s Hospital of Nanjing Medical University, Wuxi, People’s Republic of China; 4Department of Gastroenterology, The Jiangyin Clinical College of Xuzhou Medical University, Wuxi, People’s Republic of China&ast;These authors contributed equally to this workCorrespondence: Biao HuangImmunoassay Laboratory, College of Life Sciences and Medicine, Zhejiang Sci-Tech University, Hangzhou, 310016, People’s Republic of ChinaTel +86 571-86843187Email jswxhb@163.comHongming FangDepartment of Laboratory, Affiliated Xiaoshan Hospital, Hangzhou Normal University, Hangzhou, 310016, People’s Republic of ChinaTel +86 571 83865858Email 3295988078@qq.comPurpose: The present study aimed to evaluate the clinical value of the combined detection of soluble T cell immunoglobulinand mucin domain molecule 3 (sTim-3) and pepsinogen (PG) in sera for gastric cancer (GC) diagnosis.Patients and Methods: The double antibody sandwich method was used to establish a highly sensitive time-resolved fluorescence immunoassay for the detection of sTim-3. Serum sTim-3, PGI, and PGII levels in 149 GC patients (123 first-diagnosis GC patients and 26 post-GC patients), 81 patients with benign gastric disease (BGD), and 73 healthy controls were quantitatively detected. The clinical diagnostic value of the combined detection of sTim-3 and PG in GC was analyzed.Results: Serum sTim-3 levels in GC (20.41 ± 9.55 ng/mL) and BGD (16.50 ± 9.76 ng/mL) patients were significantly higher (P < 0.001) than those in healthy controls (9.22 ± 3.40 ng/mL). Combined detection of sTim-3 and PGI/PGII (AUC: 0.9330, sensitivity: 86.44%, and specificity: 91.78%) showed a high diagnostic value for GC. When the level of PGI/PGII was less than 12.11 and that of sTim-3 was greater than 14.30 ng/mL, the positive rate of the control group was reduced to 0%, and the positive detection rate of GC was 54.47%. In addition, in post-operative patients, serum sTim-3 levels in the recurrence group (33.56 ± 4.91 ng/mL) were significantly higher than those in the no recurrence group (11.95 ± 5.16 ng/mL).Conclusion: sTim-3 levels in BGD and GC sera were significantly higher than those in the control group sera. Additionally, sTim-3 serum levels can predict recurrence in post-operative patients. Compared with PG alone, the combined detection of serum PG and sTim-3 can significantly improve the detection sensitivity and specificity of BGD and GC.Keywords: T cell immunoglobulin and mucin domain molecule 3, time-resolved fluorescence immunoassay, biomarker, gastric cancer

This paper synchronizes control theory with computer vision by formalizing object tracking as a sequential decision-making process. A reinforcement learning (RL) agent successfully tracks an interface between two liquids, which is often a critical variable to track in many chemical, petrochemical, metallurgical, and oil industries. This method utilizes less than 100 images for creating an environment, from which the agent generates its own data without the need for expert knowledge. Unlike supervised learning (SL) methods that rely on a huge number of parameters, this approach requires far fewer parameters, which naturally reduces its maintenance cost. Besides its frugal nature, the agent is robust to environmental uncertainties such as occlusion, intensity changes, and excessive noise. From a closed-loop control context, an interface location-based deviation is chosen as the optimization goal during training. The methodology showcases RL for real-time object-tracking applications in the oil sands industry. Along with a presentation of the interface tracking problem, this paper provides a detailed review of one of the most effective RL methodologies: actor–critic policy.

Parkinson's disease (PD), a chronic neurodegenerative disease, is characterized by sensorimotor and cognitive deficits. Previous diffusion tensor imaging (DTI) studies found abnormal DTI metrics in white matter bundles, such as the corpus callosum, cingulate, and frontal-parietal bundles, in PD patients. These studies mainly focused on alterations in microstructural features of long-range bundles within the deep white matter (DWM) that connects pairs of distant cortical regions. However, less is known about the DTI metrics of the superficial white matter (SWM) that connects local cortical regions in PD patients. To determine whether the DTI metrics of the SWM were different between the PD patients and the healthy controls, we recruited DTI data from 34 PD patients and 29 gender- and age-matched healthy controls. Using a probabilistic tractographic approach, we first defined a population-based SWM mask across all the subjects. Using a tract-based spatial statistical (TBSS) analytic approach, we then identified the SWM bundles showing abnormal DTI metrics in the PD patients. We found that the PD patients showed significantly lower DTI metrics in the SWM bundles connecting the sensorimotor cortex, cingulate cortex, posterior parietal cortex (PPC), and parieto-occipital cortex than the healthy controls. We also found that the clinical measures in the PD patients was significantly negatively correlated with the fractional anisotropy in the SWM (FASWM) that connects core regions in the default mode network (DMN). The FASWM in the bundles that connected the PPC was significantly positively correlated with cognitive performance in the PD patients. Our findings suggest that SWM may serve as the brain structural basis underlying the sensorimotor deficits and cognitive degeneration in PD patients.

