Investigation of factors regarding the effects of COVID-19 pandemic on college students' depression by quantum annealer.

Diverse cases regarding the impact, with its related factors, of the COVID-19 pandemic on mental health have been reported in previous studies. In this study, multivariable datasets were collected from 751 college students who could be easily affected by pandemics based on the complex relationships between various mental health factors. We utilized quantum annealing (QA)-based feature selection algorithms that were executed by commercial D-Wave quantum computers to determine the changes in the relative importance of the associated factors before and after the pandemic. Multivariable linear regression (MLR) and XGBoost models were also applied to validate the QA-based algorithms. Based on the experimental results, we confirm that QA-based algorithms have comparable capabilities in factor analysis research to the MLR models that have been widely used in previous studies. Furthermore, the performance of the QA-based algorithms was validated through the important factor results from the algorithms. Pandemic-related factors (e.g., confidence in the social system) and psychological factors (e.g. decision-making in uncertain situations) were more important in post-pandemic conditions. Although the results should be validated using other mental health variables or national datasets, this study will serve as a reference for researchers regarding the use of the quantum annealing approach in factor analysis with validation through real-world survey dataset analysis.

Investigation of factors associated with mental health during the early part of the COVID-19 pandemic in South Korea based on machine learning algorithms: A cohort study.

Objective: The coronavirus disease 2019 (COVID-19) pandemic is among the most critical public health problems worldwide in the last three years. We tried to investigate changes in factors between pre- and early stages of the COVID-19 pandemic. Methods: The data of 457,309 participants from the 2019 and 2020 Community Health Survey were examined. Four mental health-related variables were selected for examination as a dependent variable (patient health questionnaire-9, depression, stress, and sleep time). Other variables without the aforementioned four variables were split into three groups based on the coefficient values of lasso and ridge regression models. The importance of each variable was calculated and compared using feature importance values obtained from three machine learning algorithms. Results: Psychiatric and sociodemographic variables were identified, both during the pre- and early pandemic periods. In contrast, during the early pandemic period, average sleep time variables ranked the highest with the dependent variables regarding the experience of depression. The difference in sleep time before and after the pandemic was validated by the results of paired t-tests, which were statistically significant (p-value < 0.05). Conclusions: Changes in the importance of mental health factors in the early pandemic period in South Korea were identified. For each mental health-dependent variable, average sleep time, experience of depression, and experience of accidents or addictions were found to be the most important factors. House type and type of residence were also found in regions with larger populations and a higher number of confirmed cases.

Validation of the influence of biosignals on performance of machine learning algorithms for sleep stage classification.

Background: Sleep stage identification is critical in multiple areas (e.g. medicine or psychology) to diagnose sleep-related disorders. Previous studies have reported that the performance of machine learning algorithms can be changed depending on the biosignals and feature-extraction processes in sleep stage classification. Methods: To compare as many conditions as possible, 414 experimental conditions were applied, considering the combination of different biosignals, biosignal length, and window length. Five biosignals in polysomnography (i.e. electrocardiogram (ECG), electroencephalogram (EEG), electromyogram (EMG), electrooculogram left, and electrooculogram right) were used to identify optimal signal combinations for classification. In addition, three different signal-length conditions and six different window-length conditions were applied. The validity of each condition was examined via classification performance from the XGBoost classifiers trained using 10-fold cross-validation. Furthermore, results considering feature importance were examined to validate the experimental results in terms of model explanation. Results: The combination of EEG + EMG + ECG with a 40 s window and 120 s signal length resulted in the best classification performance (precision: 0.853, recall: 0.855, F1-score: 0.853, and accuracy: 0.853). Compared to other conditions and feature importance results, EEG signals showed a relatively higher importance for classification in the present study. Conclusion: We determined the optimal biosignal and window conditions for the feature-extraction process in machine learning algorithm-based sleep stage classification. Our experimental results inform researchers in the future conduct of related studies. To generalize our results, more diverse methodologies and conditions should be applied in future studies.

Validation of depression determinants in caregivers of dementia patients with machine learning algorithms and statistical model.

