Geographically distributed data management to support large-scale data analysis.

Nowadays, several companies prefer storing their data on multiple data centers with replication for many reasons. The data that spans various data centers ensures the fastest possible response time for customers and workforces who are geographically separated. It also provides protecting the information from the loss in case a single data center experiences a disaster. However, the amount of data is increasing at a rapid pace, which leads to challenges in storage, analysis, and various processing tasks. In this paper, we propose and design a geographically distributed data management framework to manage the massive data stored and distributed among geo-distributed data centers. The goal of the proposed framework is to enable efficient use of the distributed data blocks for various data analysis tasks. The architecture of the proposed framework is composed of a grid of geo-distributed data centers connected to a data controller (DCtrl). The DCtrl is responsible for organizing and managing the block replicas across the geo-distributed data centers. We use the BDMS system as the installed system on the distributed data centers. BDMS stores the big data file as a set of random sample data blocks, each being a random sample of the whole data file. Then, DCtrl distributes these data blocks into multiple data centers with replication. In analyzing a big data file distributed based on the proposed framework, we randomly select a sample of data blocks replicated from other data centers on any data center. We use simulation results to demonstrate the performance of the proposed framework in big data analysis across geo-distributed data centers.

Face and Content Validity of an Artificial Eye Model for Secondary Intraocular Lens Fixation via Yamane Technique.

Purpose: To assess the face and content validity of an artificial eye model for secondary intraocular lens (IOL) fixation via the Yamane technique. Methods: Ophthalmologists and residents participated in a 90-minute simulation session on secondary IOL fixation via the Yamane technique. Hands-on practice of this technique was performed on an artificial eye, the Bioniko Okulo BR8. After, all ophthalmologists answered an 18-question survey assessing the face and content validity of the model. Survey responses were recorded on a 5-point double-headed Likert scale, ranging from strongly agree (1)-to-strongly disagree (5) (Figure 1). Results: Twenty-three surveys were completed. Respondents rated the survey with a median response of 1 (strongly agree)-to-3 (neutral). Highest ratings for the model were received for "usefulness for training residents", and "easier to set up and clean-up compared to a cadaver". Lowest ratings were received for realism of the model compared to cadaveric eyes. Statistical analysis revealed no significant difference among identified groups. Ratings for face and content validity were viewed favorably, both with an overall median response of 2.00 (agree). Conclusion: The Bioniko Okulo BR8 shows promise as a valid tool for practicing secondary IOL fixation via the Yamane technique. Considering recent guidelines in competency-based ophthalmology education programs, this model may be a valuable tool over traditional techniques for teaching and improving surgical skill amongst trainees.

Comparison of 3D heads-up display system with conventional surgical microscopy for minimally invasive glaucoma surgery on an artificial eye model.

OBJECTIVE: To survey ophthalmic surgeons' opinions comparing a novel three-dimensional (3D) heads-up display system with a conventional surgical microscopy for minimally invasive glaucoma surgery (MIGS) on an artificial eye model. MATERIALS AND METHODS: Twenty-one ophthalmologists at the 2021 Canadian Ophthalmological Society Annual Meeting in Halifax, Nova Scotia, underwent a 90-minute skills-transfer course on MIGS. Using an artificial eye model (SimulEYE iTrack Model; InsEYE LLC, Westlake Village, Calif.), participants engaged in hands-on practice of MIGS via both a 3D heads-up display system (3D HUDS) (Zeiss Artevo 800; Carl Zeiss Meditec, Jena, Germany) and a conventional surgical microscope. Following completion, participants and instructors answered a 16-question survey comparing the 2 systems (3D HUDS vs conventional surgical microscope). Survey responses were recorded on a 9-point double-headed Likert scale ranging from strongly favour 3D HUDS (1) to strongly favour conventional surgical microscopy (9). Mann-Whitney U nonparametric analysis was used to compare instructor versus participants and experts versus nonexperts. RESULTS: Survey ratings favoured the 3D HUDS over the conventional surgical microscopy, with respondent ratings for all survey questions ranging from a response of 1 (strongly favour 3D HUDS) to 5 (equal). Mann-Whitney U statistical analysis revealed no significant difference between instructor versus participant as well as between expert versus nonexpert. Most ratings for the 3D HUDS were received for ergonomic setup of the surgical modality, depth of field (or) field of view, and usefulness in training residents for MIGS. Equal ratings for the 3D HUDS and conventional surgical microscope were received for system malfunctions and lag during surgery. CONCLUSIONS: The 3D HUDS was favoured over conventional microscopy for the performance of simulated MIGS by ophthalmologists with varying levels of experience. The survey results suggest that the 3D HUDS in an artificial eye model is useful for teaching minimally invasive glaucoma surgery, particularly with the advent of competency-based ophthalmology education programs.

