Blood urea nitrogen to serum albumin ratio as a new prognostic indicator in type 2 diabetes mellitus patients with chronic kidney disease.

Chronic kidney disease (CKD) is often a common comorbidity in critically ill patients with type 2 diabetes mellitus (T2DM). This study explored the relationship between blood urea nitrogen to serum albumin ratio (BAR) and mortality in T2DM patients with CKD in intensive care unit (ICU). Patients were recruited from the Medical Information Mart database, retrospectively. The primary and secondary outcomes were 90-day mortality, the length of ICU stay, hospital mortality and 30-day mortality, respectively. Cox regression model and Kaplan-Meier survival curve were performed to explore the association between BAR and 90-day mortality. Subgroup analyses were performed to determine the consistency of this association. A total of 1920 patients were enrolled and divided into the three groups (BAR < 9.2, 9.2 ≤ BAR ≤ 21.3 and BAR > 21.3). The length of ICU stay, 30-day mortality, and 90-day mortality in the BAR > 21.3 group were significantly higher than other groups. In Cox regression analysis showed that high BAR level was significantly associated with increased greater risk of 90-day mortality. The adjusted HR (95%CIs) for the model 1, model 2, and model 3 were 1.768 (1.409-2.218), 1.934, (1.489-2.511), and 1.864, (1.399-2.487), respectively. Subgroup analysis also showed the consistency of results. The Kaplan-Meier survival curve analysis revealed similar results as well that BAR > 21.3 had lower 90-day survival rate. High BAR was significantly associated with increased risk of 90-day mortality. BAR could be a simple and useful prognostic tool in T2DM patients with CKD in ICU.

Comprehensive analysis of the clinical and prognostic significance of SFRP1 and PRKCB expression in non-small cell lung cancer: a retrospective analysis.

OBJECTIVES: Secreted frizzled-related protein 1 (SFRP1) and protein kinase C-B (PRKCB) contribute to cancer progression and angiogenesis. This study intended to detect SFRP1 and PRKCB expression in non-small-cell lung cancer (NSCLC) patients and analyze its association with clinicopathological features. METHODS: A total of 108 NSCLC patients who underwent surgical resection in our hospital between 2012 and 2017 were retrospectively analyzed. SFRP1 and PRKCB expression was detected using immunohistochemical staining. The relationships between SFRP1 and PRKCB expression and clinicopathological data were analyzed using the chi-square method. Kaplan-Meier analysis was used to investigate survival probability over time. The potential risk of NSCLC morbidity associated with SFRP1 and PRKCB levels was analyzed using univariate and multivariate Cox proportional risk models. RESULTS: SFRP1 and PRKCB expression was negative in 114 and 109 of the 180 NSCLC specimens, respectively. SFRP1 expression was significantly associated with TNM stage ( P  < 0.001) and tumor diameter ( P  < 0.001). PRKCB expression was significantly associated with the TNM stage ( P  < 0.001). The correlation between SFRP1 and PRKCB expression was evident ( P  = 0.023). SFRP1(-) or PRKCB(-) patients shows lower survival rates than SFRP1(+) or PRKCB(+) patients ( P < 0.001). SFRP1(-)/PRKCB(-) patients had the worst prognosis ( P < 0.001). Furthermore, the mortality of SFRP1(-) or PRKCB(-) patients was significantly higher than that of SFRP1(+) or PRKCB(+). CONCLUSION: SFRP1 and PRKCB expression can be used to predict prognosis in patients with NSCLC.

The necroptosis signature and molecular mechanism of lung squamous cell carcinoma.

