Backward Elimination Approach Research Articles

Prediction of helminthiases in travellers and migrants with eosinophilia: a cohort study.

The aim of this study was to determine predictors for helminthiasis among travellers and migrants with eosinophilia for which a visto to tropical regions or endemic regions for common helminthiasis had been registered. A retrospective cohort study was performed using electronic health records of 23,905 patients with eosinophilia (January-2011 to August-2021) at Karolinska University Hospital, (Stockholm), including patients tested for helminthiasis with a registered stay in a helminth endemic region. Outcomes were diagnosis of any helminthiasis, and diagnosis of schistosomiasis and strongyloidiasis. Multivariable logistic regression was used to assess associations between potential predictors and helminthiases with a backwards stepwise elimination approach until a predictive model was reached in which each variable had a p-value<0.15. Of 1,112 eligible patients with eosinophilia and documented stay in endemic regions, 219 (19.7%) had been diagnosed with helminthiasis, most frequently schistosomiasis (n=95, 43.4%) and strongyloidiasis (n=64, 29.2%). A stay in Sub-Saharan Africa (SSA),(OR 8.2 [95%CI 2.44-27.56]), malaise and fatigue (OR 2.65 [0.77-9.09]) and high-grade eosinophilia >1500cells/μl, (2.26, [1.54-3.32]) were the most important predictors for any helminthiasis (AUC: 0.77 [0.74-0.80). A SSA origin (2.97 [1.11-7.95]), malaise and fatigue (5.48 (1.13-26.63) and high-grade eosinophilia, (1.53, [0.86-2.71]) were predictors for schistosomiasis (AUC: 0.74 [0.70-0.77]; whereas SSA origin (5.68 [3.04-10.59]), itching symptom (5.05 [1.32-19.36], and high-grade eosinophilia (2.42, [1.33-4.41]) were predictors for strongyloidiasis (AUC: 0.73 [0.69-0.76]). A stay in an endemic region, specifically SSA, having high-grade eosinophilia, and malaise and fatigue were the most important predictors for helminthiasis. Itching was an additional predictor for strongyloidiasis.

Enhanced Estimation of Crown-Level Leaf Dry Biomass of Ginkgo Saplings Based on Multi-Height UAV Imagery and Digital Aerial Photogrammetry Point Cloud Data

Ginkgo is a multi-purpose economic tree species that plays a significant role in human production and daily life. The dry biomass of leaves serves as an accurate key indicator of the growth status of Ginkgo saplings and represents a direct source of economic yield. Given the characteristics of flexibility and high operational efficiency, affordable unmanned aerial vehicles (UAVs) have been utilized for estimating aboveground biomass in plantations, but not specifically for estimating leaf biomass at the individual sapling level. Furthermore, previous studies have primarily focused on image metrics while neglecting the potential of digital aerial photogrammetry (DAP) point cloud metrics. This study aims to investigate the estimation of crown-level leaf biomass in 3-year-old Ginkgo saplings subjected to different nitrogen treatments, using a synergistic approach that combines both image metrics and DAP metrics derived from UAV RGB images captured at varying flight heights (30 m, 60 m, and 90 m). In this study, image metrics (including the color and texture feature parameters) and DAP point cloud metrics (encompassing crown-level structural parameters, height-related and density-related metrics) were extracted and evaluated for modeling leaf biomass. The results indicated that models that utilized both image metrics and point cloud metrics generally outperformed those relying solely on image metrics. Notably, the combination of image metrics obtained from the 60 m flight height with DAP metrics derived from the 30 m height significantly enhanced the overall modeling performance, especially when optimal metrics were selected through a backward elimination approach. Among the regression methods employed, Gaussian process regression (GPR) models exhibited superior performance (CV-R2 = 0.79, rRMSE = 25.22% for the best model), compared to Partial Least Squares Regression (PLSR) models. The common critical image metrics for both GPR and PLSR models were found to be related to chlorophyll (including G, B, and their normalized indices such as NGI and NBI), while key common structural parameters from the DAP metrics included height-related and crown-related features (specifically, tree height and crown width). This approach of integrating optimal image metrics with DAP metrics derived from multi-height UAV imagery shows great promise for estimating crown-level leaf biomass in Ginkgo saplings and potentially other tree crops.