Data quality model for assessing public COVID-19 big datasets.

For decision-making support and evidence based on healthcare, high quality data are crucial, particularly if the emphasized knowledge is lacking. For public health practitioners and researchers, the reporting of COVID-19 data need to be accurate and easily available. Each nation has a system in place for reporting COVID-19 data, albeit these systems' efficacy has not been thoroughly evaluated. However, the current COVID-19 pandemic has shown widespread flaws in data quality. We propose a data quality model (canonical data model, four adequacy levels, and Benford's law) to assess the quality issue of COVID-19 data reporting carried out by the World Health Organization (WHO) in the six Central African Economic and Monitory Community (CEMAC) region countries between March 6,2020, and June 22, 2022, and suggest potential solutions. These levels of data quality sufficiency can be interpreted as dependability indicators and sufficiency of Big Dataset inspection. This model effectively identified the quality of the entry data for big dataset analytics. The future development of this model requires scholars and institutions from all sectors to deepen their understanding of its core concepts, improve integration with other data processing technologies, and broaden the scope of its applications.

A novel correlation Gaussian process regression-based extreme learning machine.

An obvious defect of extreme learning machine (ELM) is that its prediction performance is sensitive to the random initialization of input-layer weights and hidden-layer biases. To make ELM insensitive to random initialization, GPRELM adopts the simple an effective strategy of integrating Gaussian process regression into ELM. However, there is a serious overfitting problem in kernel-based GPRELM (kGPRELM). In this paper, we investigate the theoretical reasons for the overfitting of kGPRELM and further propose a correlation-based GPRELM (cGPRELM), which uses a correlation coefficient to measure the similarity between two different hidden-layer output vectors. cGPRELM reduces the likelihood that the covariance matrix becomes an identity matrix when the number of hidden-layer nodes is increased, effectively controlling overfitting. Furthermore, cGPRELM works well for improper initialization intervals where ELM and kGPRELM fail to provide good predictions. The experimental results on real classification and regression data sets demonstrate the feasibility and superiority of cGPRELM, as it not only achieves better generalization performance but also has a lower computational complexity.

Trajectories of Neurologic Recovery 12 Months After Hospitalization for COVID-19: A Prospective Longitudinal Study.