BACKGROUND: Given the poor prognosis of lung squamous cell carcinoma (LUSC), the aim of this study was to screen for new prognostic biomarkers. METHODS: The TGCA_LUSC dataset was used as the training set, and GSE73403 was used as the validation set. The genes involved in necroptosis-related pathways were acquired from the KEGG database, and the differential genes between the LUSC and normal samples were identified using the GSEA. A necroptosis signature was constructed by survival analysis, and its correlation with patient prognosis and clinical features was evaluated. The molecular characteristics and drug response associated with the necroptosis signature were also identified. The drug candidates were then validated at the cellular level. RESULTS: The TCGA_LUSC dataset included 51 normal samples and 502 LUSC samples. The GSE73403 dataset included 69 samples. 159 genes involved in necroptosis pathways were acquired from the KEGG database, of which most showed significant differences between two groups in terms of genomic, transcriptional and methylation alterations. In particular, CHMP4C, IL1B, JAK1, PYGB and TNFRSF10B were significantly associated with the survival (p < 0.05) and were used to construct the necroptosis signature, which showed significant correlation with patient prognosis and clinical features in univariate and multivariate analyses (p < 0.05). Furthermore, CHMP4C, IL1B, JAK1 and PYGB were identified as potential targets of trametinib, selumetinib, SCH772984, PD 325901 and dasatinib. Finally, knockdown of these genes in LUSC cells increased chemosensitivity to those drugs. CONCLUSION: We identified a necroptosis signature in LUSC that can predict prognosis and identify patients who can benefit from targeted therapies.

1000 °C High-Temperature Wetting Behaviors of Molten Metals on Laser-Microstructured Metal Surfaces.

The melting of metals at high temperatures is common and important in many fields, e.g., metallurgy, refining, casting, welding, brazing, even newly developed batteries, and nuclear fusion, which is thus of great value in modern industrialization. However, the knowledge of the wetting behaviors of molten metals on various substrate surfaces remains insufficient, especially when the temperature is over 1000 °C and with microstructured metal substrate surfaces. Herein, we selected molten cerium (Ce) on a tantalum (Ta) substrate as an example and investigated in detail its wetting at temperatures up to 1000 °C by modulating the microstructures of the substrate surfaces via laser processing. We discovered that the wetting states of molten Ce on Ta surfaces at temperatures over 900 °C could be completely altered by modifying the laser-induced surface microstructures and the surface compositions. The molten Ce turned superlyophilic with its contact angle (CA) below 10° on the only laser-microstructured surfaces, while it exhibited lyophobicity with a CA of about 135° on the laser-microstructured plus oxidized ones, which demonstrated remarkably enhanced resistance against the melt with only tiny adhesion in this circumstance. In contrast, the CA of molten Ce on Ta substrate surfaces only changed from ∼25 to ∼95° after oxidization without laser microstructuring. We proved that modulating the substrate surface microstructures via laser together with oxidization was capable of efficiently controlling various molten metals' wetting behaviors even at very high temperatures. These findings not only enrich the understanding of molten metal high-temperature wettability but also enable a novel practical approach to control the wetting states for relevant applications.

Case Report: A rare case of non-small cell lung cancer with STRN-ALK fusion in a patient in very poor condition treated with first-line ensartinib.

Several cases of STRN-ALK fusion have been reported, and some anaplastic lymphoma kinase (ALK) inhibitors have been shown to be effective for treatment. Nevertheless, no cases of COVID-19 leading to heart failure and respiratory failure have been reported in people older than 70 years treated with ALK inhibitors. The present case report describes a 70-year-old patient with usual chronic obstructive pulmonary disease, diabetes, depression, and carotid plaque disease. Next-generation sequencing of tissue obtained by puncture biopsy revealed a STRN-ALK mutation accompanied by a TP53 mutation. The patient was treated with ensartinib and developed COVID-19 leading to heart failure and respiratory failure; nevertheless, he had a good clinical outcome and exhibited high treatment tolerability.

Prognostic value of monocyte-to-lymphocyte ratio for 90-day all-cause mortality in type 2 diabetes mellitus patients with chronic kidney disease.