Relapses in canine leishmaniosis: risk factors identified through mixed-effects logistic regression

BackgroundCanine leishmaniosis (CanL), caused by Leishmania infantum, is an important vector-borne parasitic disease in dogs with implications for human health. Despite advancements, managing CanL remains challenging due to its complexity, especially in chronic, relapsing cases. Mathematical modeling has emerged as a powerful tool in various medical fields, but its application in understanding CanL relapses remains unexplored.MethodsThis retrospective study aimed to investigate risk factors associated with disease relapse in a cohort of dogs naturally infected with L. infantum. Data from 291 repeated measures of 54 dogs meeting the inclusion criteria were included. Two logistic mixed-effects models were created to identify clinicopathological variables associated with an increased risk of clinical relapses requiring a leishmanicidal treatment in CanL. A backward elimination approach was employed, starting with a full model comprising all potential predictors. Variables were iteratively eliminated on the basis of their impact on the model, considering both statistical significance and model complexity. All analyses were conducted using R software, primarily employing the lme4 package, and applying a significance level of 5% (P < 0.05).ResultsThis study identified clinicopathological variables associated with an increased risk of relapses requiring a leishmanicidal treatment. Model 1 revealed that for each 0.1 increase in the albumin/globulin ratio (A/G) ratio, the odds of requiring treatment decreased by 45%. Conversely, for each unit increase in the total clinical score (CS), the odds of requiring treatment increase by 22–30%. Indirect immunofluorescence antibody test (IFAT) was not a significant risk factor in model 1. Model 2, incorporating individual albumin and globulins values, showed that dogs with high IFAT titers, hyper beta-globulinemia, hypoalbuminemia, anemia, and high CS were at increased risk of relapse. Both models demonstrated a good fit and explained a substantial amount of variability in treatment decisions.ConclusionsDogs exhibiting higher CS, dysproteinemia, anemia, and high IFAT titers are at increased risk of requiring leishmanicidal treatment upon clinical relapse in CanL. Regular monitoring and assessment of risk factors prove essential for early detection of relapses and effective intervention in CanL cases. The contrasting findings between the two models highlight the complexity of aspects influencing treatment decisions in this disease and the importance of tailored management strategies to improve outcomes for affected dogs.Graphical

Optimization of 1,3-propanediol production from fermentation of crude glycerol by immobilized Bacillus pumilus

This study investigates the application of crude glycerol to the production of 1,3-propanediol by immobilized cells of Bacillus pumilus. This is a novel application of a naturally occurring producer obtained from a wastewater storage pond in Thailand. Crude glycerol was obtained through the methanolysis of palm oil, which was catalyzed using rice bran lipase. Ten components of the fermentation medium were screened using a Plackett-Burman design. The statistical significance of the results was determined using multiple linear regression with a backward elimination approach. The significance level was set to 5 % (p < 0.05). Only crude glycerol, (NH4)2SO4, MgSO4, and CaCl2 significantly affected 1,3-propanediol production by immobilized B. pumilus. Furthermore, preliminary screenings of environmental conditions used for 1,3-propanediol production were conducted using a Plackett-Burman design. The results showed that the temperature, time, and quantity of immobilized cells were factors that significantly affected 1,3-propanediol yield. Therefore, the quantities of crude glycerol, (NH4)2SO4, MgSO4, and CaCl2 and the temperature, time, and quantity of immobilized cells were optimized using response surface methodology based on a Box-Behnken design. The model predicted a maximum 1,3-propanediol yield of 45.68 g/L with the following conditions: 60 g/L crude glycerol, 5 g/L (NH4)2SO4, 0.55 g/L MgSO4, 0.05 g/L CaCl2, a fermentation duration of 101 h, and a temperature of 25 °C, with 250 g of immobilized cells. The validation trials confirmed a production level of 44.12 ± 1.81 g/L, indicating a 2.86-fold production increase relative to the control group. Overall, this study demonstrates the potential of using crude glycerol as a substrate to improve the yields of 1,3-propanediol produced by B. pumilus.