BACKGROUND AND OBJECTIVE: Little is known about trajectories of recovery 12 months after hospitalization for severe COVID-19. METHODS: We conducted a prospective, longitudinal cohort study of patients with and without neurologic complications during index hospitalization for COVID-19 from March 10, 2020, to May 20, 2020. Phone follow-up batteries were performed at 6 and 12 months after COVID-19 onset. The primary 12-month outcome was the modified Rankin Scale (mRS) score comparing patients with or without neurologic complications using multivariable ordinal analysis. Secondary outcomes included activities of daily living (Barthel Index), telephone Montreal Cognitive Assessment (t-MoCA), and Quality of Life in Neurologic Disorders (Neuro-QoL) batteries for anxiety, depression, fatigue, and sleep. Changes in outcome scores from 6 to 12 months were compared using nonparametric paired-samples sign test. RESULTS: Twelve-month follow-up was completed in 242 patients (median age 65 years, 64% male, 34% intubated during hospitalization) and 174 completed both 6- and 12-month follow-up. At 12 months, 197/227 (87%) had ≥1 abnormal metric: mRS >0 (75%), Barthel Index <100 (64%), t-MoCA ≤18 (50%), high anxiety (7%), depression (4%), fatigue (9%), or poor sleep (10%). Twelve-month mRS scores did not differ significantly among those with (n = 113) or without (n = 129) neurologic complications during hospitalization after adjusting for age, sex, race, pre-COVID-19 mRS, and intubation status (adjusted OR 1.4, 95% CI 0.8-2.5), although those with neurologic complications had higher fatigue scores (T score 47 vs 44; p = 0.037). Significant improvements in outcome trajectories from 6 to 12 months were observed in t-MoCA scores (56% improved, median difference 1 point; p = 0.002) and Neuro-QoL anxiety scores (45% improved; p = 0.003). Nonsignificant improvements occurred in fatigue, sleep, and depression scores in 48%, 48%, and 38% of patients, respectively. Barthel Index and mRS scores remained unchanged between 6 and 12 months in >50% of patients. DISCUSSION: At 12 months after hospitalization for severe COVID-19, 87% of patients had ongoing abnormalities in functional, cognitive, or Neuro-QoL metrics and abnormal cognition persisted in 50% of patients without a history of dementia/cognitive abnormality. Only fatigue severity differed significantly between patients with or without neurologic complications during index hospitalization. However, significant improvements in cognitive (t-MoCA) and anxiety (Neuro-QoL) scores occurred in 56% and 45% of patients, respectively, between 6 and 12 months. These results may not be generalizable to those with mild or moderate COVID-19.

FINC: An Efficient and Effective Optimization Method for Normalized Cut.

The optimization methods for solving the normalized cut model usually involve three steps, i.e., problem relaxation, problem solving and post-processing. However, these methods are problematic in both performance since they do not directly solve the original problem, and efficiency since they usually depend on the time-consuming eigendecomposition and k-means (or spectral rotation) for post-processing. In this paper, we propose a fast optimization method to speedup the classical normalized cut clustering process, in which an auxiliary variable is introduced and alternatively updated along with the cluster indicator matrix. The new method is faster than the conventional three-step optimization methods since it solves the normalized cut problem in one step. Theoretical analysis reveals that the new method is able to monotonically decrease the normalized cut objective function and converge in finite iterations. Moreover, we have proposed efficient methods for adjust two regularization parameters. Extensive experimental results show the superior performance of the new method. Moreover, it is faster than the existing methods for solving the normalized cut.

Comparison of serum neurodegenerative biomarkers among hospitalized COVID-19 patients versus non-COVID subjects with normal cognition, mild cognitive impairment, or Alzheimer's dementia.

INTRODUCTION: Neurological complications among hospitalized COVID-19 patients may be associated with elevated neurodegenerative biomarkers. METHODS: Among hospitalized COVID-19 patients without a history of dementia (N = 251), we compared serum total tau (t-tau), phosphorylated tau-181 (p-tau181), glial fibrillary acidic protein (GFAP), neurofilament light chain (NfL), ubiquitin carboxy-terminal hydrolase L1 (UCHL1), and amyloid beta (Aß40,42) between patients with or without encephalopathy, in-hospital death versus survival, and discharge home versus other dispositions. COVID-19 patient biomarker levels were also compared to non-COVID cognitively normal, mild cognitive impairment (MCI), and Alzheimer's disease (AD) dementia controls (N = 161). RESULTS: Admission t-tau, p-tau181, GFAP, and NfL were significantly elevated in patients with encephalopathy and in those who died in-hospital, while t-tau, GFAP, and NfL were significantly lower in those discharged home. These markers correlated with severity of COVID illness. NfL, GFAP, and UCHL1 were higher in COVID patients than in non-COVID controls with MCI or AD. DISCUSSION: Neurodegenerative biomarkers were elevated to levels observed in AD dementia and associated with encephalopathy and worse outcomes among hospitalized COVID-19 patients.