The role of inflammation and the correlation between inflammatory markers and type 2 diabetes mellitus (T2DM) and chronic kidney disease (CKD) have been studied. In clinical work, a large number of T2DM patients complicated with CKD, but the cause of CKD was not clear. Our study aimed to evaluate the relationship between monocyte-to-lymphocyte ratio (MLR) and mortality in T2DM patients with CKD. The data from Medical Information Mart for Intensive Care III was analyzed. The primary outcome was 90-day all-cause mortality; the secondary outcomes were the length of ICU stay, hospital mortality and 30-day all-cause mortality. Cox regression was used to evaluate the association between MLR and 90-day mortality. We performed subgroup analyses to determine the consistency of this association, and used Kaplan-Meier survival curve to analysis the survival of different levels of MLR. A total of 1830 patients were included in study retrospectively. The length of ICU stay, 30-day all-cause mortality, and 90-day all-cause mortality in the MLR > 0.71 group were significantly higher than those in the MLR < 0.28 and 0.28 ≤ MLR ≤ 0.71 group. In Cox regression analysis, high MLR level was significantly associated with increased greater risk of 90-day all-cause mortality. The adjusted HR (95%CIs) for the model 1, model 2, and model 3 were 2.429 (1.905-3.098), 2.070 (1.619-2.647), and 1.898 (1.478-2.437), respectively. Subgroup analyses also showed the consistency of association between MLR and 90-day all-cause mortality. The Kaplan-Meier survival curve analysis revealed that MLR > 0.71 had worst prognosis. In T2DM patients with CKD in the intensive care unit, high MLR was significantly associated with increased risk 90-day all-cause mortality.

Enhancing Unsupervised Anomaly Detection With Score-Guided Network.

Anomaly detection plays a crucial role in various real-world applications, including healthcare and finance systems. Owing to the limited number of anomaly labels in these complex systems, unsupervised anomaly detection methods have attracted great attention in recent years. Two major challenges faced by the existing unsupervised methods are as follows: 1) distinguishing between normal and abnormal data when they are highly mixed together and 2) defining an effective metric to maximize the gap between normal and abnormal data in a hypothesis space, which is built by a representation learner. To that end, this work proposes a novel scoring network with a score-guided regularization to learn and enlarge the anomaly score disparities between normal and abnormal data, enhancing the capability of anomaly detection. With such score-guided strategy, the representation learner can gradually learn more informative representation during the model training stage, especially for the samples in the transition field. Moreover, the scoring network can be incorporated into most of the deep unsupervised representation learning (URL)-based anomaly detection models and enhances them as a plug-in component. We next integrate the scoring network into an autoencoder (AE) and four state-of-the-art models to demonstrate the effectiveness and transferability of the design. These score-guided models are collectively called SG-Models. Extensive experiments on both synthetic and real-world datasets confirm the state-of-the-art performance of SG-Models.

Semi-Supervised Mixture Learning for Graph Neural Networks With Neighbor Dependence.

A graph neural network (GNN) is a powerful architecture for semi-supervised learning (SSL). However, the data-driven mode of GNNs raises some challenging problems. In particular, these models suffer from the limitations of incomplete attribute learning, insufficient structure capture, and the inability to distinguish between node attribute and graph structure, especially on label-scarce or attribute-missing data. In this article, we propose a novel framework, called graph coneighbor neural network (GCoNN), for node classification. It is composed of two modules: GCoNN Γ and GCoNN Γ° . GCoNN Γ is trained to establish the fundamental prototype for attribute learning on labeled data, while GCoNN [Formula: see text] learns neighbor dependence on transductive data through pseudolabels generated by GCoNN Γ . Next, GCoNN Γ is retrained to improve integration of node attribute and neighbor structure through feedback from GCoNN [Formula: see text] . GCoNN tends to convergence iteratively using such an approach. From a theoretical perspective, we analyze this iteration process from a generalized expectation-maximization (GEM) framework perspective which optimizes an evidence lower bound (ELBO) by amortized variational inference. Empirical evidence demonstrates that the state-of-the-art performance of the proposed approach outperforms other methods. We also apply GCoNN to brain functional networks, the results of which reveal response features across the brain which are physiologically plausible with respect to known language and visual functions.

