Distributional Drift Adaptation With Temporal Conditional Variational Autoencoder for Multivariate Time Series Forecasting.

Due to the nonstationary nature, the distribution of real-world multivariate time series (MTS) changes over time, which is known as distribution drift. Most existing MTS forecasting models greatly suffer from distribution drift and degrade the forecasting performance over time. Existing methods address distribution drift via adapting to the latest arrived data or self-correcting per the meta knowledge derived from future data. Despite their great success in MTS forecasting, these methods hardly capture the intrinsic distribution changes, especially from a distributional perspective. Accordingly, we propose a novel framework temporal conditional variational autoencoder (TCVAE) to model the dynamic distributional dependencies over time between historical observations and future data in MTSs and infer the dependencies as a temporal conditional distribution to leverage latent variables. Specifically, a novel temporal Hawkes attention (THA) mechanism represents temporal factors that subsequently fed into feedforward networks to estimate the prior Gaussian distribution of latent variables. The representation of temporal factors further dynamically adjusts the structures of Transformer-based encoder and decoder to distribution changes by leveraging a gated attention mechanism (GAM). Moreover, we introduce conditional continuous normalization flow (CCNF) to transform the prior Gaussian to a complex and form-free distribution to facilitate flexible inference of the temporal conditional distribution. Extensive experiments conducted on six real-world MTS datasets demonstrate the TCVAE's superior robustness and effectiveness over the state-of-the-art MTS forecasting baselines. We further illustrate the TCVAE applicability through multifaceted case studies and visualization in real-world scenarios.

COVID-19 detection from chest X-ray images using CLAHE-YCrCb, LBP, and machine learning algorithms.

BACKGROUND: COVID-19 is a disease that caused a contagious respiratory ailment that killed and infected hundreds of millions. It is necessary to develop a computer-based tool that is fast, precise, and inexpensive to detect COVID-19 efficiently. Recent studies revealed that machine learning and deep learning models accurately detect COVID-19 using chest X-ray (CXR) images. However, they exhibit notable limitations, such as a large amount of data to train, larger feature vector sizes, enormous trainable parameters, expensive computational resources (GPUs), and longer run-time. RESULTS: In this study, we proposed a new approach to address some of the above-mentioned limitations. The proposed model involves the following steps: First, we use contrast limited adaptive histogram equalization (CLAHE) to enhance the contrast of CXR images. The resulting images are converted from CLAHE to YCrCb color space. We estimate reflectance from chrominance using the Illumination-Reflectance model. Finally, we use a normalized local binary patterns histogram generated from reflectance (Cr) and YCb as the classification feature vector. Decision tree, Naive Bayes, support vector machine, K-nearest neighbor, and logistic regression were used as the classification algorithms. The performance evaluation on the test set indicates that the proposed approach is superior, with accuracy rates of 99.01%, 100%, and 98.46% across three different datasets, respectively. Naive Bayes, a probabilistic machine learning algorithm, emerged as the most resilient. CONCLUSION: Our proposed method uses fewer handcrafted features, affordable computational resources, and less runtime than existing state-of-the-art approaches. Emerging nations where radiologists are in short supply can adopt this prototype. We made both coding materials and datasets accessible to the general public for further improvement. Check the manuscript's availability of the data and materials under the declaration section for access.

Small molecule LpxC inhibitors against gram-negative bacteria: Advances and future perspectives.

Uridine diphosphate-3-O-(hydroxymyristoyl)-N-acetylglucosamine deacetylase (LpxC) is a metalloenzyme with zinc ions as cofactors and is a key enzyme in the essential structural outer membrane lipid A synthesis commitment step of gram-negative bacteria. As LpxC is extremely homologous among different Gram-negative bacteria, it is conserved in almost all gram-negative bacteria, which makes LpxC a promising target. LpxC inhibitors have been reported extensively in recent years, such as PF-5081090 and CHIR-090 were found to have broad-spectrum antibiotic activity against P. aeruginosa and E. coli. They are mainly classified into hydroxamate inhibitors and non-hydroxamate inhibitors based on their structure, but no LpxC inhibitors have been marketed due to safety and activity issues. This review, therefore, focuses on small molecule inhibitors of LpxC against gram-negative pathogenic bacteria and covers recent advances in LpxC inhibitors, focusing on their structural optimization process, structure-activity relationships, and future directions, with the aim of providing ideas for the development of LpxC inhibitors and clinical research.