QSPR/QSAR study of antiviral drugs modeled as multigraphs by using TI’s and MLR method to treat COVID-19 disease

The ongoing COVID-19 pandemic continues to pose significant challenges worldwide, despite widespread vaccination. Researchers are actively exploring antiviral treatments to assess their efficacy against emerging virus variants. The aim of the study is to employ M-polynomial, neighborhood M-polynomial approach and QSPR/QSAR analysis to evaluate specific antiviral drugs including Lopinavir, Ritonavir, Arbidol, Thalidomide, Chloroquine, Hydroxychloroquine, Theaflavin and Remdesivir. Utilizing degree-based and neighborhood degree sum-based topological indices on molecular multigraphs reveals insights into the physicochemical properties of these drugs, such as polar surface area, polarizability, surface tension, boiling point, enthalpy of vaporization, flash point, molar refraction and molar volume are crucial in predicting their efficacy against viruses. These properties influence the solubility, permeability, and bio availability of the drugs, which in turn affect their ability to interact with viral targets and inhibit viral replication. In QSPR analysis, molecular multigraphs yield notable correlation coefficients exceeding those from simple graphs: molar refraction (MR) (0.9860), polarizability (P) (0.9861), surface tension (ST) (0.6086), molar volume (MV) (0.9353) using degree-based indices, and flash point (FP) (0.9781), surface tension (ST) (0.7841) using neighborhood degree sum-based indices. QSAR models, constructed through multiple linear regressions (MLR) with a backward elimination approach at a significance level of 0.05, exhibit promising predictive capabilities highlighting the significance of the biological activity IC50\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$IC_{50}$$\\end{document} (Half maximal inhibitory concentration). Notably, the alignment of predicted and observed values for Remdesivir’s with obs pIC50=6.01\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$${pIC_{50} = 6.01}$$\\end{document},pred pIC50=6.01\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$${pIC_{50} = 6.01}$$\\end{document} (pIC50\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$pIC_{50}$$\\end{document} represents the negative logarithm of IC50\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$IC_{50}$$\\end{document}) underscores the accuracy of multigraph-based QSAR analysis. The primary objective is to showcase the valuable contribution of multigraphs to QSPR and QSAR analyses, offering crucial insights into molecular structures and antiviral properties. The integration of physicochemical applications enhances our understanding of factors influencing antiviral drug efficacy, essential for combating emerging viral strains effectively.

Development and external validation of a home-based risk predicTion modEl of natUral onseT of menopAuse -TEUTA.

To develop and externally validate a 10-year risk prediction model of natural onset of menopause using ready-to-use predictors. Population-based prospective cohort study. Community-dwelling, premenopausal women aged 28 years and older enrolled in the Swiss (CoLaus) and Dutch (PREVEND) study. Incidence of self-reported natural menopause. Based on existing literature, 11 predictors were tested in this study. The CoLaus cohort was used to develop the model by applying the backward-elimination approach and Bayesian Model Averaging. Internal validation was performed by bootstrapping. External validation was performed using data from the PREVEND cohort and recalibrating the baseline survival estimate. C-statistic, calibration slopes, and expected/observed probabilities were calculated as measures of model internal and/or external performances. The final analysis included 750 and 1032 premenopausal women from the CoLaus and the PREVEND cohort, respectively. Among them, 445 (59%) from CoLaus and 387 (38%) from PREVEND experienced menopause over a median follow-up of 10.7 and 9 years, respectively. The final model included age, alcohol consumption, smoking status, education level, and systolic blood pressure. Upon external calibration in the PREVEND cohort, the model exhibited good discrimination, with a C-statistic of 0.888 and an expected/observed probability of 0.82. We present the first internally and externally validated prediction model of natural menopause onset using readily available predictors. Validation of our model to other populations is needed.

Population Pharmacodynamic Models of Risperidone on PANSS Total Scores and Prolactin Levels in Schizophrenia.