A prospective study of long-term outcomes among hospitalized COVID-19 patients with and without neurological complications.

BACKGROUND: Little is known regarding long-term outcomes of patients hospitalized with COVID-19. METHODS: We conducted a prospective study of 6-month outcomes of hospitalized COVID-19 patients. Patients with new neurological complications during hospitalization who survived were propensity score-matched to COVID-19 survivors without neurological complications hospitalized during the same period. The primary 6-month outcome was multivariable ordinal analysis of the modified Rankin Scale(mRS) comparing patients with or without neurological complications. Secondary outcomes included: activities of daily living (ADLs;Barthel Index), telephone Montreal Cognitive Assessment and Neuro-QoL batteries for anxiety, depression, fatigue and sleep. RESULTS: Of 606 COVID-19 patients with neurological complications, 395 survived hospitalization and were matched to 395 controls; N = 196 neurological patients and N = 186 controls completed follow-up. Overall, 346/382 (91%) patients had at least one abnormal outcome: 56% had limited ADLs, 50% impaired cognition, 47% could not return to work and 62% scored worse than average on ≥1 Neuro-QoL scale (worse anxiety 46%, sleep 38%, fatigue 36%, and depression 25%). In multivariable analysis, patients with neurological complications had worse 6-month mRS (median 4 vs. 3 among controls, adjusted OR 1.98, 95%CI 1.23-3.48, P = 0.02), worse ADLs (aOR 0.38, 95%CI 0.29-0.74, P = 0.01) and were less likely to return to work than controls (41% versus 64%, P = 0.04). Cognitive and Neuro-QOL metrics were similar between groups. CONCLUSIONS: Abnormalities in functional outcomes, ADLs, anxiety, depression and sleep occurred in over 90% of patients 6-months after hospitalization for COVID-19. In multivariable analysis, patients with neurological complications during index hospitalization had significantly worse 6-month functional outcomes than those without.

Toxic Metabolic Encephalopathy in Hospitalized Patients with COVID-19.

BACKGROUND: Toxic metabolic encephalopathy (TME) has been reported in 7-31% of hospitalized patients with coronavirus disease 2019 (COVID-19); however, some reports include sedation-related delirium and few data exist on the etiology of TME. We aimed to identify the prevalence, etiologies, and mortality rates associated with TME in severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-positive patients. METHODS: We conducted a retrospective, multicenter, observational cohort study among patients with reverse transcriptase-polymerase chain reaction-confirmed SARS-CoV-2 infection hospitalized at four New York City hospitals in the same health network between March 1, 2020, and May 20, 2020. TME was diagnosed in patients with altered mental status off sedation or after an adequate sedation washout. Patients with structural brain disease, seizures, or primary neurological diagnoses were excluded. The coprimary outcomes were the prevalence of TME stratified by etiology and in-hospital mortality (excluding comfort care only patients) assessed by using a multivariable time-dependent Cox proportional hazards models with adjustment for age, race, sex, intubation, intensive care unit requirement, Sequential Organ Failure Assessment scores, hospital location, and date of admission. RESULTS: Among 4491 patients with COVID-19, 559 (12%) were diagnosed with TME, of whom 435 of 559 (78%) developed encephalopathy immediately prior to hospital admission. The most common etiologies were septic encephalopathy (n = 247 of 559 [62%]), hypoxic-ischemic encephalopathy (HIE) (n = 331 of 559 [59%]), and uremia (n = 156 of 559 [28%]). Multiple etiologies were present in 435 (78%) patients. Compared with those without TME (n = 3932), patients with TME were older (76 vs. 62 years), had dementia (27% vs. 3%) or psychiatric history (20% vs. 10%), were more often intubated (37% vs. 20%), had a longer hospital length of stay (7.9 vs. 6.0 days), and were less often discharged home (25% vs. 66% [all P < 0.001]). Excluding comfort care patients (n = 267 of 4491 [6%]) and after adjustment for confounders, TME remained associated with increased risk of in-hospital death (n = 128 of 425 [30%] patients with TME died, compared with n = 600 of 3799 [16%] patients without TME; adjusted hazard ratio [aHR] 1.24, 95% confidence interval [CI] 1.02-1.52, P = 0.031), and TME due to hypoxemia conferred the highest risk (n = 97 of 233 [42%] patients with HIE died, compared with n = 631 of 3991 [16%] patients without HIE; aHR 1.56, 95% CI 1.21-2.00, P = 0.001). CONCLUSIONS: TME occurred in one in eight hospitalized patients with COVID-19, was typically multifactorial, and was most often due to hypoxemia, sepsis, and uremia. After we adjustment for confounding factors, TME was associated with a 24% increased risk of in-hospital mortality.

