Search

Diffusion weighted magnetic resonance imaging (DW-MRI) captures tissue microarchitecture at millimeter scale. With recent advantages in data sharing, large-scale multi-site DW-MRI datasets are being made available for multi-site studies. However, DW-MRI suffers from measurement variability (e.g., inter- and intra-site variability, hardware performance, and sequence design), which consequently yields inferior performance on multi-site and/or longitudinal diffusion studies. In this study, we propose a novel, deep learning-based method to harmonize DW-MRI signals for a more reproducible and robust estimation of microstructure. Our method introduces a data-driven scanner-invariant regularization scheme to model a more robust fiber orientation distribution function (FODF) estimation. We study the Human Connectome Project (HCP) young adults test-retest group as well as the MASiVar dataset (with inter- and intra-site scan/rescan data). The 8(th) order spherical harmonics coefficients are employed as data representation. The results show that the proposed harmonization approach maintains higher angular correlation coefficients (ACC) with the ground truth signals (0.954 versus 0.942), while achieves higher consistency of FODF signals for intra-scanner data (0.891 versus 0.826), as compared with the baseline supervised deep learning scheme. Furthermore, the proposed data-driven framework is flexible and potentially applicable to a wider range of data harmonization problems in neuroimaging.

Multi-class cell segmentation in high-resolution Giga-pixel whole slide images (WSI) is critical for various clinical applications. Training such an AI model typically requires labor-intensive pixel-wise manual annotation from experienced domain experts (e.g., pathologists). Moreover, such annotation is error-prone when differentiating fine-grained cell types (e.g., podocyte and mesangial cells) via the naked human eye. In this study, we assess the feasibility of democratizing pathological AI deployment by only using lay annotators (annotators without medical domain knowledge). The contribution of this paper is threefold: (1) We proposed a molecular-empowered learning scheme for multi-class cell segmentation using partial labels from lay annotators; (2) The proposed method integrated Giga-pixel level molecular-morphology cross-modality registration, molecular-informed annotation, and molecular-oriented segmentation model, so as to achieve significantly superior performance via 3 lay annotators as compared with 2 experienced pathologists; (3) A deep corrective learning (learning with imperfect label) method is proposed to further improve the segmentation performance using partially annotated noisy data. From the experimental results, our learning method achieved F1 = 0.8496 using molecular-informed annotations from lay annotators, which is better than conventional morphology-based annotations (F1 = 0.7051) from experienced pathologists. Our method democratizes the development of a pathological segmentation deep model to the lay annotator level, which consequently scales up the learning process similar to a non-medical computer vision task. The official implementation and cell annotations are publicly available at https://github.com/hrlblab/MolecularEL.

Diffusion-weighted magnetic resonance imaging (DW-MRI) is a critical imaging method for capturing and modeling tissue microarchitecture at a millimeter scale. A common practice to model the measured DW-MRI signal is via fiber orientation distribution function (fODF). This function is the essential first step for the downstream tractography and connectivity analyses. With recent advantages in data sharing, large-scale multi-site DW-MRI datasets are being made available for multi-site studies. However, measurement variabilities (e.g., inter- and intra-site variability, hardware performance, and sequence design) are inevitable during the acquisition of DW-MRI. Most existing model-based methods (e.g., constrained spherical deconvolution (CSD)) and learning based methods (e.g., deep learning (DL)) do not explicitly consider such variabilities in fODF modeling, which consequently leads to inferior performance on multi-site and/or longitudinal diffusion studies. In this paper, we propose a novel data-driven deep constrained spherical deconvolution method to explicitly constrain the scan-rescan variabilities for a more reproducible and robust estimation of brain microstructure from repeated DW-MRI scans. Specifically, the proposed method introduces a new 3D volumetric scanner-invariant regularization scheme during the fODF estimation. We study the Human Connectome Project (HCP) young adults test-retest group as well as the MASiVar dataset (with inter- and intra-site scan/rescan data). The Baltimore Longitudinal Study of Aging (BLSA) dataset is employed for external validation. From the experimental results, the proposed data-driven framework outperforms the existing benchmarks in repeated fODF estimation. The proposed method is assessing the downstream connectivity analysis and shows increased performance in distinguishing subjects with different biomarkers., Comment: 33 pages, 7 figures

Sensor fusion is a crucial augmentation technique for improving the accuracy and reliability of perception systems for automated vehicles under diverse driving conditions. However, adverse weather and low-light conditions remain challenging, where sensor performance degrades significantly, exposing vehicle safety to potential risks. Advanced sensors such as LiDARs can help mitigate the issue but with extremely high marginal costs. In this paper, we propose a novel transformer-based 3D object detection model "REDFormer" to tackle low visibility conditions, exploiting the power of a more practical and cost-effective solution by leveraging bird's-eye-view camera-radar fusion. Using the nuScenes dataset with multi-radar point clouds, weather information, and time-of-day data, our model outperforms state-of-the-art (SOTA) models on classification and detection accuracy. Finally, we provide extensive ablation studies of each model component on their contributions to address the above-mentioned challenges. Particularly, it is shown in the experiments that our model achieves a significant performance improvement over the baseline model in low-visibility scenarios, specifically exhibiting a 31.31% increase in rainy scenes and a 46.99% enhancement in nighttime scenes.The source code of this study is publicly available.