Reinforcement learning (RL) for continuous state/action space systems has remained a challenge for nonlinear multivariate dynamical systems even at a simulation level. Implementing such schemes for real-time control is still of a difficulty and remains largely unanswered. In this study, several critical strategies for practical implementation of RL are developed, and a multivariable, multi-modal, hybrid three-tank (HTT) physical process is utilized to illustrate the proposed strategies. A successful real-time implementation of RL is reported. The first step is a meta-heuristic first principles model parameter optimization, where a custom pseudo random binary signal (PRBS) is used to obtain open-loop experimental data. This is followed by in silico asynchronous advantage actor–critic (A3C/A-A2C) based policy learning. In the second step, three different approaches (namely proximal learning, single trajectory learning, and multiple trajectory learning) are utilized to explore the state/action space. In the final step, online learning (A2C) using the best in silico policy on the real process using a socket connection is established. The extent of exploration (EoE, a measure of exploration) is proposed as a parameter for quantifying exploration of the state/action space. While the online sample efficiency of RL application is enhanced, a soft constraint based constrained learning is proposed and validated. With considerations of the proposed strategies, this work demonstrates the possibility of applying RL to solve practical control problems.

Storm geysers increasingly occur in sewer systems under climate change and rapid urbanization. Mitigation measures are in great demand to avoid safety problems. In this study, three-dimensional computational fluid dynamics models of single-inlet and multi-inlet systems were established to investigate geysering induced by rapid filling and assess the effectiveness of potential mitigation methods. The modeling results suggest that increasing the capacity of the downstream pipe before the inflow front reaches the chamber can effectively reduce the maximum geyser pressure. The peak pressure can be significantly mitigated when the chamber size is designed with care and the drop height between the upstream and downstream pipes is reduced. A diversion deflector with air vents and an orifice plate at the riser top end can alleviate the maximum pressure by about 65% with about 75% of the entrapped air being released. The peak pressure during the geyser event in the multi-inlet model is less than that of a single-inlet model under the same total inflow condition, but more water can be released. HIGHLIGHTS A small drop height between the connecting pipes can effectively reduce the maximum pressure during rapid filling.; Mitigation measures including flow deflectors and orifice plates were studied to alleviate geyser intensity.; Intermittent air–water jets can be generated much more easily in a system with multiple inlets, and the peak pressure is generally small but more water can be released.

In this paper, a novel and effective stiction detection method is proposed by combining K-means clustering and the moving window approach. As a byproduct, the proposed stiction detection method off...

Aims: Although preeclampsia (PE) has been attributed to excessive oxidative stress (OS) in the placenta, mild antioxidants failed to prevent PE in clinical trials. As mitochondria are a major sourc...

In oil sands industry, primary separation vessel (PSV) is a critical component to recover bitumen from oil sands slurry. Accurate interface level estimation between froth and middlings layers ensures economical and environmental benefits of bitumen recovery. Nuclear density profiler, differential pressure (DP) cell, and image processing based computer vision system are usually used to estimate the interface level. The computer vision system, which uses a camera to capture sight glass vision frames, is considered to be the most accurate. Although the accuracy of computer vision system is high in normal operational conditions, its qualities are influenced by abnormalities, such as sight glass vision blocking, stains, and level switching between sight glasses. A sensor fusion approach, which recursively updates fusion parameters according to accurate computer vision results whenever they are reliable, is proposed. The fused results can then be used to provide reliable interface level estimation under abnormal scenarios. The sensor fusion algorithm is further integrated with computer vision system to improve froth-middlings interface level estimation accuracy and robustness. Industrial environment simulations and factory accepted test (FAT) demonstrate the advantages and effectiveness of the sensor fusion and computer vision integrated system, which is applied in the industry.

Bitumen in the oil sands industry is separated from sand using a water-based gravity separation process in a primary separation vessel (PSV). The interface between the froth and the middlings layer is an important parameter to control for optimal operation of the PSV unit. In this paper, a method using computer vision based on convolutional neural networks (CNNs) and Kalman filter (KF) is designed for the detection of the interface in PSV, with the CNN estimating both the interface level and the image quality. The proposed method consists of two parts: an offline and an online stage, wherein the parameters of CNN and KF are trained in the offline stage using the available data. The algorithm is made robust to any new type of occlusions, not present in the training dataset, in the online stage. Experimental results demonstrate that the proposed method is accurate and robust to different abnormalities in the process.