Currently, research predominantly focuses on evaluating clinical effects at specific time points while neglecting underlying patterns within the treatment process. This study aims to analyze the dynamic alterations in PANSS total scores and prolactin levels in patients with schizophrenia treated with risperidone, along with the influencing covariates. Using data from an 8-week randomized, double-blind, multicenter clinical trial, a population pharmacodynamic model was established for the PANSS total scores of and prolactin levels in patients treated with risperidone. The base model employed was the Emax model. Covariate selection was conducted using a stepwise forward inclusion and backward elimination approach. A total of 144 patients were included in this analysis, with 807 PANSS total scores and 531 prolactin concentration values. The PANSS total scores of the patients treated with risperidone decreased over time, fitting a proportionally parameterized sigmoid Emax model with covariates including baseline score, course of the disease, gender, plasma calcium ions, and lactate dehydrogenase levels. The increase in prolactin levels conformed to the ordinary Emax model, with covariates encompassing course of the disease, gender, weight, red blood cell count, and triglyceride levels. The impacts of the baseline scores and the course of the disease on the reduction of the PANSS scores, as well as the influence of gender on the elevation of prolactin levels, each exceeded 20%. This study provides valuable quantitative data regarding PANSS total scores and prolactin levels among patients undergoing risperidone treatment across various physiological conditions.

Spatial patterns and sociodemographic predictors of chronic obstructive pulmonary disease in Florida.

Chronic obstructive pulmonary disease (COPD) is a chronic, inflammatory respiratory disease that obstructs airflow and decreases lung function and is a leading cause death globally. In the United States (US), the prevalence among adults is 6.2%, but increases with age to 12.8% among those 65 years or older. Florida has one of the largest populations of older adults in the US, accounting for 4.5 million adults 65 years or older. This makes Florida an ideal geographic location for investigating COPD as disease prevalence increases with age. Understanding the geographic disparities in COPD and potential associations between its disparities and environmental factors as well as population characteristics is useful in guiding intervention strategies. Thus, the objectives of this study are to investigate county-level geographic disparities of COPD prevalence in Florida and identify county-level socio-demographic predictors of COPD prevalence. This ecological study was performed in Florida using data obtained from the US Census Bureau, Florida Health CHARTS, and County Health Rankings and Roadmaps. County-level COPD prevalence for 2019 was age-standardized using the direct method and 2020 US population as the standard population. High-prevalence spatial clusters of COPD were identified using Tango's flexible spatial scan statistics. Predictors of county-level COPD prevalence were investigated using multivariable ordinary least squares model built using backwards elimination approach. Multicollinearity of regression coefficients was assessed using variance inflation factor. Shapiro-Wilks, Breusch Pagan, and robust Lagrange Multiplier tests were used to assess for normality, homoskedasticity, and spatial autocorrelation of model residuals, respectively. County-level age-adjusted COPD prevalence ranged from 4.7% (Miami-Dade) to 16.9% (Baker and Bradford) with a median prevalence of 9.6%. A total of 6 high-prevalence clusters with prevalence ratios >1.2 were identified. The primary cluster, which was also the largest geographic cluster that included 13 counties, stretched from Nassau County in north-central Florida to Charlotte County in south-central Florida. However, cluster 2 had the highest prevalence ratio (1.68) and included 10 counties in north-central Florida. Together, the primary cluster and cluster 2 covered most of the counties in north-central Florida. Significant predictors of county-level COPD prevalence were county-level percentage of residents with asthma and the percentage of current smokers. There is evidence of spatial clusters of COPD prevalence in Florida. These patterns are explained, in part, by differences in distribution of some health behaviors (smoking) and co-morbidities (asthma). This information is important for guiding intervention efforts to address the condition, reduce health disparities, and improve population health.

Dosing recommendation of nirmatrelvir/ritonavir using an integrated population pharmacokinetic analysis.