Additional file 1. Phantom evaluation.

Many anomaly detection approaches, especially deep learning methods, have been recently developed to identify abnormal image morphology by only employing normal images during training. Unfortunately, many prior anomaly detection methods were optimized for a specific "known" abnormality (e.g., brain tumor, bone fraction, cell types). Moreover, even though only the normal images were used in the training process, the abnormal images were oftenly employed during the validation process (e.g., epoch selection, hyper-parameter tuning), which might leak the supposed ``unknown" abnormality unintentionally. In this study, we investigated these two essential aspects regarding universal anomaly detection in medical images by (1) comparing various anomaly detection methods across four medical datasets, (2) investigating the inevitable but often neglected issues on how to unbiasedly select the optimal anomaly detection model during the validation phase using only normal images, and (3) proposing a simple decision-level ensemble method to leverage the advantage of different kinds of anomaly detection without knowing the abnormality. The results of our experiments indicate that none of the evaluated methods consistently achieved the best performance across all datasets. Our proposed method enhanced the robustness of performance in general (average AUC 0.956).

Supplemental material, sj-docx-1-uar-10.1177_10780874231169920 for Leadership Transfer Networks and Regional Environmental Governance Performance by Can Cui, Wenna Chen and Hongtao Yi in Urban Affairs Review

Supplemental material, sj-docx-1-uar-10.1177_10780874231169920 for Leadership Transfer Networks and Regional Environmental Governance Performance by Can Cui, Wenna Chen and Hongtao Yi in Urban Affairs Review

Deep learning has made great strides in medical imaging, enabled by hardware advances in GPUs. One major constraint for the development of new models has been the saturation of GPU memory resources during training. This is especially true in computational pathology, where images regularly contain more than 1 billion pixels. These pathological images are traditionally divided into small patches to enable deep learning due to hardware limitations. In this work, we explore whether the shared GPU/CPU memory architecture on the M1 Ultra systems-on-a-chip (SoCs) recently released by Apple, Inc. may provide a solution. These affordable systems (less than \$5000) provide access to 128 GB of unified memory (Mac Studio with M1 Ultra SoC). As a proof of concept for gigapixel deep learning, we identified tissue from background on gigapixel areas from whole slide images (WSIs). The model was a modified U-Net (4492 parameters) leveraging large kernels and high stride. The M1 Ultra SoC was able to train the model directly on gigapixel images (16000$\times$64000 pixels, 1.024 billion pixels) with a batch size of 1 using over 100 GB of unified memory for the process at an average speed of 1 minute and 21 seconds per batch with Tensorflow 2/Keras. As expected, the model converged with a high Dice score of 0.989 $\pm$ 0.005. Training up until this point took 111 hours and 24 minutes over 4940 steps. Other high RAM GPUs like the NVIDIA A100 (largest commercially accessible at 80 GB, $\sim$\$15000) are not yet widely available (in preview for select regions on Amazon Web Services at \$40.96/hour as a group of 8). This study is a promising step towards WSI-wise end-to-end deep learning with prevalent network architectures.

The segment anything model (SAM) was released as a foundation model for image segmentation. The promptable segmentation model was trained by over 1 billion masks on 11M licensed and privacy-respecting images. The model supports zero-shot image segmentation with various segmentation prompts (e.g., points, boxes, masks). It makes the SAM attractive for medical image analysis, especially for digital pathology where the training data are rare. In this study, we evaluate the zero-shot segmentation performance of SAM model on representative segmentation tasks on whole slide imaging (WSI), including (1) tumor segmentation, (2) non-tumor tissue segmentation, (3) cell nuclei segmentation. Core Results: The results suggest that the zero-shot SAM model achieves remarkable segmentation performance for large connected objects. However, it does not consistently achieve satisfying performance for dense instance object segmentation, even with 20 prompts (clicks/boxes) on each image. We also summarized the identified limitations for digital pathology: (1) image resolution, (2) multiple scales, (3) prompt selection, and (4) model fine-tuning. In the future, the few-shot fine-tuning with images from downstream pathological segmentation tasks might help the model to achieve better performance in dense object segmentation.