Altered temporoparietal junction activity during reflected self-evaluation in sub-clinical depression.

Negative self-schema is a core symptom of depression. According to social psychological theories, two types of self-evaluations play important roles in forming the negative self-view: direct self-evaluation (that is, evaluating the self directly through one's first-person perspective introspection) and reflected self-evaluation (which requires theory of mind (ToM) ability, and is evaluating the self through reflecting on a third person's perspective). Although many previous studies have investigated the processing of the direct self-evaluation in depression, few have extended research on the reflected self-evaluation. In the current study, functional magnetic resonance imaging scans were acquired in 26 dysphoric (individuals with elevated number of depressive symptoms) and 28 control participants during both direct and reflected self-evaluation tasks. Two regions of interest were defined within bilateral temporoparietal junctions (TPJs) because their significant role in ToM. Results showed that the dysphoric participants evaluated themselves more negatively than the control participants, regardless of whose perspective they were taking. More importantly, the enhanced TPJs' activations were observed in the control group during the reflected self-evaluation task versus the direct self-evaluation task, whereas no such difference was observed in the dysphoric participants. The results are interpreted in the framework of impaired ToM ability in sub-clinical depression.  General Scientific Summary (GSS) Negative self-schema is one of the core symptoms of depression. This study suggests that the negative self-schema reflects not only in direct self-evaluation (i.e. evaluating the self via one's own introspection) but also in reflected self-evaluation (i.e. evaluating the self via others' perspective). Importantly, altered TPJ activity was found during a reflected self-evaluation task among individuals with depressive symptoms. These changes in brain function might be associated with impaired ToM ability in sub-clinical depression.

A nomogram for predicting the mortality of patients with type 2 diabetes mellitus complicated with acute kidney injury in the intensive care unit.

BACKGROUND: There is no predictive tool for type 2 diabetes mellitus (T2DM) patients with acute kidney injury (AKI). Our study aimed to establish an effective nomogram model for predicting mortality in T2DM patients with AKI. METHOD: Data on T2DM patients with AKI were obtained from the Medical Information Mart for Intensive Care III. 70% and 30% of the patients were randomly selected as the training and validation cohorts, respectively. Univariate and multivariate logistic regression analyses were used to identify factors associated with death in T2DM patients with AKI. Factors significantly associated with survival outcomes were used to construct a nomogram predicting 90-day mortality. The nomogram effect was evaluated by receiver operating characteristic curve analysis, Hosmer‒Lemeshow test, calibration curve, and decision curve analysis (DCA). RESULTS: There were 4375 patients in the training cohort and 1879 in the validation cohort. Multivariate logistic regression analysis showed that age, BMI, chronic heart failure, coronary artery disease, malignancy, stages of AKI, white blood cell count, blood urea nitrogen, arterial partial pressure of oxygen and partial thromboplastin time were independent predictors of patient survival. The results showed that the nomogram had a higher area under the curve value than the sequential organ failure assessment score and simplified acute physiology score II. The Hosmer‒Lemeshow test and calibration curve suggested that the nomogram had a good calibration effect. The DCA curve showed that the nomogram model had good clinical application value. CONCLUSION: The nomogram model accurately predicted 90-day mortality in T2DM patients with AKI. It may provide assistance for clinical decision-making and treatment, thereby reducing the medical burden.

Formation Tracking of Multiagent Systems With Time-Varying Actuator Faults and Its Application to Task-Space Cooperative Tracking of Manipulators.

This article is concerned with a fault-tolerant formation tracking problem of nonlinear systems under unknown faults, where the leader's states are only accessible to a small set of followers via a directed graph. Under these faults, not only the amplitudes but also the signs of control coefficients become time-varying and unknown. The current setting will enhance the investigated problem's practical relevance and at the same time, it poses nontrivial design challenges of distributed control algorithms and convergence analysis. To solve this problem, a novel distributed control algorithm is developed by incorporating an estimation-based control framework together with a Nussbaum gain approach to guarantee an asymptotic cooperative formation tracking of nonlinear networked systems under unknown and dynamic actuator faults. Moreover, the proposed control framework is extended to ensure an asymptotic task-space coordination of multiple manipulators under unknown actuator faults, kinematics, and dynamics. Lastly, numerical simulation results are provided to validate the effectiveness of the proposed distributed designs.