A Prospective Study of Neurologic Disorders in Hospitalized Patients With COVID-19 in New York City.

OBJECTIVE: To determine the prevalence and associated mortality of well-defined neurologic diagnoses among patients with coronavirus disease 2019 (COVID-19), we prospectively followed hospitalized severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-positive patients and recorded new neurologic disorders and hospital outcomes. METHODS: We conducted a prospective, multicenter, observational study of consecutive hospitalized adults in the New York City metropolitan area with laboratory-confirmed SARS-CoV-2 infection. The prevalence of new neurologic disorders (as diagnosed by a neurologist) was recorded and in-hospital mortality and discharge disposition were compared between patients with COVID-19 with and without neurologic disorders. RESULTS: Of 4,491 patients with COVID-19 hospitalized during the study timeframe, 606 (13.5%) developed a new neurologic disorder in a median of 2 days from COVID-19 symptom onset. The most common diagnoses were toxic/metabolic encephalopathy (6.8%), seizure (1.6%), stroke (1.9%), and hypoxic/ischemic injury (1.4%). No patient had meningitis/encephalitis or myelopathy/myelitis referable to SARS-CoV-2 infection and 18/18 CSF specimens were reverse transcriptase PCR negative for SARS-CoV-2. Patients with neurologic disorders were more often older, male, white, hypertensive, diabetic, intubated, and had higher sequential organ failure assessment (SOFA) scores (all p < 0.05). After adjusting for age, sex, SOFA scores, intubation, history, medical complications, medications, and comfort care status, patients with COVID-19 with neurologic disorders had increased risk of in-hospital mortality (hazard ratio [HR] 1.38, 95% confidence interval [CI] 1.17-1.62, p < 0.001) and decreased likelihood of discharge home (HR 0.72, 95% CI 0.63-0.85, p < 0.001). CONCLUSIONS: Neurologic disorders were detected in 13.5% of patients with COVID-19 and were associated with increased risk of in-hospital mortality and decreased likelihood of discharge home. Many observed neurologic disorders may be sequelae of severe systemic illness.

The knowledge transfer potential of online data pools on nature-based solutions.

Improving the adoption of Nature-based Solutions (NBS) requires learning from successes and failures. Knowledge derived from implemented cases helps to identify for instance drivers and barriers of NBS implementation, generates lessons learned, and supports their upscaling. Online data pools that catalogue information from NBS case studies may help scientists and practitioners to create this knowledge. The aim of this review is to assess the knowledge transfer potential of online data pools for implementing and upscaling NBS. For that, we compared 21 online data pools that report on NBS case studies in terms of topics, availability and quality of information on NBS. We found a high variability in quantity, type and quality of the information documented, hindering comparability and limiting knowledge transfer. Our results show that the most common knowledge provided was on actions undertaken on NBS, their outcomes, case study site descriptions, specific challenges and information on responsible entities and partners. Information on key attributes of NBS, such as on ecosystem processes and services as well as on governance and financing issues, was often omitted. The missing information however would be important for further comparative research to overcome implementation gaps for NBS. Based on the discussion of our findings we propose categories for a more efficient online data pool and give recommendations for further research on NBS.