Aims The aim of this research was to investigate the alterations in functional brain networks and to assess the relationship between depressive impairment and topological network changes in Parkinson's disease (PD) patients with depression (DPD). Methods Twenty‐two DPD patients, 23 PD patients without depression (NDPD), and 25 matched healthy controls (HCs) were enrolled. All participants were examined by resting‐state functional magnetic resonance imaging scans. Graph theoretical analysis and network‐based statistic methods were used to analyze brain network topological properties and abnormal subnetworks, respectively. Results The DPD group showed significantly decreased local efficiency compared with the HC group (P = .008, FDR corrected). In nodal metrics analyses, the degree of the right inferior occipital gyrus (P = .0001, FDR corrected) was positively correlated with the Hamilton Depression Rating Scale scores in the DPD group. Meanwhile, the temporal visual cortex, including the bilateral middle temporal gyri and right inferior temporal gyrus in the HC and NDPD groups and the left posterior cingulate gyrus in the NDPD group, was defined as hub region, but not in the DPD group. Compared with the HC group, the DPD group had extensive weakening of connections between the temporal‐occipital visual cortex and the prefrontal‐limbic network. Conclusions These results suggest that PD depression is associated with disruptions in the topological organization of functional brain networks, mainly involved the temporal‐occipital visual cortex and the posterior cingulate gyrus and may advance our current understanding of the pathophysiological mechanisms underlying DPD., The functional brain networks of the patients with Parkinson's disease (PD) with depression exhibite small‐world architecture. PD depression is associated with disruptions in the topological organization of functional brain networks, mainly involved the temporal‐occipital visual cortex and the posterior cingulate gyrus.

Zhejiang Province achieved one of the best records in containing the COVID‐19 pandemic in China, what lessons can the world learn from it? What roles do community‐based organizations play in its success story? Based on more than 100 interviews during and after the outbreak in Zhejiang, this article provides a roadmap of how community‐based organizations were involved in the three distinct stages of Zhejiang's responses to COVID‐19. We recommend that public sector leaders strategically leverage the strengths of community‐based organizations in multiple stages of COVID‐19 responses; incentivize volunteers to participate in epidemic prevention and control; provide data infrastructure and digital tracking platforms; and build trust and long‐term capacity of community‐based organizations. This article is protected by copyright. All rights reserved.

Main challenges for developing data-based models lie in the existence of high-dimensional and possibly missing observations that exist in stored data from industry process. Variational autoencoder (VAE) as one of the deep learning methods has been applied for extracting useful information or features from high-dimensional dataset. Considering that existing VAE is unsupervised, an output-relevant VAE is proposed for extracting output-relevant features in this work. By using correlation between process variables, different weight is correspondingly assigned to each input variable. With symmetric Kullback–Leibler (SKL) divergence, the similarity is evaluated between the stored samples and a query sample. According to the values of the SKL divergence, data relevant for modeling are selected. Subsequently, Gaussian process regression (GPR) is utilized to establish a model between the input and the corresponding output at the query sample. In addition, owing to the common existence of missing data in output data set, the parameters and missing data in the GPR are estimated simultaneously. A practical debutanizer industrial process is utilized to illustrate the effectiveness of the proposed method.

Conventional adaptive monitoring strategies detect anomalies in time-varying process by frequently updating models, which requires high computation complexity and may falsely include abnormal samples. Cointegration analysis (CA) based monitoring strategies can be implemented with less model updating since they are developed based on the extracted long-term equilibrium relationship. However, once the cointegration relationship changes, the previous CA model cannot accurately reflect the operation status of future nonstationary process. In this study, an adaptive monitoring scheme based on recursive CA is proposed to address the aforementioned issues for nonstationary processes. First, a recursive strategy is developed for CA to effectively update the monitoring model. After that, three monitoring statistics are developed to reflect the operation status of the industrial process with representation of both static deviation and dynamic fluctuation. Finally, an adaptive monitoring strategy is constructed based on the proposed recursive CA using the aforementioned monitoring statistics. Experimental results of two real industrial processes show that the adaptive monitoring strategy based on recursive CA can effectively adapt to normal process changes without frequent model updating.

This paper deals with the normal utilization rate in the Sraffa framework. It further presents two arguments about the utilization controversy surrounding the neo‐Kaleckian theory of growth and distribution: First, it shows that the normal utilization rate is generally not independent of demand and the growth rate, and this result does not rely on the neo‐classical production function and other restrictive assumptions. Second, it argues that endogenizing the normal utilization rate is not a general solution for the utilization controversy.

Deep learning is a recently developed feature representation technique for data with complicated structures, which has great potential for soft sensing of industrial processes. However, most deep networks mainly focus on hierarchical feature learning for the raw observed input data. For soft sensor applications, it is important to reduce irrelevant information and extract quality-relevant features from the raw input data for quality prediction. To deal with this problem, a novel deep learning network is proposed for quality-relevant feature representation in this article, which is based on stacked quality-driven autoencoder (SQAE). First, a quality-driven autoencoder (QAE) is designed by exploiting the quality data to guide feature extraction with the constraint that the potential features should largely reconstruct the input layer data and the quality data at the output layer. In this way, quality-relevant features can be captured by QAE. Then, by stacking multiple QAEs to construct the deep SQAE network, SQAE can gradually reduce irrelevant features and learn hierarchical quality-relevant features. Finally, the high-level quality-relevant features can be directly applied for soft sensing of the quality variables. The effectiveness and flexibility of the proposed deep learning model are validated on an industrial debutanizer column process.