Application of Statistical K-Means Algorithm for University Academic Evaluation.

With the globalization of higher education, academic evaluation is increasingly valued by the scientific and educational circles. Although the number of published papers of academic evaluation methods is increasing, previous research mainly focused on the method of assigning different weights for various indicators, which can be subjective and limited. This paper investigates the evaluation of academic performance by using the statistical K-means (SKM) algorithm to produce clusters. The core idea is mapping the evaluation data from Euclidean space to Riemannian space in which the geometric structure can be used to obtain accurate clustering results. The method can adapt to different indicators and make full use of big data. By using the K-means algorithm based on statistical manifolds, the academic evaluation results of universities can be obtained. Furthermore, through simulation experiments on the top 20 universities of China with the traditional K-means, GMM and SKM algorithms, respectively, we analyze the advantages and disadvantages of different methods. We also test the three algorithms on a UCI ML dataset. The simulation results show the advantages of the SKM algorithm.

Back on the scene: Advances and challenges in CD3-related drugs in tumor therapy.

CD3 molecules are mainly distributed on the membrane of mature T cells. They are involved in T cell antigen recognition, signal transduction, and regulation of T cell development. CD3-related monoclonal antibodies (mAbs) are mainly used in the treatment of autoimmune diseases. Nearly half of all bispecific antibodies developed are used in tumor therapy, one of which is CD3 antigen. In this review, we discuss the importance of biological function and the crucial role of CD3 in tumor therapy. We highlight the research status of antibodies and small molecules targeting CD3 to provide guidance for future drug research.

Monitoring cardiopulmonary resuscitation quality in emergency departments: a national survey in China on current knowledge, attitudes, and practices.

BACKGROUND: To investigate current knowledge, attitudes, and practices for CPR quality control among emergency physicians in Chinese tertiary hospitals. METHODS: Anonymous questionnaires were distributed to physicians in 75 tertiary hospitals in China between January and July 2018. RESULTS: A total of 1405 respondents answered the survey without obvious logical errors. Only 54.4% respondents knew all criteria of high-quality CPR. A total of 91.0% of respondents considered CPR quality monitoring should be used, 72.4% knew the objective method for monitoring, and 63.2% always/often monitored CPR quality during actual resuscitation. The main problems during CPR were related to chest compression: low quality due to fatigue (67.3%), inappropriate depth (57.3%) and rate (54.1%). The use of recommended monitoring methods was reported as follows, ETCO2 was 42.7%, audio-visual feedback devices was 10.1%, coronary perfusion pressure was 17.9%, and invasive arterial pressure was 31.1%. A total of 96.3% of respondents considered it necessary to participate in regular CPR retraining, but 21.4% did not receive any retraining. The ideal retraining interval was considered to be 3 to 6 months, but the actual interval was 6 to 12 months. Only 49.7% of respondents reported that feedback devices were always/often used in CPR training. CONCLUSION: Chinese emergency physicians were very concerned about CPR quality, but they did not fully understand the high-quality criteria and their impact on prognosis. CPR quality monitoring was not a routine procedure during actual resuscitation. The methods recommended in guidelines were rarely used in practice. Many physicians had not received retraining or received retraining at long intervals. Feedback devices were not commonly used in CPR training.

Injury Assessment of Individuals Wounded in the Lushan Earthquake and the Emergency Department Workload: A Corresponding Correlation Study.