SKA1/2/3 is a biomarker of poor prognosis in human hepatocellular carcinoma.

Background: Spindle and kinetochore-associated complex subunits 1-3 (SKA1-3) stabilize the kinetochore-attached spindle microtubules in metaphase. Due to the dysregulation in multiple cancers, SKA1-3 is considered a predictor for the prognosis of the patients. However, the potential clinical applications of SKA1-3, particularly in hepatocellular carcinoma (HCC) prognosis and progression, have completely unknown yet. Methods: For the analysis of SKA1-3 expression and applications in clinics in HCC patients, several databases, such as STRING, UALCAN, GEO, and TCGA, were searched. In addition, the underlying mechanisms of SKA for the regulation of HCC occurrence, development, and progression were also explored. Results: Compared to the normal controls, HCC patients showed dramatically elevated SKA1-3 expression at the mRNA level, and the values of the area under the curve (AUC) were 0.982, 0.887, and 0.973, respectively. Increased SKA1-3 expression levels were associated with the clinical stage, age, body mass index, tumor grade, tissue subtype, and Tp53 mutation status in HCC patients. The analyses of Kyoto Encyclopedia of Genes and Genome (KEGG) and Gene ontology (GO) demonstrated that SKA1-3 are enriched mainly in the Fanconi anemia, homologous recombination, spliceosome, DNA replication, and cell cycle signaling pathways. The hub genes, such as CDK1, CCNB1, CCNA2, TOP2A, BUB1, AURKB, CCNB2, BUB1B, NCAPG, and KIF11, were identified in protein-protein interactions (PPIs). The expression levels of hub genes were increased in HCC patients and predictive of a poor prognosis. Finally, the expression levels of SKA1-3 were determined using the GEO database. Conclusions: SKA1-3 are potential prognostic biomarkers of and targets for HCC. In addition, SKA1-3 may affect HCC prognosis via the Fanconi anemia pathway, homologous recombination, spliceosome, DNA replication, and cell cycle signaling pathway.

Discovering hidden brain network responses to naturalistic stimuli via tensor component analysis of multi-subject fMRI data.

The study of brain network interactions during naturalistic stimuli facilitates a deeper understanding of human brain function. To estimate large-scale brain networks evoked with naturalistic stimuli, a tensor component analysis (TCA) based framework was used to characterize shared spatio-temporal patterns across subjects in a purely data-driven manner. In this framework, a third-order tensor is constructed from the timeseries extracted from all brain regions from a given parcellation, for all participants, with modes of the tensor corresponding to spatial distribution, time series and participants. TCA then reveals spatially and temporally shared components, i.e., evoked networks with the naturalistic stimuli, their time courses of activity and subject loadings of each component. To enhance the reproducibility of the estimation with the adaptive TCA algorithm, a novel spectral clustering method, tensor spectral clustering, was proposed and applied to evaluate the stability of the TCA algorithm. We demonstrated the effectiveness of the proposed framework via simulations and real fMRI data collected during a motor task with a traditional fMRI study design. We also applied the proposed framework to fMRI data collected during passive movie watching to illustrate how reproducible brain networks are evoked by naturalistic movie viewing.

Time-Varying Optimization-Based Approach for Distributed Formation of Uncertain Euler-Lagrange Systems.

We investigate a distributed time-varying formation control problem for an uncertain Euler-Lagrange system. A time-varying optimization-based approach is proposed. Based on this approach, the robots can achieve the expected formation configuration and meanwhile optimize a global objective function using only neighboring and local information. We consider the time-varying optimization where the objective functions can change in real time. In this case, the consensus-based formation tracking control issues and formation containment tracking control issues in the literature can be solved by the proposed approach. By a penalty-based method, the robots' states asymptotically converge to the estimated optimal solution to an equivalent time-varying optimization problem, whose optimal solution can achieve simultaneous formation and optimization. Furthermore, we consider two more general scenarios: 1) the local objective functions can have non-neighbor's information and 2) the optimization problems can have inequality constraints.