EEG and behavioral data and pre-registration plan

The development of city–industry integration is crucial for modern cities and is a core element of city competitiveness enhancement and sustainable development. This study considers system dynamics theory to examine city–industry integration and constructs an index system to measure the degree of integration. For this purpose, 31 regions in China (including provinces, autonomous regions, and municipalities directly under the central government) are considered as research samples. Objective evaluation methods such as factor analysis and entropy methods are applied to evaluate the target value. The research results reveal a wide gap in the levels of city–industry integration in various regions of China. Furthermore, the Middle East outperforms the Western and Northeastern regions. Accordingly, the advantages of the Central and Eastern regions should be combined, and a leading and radiation-driven role should be played. Moreover, capital investment in the Western and Northeastern regions should be increased, and emphasis should be placed on local characteristics. Moreover, urban economic development, industrial transformation, and industrial upgrading should be promoted, and the sustainable development capacity of cities should be enhanced.

A severely comatose female patient was diagnosed with Japanese encephalitis (JE). Her condition was complicated by Hashimoto’s thyroiditis (HT) and Guillain-Barré syndrome (GBS). After antiviral, glucocorticoid, and immunoglobulin treatment, the patient’s consciousness was restored, and she could breathe spontaneously. Following this, new-onset, primarily demyelinating GBS developed, which progressed to demyelination combined with axonal injury. The patient was switched to protein A immunoadsorption (PAIA) therapy, and her Hughes score decreased rapidly, from 4 to 1 after 6 months. This patient is the first to receive PAIA combined with an antiviral-glucocorticoid-immunoglobulin regimen to treat encephalitis, meningitis, HT, and GBS caused by JE infection, thereby reflecting the importance of clinical application of PAIA in the treatment of immunological complications of JE.

The rapid development of diagnostic technologies in healthcare is leading to higher requirements for physicians to handle and integrate the heterogeneous, yet complementary data that are produced during routine practice. For instance, the personalized diagnosis and treatment planning for a single cancer patient relies on the various images (e.g., radiological, pathological, and camera images) and non-image data (e.g., clinical data and genomic data). However, such decision-making procedures can be subjective, qualitative, and have large inter-subject variabilities. With the recent advances in multi-modal deep learning technologies, an increasingly large number of efforts have been devoted to a key question: how do we extract and aggregate multi-modal information to ultimately provide more objective, quantitative computer-aided clinical decision making? This paper reviews the recent studies on dealing with such a question. Briefly, this review will include the (1) overview of current multi-modal learning workflows, (2) summarization of multi-modal fusion methods, (3) discussion of the performance, (4) applications in disease diagnosis and prognosis, and (5) challenges and future directions.

Additional file 2: Fig. S2. GO analysis of differentially expressed miRNA target genes.

Documentation of searchstrategies.

Automatic methods to segment the vestibular schwannoma (VS) tumors and the cochlea from magnetic resonance imaging (MRI) are critical to VS treatment planning. Although supervised methods have achieved satisfactory performance in VS segmentation, they require full annotations by experts, which is laborious and time-consuming. In this work, we aim to tackle the VS and cochlea segmentation problem in an unsupervised domain adaptation setting. Our proposed method leverages both the image-level domain alignment to minimize the domain divergence and semi-supervised training to further boost the performance. Furthermore, we propose to fuse the labels predicted from multiple models via noisy label correction. In the MICCAI 2021 crossMoDA challenge, our results on the final evaluation leaderboard showed that our proposed method has achieved promising segmentation performance with mean dice score of 79.9% and 82.5% and ASSD of 1.29 mm and 0.18 mm for VS tumor and cochlea, respectively. The cochlea ASSD achieved by our method has outperformed all other competing methods as well as the supervised nnU-Net., Comment: Accepted by MICCAI 2021 BrainLes Workshop. arXiv admin note: substantial text overlap with arXiv:2109.06274

Integrating cross-department multi-modal data (e.g., radiological, pathological, genomic, and clinical data) is ubiquitous in brain cancer diagnosis and survival prediction. To date, such an integration is typically conducted by human physicians (and panels of experts), which can be subjective and semi-quantitative. Recent advances in multi-modal deep learning, however, have opened a door to leverage such a process to a more objective and quantitative manner. Unfortunately, the prior arts of using four modalities on brain cancer survival prediction are limited by a "complete modalities" setting (i.e., with all modalities available). Thus, there are still open questions on how to effectively predict brain cancer survival from the incomplete radiological, pathological, genomic, and demographic data (e.g., one or more modalities might not be collected for a patient). For instance, should we use both complete and incomplete data, and more importantly, how to use those data? To answer the preceding questions, we generalize the multi-modal learning on cross-department multi-modal data to a missing data setting. Our contribution is three-fold: 1) We introduce optimal multi-modal learning with missing data (MMD) pipeline with optimized hardware consumption and computational efficiency; 2) We extend multi-modal learning on radiological, pathological, genomic, and demographic data into missing data scenarios; 3) a large-scale public dataset (with 962 patients) is collected to systematically evaluate glioma tumor survival prediction using four modalities. The proposed method improved the C-index of survival prediction from 0.7624 to 0.8053.