OBJECTIVE: In this study, we aimed to evaluate the correlation between the trauma score of individuals wounded in the Lushan earthquake and emergency workload for treatment. We further created a trauma score-emergency workload calculation model. METHODS: We included data from patients wounded in the Lushan earthquake and treated at West China Hospital, Sichuan University. We calculated scores per the following models separately: Revised Trauma Score (RTS), Prehospital Index (PHI), Circulation Respiration Abdominal Movement Speech (CRAMS), Therapeutic Intervention Scoring System (TISS-28), and Nursing Activities Score (NAS). We assessed the association between values for CRAMS, PHI, and RTS and those for TISS-28 and NAS. Subsequently, we built a trauma score-emergency workload calculation model to quantitative workload estimation. RESULTS: Significant correlations were observed for all pairs of trauma scoring models with emergency workload scoring models. TISS-28 score was significantly associated with PHI score and RTS; however, no significant correlation was observed between the TISS-28 score and CRAMS score. CONCLUSIONS: CRAMS, PHI, and RTS were consistent in evaluating the injury condition of wounded individuals; TISS-28 and NAS scores were consistent in evaluating the required treatment workload. Dynamic changes in emergency workload in unit time were closely associated with wounded patient visits.

Neural Time-Aware Sequential Recommendation by Jointly Modeling Preference Dynamics and Explicit Feature Couplings.

In recommendation, both stationary and dynamic user preferences on items are embedded in the interactions between users and items (e.g., rating or clicking) within their contexts. Sequential recommender systems (SRSs) need to jointly involve such context-aware user-item interactions in terms of the couplings between the user and item features and sequential user actions on items over time. However, such joint modeling is non-trivial and significantly challenges the existing work on preference modeling, which either only models user-item interactions by latent factorization models but ignores user preference dynamics or only captures sequential user action patterns without involving user/item features and context factors and their coupling and influence on user actions. We propose a neural time-aware recommendation network (TARN) with a temporal context to jointly model 1) stationary user preferences by a feature interaction network and 2) user preference dynamics by a tailored convolutional network. The feature interaction network factorizes the pairwise couplings between non-zero features of users, items, and temporal context by the inner product of their feature embeddings while alleviating data sparsity issues. In the convolutional network, we introduce a convolutional layer with multiple filter widths to capture multi-fold sequential patterns, where an attentive average pooling (AAP) obtains significant and large-span feature combinations. To learn the preference dynamics, a novel temporal action embedding represents user actions by incorporating the embeddings of items and temporal context as the inputs of the convolutional network. The experiments on typical public data sets demonstrate that TARN outperforms state-of-the-art methods and show the necessity and contribution of involving time-aware preference dynamics and explicit user/item feature couplings in modeling and interpreting evolving user preferences.

University Academic Performance Development Prediction Based on TDA.

With the rapid development of higher education, the evaluation of the academic growth potential of universities has received extensive attention from scholars and educational administrators. Although the number of papers on university academic evaluation is increasing, few scholars have conducted research on the changing trend of university academic performance. Because traditional statistical methods and deep learning techniques have proven to be incapable of handling short time series data well, this paper proposes to adopt topological data analysis (TDA) to extract specified features from short time series data and then construct the model for the prediction of trend of university academic performance. The performance of the proposed method is evaluated by experiments on a real-world university academic performance dataset. By comparing the prediction results given by the Markov chain as well as SVM on the original data and TDA statistics, respectively, we demonstrate that the data generated by TDA methods can help construct very discriminative models and have a great advantage over the traditional models. In addition, this paper gives the prediction results as a reference, which provides a new perspective for the development evaluation of the academic performance of colleges and universities.

The first national survey on practices of neurological prognostication after cardiac arrest in China, still a lot to do.