Governance models for nature-based solutions: Seventeen cases from Germany.

Nature-based solutions (NBS) for mitigating climate change are gaining popularity. The number of NBS is increasing, but research gaps still exist at the governance level. The objectives of this paper are (i) to give an overview of the implemented NBS for flood risk management and mitigation in Germany, (ii) to identify governance models that are applied, and (iii) to explore the differences between these models. The results of a hierarchical clustering procedure and a qualitative analysis show that while no one-size-fits-all governance model exists, polycentricism is an important commonality between the projects. The study concludes by highlighting the need for further research on traditional governance model reconversion and paradigm changes. We expect the findings to identify what has worked in the past, as well as what is important for the implementation of NBS for flood risk management in future projects.

Activeness and Loyalty Analysis in Event-Based Social Networks.

Event-based social networks (EBSNs) are widely used to create online social groups and organize offline events for users. Activeness and loyalty are crucial characteristics of these online social groups in terms of determining the growth or inactiveness of the social groups in a specific time frame. However, there is less research on these concepts to clarify the existence of groups in event-based social networks. In this paper, we study the problem of group activeness and user loyalty to provide a novel insight into online social networks. First, we analyze the structure of EBSNs and generate features from the crawled datasets. Second, we define the concepts of group activeness and user loyalty based on a series of time windows, and propose a method to measure the group activeness. In this proposed method, we first compute a ratio of a number of events between two consecutive time windows. We then develop an association matrix to assign the activeness label for each group after several consecutive time windows. Similarly, we measure the user loyalty in terms of attended events gathered in time windows and treat loyalty as a contributive feature of the group activeness. Finally, three well-known machine learning techniques are used to verify the activeness label and to generate features for each group. As a consequence, we also find a small group of features that are highly correlated and result in higher accuracy as compared to the whole features.

Treatment with Zinc is Associated with Reduced In-Hospital Mortality Among COVID-19 Patients: A Multi-Center Cohort Study.

Background: Zinc impairs replication of RNA viruses such as SARS-CoV-1, and may be effective against SARS-CoV-2. However, to achieve adequate intracellular zinc levels, administration with an ionophore, which increases intracellular zinc levels, may be necessary. We evaluated the impact of zinc with an ionophore (Zn+ionophore) on COVID-19 in-hospital mortality rates. Methods: A multicenter cohort study was conducted of 3,473 adult hospitalized patients with reverse-transcriptase-polymerase-chain-reaction (RT-PCR) positive SARS-CoV-2 infection admitted to four New York City hospitals between March 10 through May 20, 2020. Exclusion criteria were: death or discharge within 24h, comfort-care status, clinical trial enrollment, treatment with an IL-6 inhibitor or remdesivir. Patients who received Zn+ionophore were compared to patients who did not using multivariable time-dependent cox proportional hazards models for time to in-hospital death adjusting for confounders including age, sex, race, BMI, diabetes, week of admission, hospital location, sequential organ failure assessment (SOFA) score, intubation, acute renal failure, neurological events, treatment with corticosteroids, azithromycin or lopinavir/ritonavir and the propensity score of receiving Zn+ionophore. A sensitivity analysis was performed using a propensity score-matched cohort of patients who did or did not receive Zn+ionophore matched by age, sex and ventilator status. Results: Among 3,473 patients (median age 64, 1947 [56%] male, 522 [15%] ventilated, 545[16%] died), 1,006 (29%) received Zn+ionophore. Zn+ionophore was associated with a 24% reduced risk of in-hospital mortality (12% of those who received Zn+ionophore died versus 17% who did not; adjusted Hazard Ratio [aHR] 0.76, 95% CI 0.60-0.96, P=0.023). More patients who received Zn+ionophore were discharged home (72% Zn+ionophore vs 67% no Zn+ionophore, P=0.003) Neither Zn nor the ionophore alone were associated with decreased mortality rates. Propensity score-matched sensitivity analysis (N=1356) validated these results (Zn+ionophore aHR for mortality 0.63, 95%CI 0.44-0.91, P=0.015). There were no significant interactions for Zn+ionophore with other COVID-19 specific medications. Conclusions: Zinc with an ionophore was associated with increased rates of discharge home and a 24% reduced risk of in-hospital mortality among COVID-19 patients, while neither zinc alone nor the ionophore alone reduced mortality. Further randomized trials are warranted.