Additional file 1: Fig. S1. miRNA expression TPM density distribution map.

Multiple Sclerosis (MS) is a chronic neuroinflammatory disease and multi-modality MRIs are routinely used to monitor MS lesions. Many automatic MS lesion segmentation models have been developed and have reached human-level performance. However, most established methods assume the MRI modalities used during training are also available during testing, which is not guaranteed in clinical practice. Previously, a training strategy termed Modality Dropout (ModDrop) has been applied to MS lesion segmentation to achieve the state-of-the-art performance with missing modality. In this paper, we present a novel method dubbed ModDrop++ to train a unified network adaptive to an arbitrary number of input MRI sequences. ModDrop++ upgrades the main idea of ModDrop in two key ways. First, we devise a plug-and-play dynamic head and adopt a filter scaling strategy to improve the expressiveness of the network. Second, we design a co-training strategy to leverage the intra-subject relation between full modality and missing modality. Specifically, the intra-subject co-training strategy aims to guide the dynamic head to generate similar feature representations between the full- and missing-modality data from the same subject. We use two public MS datasets to show the superiority of ModDrop++. Source code and trained models are available at https://github.com/han-liu/ModDropPlusPlus., Comment: MICCAI 2022

In this paper, porous NiO with three-dimensional (3D) nanostructures are prepared by a chemical bath deposition (CBD) process in oxalic acid solution at a fixed temperature using Ni foam (NF) as the substrate and Ni source, and followed by an electrochemical activation treatment in an alkaline medium. The special 3D porous structure of the as-prepared NF based NiO (NiO/NF) electrode is benefit for the diffusion of fuels (CH3OH/C2H5OH and urea) during the electrochemical reaction. Besides, the obtained NiO may combine with the remained nickel oxalate (NiC2O4) to further enhance the catalytic performance due to its synergistic effect. At 0.60 V, the electrooxidation current density reached 257 mA cm−2 in 1.0 mol dm−3 NaOH and 0.40 mol dm−3 CH3OH, and a high performance of 120 mA cm−2 was achieved in 1.0 mol dm−3 NaOH and 0.10 mol dm−3 C2H5OH. The oxidation current density reached 222 mA cm−2 in 0.1 mol dm−3 urea with the NaOH concentration of 8.0 mol dm−3 at around 0.48 V. The facile preparation method, high electrocatalytic activity, low cost and various applications make the NiO/NF be a promising electrode material.

In recent years, the construction and innovation of distribution network have been accelerating, and the overall investment in distribution network has been increasing year by year. Therefore, the investment efficiency of distribution network has become a key concern. At present, the evaluation system of distribution network often focuses on the technical aspects such as planning effect, safety and reliability, while lacks the evaluation of investment efficiency. In terms of investment decision, it usually makes a simple proportional allocation according to historical data. To solve the problems of the investment decision in distribution network construction, this paper puts forward an multi-factorial investment decision model, which is guided by the investment efficiency evaluation while satisfying the demand of distribution network. Firstly, the index system of input-output assessment, scoring function and investment benefit score evaluation model of regional distribution network are proposed. Secondly, the investment decision-making model based on genetic algorithm is proposed, which coordinate the balance between economy and safety of distribution network. Finally, a regional distribution network is taken as an example to test and verify the scientific usefulness of the model. The practical application results show that the multi-factorial investment decision model can promote the development and improvement of the distribution network more reasonably and economical efficiently.

As the construction sector is a major energy consumer and thus a significant contributor of CO2 emissions in China, it is important to consider carbon reduction in this industry. This study analyzed six life-cycle stages and calculated the life-cycle CO2 emissions of the construction sector in 30 Chinese provincial jurisdictions to understand the disparity among them. Results show that building materials production was the key stage for carbon reduction in the construction sector, followed by the building operation stage. External variables, e.g., economic growth, industrial structure, urbanization, price fluctuation, and marketization, were significantly correlated with the emission intensity of the construction sector. Specifically, economic growth exhibited an inverted U-shaped relation with CO2 emissions per capita and per area during the period examined. Secondary industry and land urbanization were negatively correlated with CO2 emission intensity indicators from the construction sector, whereas tertiary industry and urbanization were positively correlated. Price indices and marketization had negative effects on CO2 emission intensity. The policy implications of our findings are that cleaner technologies should be encouraged for cement providers, and green purchasing rules for the construction sector should also be established. Pricing tools (e.g., resource taxes) could help to adjust the demand for raw materials and energy.