Introduction: Due to its increasing prevalence, dementia is currently one of the most extensively studied health issues. Although it represents a comparatively less-addressed issue, the caregiving burden for dementia patients is likewise receiving attention. Methods: To identify determinants of depression in dementia caregivers, using Community Health Survey (CHS) data collected by the Korea Disease Control and Prevention Agency (KDCA). By setting "dementia caregiver's status of residence with patient" as a standard variable, we selected corresponding CHS data from 2011 to 2019. After refining the data, we split dementia caregiver and general population groups among the dataset (n = 15,708; common variables = 34). We then applied three machine learning algorithms: Extreme Gradient Boosting (XGBoost), Logistic Regression (LR), and Support Vector Classifier (SVC). Subsequently, we selected XGBoost, as it exhibited superior performance to the other algorithms. On the feature importance of XGBoost, we performed a multivariate hierarchical regression analysis to validate the depression causes experienced in each group. We validated the results of the statistical model analysis by performing Welch's t-test on the main determinants exhibited within each group. Results: By verifying the results from machine learning via statistical model analysis, we found "sex" to highly impact depression in dementia caregivers, whereas "status of economic activities" is significantly associated with depression in the general population. Discussion: The evident difference in causes of depression between the two groups may serve as a basis for policy development to improve the mental health of dementia caregivers.

Associations of actigraphy derived rest activity patterns and circadian phase with clinical symptoms and polysomnographic parameters in chronic insomnia disorders.

We explored the associations of actigraphy-derived rest-activity patterns and circadian phase parameters with clinical symptoms and level 1 polysomnography (PSG) results in patients with chronic insomnia to evaluate the clinical implications of actigraphy-derived parameters for PSG interpretation. Seventy-five participants underwent actigraphy assessments and level 1 PSG. Exploratory correlation analyses between parameters derived from actigraphy, PSG, and clinical assessments were performed. First, participants were classified into two groups based on rest-activity pattern variables; group differences were investigated following covariate adjustment. Participants with poorer rest-activity patterns on actigraphy (low inter-day stability and high intra-daily variability) exhibited higher insomnia severity index scores than participants with better rest-activity patterns. No between-group differences in PSG parameters were observed. Second, participants were classified into two groups based on circadian phase variables. Late-phase participants (least active 5-h and most active 10-h onset times) exhibited higher insomnia severity scores, longer sleep and rapid eye movement latency, and lower apnea-hypopnea index than early-phase participants. These associations remained significant even after adjusting for potential covariates. Some actigraphy-derived rest-activity patterns and circadian phase parameters were significantly associated with clinical symptoms and PSG results, suggesting their possible adjunctive role in deriving plans for PSG lights-off time and assessing the possible insomnia pathophysiology.

EEG-Based Identification of Emotional Neural State Evoked by Virtual Environment Interaction.

Classifying emotional states is critical for brain-computer interfaces and psychology-related domains. In previous studies, researchers have tried to identify emotions using neural data such as electroencephalography (EEG) signals or brain functional magnetic resonance imaging (fMRI). In this study, we propose a machine learning framework for emotion state classification using EEG signals in virtual reality (VR) environments. To arouse emotional neural states in brain signals, we provided three VR stimuli scenarios to 15 participants. Fifty-four features were extracted from the collected EEG signals under each scenario. To find the optimal classification in our research design, three machine learning algorithms (XGBoost classifier, support vector classifier, and logistic regression) were applied. Additionally, various class conditions were used in machine learning classifiers to validate the performance of our framework. To evaluate the classification performance, we utilized five evaluation metrics (precision, recall, f1-score, accuracy, and AUROC). Among the three classifiers, the XGBoost classifiers showed the best performance under all experimental conditions. Furthermore, the usability of features, including differential asymmetry and frequency band pass categories, were checked from the feature importance of XGBoost classifiers. We expect that our framework can be applied widely not only to psychological research but also to mental health-related issues.

Identification of Risk Factors for Suicidal Ideation and Attempt Based on Machine Learning Algorithms: A Longitudinal Survey in Korea (2007-2019).