Prevalence and Impact of Hyponatremia in Patients With Coronavirus Disease 2019 in New York City.

OBJECTIVES: Hyponatremia occurs in up to 30% of patients with pneumonia and is associated with increased morbidity and mortality. The prevalence of hyponatremia associated with coronavirus disease 2019 and the impact on outcome is unknown. We aimed to identify the prevalence, predictors, and impact on outcome of mild, moderate, and severe admission hyponatremia compared with normonatremia among coronavirus disease 2019 patients. DESIGN: Retrospective, multicenter, observational cohort study. SETTING: Four New York City hospitals that are part of the same health network. PATIENTS: Hospitalized, laboratory-confirmed adult coronavirus disease 2019 patients admitted between March 1, 2020, and May 13, 2020. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Hyponatremia was categorized as mild (sodium: 130-134 mmol/L), moderate (sodium: 121-129 mmol/L), or severe (sodium: ≤ 120 mmol/L) versus normonatremia (135-145 mmol/L). The primary outcome was the association of increasing severity of hyponatremia and in-hospital mortality assessed using multivariable logistic regression analysis. Secondary outcomes included encephalopathy, acute renal failure, mechanical ventilation, and discharge home compared across sodium levels using Kruskal-Wallis and chi-square tests. In exploratory analysis, the association of sodium levels and interleukin-6 levels (which has been linked to nonosmotic release of vasopressin) was assessed. Among 4,645 patient encounters, hyponatremia (sodium < 135 mmol/L) occurred in 1,373 (30%) and 374 of 1,373 (27%) required invasive mechanical ventilation. Mild, moderate, and severe hyponatremia occurred in 1,032 (22%), 305 (7%), and 36 (1%) patients, respectively. Each level of worsening hyponatremia conferred 43% increased odds of in-hospital death after adjusting for age, gender, race, body mass index, past medical history, admission laboratory abnormalities, admission Sequential Organ Failure Assessment score, renal failure, encephalopathy, and mechanical ventilation (adjusted odds ratio, 1.43; 95% CI, 1.08-1.88; p = 0.012). Increasing severity of hyponatremia was associated with encephalopathy, mechanical ventilation, and decreased probability of discharge home (all p < 0.001). Higher interleukin-6 levels correlated with lower sodium levels (p = 0.017). CONCLUSIONS: Hyponatremia occurred in nearly a third of coronavirus disease 2019 patients, was an independent predictor of in-hospital mortality, and was associated with increased risk of encephalopathy and mechanical ventilation.

LABIN: Balanced Min Cut for Large-Scale Data.

Although many spectral clustering algorithms have been proposed during the past decades, they are not scalable to large-scale data due to their high computational complexities. In this paper, we propose a novel spectral clustering method for large-scale data, namely, large-scale balanced min cut (LABIN). A new model is proposed to extend the self-balanced min-cut (SBMC) model with the anchor-based strategy and a fast spectral rotation with linear time complexity is proposed to solve the new model. Extensive experimental results show the superior performance of our proposed method in comparison with the state-of-the-art methods including SBMC.