Protease inhibitor nirmatrelvir coadministered with ritonavir as a pharmacokinetic enhancer (PAXLOVID™; Pfizer Inc) became the first orally bioavailable antiviral agent granted Emergency Use Authorization in the United States in patients ≥12 years old with mild to moderate coronavirus disease 2019 (COVID-19). This population pharmacokinetic analysis used pooled plasma nirmatrelvir concentrations from eight completed phase I and II/III studies to characterize nirmatrelvir pharmacokinetics when coadministered with ritonavir in adults with/without COVID-19. Influence of covariates (e.g., formulation, dose, COVID-19) was examined using a stepwise forward selection (α = 0.05) and backward elimination (α = 0.001) approach. Simulations with 5000 subjects for each age and weight group and renal function category were performed to support dosing recommendations of nirmatrelvir/ritonavir for adults with COVID-19 and guide dose adjustments for specific patient populations (e.g., renal insufficiency, pediatrics). The final model was a two-compartment model with first-order absorption, including allometric scaling of body weight and dose-dependent absorption (power function on relative bioavailability). Nirmatrelvir clearance (CL) increased proportionally to body surface area-normalized creatinine CL (nCLCR) up to 70 ml/min/1.73 m2 and was independent of nCLCR above the breakpoint. Significant covariates included carbamazepine or itraconazole coadministration as markers for drug interactions, COVID-19 on CL, formulation on relative bioavailability, and age on central volume of distribution. Simulation results support current dosing recommendations of nirmatrelvir/ritonavir 300/100 mg twice daily (b.i.d.) in adults with normal renal function or mild impairment and pediatrics (12 to <18 years) weighing ≥40 kg and nirmatrelvir/ritonavir 150/100 mg b.i.d. in adults with moderate renal impairment.

Modelling the effects of mixing ratio and temperature on the thermal conductivity of GNP-Alumina hybrid nanofluids: A comparison of ANN, RSM, and linear regression methods

This research aimed to evaluate and compare the efficacy of three distinct methods for forecasting the thermal conductivity of GNP-Alumina hybrid nanofluids. The methods under consideration were artificial neural network (ANN), response surface methodology (RSM), and linear regression (LR). The predictive performance of the ANN model was investigated in relation to the number of neurons in the hidden layer. The findings revealed that the optimal number of neurons was 7, which produced the best performance with an overall mean square error (MSE) of 1.08E-06. The correlation coefficient was also high at 0.99799. The RSM approach involved testing linear, quadratic, cubic, and quartic models, with the quadratic model showing the highest predicted R2 (0.9721) values, indicating that it provided the best fit to the data. Finally, the LR model was developed using a backward elimination approach, with temperature and volume fraction being the significant variables in the final model. Overall, the ANN model produced the most accurate predictions, followed by the RSM and LR models. These findings suggest that the ANN and RSM techniques can be effective tools for forecasting the thermal conductivity of nanofluids, and highlight the importance of selecting appropriate model parameters for optimal performance.

Engagement and outcomes of cancer patients referred to a tobacco cessation program at a National Cancer Institute-designated cancer center.

Tobacco cessation is a critical but challenging intervention for cancer patients. Our National Cancer Institute-designated Comprehensive Cancer Center instituted a tobacco cessation program in 2019. This manuscript reports on the first 2 years of our experience. Patients were referred to the program by their care team, and a certified tobacco treatment specialist contacted patients remotely and provided behavioral therapy and coordinated pharmacotherapy. We retrospectively captured data from patients with a cancer diagnosis referred to the tobacco cessation program. Univariate and multivariable logistic regression analyses with the backward elimination approach were performed to determine factors associated with patient acceptance of referral to the tobacco cessation program. Tobacco cessation rates after referral to the program were also captured. Between July 2019 and August 2021, 194 patients were referred to the tobacco cessation program. Of the 194 patients referred, 93 agreed to enroll in the tobacco cessation program (47.9%), of which 84 requested pharmacotherapy (90.3%). Twenty-four were able to cease tobacco use (25.8%). Only 7 patients out of the 101 patients (6.9%) who declined cessation services were successful (p < 0.001). On univariate logistic regression, race (p=0.027) and marital status (p=0.020) were associated with referral acceptance. On multivariable analysis, single patients (odds ratio [OR]=0.33) and Caucasian patients (OR=0.43) were less likely to accept a referral. Access to tobacco cessation services is a critical component of comprehensive cancer care. Our experience highlights the need to understand patient-specific factors associated with engagement with a tobacco cessation program during cancer treatment. The use of pharmacotherapy is also a critical component of successful tobacco cessation.