Lead sulfide (PbS) quantum dots (QDs) are important near infrared (NIR) luminescent materials with tunable and strong emission covering a broad NIR region. However, their optical properties are quite sensitive to air, water, and high temperature due to the surface oxidation, thus limiting their applications in optoelectronic devices and biological imaging. Herein, a cation-doping strategy is presented to make a series of high-quality Zn-doped PbS QDs with strong emission covering whole second near-infrared window (NIR-II, 1,000-1,700 nm). First-principle calculations confirmed that Zn dopants formed dopant states and decreased the recombination energy gap of host PbS. Notably, the Zn dopants significantly improved the quantum yield, photoluminescence lifetime and thermal stability of PbS QDs. Moreover, the PEGylated Zn-doped PbS QDs emitting in the NIR-llb window (1,500-1,700 nm) realized the noninvasive imaging of cerebral vascular of mouse with high resolution, being able to distinguish blood capillary. This material not only provides a new tool for deep tissue fluorescence imaging, but is also promising for the development of other NIR related devices.

Gaussian process regression (GPR) is one of the most important data analytic tools in modelling processes. It has attracted increasing interest in chemical engineering applications due to its superior performance in dealing with complex modelling problems such as high-dimensional and nonlinear data. However, traditional GPR has the main limitation in that it considers an independent identically distributed (i.i.d.) noise at every sample point. Modern chemical processes typically have a more complex data structure and noise properties. The assumption of i.i.d. noise is not realistic. Thus, there is a growing interest in solving a heteroscedastic noise problem that does not satisfy the i.i.d. condition. The most common heteroscedastic noise is the noise with varying variance. This paper proposes a novel machine learning variance prediction method to solve the heteroskedastic GPR problem. By considering not only the input-dependent noise variance but also the input-output-dependent noise variance, a regression model based on support vector regression (SVR) and extreme learning machine (ELM) method is proposed for both noise variance prediction and smoothing. Compared with the existing weighted Gaussian process regression (W-GPR) of the literature, the proposed method not only expands the use of W-GPR but also improves the prediction performance of heteroscedastic GPR models. Finally, the proposed algorithm is verified by two numerical examples and tested in a real polyester polymerization process. The results all demonstrate the effectiveness of the proposed approach.

Autoencoder (AE) is a deep neural network that has been widely utilized in process industry owing to its superior abilities of feature extraction and data reconstruction. Recently, assuming the latent variables to be random variables, a probabilistic variant of it called variational autoencoder (VAE) has achieved a major success in different applications. In this article, we develop two novel submodels based on deep VAEs (DVAE), which are further utilized to establish a soft sensor framework. By the use of our first submodel known as supervised DVAE (SDVAE), the distribution information of latent features can be obtained. This is used as a prior of the second submodel known as the modified unsupervised DVAE (MUDVAE). Then, a new soft sensor framework can be constructed by combing the encoder of SDVAE with the decoder of MUDVAE. Since our designed VAE has superior ability in data reconstruction, it also works well under the missing data situation which is common in process industries due to sensor failures. Thus, we extend the proposed soft sensor framework to handle the missing data situation. The effectiveness of our proposed soft sensor frameworks is finally demonstrated via an industrial polymerization dataset.

Phthalate pollution in plastic greenhouses (PGs) has aroused concerns. However, mechanisms and factors of vegetables planted in PGs (VPGs) accumulating phthalates from soil and air are unclear. To fill the gap, 19 PGs in Shaanxi, the largest vegetable production province in northwestern China, were selected to probe this issue. 35 soil samples, 48 air samples, and 26 VPG samples were collected in winter and summer. Medians of sum of 7 phthalate concentrations (∑7 PAEs) in PG soil, air, and VPGs were 73.9 μg kg−1, 5300 ng m−3, and 1053 μg kg−1 dry weight, respectively. ∑7 PAE concentrations in PG environmental media in winter were higher than summer, with the significant difference in VPGs. Sum concentrations of bis (2-ethylhexyl) phthalate (DEHP) and di-n-butyl phthalate (DnBP) accounted for 76.8% and 82.3% of the ∑7 PAEs in soil and VPGs. DnBP and DEHP concentrations in VPGs were significantly correlated to those in air and soil, with correlation coefficients (R) of 0.89 and 0.96 to air and 0.68 and 0.59 to soil. Log-transformed soil-air partition coefficient (log KSA) and fugacity fraction (log ff) of DnBP decreased while log KSA and log ff of DEHP increased from winter to summer, though DnBP in soil volatilized to air while DEHP in air sank to soil within the year. These issues were caused by air temperature changes and the application of plastic films. Furthermore, DnBP concentrations in VPGs were positively correlated to KSA values of DnBP (R = 0.87) while those of DEHP were negative (R = −0.82). Therefore, VPGs could uptake more phthalates from air than from soil, especially for edible parts of leafy and solanaceous VPGs. Applying phthalates free agricultural films and precision management such as adjusting air temperature in PGs could be considered to ensure VPG safeties.