SS-31 Peptide Reverses the Mitochondrial Fragmentation Present in Fibroblasts From Patients With DCMA, a Mitochondrial Cardiomyopathy.

We used patient dermal fibroblasts to characterize the mitochondrial abnormalities associated with the dilated cardiomyopathy with ataxia syndrome (DCMA) and to study the effect of the mitochondrially-targeted peptide SS-31 as a potential novel therapeutic. DCMA is a rare and understudied autosomal recessive disorder thought to be related to Barth syndrome but caused by mutations in DNAJC19, a protein of unknown function localized to the mitochondria. The clinical disease is characterized by 3-methylglutaconic aciduria, dilated cardiomyopathy, abnormal neurological development, and other heterogeneous features. Until recently no effective therapies had been identified and affected patients frequently died in early childhood from intractable heart failure. Skin fibroblasts from four pediatric patients with DCMA were used to establish parameters of mitochondrial dysfunction. Mitochondrial structure, reactive oxygen species (ROS) production, cardiolipin composition, and gene expression were evaluated. Immunocytochemistry with semi-automated quantification of mitochondrial structural metrics and transmission electron microscopy demonstrated mitochondria to be highly fragmented in DCMA fibroblasts compared to healthy control cells. Live-cell imaging demonstrated significantly increased ROS production in patient cells. These abnormalities were reversed by treating DCMA fibroblasts with SS-31, a synthetic peptide that localizes to the inner mitochondrial membrane. Levels of cardiolipin were not significantly different between control and DCMA cells and were unaffected by SS-31 treatment. Our results demonstrate the abnormal mitochondria in fibroblasts from patients with DCMA and suggest that SS-31 may represent a potential therapy for this devastating disease.

Subspace Weighting Co-Clustering of Gene Expression Data.

Microarray technology enables the collection of vast amounts of gene expression data from biological experiments. Clustering algorithms have been successfully applied to exploring the gene expression data. Since a set of genes may be only correlated to a subset of samples, it is useful to use co-clustering to recover co-clusters in the gene expression data. In this paper, we propose a novel algorithm, called Subspace Weighting Co-Clustering (SWCC), for high dimensional gene expression data. In SWCC, a gene subspace weight matrix is introduced to identify the contribution of gene objects in distinguishing different sample clusters. We design a new co-clustering objective function to recover the co-clusters in the gene expression data, in which the subspace weight matrix is introduced. An iterative algorithm is developed to solve the objective function, in which the subspace weight matrix is automatically computed during the iterative co-clustering process. Our empirical study shows encouraging results of the proposed algorithm in comparison with six state-of-the-art clustering algorithms on ten gene expression data sets. We also propose to use SWCC for gene clustering and selection. The experimental results show that the selected genes can improve the classification performance of Random Forests.

An Algorithm for Clustering Categorical Data With Set-Valued Features.

In data mining, objects are often represented by a set of features, where each feature of an object has only one value. However, in reality, some features can take on multiple values, for instance, a person with several job titles, hobbies, and email addresses. These features can be referred to as set-valued features and are often treated with dummy features when using existing data mining algorithms to analyze data with set-valued features. In this paper, we propose an SV- $k$ -modes algorithm that clusters categorical data with set-valued features. In this algorithm, a distance function is defined between two objects with set-valued features, and a set-valued mode representation of cluster centers is proposed. We develop a heuristic method to update cluster centers in the iterative clustering process and an initialization algorithm to select the initial cluster centers. The convergence and complexity of the SV- $k$ -modes algorithm are analyzed. Experiments are conducted on both synthetic data and real data from five different applications. The experimental results have shown that the SV- $k$ -modes algorithm performs better when clustering real data than do three other categorical clustering algorithms and that the algorithm is scalable to large data.