The Frequency of Alternative Splicing in Multiple Myeloma Is Determined By the Non-Homologous End Joining Pathway

Introduction: Alternative Splicing (AS) plays a pivotal role in regulating cellular processes in both normal and malignant cells. Previous studies indicated mutations in the spliceosome complex, such as SF3B1, can cause aberrant AS in acute myeloid leukemia and multiple myeloma (MM) patients. However, there is also a known interaction between AS and the DNA damage response in other cancer types, which we investigate here. Materials and Methods: RNA-sequencing data of newly diagnosed MM patients (n=598) from the MMRF CoMMpass dataset were utilized to generate whole transcriptome AS profiles. For each patient, RNA-seq data was aligned to Hg38 reference genome using STAR while expression level was measured by Salmon. AS profiles were established from their alignment and expression profile using SUPPA2. Splicing level was defined by Percentage of Spliced-In (PSI). UMAP plots were generated for all patients using top 20,000 AS events ranked by variation in the sample set. High-quality differential AS events between groups of patients were obtained from the following thresholds: significant splicing level difference (|PSI|>10% and P <0.05 from T-test), sufficient expression level (TPM>1), not similarly spliced (|PSI|>10% and P-value<0.05 from T-test) in normal bone marrow plasma cells (N=5). Patient samples were annotated for cytogenetic, mutation, copy number and structural variation status. Pathway analysis was conducted for differentially expressed genes (DEGs). Patients' overall survival and progression-free survival was displayed with Kaplan-Meier curves. Survival difference was measured by log-rank test and cox-regression (when other well-known risk factors were incorporated). Results: We examined differential AS by defining and comparing high and low activity of key myeloma KEGG pathways including the spliceosome, proteasome, DNA repair and ubiquitination pathways. The DNA repair pathway generated significantly more (P<0.05) differential AS events (N=1,197) than the other pathways. To examine this further, we performed the analysis for the three main DNA repair mechanisms: homologous recombination (HR), non-homologous end joining (NHEJ) and microhomology-mediated end joining (MMEJ) pathways. The number of differential AS events (N=2,999) between NHEJ-high (N=121) and -low (N=36) samples was significantly more than for the other DNA repair pathways (P<0.05). NHEJ-high samples were significantly enriched for TP53 mutations (15% vs. 6%, P=2x10-5, Fisher's exact test), biallelic-TP53 events (10% vs. 4%, P=3x10-5), and biallelic-TENT5C (21% vs. 10%, P=1x10-9) events (Figure 1). Additionally, expression of LIG4 and XRCC4, encoding DNA IV ligase complex, a key component of the NHEJ pathway, correlated highly (, Spearman correlation) with the number of differential AS events. The NHEJ-high group also harbored significantly more structural abnormalities than the NHEJ-low group (mean 87 vs. 19, P=6x10-4, Mann-Whitney U test), indicating their different status of DNA stability that was highly associated with this pathway. Survival analysis indicated that patients in the NHEJ-high group showed a significant worse progression-free survival (PFS) compared to others (Logrank statistic=7.3, P=0.007). In a multivariate analysis including other high-risk factors such as MGP-Double-Hit, t(4;14), t(14;16), TP53 mutation, 17p del, 1q gain/amp, and ISS=III, the NHEJ-high group retained significance (Hazard Ratio=1.8, P=1.6x10-5). To incorporate AS events we further applied a backward elimination approach to automatically select an optimal prognostic model from top 200 highly variant AS events ranked by variation and 7 well-known risk factors along with high NHEJ activity in PFS for the entire population. This model contained 16 significant risk factors (P<0.05, Multivariate cox regression) and 12 of them were AS events. NHEJ-high samples were significantly enriched (P<0.05, Fisher exact test) for eight of the AS events that were associated with poor prognosis. Conclusion: AS frequency is associated with the NHEJ pathway in MM. NHEJ activity is also associated with DNA instability through increased number of structural rearrangements and is a high-risk factor. Figure 1. High expression of the NHEJ pathways is associated with TP53 and TENT5C abnormalities. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Prediction of Household Food Security Status using Ensemble Learning Models