Although studies have investigated the associations between PM2.5 and mortality risk, evidence from rural areas is scarce. We aimed to compare the PM2.5-mortality associations between urban cities and rural areas in China. Daily mortality and air pollution data were collected from 215 locations during 2014–2017 in China. A two-stage approach was employed to estimate the location-specific and combined cumulative associations between short-term exposure to PM2.5 (lag 0–3 days) and mortality risks. The excess risks (ER) of all-cause, respiratory disease (RESP), cardiovascular disease (CVD), and cerebrovascular disease (CED) mortality for each 10 μg/m3 increment in PM2.5 across all locations were 0.54% (95% confidence interval [CI]: 0.38%, 0.70%), 0.51% (0.10%, 0.93%), 0.74% (0.50%, 0.97%), and 0.52% (0.20%, 0.83%), respectively. Slightly stronger associations for CVD (0.80% versus 0.60%) and CED (0.61% versus 0.26%) mortality were observed in urban cities than in rural areas, and slightly greater associations for RESP mortality (0.51% versus 0.43%) were found in rural areas than in urban cities. A mean of 2.11% (attributable fraction [AF], 95% CI: 1.48%, 2.76%) of all-cause mortality was attributable to PM2.5 exposure in China, with a larger AF in urban cities (2.89% [2.12%, 3.67%]) than in rural areas (0.61% [−0.60%, 1.84%]). Disparities in PM2.5-mortality associations between urban cities and rural areas were also found in some subgroups classified by sex and age. This study provided robust evidence on the associations of PM2.5 with mortality risks in China and demonstrated urban-rural disparities of PM2.5-mortality associations for various causes of death.

Colorectal cancer is an aggressive malignancy for which there are limited treatment options. Oncolytic vaccinia virus is being developed as a novel strategy for cancer therapy. Arming vaccinia virus with immunostimulatory cytokines can enhance the tumor cell-specific replication and antitumor efficacy. Interleukin-24 (IL-24) is an important immune mediator, as well as a broad-spectrum tumor suppressor. We constructed a targeted vaccinia virus of Guang9 strain harboring IL-24 (VG9-IL-24) to evaluate its antitumor effects. In vitro, VG9-IL-24 induced an increased number of apoptotic cells and blocked colorectal cancer cells in the G2/M phase of the cell cycle. VG9-IL-24 induced apoptosis in colorectal cancer cells via multiple apoptotic signaling pathways. In vivo, VG9-IL-24 significantly inhibited the tumor growth and prolonged the survival both in human and murine colorectal cancer models. In addition, VG9-IL-24 stimulated multiple antitumor immune responses and direct bystander antitumor activity. Our results indicate that VG9-IL-24 can inhibit the growth of colorectal cancer tumor by inducing oncolysis and apoptosis as well as stimulating the antitumor immune effects. These findings indicate that VG9-IL-24 may exert a potential therapeutic strategy for combating colorectal cancer.

The Brain Tumor Reporting and Data System (BT-RADS) category 3 is suitable for identifying cases with intermediate probability of tumor recurrence that do not meet the Response Assessment in Neuro-Oncology (RANO) criteria for progression. The aim of this study was to evaluate the added value of dynamic susceptibility contrast-enhanced perfusion-weighted imaging (DSC PWI) and diffusion-weighted imaging (DWI) to BT-RADS for differentiating tumor recurrence from non-recurrence in postoperative high-grade glioma (HGG) patients with category 3 lesions. Patients with BT-RADS category 3 lesions were included. The maximal relative cerebral blood volume (rCBVmax) and the mean apparent diffusion coefficient (ADCmean) values were measured. The added value of DSC PWI and DWI to BT-RADS was evaluated by receiver operating characteristic (ROC) curve analysis. Fifty-one of 91 patients had tumor recurrence, and 40 patients did not. There were significant differences in rCBVmax and ADCmean between the tumor recurrence group and non-recurrence group. Compared to BT-RADS alone, the addition of DSC PWI to BT-RADS increased the area under curve (AUC) from 0.76 (95% confidence interval [CI] 0.66–0.84) to 0.90 (95% CI 0.81–0.95) for differentiating tumor recurrence from non-recurrence. The addition of DWI to BT-RADS increased the AUC from 0.76 (95% CI 0.66–0.84) to 0.88 (95% CI 0.80–0.94). The combination of BT-RADS, DSC PWI, and DWI exhibited the best diagnostic performance (AUC = 0.95; 95% CI 0.88–0.98) for differentiating tumor recurrence from non-recurrence. Adding DSC PWI and DWI to BT-RADS can significantly improve the diagnostic performance for differentiating tumor recurrence from non-recurrence in BT-RADS category 3 lesions.

By taking the dynamic features of the multi-zone heating ventilation air conditioning (HVAC) system into consideration, a distributed iterative learning temperature control (DILTC) method is proposed for building rooms where the exact values of the thermal capacitances, as the key parameters of the HVAC system, are unavailable. At first, an iterative dynamic linearization is applied to reformulate the HVAC system into an all-purposed linear incremental form virtually to facilitate the controller design without relying on model information. On this basis, a data-driven DILTC method is proposed with rigorous convergence analysis where the topology among rooms is considered via a thermal resistance matrix. Under the developed convergence conditions, the tracking error is guaranteed to be iteratively decreased to a small bound with nonrepetitive disturbances. If the disturbances are completely repeatable, a perfect tracking performance can be achieved. The results have also been extended to the HVAC systems subjected to I/O constraints to address the limited heating and cooling capacities. Through extensive simulations, we confirm the good efficiency and applicability of the proposed methods.