In the era of big data and artificial intelligence development, data has become an important asset. Open data ecology has become an inevitable choice to promote business success. In the process of data open sharing, data provenance technology can supervise the data sharing path among multiple entities, and protect the interests of data owners. This paper proposes a data provenance technology for dispatching and control data based on blockchain. Firstly, a data provenance model is designed based on blockchain, which describes the provenance process of dispatching and control model data.Secondly, the distributed ledger of blockchain is used to record the data sharing path described by the data traceability model.Finally, the automatic implementation of provenance process is realized by using smart contract of blockchain.

Additional file 2: Supplementary file 2. Interview guide for the questionnaire in this survey.

Computer-assisted quantitative analysis on Giga-pixel pathology images has provided a new avenue in histology examination. The innovations have been largely focused on cancer pathology (i.e., tumor segmentation and characterization). In non-cancer pathology, the learning algorithms can be asked to examine more comprehensive tissue types simultaneously, as a multi-label setting. The prior arts typically needed to train multiple segmentation networks in order to match the domain-specific knowledge for heterogeneous tissue types (e.g., glomerular tuft, glomerular unit, proximal tubular, distal tubular, peritubular capillaries, and arteries). In this paper, we propose a dynamic single segmentation network (Omni-Seg) that learns to segment multiple tissue types using partially labeled images (i.e., only one tissue type is labeled for each training image) for renal pathology. By learning from ~150,000 patch-wise pathological images from six tissue types, the proposed Omni-Seg network achieved superior segmentation accuracy and less resource consumption when compared to the previous the multiple-network and multi-head design. In the testing stage, the proposed method obtains "completely labeled" tissue segmentation results using only "partially labeled" training images. The source code is available at https://github.com/ddrrnn123/Omni-Seg

Optical coherence tomography (OCT) is a non-invasive imaging technique widely used for ophthalmology. It can be extended to OCT angiography (OCT-A), which reveals the retinal vasculature with improved contrast. Recent deep learning algorithms produced promising vascular segmentation results; however, 3D retinal vessel segmentation remains difficult due to the lack of manually annotated training data. We propose a learning-based method that is only supervised by a self-synthesized modality named local intensity fusion (LIF). LIF is a capillary-enhanced volume computed directly from the input OCT-A. We then construct the local intensity fusion encoder (LIFE) to map a given OCT-A volume and its LIF counterpart to a shared latent space. The latent space of LIFE has the same dimensions as the input data and it contains features common to both modalities. By binarizing this latent space, we obtain a volumetric vessel segmentation. Our method is evaluated in a human fovea OCT-A and three zebrafish OCT-A volumes with manual labels. It yields a Dice score of 0.7736 on human data and 0.8594 +/- 0.0275 on zebrafish data, a dramatic improvement over existing unsupervised algorithms., Comment: Accepted by International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2021

In neuromorphic computing, artificial synapses provide a multi-weight conductance state that is set based on inputs from neurons, analogous to the brain. Additional properties of the synapse beyond multiple weights can be needed, and can depend on the application, requiring the need for generating different synapse behaviors from the same materials. Here, we measure artificial synapses based on magnetic materials that use a magnetic tunnel junction and a magnetic domain wall. By fabricating lithographic notches in a domain wall track underneath a single magnetic tunnel junction, we achieve 4-5 stable resistance states that can be repeatably controlled electrically using spin orbit torque. We analyze the effect of geometry on the synapse behavior, showing that a trapezoidal device has asymmetric weight updates with high controllability, while a straight device has higher stochasticity, but with stable resistance levels. The device data is input into neuromorphic computing simulators to show the usefulness of application-specific synaptic functions. Implementing an artificial neural network applied on streamed Fashion-MNIST data, we show that the trapezoidal magnetic synapse can be used as a metaplastic function for efficient online learning. Implementing a convolutional neural network for CIFAR-100 image recognition, we show that the straight magnetic synapse achieves near-ideal inference accuracy, due to the stability of its resistance levels. This work shows multi-weight magnetic synapses are a feasible technology for neuromorphic computing and provides design guidelines for emerging artificial synapse technologies., Comment: 27 pages 6 figures 1 table