Background:This research uses the Ethiopian HICE survey dataset. Predicting food insecurity is critical in presenting the household's situation to the appropriate agencies that take preventative and intervention measures.Objective:This research paper's primary goal is to predict households' food security status using ensemble learning models.Methods:We use five base classifiers and a voting strategy for ensemble classification to enhance the performance of different base classifiers. Backward feature elimination and hard and soft voting-based ensemble learning are used to evaluate household food security. The training set for the basic classifiers is composed of the features that have been selected. Each ML classifier makes its prediction about the class label with the help of an ensemble learning method. For making decisions, hard voting uses a simple majority, whereas soft vote employs a weighted probability. To determine the final prediction. Ethiopian household income, consumption, and expenditure dataset are used to test the proposed ensemble learning approach. The backward feature elimination approach improved the model's performance by removing irrelevant and redundant features. Random forest, gradient boosting, multi-layer perceptron, K-nearest Neighbor, and Extra Tree classifiers were used to predict the family's level of food security. Finally, the authors compare the accuracy of ensemble and base classifiers.Result:The experiment result shows that the RF classifier surpasses the other base and ensemble classifiers and scored 99.98% accuracy. Because a Random forest classifier is an ensemble learning classifier that uses several decision trees, the final prediction is computed based on the majority vote of the several trees. The comparison result of hard and soft voting reveals that soft voting outperforms hard voting before and after feature selection with accuracies of 99.79% and 99.77%, respectively.Conclusion:Based on the result obtained, ensemble learning plays a significant role in predicting household food security status and implementing hard and soft voting. The RF classifier surpasses the other base and ensemble classifiers with an accuracy of 99.98%. From ensemble methods, soft voting surpasses hard voting with an accuracy score of 99.79%.

Assessing the potential determinants of national vitamin A supplementation among children aged 6–35 months in Ethiopia: further analysis of the 2019 Ethiopian Mini Demographic and Health Survey

BackgroundVitamin A is a nutrient that is required in a small amount for normal visual system function, growth and development, epithelia’s cellular integrity, immune function, and reproduction. Vitamin A has a significant and clinically important effect since it has been associated with a reduction in all-cause and diarrhea mortality. The aim of this study was to determine factors associated with national vitamin A supplementation among children aged 6–35 months.MethodThe data for this study was extracted from the 2019 Ethiopian Mini Demographic and Health Survey. A total weighted sample of 2242 women with children aged 6–35 months was included in the study. The analysis was performed using Stata version 14.2 software. Applying sampling weight for descriptive statistics and complex sample design for inferential statistics, a manual backward stepwise elimination approach was applied. Finally, statistical significance declared at the level of p value < 0.05.ResultThe overall coverage of vitamin A supplementation among children aged 6–35 months for the survey included was 44.4 95% CI (40.15, 48.74). In the multivariable analysis, mothers who had four or more antenatal visits [AOR = 2.02 (95% CI: 1.34, 3.04)] were two times more likely to receive vitamin A capsules for their children than mothers who had no antenatal visits. Children from middle-wealth quintiles had higher odds of receiving vitamin A capsules in comparison to children from the poorest wealth quintile [AOR = 1.77 (95% CI: 1.14, 2.73)]. Older children had higher odds of receiving vitamin A capsules than the youngest ones. Other factors that were associated with vitamin A supplementation were mode of delivery and region.ConclusionThe coverage of vitamin A supplementation in Ethiopia remains low and it is strongly associated with antenatal visit, household wealth index and age of child. Expanding maternal health services like antenatal care visits should be prioritized.