AIMS: To investigate current awareness and practices of neurological prognostication in comatose cardiac arrest (CA) patients. METHODS: An anonymous questionnaire was distributed to 1600 emergency physicians in 75 hospitals which were selected randomly from China between January and July 2018. RESULTS: 92.1% respondents fulfilled the survey. The predictive value of brain stem reflex, motor response and myoclonus was confirmed by 63.5%, 44.6% and 31.7% respondents, respectively. Only 30.7% knew that GWR value < 1.1 indicated poor prognosis and only 8.1% know the most commonly used SSEP N20. Status epilepticus, burst suppression and suppression were considered to predict poor outcome by only 35.0%, 27.4% and 20.9% respondents, respectively. Only 46.7% knew NSE and only 24.7% knew S-100. Only a few respondents knew that neurological prognostication should be performed later than 72 hours from CA either in TTM or non-TTM patients. In practice, the most commonly used method was clinical examination (85.4%). Only 67.9% had used brain CT for prognosis and 18.4% for MRI. NSE (39.6%) was a little more widely used than S-100ß (18.0%). However, SSEP (4.4%) and EEG (11.4%) were occasionally performed. CONCLUSIONS: Neurological prognostication in CA survivors had not been well understood and performed by emergency physicians in China. They were more likely to use clinical examination rather than objective tools, especially SSEP and EEG, which also illustrated that multimodal approach was not well performed in practice.

Poly(I:C) preconditioning protects the heart against myocardial ischemia/reperfusion injury through TLR3/PI3K/Akt-dependent pathway.

Emerging evidence suggests that Toll-like receptors (TLRs) ligands pretreatment may play a vital role in the progress of myocardial ischemia/reperfusion (I/R) injury. As the ligand of TLR3, polyinosinic-polycytidylic acid (poly(I:C)), a synthetic double-stranded RNA, whether its preconditioning can exhibit a cardioprotective phenotype remains unknown. Here, we report the protective effect of poly(I:C) pretreatment in acute myocardial I/R injury by activating TLR3/PI3K/Akt signaling pathway. Poly(I:C) pretreatment leads to a significant reduction of infarct size, improvement of cardiac function, and downregulation of inflammatory cytokines and apoptotic molecules compared with controls. Subsequently, our data demonstrate that phosphorylation of TLR3 tyrosine residue and its interaction with PI3K is enhanced, and protein levels of phospho-PI3K and phospho-Akt are both increased after poly(I:C) pretreatment, while knock out of TLR3 suppresses the cardioprotection of poly(I:C) preconditioning through a decreased activation of PI3K/Akt signaling. Moreover, inhibition of p85 PI3K by the administration of LY294002 in vivo and knockdown of Akt by siRNA in vitro significantly abolish poly(I:C) preconditioning-induced cardioprotective effect. In conclusion, our results reveal that poly(I:C) preconditioning exhibits essential protection in myocardial I/R injury via its modulation of TLR3, and the downstream PI3K/Akt signaling, which may provide a potential pharmacologic target for perioperative cardioprotection.

RNA-seq analysis of the key long noncoding RNAs and mRNAs related to cognitive impairment after cardiac arrest and cardiopulmonary resuscitation.

Cardiac arrest (CA) is the leading cause of death around the world. Survivors after CA and cardiopulmonary resuscitation (CPR) develop moderate to severe cognitive impairment up to 60% at 3 months. Accumulating evidence demonstrated that long non-coding RNAs (lncRNAs) played a pivotal role in ischemic brain injury. This study aimed to identify potential key lncRNAs associated with early cognitive deficits after CA/CPR. LncRNA and mRNA expression profiles of the hippocampus in CA/CPR or sham group were analyzed via high-throughput RNA sequencing, which exhibited 1920 lncRNAs and 1162 mRNAs were differentially expressed. These differentially expressed genes were confirmed to be primarily associated with inflammatory or apoptotic signaling pathways through GO and KEGG pathway enrichment analysis and coding-noncoding co-expression network analysis. Among which, five key pairs of lncRNA-mRNA were further analyzed by qRT-PCR and western blot. We found that the lncRNANONMMUT113601.1 and mRNA Shc1, an inflammation and apoptosis-associated gene, exhibited the most significant changes in hippocampus of CA/CPR mice. Furthermore, we found that the correlations between this lncRNA and mRNA mainly happened in neurons of hippocampus by in situ hybridization. These results suggested that the critical pairs of lncRNA-mRNA may act as essential regulators in early cognitive deficits after resuscitation.