Idiopathic rapid eye movement sleep behavior disorder (iRBD) is an important risk factor for α-synucleinopathy. We investigated alterations in the cerebral blood flow (CBF) based on arterial spin-labeled (ASL) imaging in patients with iRBD to determine brain perfusion changes associated with the disorder. Fifteen patients with iRBD and twenty age–gender-matched healthy controls were enrolled. Cortical perfusions were compared between the two groups after the ASL data was co-registered to the high-resolution T1-weighted images. No significant differences were detected between the groups in regard to age, gender, education, or UPDRS-III score. The iRBD group showed a lower MMSE score than the healthy controls (27.07 ± 2.25 vs. 28.55 ± 1.23, p

For the process data analytics, numerous statistical methods are designed to extract informative features to reveal latent characteristics and correlation patterns among the variables. As a classic statistical method, independent component analysis extracts mutually independent components by searching for the maximum direction of non-Gaussianity from the mixed signals. However, the algorithm does not work for the latent Gaussian components, and the disorder of extracted features also brings trouble to the practical applications. In the present work, a slow independent component analysis algorithm is proposed to tackle the problems by concurrently considering the high-order statistic and slowness. It constructs a dual-objective optimization criterion function, simultaneously incorporating independence and slow varying characteristics into the feature extraction procedure, which ties together the ideas of independent component analysis and slow feature analysis under the same mathematical umbrella. Through adjusting the controllable sub-optimization objective weight, the proposed slow independent component analysis adds insight into the different roles of independence and slow varying characteristics in calibration modeling, and, thus, provides possibilities to combine the advantages of independent component analysis and slow feature analysis. Compared with the traditional method, the new algorithm no longer relies solely on higher-order statistic but also utilizes the slow varying characteristic to extract latent Gaussian or non-Gaussian components from mixed signals. Besides the unsupervised paradigm, the proposed algorithm is extended to the supervised form for classification task by utilizing the temporal structure as well as the label information. The effectiveness of the proposed algorithm is illustrated by blind source separation task and industrial fault diagnosis task using real-world datasets including speech signals and multiphase flow facility.

Slow feature analysis (SFA) is an approach that extracts features from a time series dataset in an unsupervised manner based on the temporal slowness principle. These slow features contain relevant information about the dynamics of the process and hence are useful for developing models. For supervised learning objectives, these slow features need to be relevant to the outputs. Partial least squares (PLS) is a method used to perform supervised feature extraction and build models. For time series datasets this approach cannot be used directly as it assumes each sample to be sequentially independent of each other. We propose an approach to perform feature extraction which combines the temporal slowness element of SFA and the output relevance element of PLS. The proposed approach extracts temporally slow features that are relevant to the outputs, which is an essential aspect of supervised learning. The proposed formulations can be solved using the existing PLS algorithms like NIPALS and SIMPLS with proposed modifications. The proposed methods are applied to three case studies: simulated, industrial and experimental case studies to demonstrate their efficacy.

Erythermalgia is a rare cutaneous disease characterized by episodic attacks of burning pain, erythema, and increased temperature. It primarily involves the extremities, with possible extensions to the ears, face, neck, and scrotum; in rare instances, it may afflict the ears, face, or the scrotum alone. Although various medications alone or in different combinations have been tried with significant variations in response, no recommended therapeutics have been established until very recently. This report presents two case histories of a 20-year-old and a 46-year-old woman, respectively, who displayed intermittent facial erythema associated with warmth and pain during the episodes and who presented normal between episodes. Both had good response to combinative treatments of systemic medication and topical lidocaine compounds. The younger was disease-free after more than 4 years of follow-up, and the older had recurrence after stopping the treatment. Conclusion: Erythermalgia may involve the face alone, and combinative approaches may be of choice for its treatment. Topical lidocaine compounds are considered to be a good option for palliative treatment.

Von Neumann (Review of Economics Studies 13(1):1–9, 1945) and Sraffa (Production of commodities by means of commodities. Prelude to a critique of economic theory. Cambridge University Press, Cambridge, 1960) revived the classical ideas of treating fixed capital as a special case of joint production. Sraffa’s model with a single machine that has constant efficiency has been widely generalised. Among such generalisations, Salvadori’s contribution (Value, distribution and capital. Essays in honour of Pierangelo Garegnani, Routledge, London, pp 270–285, 1999) not only re-ignited research interest in the field, but also reshaped scholastic understanding of the importance of machines’ efficiencies with regard to fixed capital models. Such novel insights make it necessary to revisit the development of fixed capital models in the Sraffa framework. In this paper these models are surveyed with a focus on the properties of the cost-minimising technique in each model.

Bitumen from the surface-mined oil sands is separated through a water-based gravity separation process in the Primary Separation Vessels (PSV). The froth-middlings interface level in the PSVs is the most important control variable in the process. Bitumen recovery and downstream operations are critically dependent on the interface level measurement and control. In this paper, we describe a robust computer vision system that uses image and data processing techniques to estimate the froth-middlings interface level. The algorithm in the computer vision system processes the online video frames of a camera mounted to the PSV sight glasses, and it is based on a combination of static and dynamic image processing steps. Industrial results show that the computer vision algorithm is more accurate and reliable when compared to the other instruments, as well as robust against process and environmental abnormalities.