Investigating suicide risk factors is critical for socioeconomic and public health, and many researchers have tried to identify factors associated with suicide. In this study, the risk factors for suicidal ideation were compared, and the contributions of different factors to suicidal ideation and attempt were investigated. To reflect the diverse characteristics of the population, the large-scale and longitudinal dataset used in this study included both socioeconomic and clinical variables collected from the Korean public. Three machine learning algorithms (XGBoost classifier, support vector classifier, and logistic regression) were used to detect the risk factors for both suicidal ideation and attempt. The importance of the variables was determined using the model with the best classification performance. In addition, a novel risk-factor score, calculated from the rank and importance scores of each variable, was proposed. Socioeconomic and sociodemographic factors showed a high correlation with risks for both ideation and attempt. Mental health variables ranked higher than other factors in suicidal attempts, posing a relatively higher suicide risk than ideation. These trends were further validated using the conditions from the integrated and yearly dataset. This study provides novel insights into suicidal risk factors for suicidal ideations and attempts.

Associations of rest-activity patterns with amyloid burden, medial temporal lobe atrophy, and cognitive impairment.

BACKGROUND: We sought to investigate the possible associations of rest-activity patterns with cortical amyloid burden, medial temporal lobe (MTL) neurodegeneration, and cognitive function in patients in the early stage of cognitive impairment. METHODS: Rest-activity patterns were assessed in 100 participants (70 with mild cognitive impairment and 30 with mild dementia) using wrist actigraphy. All participants underwent 18F-flutemetamol positron emission tomography (PET) imaging to quantify cortical amyloid burden, structural brain magnetic resonance imaging (MRI) to quantify MTL grey matter volume, neuropsychological testing, and clinical diagnosis. We used multiple linear regression models adjusted for covariates, including demographics, diabetes, hypertension, depressive symptom, psychotropic medication, sleep medication, weekend effect, and apolipoprotein-ε allele status. FINDINGS: After adjusting for possible confounders, we found that the midline estimation of statistic of rhythm (MESOR) associated positively with frontal/executive function (estimate = 1.17, standard error [SE] = 0.37, p = 0.002). The least active 5-h (L5) onset time associated positively with MTL grey matter volume and memory function (estimate = 1.24, SE = 0.33, p = 0.001, and estimate = 3.77, SE = 1.22, p = 0.003, respectively), particularly in amyloid-negative participants. Additional path analysis revealed that MTL grey matter volume partially mediated the association between L5 onset time and memory function in amyloid-negative participants. INTERPRETATION: Decreased MESOR and advanced L5 onset time may be useful as early signs of cognitive decline or MTL neurodegeneration. Furthermore, amyloid pathology may act as a moderator of the relationships between rest-activity patterns, neurodegeneration, and cognitive function. FUNDING: Korea Centres for Disease Control and Prevention (#4845-303); National Research Foundation of Korea (2019M3C7A1031905, 2019R1A5A2026045).

Deep Learning Approach for Imputation of Missing Values in Actigraphy Data: Algorithm Development Study.

BACKGROUND: Data collected by an actigraphy device worn on the wrist or waist can provide objective measurements for studies related to physical activity; however, some data may contain intervals where values are missing. In previous studies, statistical methods have been applied to impute missing values on the basis of statistical assumptions. Deep learning algorithms, however, can learn features from the data without any such assumptions and may outperform previous approaches in imputation tasks. OBJECTIVE: The aim of this study was to impute missing values in data using a deep learning approach. METHODS: To develop an imputation model for missing values in accelerometer-based actigraphy data, a denoising convolutional autoencoder was adopted. We trained and tested our deep learning-based imputation model with the National Health and Nutrition Examination Survey data set and validated it with the external Korea National Health and Nutrition Examination Survey and the Korean Chronic Cerebrovascular Disease Oriented Biobank data sets which consist of daily records measuring activity counts. The partial root mean square error and partial mean absolute error of the imputed intervals (partial RMSE and partial MAE, respectively) were calculated using our deep learning-based imputation model (zero-inflated denoising convolutional autoencoder) as well as using other approaches (mean imputation, zero-inflated Poisson regression, and Bayesian regression). RESULTS: The zero-inflated denoising convolutional autoencoder exhibited a partial RMSE of 839.3 counts and partial MAE of 431.1 counts, whereas mean imputation achieved a partial RMSE of 1053.2 counts and partial MAE of 545.4 counts, the zero-inflated Poisson regression model achieved a partial RMSE of 1255.6 counts and partial MAE of 508.6 counts, and Bayesian regression achieved a partial RMSE of 924.5 counts and partial MAE of 605.8 counts. CONCLUSIONS: Our deep learning-based imputation model performed better than the other methods when imputing missing values in actigraphy data.