Complementarity in Requirements Tracing.

Complementarity between activities reveals that doing any one of them increases the returns to doing the others. In other words, complementarity leads to the synergistic effect that the whole is greater than the sum of its parts. Identifying and exploiting complementarity can benefit many cybernetic activities, where human-machine interactions are inherent and dominant. One such activity is requirements tracing that helps stakeholders to track the status of their goals. Although various kinds of support for human analysts in requirements tracing have been proposed, little is known about the nature of complementarity when different tracing practices are involved. In this paper, we explore the role of complementarity by considering together the tagging-to-trace (T2T) and learning-to-trace (L2T) activities. We present a novel approach to examining which T2T and L2T practices enhance the qualities of each other. Our approach also uncovers the environmental factors which the complementarity is sensitive to. Applying our approach to the logs of 140 analyst-tracing units offers operational insights into the rigorous detection of complementarity and shows the importance of understanding the cybernetic conditions under which the requirements tracing practices may in fact be complementary.

PHD2 exerts anti-cancer and anti-inflammatory effects in colon cancer xenografts mice via attenuating NF-κB activity.

Recent studies suggested that prolyl hydroxylase 2 (PHD2) functions as an important regulator in vascular inflammation and Streptococcus pneumonia infection. However, whether PHD2 contributed to tumor progression prompted by intratumoral inflammation remains elusive. In this study, the effects of PHD2 in colon cancer were evaluated, and the underlying molecular mechanisms were investigated. The results showed that overexpressing PHD2 exerted proliferative and migratory inhibition in colon cancer cells. The expression of cell cycle and epithelial-mesenchymal transition (EMT)-associated proteins were changed: CyclinD1, CDK4, N-cadherin, and Vimentin were down-regulated, while E-cadherin was up-regulated in PHD2-overexpressing colon cancer cells. Moreover, in colon cancer xenograft mice, PHD2 overexpression suppressed tumor growth accompanied by decreased Ki67 expression. Importantly, we further demonstrated that overexpressing PHD2 attenuated inflammation in colon cancer xenograft mice through weakening accumulation of myeloid-derived suppressor cells (MDSCs) and M2-like tumor-associated macrophages (TAMs), as well as secretions of pro-inflammatory cytokines including G-CSF, TNF-α, IL-6, IL-8, IL-1ß, and IL-4. Mechanistically, PHD2 overexpression obviously suppressed NF-κB activity through decreasing phosphorylated IκB-α while increasing cytoplasmic NF-κB p65 levels in colon cancer. Our findings support the anti-cancer and anti-inflammatory roles of PHD2 and offer a preclinical proof of tumor progression regulated by cancer cells and inflammation.

Predicting the Post-therapy Severity Level (UPDRS-III) of Patients With Parkinson's Disease After Drug Therapy by Using the Dynamic Connectivity Efficiency of fMRI.

Parkinson's disease (PD) is a multi-systemic disease in the brain arising from the dysfunction of several neural networks. The diagnosis and treatment of PD have gained more attention for clinical researchers. While there have been many fMRI studies about functional topological changes of PD patients, whether the dynamic changes of functional connectivity can predict the drug therapy effect is still unclear. The primary objective of this study was to assess whether large-scale functional efficiency changes of topological network are detectable in PD patients, and to explore whether the severity level (UPDRS-III) after drug treatment can be predicted by the pre-treatment resting-state fMRI (rs-fMRI). Here, we recruited 62 Parkinson's disease patients and calculated the dynamic nodal efficiency networks based on rs-fMRI. With connectome-based predictive models using the least absolute shrinkage and selection operator, we demonstrated that the dynamic nodal efficiency properties predict drug therapy effect well. The contributed regions for the prediction include hippocampus, post-central gyrus, cingulate gyrus, and orbital gyrus. Specifically, the connections between hippocampus and cingulate gyrus, hippocampus and insular gyrus, insular gyrus, and orbital gyrus are positively related to the recovery (post-therapy severity level) after drug therapy. The analysis of these connection features may provide important information for clinical treatment of PD patients.