Two effective clinical prediction models to screen for obstructive sleep apnoea based on body mass index and other parameters.

BACKGROUND AND OBJECTIVE: The diagnosis of obstructive sleep apnea (OSA) relies on polysomnography which is time-consuming and expensive. We therefore aimed to develop two simple, non-invasive models to screen adults for OSA. METHODS: The effectiveness of using body mass index (BMI) and a new visual prediction model to screen for OSA was evaluated using a development set (1769 participants) and confirmed using an independent validation set (642 participants). RESULTS: Based on the development set, the best BMI cut-off value for diagnosing OSA was 26.45 kg/m2, with an area under the curve (AUC) of 0.7213 (95% confidence interval (CI), 0.6861-0.7566), a sensitivity of 57% and a specificity of 78%. Through forward conditional logistic regression analysis using a stepwise selection model developed from observed data, seven clinical variables were evaluated as independent predictors of OSA: age, BMI, sex, Epworth Sleepiness Scale score, witnessed apnoeas, dry mouth and arrhythmias. With this new model, the AUC was 0.7991 (95% CI, 0.7668-0.8314) for diagnosing OSA (sensitivity, 75%; specificity, 71%). The results were confirmed using the validation set. A nomogram for predicting OSA was generated based on this new model using statistical software. CONCLUSIONS: BMI can be used as an indicator to screen for OSA in the community. We created an internally validated, highly distinguishable, visual and parsimonious prediction model comprising BMI and other parameters that can be used to identify patients with OSA among outpatients. Use of this prediction model may help to improve clinical decision-making.

Asymmetric Input-Output Constraint Control of a Flexible Variable-Length Rotary Crane Arm.

This article demonstrates the realization of angle tracking and deformation suppression by developing two boundary controllers for a flexible variable-length rotary crane arm with extraneous disturbances and asymmetric input-output constraints. The dynamic model description of this kind of crane arm system is several partial differential equations integrated into few ordinary differential equations. The S-curve acceleration and deceleration scheme is utilized to adjust the elongation rate of the arm. A kind of novel observer is put forward to tackle unknown extraneous disturbances. Auxiliary systems and barrier Lyapunov functions are introduced to meet the asymmetric input-output constraints. With the help of Lyapunov's theory, the global exponential stability and uniform boundedness are analyzed. The numerical simulations are finally provided to illuminate its availability of the designed control schemes.

An efficient functional magnetic resonance imaging data reduction strategy using neighborhood preserving embedding algorithm.

High dimensionality data have become common in neuroimaging fields, especially group-level functional magnetic resonance imaging (fMRI) datasets. fMRI connectivity analysis is a widely used, powerful technique for studying functional brain networks to probe underlying mechanisms of brain function and neuropsychological disorders. However, data-driven technique like independent components analysis (ICA), can yield unstable and inconsistent results, confounding the true effects of interest and hindering the understanding of brain functionality and connectivity. A key contributing factor to this instability is the information loss that occurs during fMRI data reduction. Data reduction of high dimensionality fMRI data in the temporal domain to identify the important information within group datasets is necessary for such analyses and is crucial to ensure the accuracy and stability of the outputs. In this study, we describe an fMRI data reduction strategy based on an adapted neighborhood preserving embedding (NPE) algorithm. Both simulated and real data results indicate that, compared with the widely used data reduction method, principal component analysis, the NPE-based data reduction method (a) shows superior performance on efficient data reduction, while enhancing group-level information, (b) develops a unique stratagem for selecting components based on an adjacency graph of eigenvectors, (c) generates more reliable and reproducible brain networks under different model orders when the outputs of NPE are used for ICA, (d) is more sensitive to revealing task-evoked activation for task fMRI, and (e) is extremely attractive and powerful for the increasingly popular fast fMRI and very large datasets.