Risk Evaluation of Azithromycin-Induced QT Prolongation in Real-World Practice.

BACKGROUND: Azithromycin exposure has been reported to increase the risk of QT prolongation and cardiovascular death. However, findings on the association between azithromycin and cardiovascular death are controversial, and azithromycin is still used in actual practice. Additionally, quantitative assessments of risk have not been performed, including the risk of QT prolongation when patients are exposed to azithromycin in a real-world clinical setting. Therefore, in this study, we aimed to evaluate the risk of exposure to azithromycin on QT prolongation in a real-world clinical setting using a 21-year medical history database of a tertiary medical institution. METHODS: We analyzed the electrocardiogram results and relevant electronic health records of 402,607 subjects in a tertiary teaching hospital in Korea from 1996 to 2015. To evaluate the risk of QT prolongation of azithromycin, we conducted a case-control analysis using amoxicillin for comparison. Multiple logistic regression analysis was performed to correct for age, sex, accompanying drugs, and disease. RESULTS: The odds ratio (OR) for QT prolongation (QTc>450 ms in male and >460 ms in female) on azithromycin exposure was 1.40 (95% confidence interval [CI], 1.23-1.59), and the OR for severe QT prolongation (QTc>500 ms) was 1.43 (95% CI, 1.13-1.82). On the other hand, the ORs on exposure to amoxicillin were 1.06 (95% CI, 0.97-1.15) and 0.88 (95% CI, 0.70-1.09). In a subgroup analysis, the risk of QT prolongation in patients aged between 60 and 80 years was significantly higher when they are exposed to azithromycin. CONCLUSIONS: The risk of QT prolongation was increased when patients, particularly the elderly aged 60-79 years, were exposed to azithromycin. Therefore, clinicians should pay exercise caution using azithromycin or consider using other antibiotics, such as amoxicillin, instead of azithromycin.

In vitro and in vivo defensive effect of probiotic LAB against Pseudomonas aeruginosa using Caenorhabditis elegans model.

This study aimed to investigate in vitro and in vivo the probiotic characteristics of lactic acid bacteria (LAB) isolated from Korean traditional fermented foods. Caenorhabditis elegans  (C. elegans) was used for analytical assays of fertility, chemotaxis, life-span, worm-killing and bacterial colonization in the intestinal lumen of the worm. All 35 strains of LAB reduced fertility and slowed development in the worms. The worm-killing assay showed that LAB significantly increased the lifespan (P < 0.05) and reduced the susceptibility to virulent PA14; however, the heat-killed LAB did not. The bacterial colonization assay revealed that LAB proliferated and protected the gut of the worm against infection by Pseudomonas aeruginosa PA14. In addition, specific LAB Pediococcus acidilactici(P. acidilactici DM-9), Pediococcus brevis (L. brevis SDL1411), and Pediococcus pentosaceus (P. pentosaceus SDL1409) strains showed acid resistance (66-91%), resistance to pepsin (64-67%) and viability in simulated intestinal fluid (67-73%) based on in vitro probiotic analyses. Taken together, these results suggest that C. elegans may be a tractable model for screening efficient probiotics.

Construction of an Electrocardiogram Database Including 12 Lead Waveforms.