Topological reorganization after partial auditory deprivation-a structural connectivity study in single-sided deafness.

Growing evidence shows that partial auditory deprivation leads to extensive neural functional plasticity, which occurs not only in the auditory cortex but also in other sensory regions and cognitive areas. However, studies in structural topological properties are still limited, especially those investigating the relationship between structural connectome alterations and auditory abilities. To clarify this, we investigated white matter structural connectivity changes and the relationship between connection strength and hearing abilities in individuals with long-term single-sided deafness (SSD), a common form of partial hearing deprivation, using diffusion tensor imaging and network-based analysis. The results showed globally improved connection efficiency, locally weakened visual networks, and strengthened fronto-parietal sub-networks in SSD compared to normal hearing controls. Furthermore, a strong positive correlation between hearing abilities (including speech recognition in noise and sound localization) and connection strength, mainly in the fronto-parietal areas, was found in SSD. Our study reveals alteration of the structural network connections in SSD, especially in cognitive related networks, which showed close correlation with hearing abilities. Our findings provide new insights into topological white matter reorganization of the brain after partial sensory deprivation.

Sensing Urban Transportation Events from Multi-Channel Social Signals with the Word2vec Fusion Model.

Social sensors perceive the real world through social media and online web services, which have the advantages of low cost and large coverage over traditional physical sensors. In intelligent transportation researches, sensing and analyzing such social signals provide a new path to monitor, control and optimize transportation systems. However, current research is largely focused on using single channel online social signals to extract and sense traffic information. Clearly, sensing and exploiting multi-channel social signals could effectively provide deeper understanding of traffic incidents. In this paper, we utilize cross-platform online data, i.e., Sina Weibo and News, as multi-channel social signals, then we propose a word2vec-based event fusion (WBEF) model for sensing, detecting, representing, linking and fusing urban traffic incidents. Thus, each traffic incident can be comprehensively described from multiple aspects, and finally the whole picture of unban traffic events can be obtained and visualized. The proposed WBEF architecture was trained by about 1.15 million multi-channel online data from Qingdao (a coastal city in China), and the experiments show our method surpasses the baseline model, achieving an 88.1% F1 score in urban traffic incident detection. The model also demonstrates its effectiveness in the open scenario test.

RNase alleviates neurological dysfunction in mice undergoing cardiac arrest and cardiopulmonary resuscitation.

Cardiac arrest (CA) is one of the leading lethal factors. Despite cardiopulmonary resuscitation (CPR) procedure has been consecutively improved and lots of new strategies have been developed, neurological outcome of the patients experienced CPR is still disappointing. Ribonuclease (RNase) has been demonstrated to have neuroprotective effects in acute stroke and postoperative cognitive impairment, possibly through acting against endogenous RNA that released from damaged tissue. However, the role of RNase in post-cardiac arrest cerebral injury is unknown. In the present study, we investigated the role of RNase in neurological outcome of mice undergoing 5 minutes of CA and followed by CPR. RNase or the same dosage of normal saline was administrated. We found that RNase administration could: 1) improve neurologic score on day 1 and day 3 after CA/CPR performance; 2) improve memory and learning ability on day 3 after training in contextual fear-conditioning test; 3) reduce extracellular RNA (exRNA) level in plasma and hippocampus tissue, and hippocampal cytokines mRNA production on day 3 after CA/CPR procedure; 4) attenuate autophagy levels in hippocampus tissue on day 3 after CA/CPR procedure. In conclusion, RNase could improve neurological function by reducing inflammation response and autophagy in mice undergoing CA/CPR.