Ki67 is a significant biomarker in the diagnosis and prognosis of cancer, whose index can be evaluated by quantifying its expression in Ki67 immunohistochemistry (IHC) stained images. However, quantitative analysis on multi-source Ki67 images is yet a challenging task in practice due to cross-domain distribution differences, which result from imaging variation, staining styles, and lesion types. Many recent studies have made some efforts on domain generalization (DG), whereas there are still some noteworthy limitations. Specifically in the case of Ki67 images, learning invariant representation is at the mercy of the insufficient number of domains and the cell categories mismatching in different domains. In this paper, we propose a novel method to improve DG by searching the domain-agnostic subnetwork in a domain merging scenario. Partial model parameters are iteratively pruned according to the domain gap, which is caused by the data converting from a single domain into merged domains during training. In addition, the model is optimized by fine-tuning on merged domains to eliminate the interference of class mismatching among various domains. Furthermore, an appropriate implementation is attained by applying the pruning method to different parts of the framework. Compared with known DG methods, our method yields excellent performance in multiclass nucleus recognition of Ki67 IHC images, especially in the lost category cases. Moreover, our competitive results are also evaluated on the public dataset over the state-of-the-art DG methods., Comment: Accepted by MICCAI2021

Additional file 1: Supplementary file 1. English version of questionnaire in this survey, including informed consent, demographic characteristics, GAD-7, GSE and RS-14.

With the continuous development of container technology, the scale of dispatching control cloud is expanding, and the relationship between container applications on cloud platform is becoming more and more complex. The intelligent probe technology is used to detect the health status of the container to ensure the normal operation of the container application, thus ensuring the safe and stable operation of the dispatching control cloud platform, and providing strong support for the scheduling of the smart grid. In this paper, the health status of containerized application is detected by the container probe detection technology. The results show that the container probe detection technology has an obvious role in prolonging the life cycle of containerized application.

This paper presents a cascade utilization method of multi-grade energy based on thermal coupling analysis. This paper analyzes the principle of energy cascade utilization, constructs the cascade utilization energy flow structure of electrical thermal coupling transformation, and analyzes the utilization form and energy grade rise and fall method of electric power, steam, high temperature hot water, medium temperature hot water and low temperature hot water from the Angle of energy cascade utilization. Based on the operation constraints of the energy cascade optimization utilization model, the optimized economic operation of the electrical thermal micro-grid system is realized by coordinating and scheduling the coupling equipment of each energy flow. The method takes into full account the coupling and complementation of various forms of energy and the principle of multi-grade cascade optimization, and makes full use of the energy of different grades in the production process to realize the energy efficiency optimization of the system.

Data processing and analytics are fundamental and pervasive. Algorithms play a vital role in data processing and analytics where many algorithm designs have incorporated heuristics and general rules from human knowledge and experience to improve their effectiveness. Recently, reinforcement learning, deep reinforcement learning (DRL) in particular, is increasingly explored and exploited in many areas because it can learn better strategies in complicated environments it is interacting with than statically designed algorithms. Motivated by this trend, we provide a comprehensive review of recent works focusing on utilizing DRL to improve data processing and analytics. First, we present an introduction to key concepts, theories, and methods in DRL. Next, we discuss DRL deployment on database systems, facilitating data processing and analytics in various aspects, including data organization, scheduling, tuning, and indexing. Then, we survey the application of DRL in data processing and analytics, ranging from data preparation, natural language processing to healthcare, fintech, etc. Finally, we discuss important open challenges and future research directions of using DRL in data processing and analytics., Comment: 39 pages, 3 figures and 3 tables

In order to further solve the multiple pressures of economy, energy and environment, and improve the utilization efficiency of energy, Integrated energy system has become an important way of energy utilization. The difference of users' energy demand makes the connection between electricity, heat, cold and gas system more close. Accurate energy demand forecasting will be the theoretical and data basis for integrated regional energy system design and energy efficiency assessment. Thus, a regional Integrated energy load forecasting method based on the CNN-LSTM model with user energy label differentiation is proposed in this paper. Analysis based on differential users can use behavior and make the user use the label, and then based on the energy type, weather conditions, such as time period 3 users can use the label as input data, CNN - LSTM model for load forecasting and verify, in which CNN extracts effective input features and LSTM is good at processing time-series data. According to the results based on user can use the label differentiation of CNN - LSTM model at the regional level of comprehensive power load forecasting has good results.

Neural networks have made significant achievements in the field of image restoration. To efficiently repair facial images with large areas damaged, a decoder-encoder structured convolutional neural network is used as a generative model and skip-connection is added between some of its layers to enhance the structure prediction ability of the generated model and well suppressed the problem that the repair network is easy to over-fitting. The global discrimination network mostly uses the image’s edge structure and feature information to ensure that the repaired image, which is the output from the repair network, conforms to visual connectivity, while the local discriminators, not only recognize local consistency but also optimize more details. The network structure proposed in this paper combines the encoder-decoder, skip-connection, and dual discriminator networks to improve the effect of face completion. The experimental results on the CelebA show that the proposed method is superior to other methods in repairing images with large areas of damage.