OBJECTIVES: Electrocardiogram (ECG) data are important for the study of cardiovascular disease and adverse drug reactions. Although the development of analytical techniques such as machine learning has improved our ability to extract useful information from ECGs, there is a lack of easily available ECG data for research purposes. We previously published an article on a database of ECG parameters and related clinical data (ECG-ViEW), which we have now updated with additional 12-lead waveform information. METHODS: All ECGs stored in portable document format (PDF) were collected from a tertiary teaching hospital in Korea over a 23-year study period. We developed software which can extract all ECG parameters and waveform information from the ECG reports in PDF format and stored it in a database (meta data) and a text file (raw waveform). RESULTS: Our database includes all parameters (ventricular rate, PR interval, QRS duration, QT/QTc interval, P-R-T axes, and interpretations) and 12-lead waveforms (for leads I, II, III, aVR, aVL, aVF, V1, V2, V3, V4, V5, and V6) from 1,039,550 ECGs (from 447,445 patients). Demographics, drug exposure data, diagnosis history, and laboratory test results (serum calcium, magnesium, and potassium levels) were also extracted from electronic medical records and linked to the ECG information. CONCLUSIONS: Electrocardiogram information that includes 12 lead waveforms was extracted and transformed into a form that can be analyzed. The description and programming codes in this case report could be a reference for other researchers to build ECG databases using their own local ECG repository.

In vivo screening platform for shiga toxin-producing Escherichia coli (STEC) using Caenorhabditis elegans as a model.

Shiga toxin-producing Escherichia coli (STEC) strains are the main cause of bacillary dysentery, although STEC strains generally induce milder disease symptoms compared to Shigella species. This study aimed to determine the virulence of STEC using the nematode Caenorhabditis elegans as a model host. Worm killing, fertility and bacterial colonisation assays were performed to examine the potential difference in the virulence of STEC strains compared to that of the control E. coli OP50 strains on which worms were fed. A statistically significant difference in the survival rates of C. elegans was observed in that the STEC strains caused death in 8-10 days and the E. coli OP50 strains caused death in 15 days. STEC strains severely reduced the fertility of the worms. The intestinal load of bacteria in the adult stage nematodes harbouring the E. coli OP50 strains was found to be 3.5 log CFU mL-1. In contrast, the STEC strains E15, E18 and E22 harboured 4.1, 4.2 and 4.7 log CFU ml-1 per nematode, respectively. The heat-killed STEC strains significantly increased the longevity of the worms compared to the non-heated STEC strains. In addition, PCR-based genomic profiling of shiga toxin genes, viz., stx1 and stx2, identified in selected STEC strains revealed that these toxins may be associated with the virulence of the STEC strains. This study demonstrated that C. elegans is an effective model to examine and compare the pathogenicity and virulence variation of STEC strains to that of E. coli OP50 strains.

Quantitative Evaluation of the Relationship between T-Wave-Based Features and Serum Potassium Level in Real-World Clinical Practice.

BACKGROUND: Proper management of hyperkalemia that leads to fatal cardiac arrhythmia has become more important because of the increased prevalence of hyperkalemia-prone diseases. Although T-wave changes in hyperkalemia are well known, their usefulness is debatable. We evaluated how well T-wave-based features of electrocardiograms (ECGs) are correlated with estimated serum potassium levels using ECG data from real-world clinical practice. METHODS: We collected ECGs from a local ECG repository (MUSE™) from 1994 to 2017 and extracted the ECG waveforms. Of about 1 million reports, 124,238 were conducted within 5 minutes before or after blood collection for serum potassium estimation. We randomly selected 500 ECGs and two evaluators measured the amplitude (T-amp) and right slope of the T-wave (T-right slope) on five lead waveforms (V3, V4, V5, V6, and II). Linear correlations of T-amp, T-right slope, and their normalized feature (T-norm) with serum potassium levels were evaluated using Pearson correlation coefficient analysis. RESULTS: Pearson correlation coefficients for T-wave-based features with serum potassium between the two evaluators were 0.99 for T-amp and 0.97 for T-right slope. The coefficient for the association between T-amp, T-right slope, and T-norm, and serum potassium ranged from -0.22 to 0.02. In the normal ECG subgroup (normal ECG or otherwise normal ECG), there was no correlation between T-wave-based features and serum potassium level. CONCLUSIONS: T-wave-based features were not correlated with serum potassium level, and their use in real clinical practice is currently limited.