Background The recovery of upper extremity motor deficits represents a functional challenge for persons with stroke. The aim of this study was to evaluate the mechanisms of Unilateral task-oriented training combined with dynamic hand orthosis for the recovery in subacute stroke by using functional near-infrared spectroscopy (fNIRS). Methods This was a pilot randomized, controlled, assessor-blinded trial. A total sample of 30 subacute persons with stroke were randomized into two groups: the orthosis group and the usual exercise group. All participants received four weeks (60 min/day, 5 days/week) of unilateral upper limb training. The orthosis group wore a dynamic hand orthosis during half the training (30 min/day). O2Hb concentrations and a laterality index for bilateral premotor cortex, sensorimotor cortex by fNIRS were calculated as outcome indicators that were measured at baseline and 4 weeks after the interventions. Clinical outcomes were collected for correlation analysis. Results No significant difference in O2Hb concentrations change between two groups. Cortical activation shifted from the ipsilateral to the contralateral premotor cortex and from the contralateral to the ipsilateral sensorimotor cortex in both groups. In the Orthosis group, the change of Fugl-Meyer Assessment of arm score was positively correlated to the shift to contralateral premotor cortex activation and the change of Action Research Arm Test score was positively correlated to the shift to ipsilateral sensorimotor cortex activation. In the UL group, a negative correlation was observed between the shift to ipsilateral sensorimotor cortex and changes of Fugl-Meyer Assessment of arm score. Conclusion This is the first fNIRS study to explore the mechanism of recovery in subacute stroke after utilizing unilateral task-oriented training and a dynamic orthosis. The current results suggest that fNIRS combined with clinical tests may be useful to evaluate the mechanism of stroke recovery after different rehabilitation interventions. Trial registration: Registered on January 11, 2018 at ClinicalTrials.gov (NCT03396939). https://clinicaltrials.gov/ct2/home.

Excessive accumulation of soil heavy metals (HMs) result in the deterioration of soil quality and reduction of agricultural productivity and safety. The accumulation status, temporal change, and sources of soil HMs were determined by large-scale field surveys in 2014 and 2019 in rapid urbanization and industrialization area along the lower reaches of the Yangtze River, China. Eighty-two surface soil samples were collected in 2014 and ninety-five surface soil samples and seven soil profiles (0-100 cm) were collected in 2019. The mean concentrations (in, mg kg-1) of As (10.17), Cd (0.33), Cr (86.38), Cu (38.22), Hg (0.11), Ni (37.67), Pb (43.95), and Zn (113.15) were greater than the corresponding background values. The concentrations of these 8 HMs significantly varied with site-specific distributions depending on nearby landscape patterns with decreasing order: agricultural soil around industrial > agricultural soil > fallow soil. Cd and Hg were found to be priority pollutants due to their greater accumulations in this study area. Combined analyses of principal component analysis and positive matrix factorization model addressed source apportionment of soil HMs. Industrial activities, parent materials, and agricultural and traffic activities were three major sources and their contributions were 35.56%, 35.20%, and 29.23%, respectively. The concentrations of soil As, Cd, Cr and Pb increased with time. This study elucidates how changes in land uses and time affect soil HMs and provides reasonable suggestions for the effective reduction of HM contamination in economically and industrially developed areas of China, and elsewhere.

An ongoing outbreak of severe respiratory illness and pneumonia caused by the severe acute respiratory coronavirus 2 (SARS-CoV-2) commenced in December 2019, and the disease was named as coronavirus disease 2019 (COVID-19). Soon after, scientists identified the characteristics of SARS-CoV-2, including its genome sequence and protein structure. The clinical manifestations of COVID-19 have now been established; and nucleic acid amplification is used for the direct determination of the virus, whereas immunoassays can determine the antibodies against SARS-CoV-2. Clinical trials of several antiviral drugs are ongoing. However, there is still no specific drugs to treat COVID-19. Traditional Chinese medicine (TCM) was used in the treatment of COVID-19 during the early stages of the outbreak in China. Some ancient TCM prescriptions, which were efficacious in the treatment of severe acute respiratory syndrome (SARS) in 2002–03 and the influenza pandemic (H1N1) of 2009, have been improved by experienced TCM practitioners for the treatment of COVID-19 based on their clinical symptoms. These developed new prescriptions include Lianhua Qingwen capsules/granules, Jinhua Qinggan granules and XueBiJing injection, among others. In this review, we have summarized the presenting features of SARS-CoV-2, the clinical characteristics of COVID-19, and the progress in the treatment of COVID-19 using TCMs.