With the wide spread of power quality monitors in smart grid, information-oriented use of power quality disturbances has become a concerned research area. In this paper, a two-step strategy classification system for various voltage sags is developed. A typical classification system is composed of one selector of discriminative feature and one classifier which the former is fundamental and critical. Further research on classification ability of physical features with respect to a given classification task is necessary. By studying the intrinsic relations between features and classification task, an automatic feature selection based classification system is presented. The main advantages of the proposed method are its ability to select features highly related to the given classification task, and its ability to reduce the dimensionality of the feature set simultaneously. In order to validate the proposed method, three common voltage sags causes, namely grid fault, large motor starting and transformer energizing, are considered. Simulating and practical recorded data are tested. Experimental results have shown that the proposed method is accurate, time-saving, and robust.

Micro simulation asynchronous motor unit is used to simulate the real asynchronous motor load, which is an important component of the micro dynamic simulation laboratory of power system. The real-time and accuracy of monitoring and control of micro simulation asynchronous motor will affect the results of micro dynamic simulation experiments directly. This paper designs and implements an integrated control instrument for micro simulation asynchronous motor unit (ICIMSAM). ICIMSAM consists of data communication unit, power amplifier unit, data processing unit, power supply unit, operation execution unit and liquid crystal display unit, realizing the functions of real-time running state monitoring, automatic starting/stopping and automatic power regulation for the motor unit. Light synchronized phasor measurement unit (PMU Light) is employed to get the real-time operation information of the micro simulation asynchronous motor unit. Real-time operation information includes three-phase frequency, voltage amplitude, voltage phasor, current amplitude, current phasor, which would be sent from the PMU Light to ICIMSG through the ethernet network. Appropriate software was developed on server to communicate with ICIMSG in a two-way manner through the network. Four group of micro dynamic simulation experiments have been completed to test the functions of ICIMSAM.

Whether the design scheme of the distributed energy system is reasonable, it is necessary to comprehensively consider the benefits of various aspects for evaluation, and introduce appropriate evaluation methods. This paper comprehensively considers the economic, energy and environmental factors of distributed energy system to construct an evaluation index system. Based on the measurement mathematics method, the subjective and objective weighting methods are combined to make up for the deficiency of using the empowerment alone, and a more scientific and reasonable AHP-entropy is proposed. The power empowerment method establishes a comprehensive benefit evaluation model for distributed energy systems. According to the calculated closeness of the scheme, considering the economic, energy and environmental factors affecting the distributed energy system, giving the advantages and disadvantages of each scheme, it has very important reference significance for realizing the economic operation of the distributed energy system.

Additional file 1. Supplemental Table 1. Patient/lesion characteristics of the cohort and sub-cohort (HCC and metastasis). Supplemental Table 2. Spearman correlation coefficients between lesion-level overall response and all radiomics features/absorbed dose with corresponding p-values (with Bonferroni correction). Univariate Cox regression with c-index, hazard ratio and corresponding p-values for progression are also indicated. Supplemental Table 3. Mean CCC for 5 repeat scans of the liver phantom, OS-EM iterations 1/2, with/without Gaussian filtering and across all conditions. Supplemental Fig. 1 Radiomics_all+dose model order determination for OR (left) and PFS (right). Average AUC/c-index vs. number of top features included. When using all the radiomics features, the average model order calculated using nested cross validation is 2 for OR classification and 3 for PFS, with top 2 features being variance and absorbed dose and top 3 features being variance, absorbed dose and LRHGE, respectively. The nested CV AUCs for radiomics_all+dose is 0.672 (0.620-0.716) for OR and 0.791 (95%CI: 0.740-0.825) for progression. The top 5 features are shown in supplemental table 4 for OR and progression models. Supplemental Table 4. Top 5 features for the combined models with all radiomics features, volume and absorbed dose.

Additional file 1: Table S1. Baseline characteristics between plaque rupture and plaque erosion. Table S2. Presictive models of PR established using ARPG/GVT single or combined with traditional risk factors. Table S3. The correlation between glucose-related variables and plaque morphological characteristics using Spearman correlation.