Delay-Dependent Distributed Kalman Fusion Estimation With Dimensionality Reduction in Cyber-Physical Systems.

This article studies the distributed dimensionality reduction fusion estimation problem with communication delays for a class of cyber-physical systems (CPSs). The raw measurements are preprocessed in each sink node to obtain the local optimal estimate (LOE) of a CPS, and the compressed LOE under dimensionality reduction encounters with communication delays during the transmission. Under this case, a mathematical model with compensation strategy is proposed to characterize the dimensionality reduction and communication delays. This model also has the property of reducing the information loss caused by the dimensionality reduction and delays. Based on this model, a recursive distributed Kalman fusion estimator (DKFE) is derived by optimal weighted fusion criterion in the linear minimum variance sense. A stability condition for the DKFE, which can be easily verified by the exiting software, is derived. In addition, this condition can guarantee that the estimation error covariance matrix of the DKFE converges to the unique steady-state matrix for any initial values and, thus, the steady-state DKFE (SDKFE) is given. Note that the computational complexity of the SDKFE is much lower than that of the DKFE. Moreover, a probability selection criterion for determining the dimensionality reduction strategy is also presented to guarantee the stability of the DKFE. Two illustrative examples are given to show the advantage and effectiveness of the proposed methods.

LightSleepNet: A Lightweight Deep Model for Rapid Sleep Stage Classification with Spectrograms.

Deep learning has achieved unprecedented success in sleep stage classification tasks, which starts to pave the way for potential real-world applications. However, due to its enormous size, deployment of deep neural networks is hindered by high cost at various aspects, such as computation power, storage, network bandwidth, power consumption, and hardware complexity. For further practical applications (e.g., wearable sleep monitoring devices), there is a need for simple and compact models. In this paper, we propose a lightweight model, namely LightSleepNet, for rapid sleep stage classification based on spectrograms. Our model is assembled by a much fewer number of model parameters compared to existing ones. Furthermore, we convert the raw EEG data into spectrograms to speed up the training process. We evaluate the model performance on several public sleep datasets with different characteristics. Experimental results show that our lightweight model using spectrogram as input can achieve comparable overall accuracy and Cohen's kappa (SHHS100: 86.7%-81.3%, Sleep-EDF: 83.7%-77.5%, Sleep-EDF-v1: 88.3%-84.5%) compared to the state-of-the-art methods on experimental datasets.

The key role of pulse wave transit time to predict blood pressure variation during anaesthesia induction.

OBJECTIVE: To establish the relationship between pulse wave transit time (PWTT) before anaesthesia induction and blood pressure variability (BPV) during anaesthesia induction. METHODS: This prospective observational cohort study enrolled consecutive patients that underwent elective surgery. Invasive arterial pressure, electrocardiography, pulse oximetry, heart rate and bispectral index were monitored. PWTT and BPV were measured with special software. Anaesthesia was induced with propofol, sufentanil and rocuronium. RESULTS: A total of 54 patients were included in this study. There was no correlation between BPV and the dose of propofol, sufentanil and rocuronium during anaesthesia induction. Bivariate linear regression analysis demonstrated that PWTT (r = -0.54), age (r = 0.34) and systolic blood pressure (r = 0.31) significantly correlated with systolic blood pressure variability (SBPV). Only PWTT (r = -0.38) was significantly correlated with diastolic blood pressure variability (DBPV). Patients were stratified into high PWTT and low PWTT groups according to the mean PWTT value (96.8 ± 17.2 ms). Compared with the high PWTT group, the SBPV of the low PWTT group increased significantly by 3.4%. The DBPV of the low PWTT group increased significantly by 2.1% compared with the high PWTT group. CONCLUSIONS: PWTT, assessed before anaesthesia induction, may be an effective predictor of haemodynamic fluctuations during anaesthesia induction.