Efficient charge storage media play a pivotal role in transistor‐based memories and thus are under intense research. In this work, the charge storage ability of type‐I InP/ZnS core/shell quantum dots is well revealed through studying a pentacene‐based organic transistor with the quantum dots (QDs) integrated. The quantum well‐like energy band structure enables the QDs to directly confine either holes or electrons in the core, signifying a dielectric layer‐free nonvolatile memory. Especially, the QDs in this device can be charged by electrons using light illumination as the exclusive method. The electron charging process is ascribed to the photoexcitation process in the InP‐core and the hot holes induced. The QDs layer demonstrates an electron storage density of ≈5.0 × 1011 cm−2 and a hole storage density of ≈6.4 × 1011 cm−2. Resultingly, the output device shows a fast response speed to gate voltage (10 µs), large memory window (42 V), good retention (>4.0 × 104 s), and reliable endurance. This work suggests that the core/shell quantum dot as a kind of charge storage medium is of great promise for optoelectronic memories., Type‐I InP/ZnS core/shell quantum dots are implanted into a pentacene organic transistor. The ambipolar charge storage ability and excellent optoelectronic property of the QDs enable the device to be optically programmed and electrically erased, signify a nonvolatile optoelectronic memory. The device demonstrates good retention, endurance, and fast response speed to the electric bias.

In the paper, A polarization-insensitive, passive micro-chipless radio frequency identification tag is proposed. The designed tag is based on nested L-shaped arc-angle resonators and is deployed on a trapezoidal substrate with an area of only . The tag can encode 5 bits of effective data, occupying a bandwidth of 5GHz-10GHz, and can change the encoding sequence by adding or eliminating resonators, thereby providing 1:1 data bits for resonance, and the data density reaches . The structure of data encoding is analyzed for three easily available substrates, which are TaconicTLX-0, TaconicTLX-8 and Rogers RT/Duroid®/5880. The compact tag structure has the ability to generate different IDs, and with its compact size and good robustness, it has become a potential candidate for low-cost and small-volume item-level tagging in domestic, medical and industrial applications of the Internet of Things.

A compact, frequency domain-based, data encoding structure amenable to be designed as a novel chipless RFID tag is proposed. The tag is a tiny trapezoidal lamina on which multiple L-shaped arc-angle resonators of different sizes are deployed to generate distinct one-to-one resonant frequencies corresponding to resonators. The material of the substrate is Rogers RT/Duroid®/5880 suitable for RFID tags. The 5 bits encoding is contained in a tag with OOK coding method and 25 different items can be coded. The angle insensitivity (including polarization insensitivity and incident angle insensitivity) of the tag is verified to ensure that the tag can operate smoothly in different environments. Even more important, with FSK coding, the data capacity and bit density of the tag are doubled, and the number of encoded IDs increases exponentially. Hence, the realized tag with its small size, low cost, high robustness, and flexibility has the potential to replace the optical codes for the perception layer of IoT based smart applications.

Aim To establish a highly sensitive time-resolved fluorescence immunoassay of heparin-binding protein (HBP-TRFIA) and evaluate its application value for bacterial or fungal infections in tumor patients. Methods Two types of HBP monoclonal specific antibodies against different epitopes of the antigen molecule were used as coating antibodies and Eu3+-labeled antibodies, respectively. The double-antibody sandwich method was used in establishing HBP-TRFIA, and the methodology was evaluated. The established HBP-TRFIA was used in detecting HBP concentration in the plasma samples of healthy individuals, patients with bacterial or fungal infections, and infected or uninfected patients with various types of tumors. Results The linear range of HBP-TRFIA was (0.11–530 ng/mL). Plasma HBP concentrations detected through HBP-TRFIA were consistent with the results of fluorescence quantitative immunochromatography (ρ = 0.964). The plasma HBP concentrations of infected tumor patients were significantly higher than those of uninfected tumor patients (P Conclusion This study successfully established a highly sensitive HBP-TRFIA, which was highly comparable to commercially available fluorescent quantitative immunochromatographic kits and was able to facilitate the timely diagnosis of bacterial or fungal infections in patients with tumor.

This study aims to explore the freight demand network spatial patterns in six provinces of central China from the perspective of the spread of the epidemic and the freight imbalance and breakout. To achieve this purpose, the big data of “cart search” demand information provided by small and medium freight enterprises on the freight information platform are analyzed. 343,690 pieces of freight demand big data on the freight information platform and Python, ArcGIS, UCINET, and Gephi software are used. The results show that: (1) The choke-point of unbalanced freight demand network is Wuhan, and the secondary choke-points are Hefei and Zhengzhou. (2) In southern China, a chain reaction circle of freight imbalance is formed with Wuhan, Hefei, and Nanchang as the centers. In northern China, a chain reaction circle of freight imbalance is formed with Zhengzhou and Taiyuan as the centers. (3) The freight demand of the six provinces in central China exhibits typical characteristics of long tail distribution with large span and unbalanced distribution. (4) The import and export of freight in different cities vary greatly, and the distribution is unbalanced. This study indicates the imbalance difference, chain reaction, keys and hidden troubles posed by the freight demand network. From the perspectives of freight transfer breakout, freight balance breakout, freight strength breakout, and breakout of freight periphery cities, we propose solutions to breakouts in the freight market in six provinces of central China in the post-epidemic era.