Objective: To assess the associations of 43 autoimmune diseases with the subsequent risk of ALS and further evaluate the contribution of familial confounding to these associations. Methods: We conducted a nationwide register-based nested case-control study including 3561 ALS patients diagnosed during 1990–2013 in Sweden and 35,610 controls that were randomly selected from the general population and individually matched to the cases on age, sex, and county of birth. To evaluate the contribution of familial factors on the studied association, we additionally studied the first-degree relatives (siblings and children) of ALS patients and their controls. Results: Patients with ALS had a 47% higher risk of being previously diagnosed with autoimmune disease (OR 1.47, 95% confidence interval [CI] 1.31–1.64), compared with controls. A positive association was noted for several autoimmune diseases, including myasthenia gravis, polymyositis or dermatomyositis, Guillain-Barre syndrome, type 1 diabetes diagnosed younger than 30 years, multiple sclerosis, and hypothyreosis. The increased risk of any autoimmune disease was greatest during the year before ALS diagnosis, likely due to misdiagnosis. A statistically significantly increased risk was also noted during 2–5 years, but not earlier, before ALS diagnosis. First-degree relatives of ALS patients had however no increased risk of autoimmune diseases compared with first-degree relatives of controls. Conclusions: Although it is difficult to completely remove the potential effects of misdiagnosis, there is likely a positive association between autoimmune disease (such as type 1 diabetes and multiple sclerosis) and ALS, which is not fully explained by shared familial confounding factors.

Computer-generated multi-view (MV) image for the three-dimensional (3D) display with lenticular lens arrays based on the backward ray-tracing technique is presented. To overcome the limitations of the traditional ray-tracing technique, whose minimum operating unit is the pixel, the proposed method can be applied at the sub-pixel level. The quality of the generated 3D image is not reduced, in contrast to existing methods for converting 2D images to 3D. Compared to the previous each camera viewpoint independent rendering method, the backward ray-tracing technique directly renders the final MV image rather than all the single-viewpoint images. Accordingly, the total render time is substantially reduced. Experimental results show that our proposed method significantly increases the simplicity and efficiency of the rendering process, especially ensuring the resolution of the final MV 3D image. Experimental results for 3D display with the generated 32-viewpoint 3D images are presented.

A three-electrode electrochemical study concerning electrically conductive materials submitted to supercritical fluids above 550 °C has not been reported in spite of its significance in many related fields. In this work, the design of a novel experimental device allowing three-electrode electrochemical measurements in high temperature and pressure fluids up to 700 °C/1000 bar is detailed described. In this device, pyrophyllite is used as a new electrode sealing material and the electrodes are sealed using a conical self-energizing sealing structure, which ensures the water tightness of the electrodes at pressures up to 1000 bar. The reference electrode in this device can either be an Ag/AgCl external pressure balanced reference electrode (EPBRE) or a platinum reference electrode. The potentiodynamic polarization plots and electrochemical impedance spectroscopy (EIS) of nickel-based superalloy Inconel 740H, a candidate material for Generation IV nuclear reactors, in deaerated water and chloride containing fluid (600–700 °C, 600–1000 bar) are presented using this device.

Cobalt has been extensively studied as an effective catalyst for the electrooxidation of sodium borohydride. However, there is still the problem of the low catalytic activity of cobalt catalysts. The improvement in the catalytic activity of cobalt for the electrooxidation of sodium borohydride has become a challenge. In this paper, a novel catalyst Co–P alloy was supported on a copper foam (CF) substrate by a one-step electrodeposition method. In the electrodeposition process, a Co–P/CF electrode with a three-dimensional (3D) porous structure was successfully prepared using hydrogen bubbles as a dynamic template. Through the physical characterization of Co–P/CF, it was proven that the doping of P element produced more Co clusters on the surface of the electrode, which resulted in more active sites. These characteristics were very beneficial for NaBH4 electrooxidation. The electrochemical characterization confirmed that the electrocatalytic activity of Co–P/CF for NaBH4 was better than that of Co/CF. At room temperature, for the Co–P/CF in 2.0 mol dm−3 NaOH + 0.20 mol dm−3 NaBH4 solution, the oxidation current density reached 1795 mA cm−2 at −0.20 V (vs. Ag/AgCl), which was higher than that previously reported the literature.

At present, the service scale of the power grid dispatching and control system is rapidly expanding and the multi-level power grid dispatching and control services possess features, including a wide variety of data and scattered storage. In order to manage the data of power grid more efficiently and meet the more efficient, flexible and accurate information acquisition requirements of multi-type users, we apply intelligent search technology to multi-level power grid dispatching and control system.. Facing with the drawbacks of single database in storage and search of massive data, including model data historical data and real-time data, a multi database hybrid storage based intelligent search method and service has been proposed to unified management and application service for multiple types of power dispatching and control big data. On this basis, performance of collection, storage, search, utilization and display of dispatching and control big data for different services has been able to be enhanced.