Factors associated with early initiation of breastfeeding among children less than 24 months old: the 2019 Ethiopian mini demographic and health survey

BackgroundThe world Health Organization recommended that all mothers be supported to initiate breastfeeding as soon as possible after birth, within the first hour. This study examined the determinants of early initiation of breastfeeding in Ethiopia by using data from the 2019 Ethiopia Mini Demographic and Health Survey.MethodsThe data for this study was extracted from the 2019 Ethiopia Mini Demographic and Health Survey. A total of 1948 children aged less than 24 months at the time of the survey were included for analysis from the nine regional states and two city administrations. The analysis was carried out in STATA Version 14.2 software using survey commands to account for the complex sample design and apply sampling weights. A manual backward stepwise elimination approach was applied.ResultThe prevalence of early initiation of breastfeeding in Ethiopia was noted to be 75.2% [95% CI (71.9, 78.5%)]. In the multivariable analysis, mothers who had vaginal deliveries [AOR = 3.02 (95% CI: 1.55, 5.88)] had 3 times higher odds practicing early initiation of breastfeeding compared to mothers who had a cesarian section. Mothers aged between 35 and 49 years [AOR = 2.40, 95% CI: 1.20, 4.49] had 2.4 times higher odds of practicing early initiation of breastfeeding compared to mothers aged below 20 years. In addition, early initiation of breastfeeding was also associated with the region where mothers resided, in particular mothers in Amhara and Somali region, had lower odds of practicing early initiation of breastfeeding as compared with mothers residing in Tigray region.ConclusionEarly initiation of breastfeeding in Ethiopia was found to be significantly associated with mode of delivery, mother’s age, and region. As a result, raising awareness about early initiation of breastfeeding is especially important for mothers who have had a cesarean section, which could be accomplished with the help of the health extension workforce.

The incidence of early neurological complications after on-pump cardiac surgery: a retrospective study

Background: Cardiac surgery with cardiopulmonary bypass (CPB) is associated with a significant risk for neurological complications. Reported incidence and risk factors for these complications vary significantly. Identifying risk factors could lead to preventive strategies to reduce complications and improve patient’s outcome. Objective: The study aims to assess the overall incidence and risk factors for severe early postoperative neurological complications after elective on-pump cardiac surgery. We specifically analyzed the incidence of stroke, global cerebral ischemia (GCI) and epilepsy in these patients. Methods: After getting approval from the Ethics Committee Research UZ/KU Leuven, on 14/12/2021 (s65871), we retrospectively evaluated data of 1080 adult patients after cardiac surgery with CPB between 06/2019 and 06/2021 at the University Hospitals Leuven. After exclusion of emergency procedures and patients who died before neurological evaluation, 977 patients remained for primary analysis. All data were collected from the electronic patient’s file. Primary objective was to identify the incidence of stroke, GCI and epilepsy. We defined stroke and GCI according to the American Stroke Association. Secondary endpoints were identifying independent risk-factors and assessing the impact of early neurological complications on mortality. Statistical analysis was performed using econometric and statistical modeling with python. We performed univariate logistic regression with Bonferonni correction and multivariable logistic regression with backwards elimination approach and p-value set to be &lt;0.05. Results: The overall incidence of defined neurological complications after elective on-pump cardiac surgery at our institution was 3.17% (n=31) (stroke 2.35% (n=23), epilepsy 0.61% (n=6) and GCI 0.31% (n=3)). No statistically significant risk factors for these complications were found. In secondary analysis, patients with stroke and GCI had a higher risk of in-hospital mortality (Fisher’s exact test resulted in odds ratio 7.23 with p=0.005 and odds ratio 65.17 with p=0.003 respectively) Diabetes mellitus, preoperative atrial fibrillation, and endocarditis were also significantly related to in-hospital mortality. Conclusions: The incidence of early neurological complications after elective on-pump cardiac surgery at our institution was comparable to that reported in earlier studies. No independent risk factors for these neurological complications were found. The occurrence of stroke and GCI significantly increased in-hospital mortality which emphasizes the importance of these complications, with possible mortality benefit of early recognition and management of stroke.

A simple mortality risk prediction score for viper envenoming in India (VENOMS): A model development and validation study.

Snakebite is a neglected problem with a high mortality in India. There are no simple clinical prognostic tools which can predict mortality in viper envenomings. We aimed to develop and validate a mortality-risk prediction score for patients of viper envenoming from Southern India. We used clinical predictors from a prospective cohort of 248 patients with syndromic diagnosis of viper envenoming and had a positive 20-minute whole blood clotting test (WBCT 20) from a tertiary-care hospital in Puducherry, India. We applied multivariable logistic regression with backward elimination approach. External validation of this score was done among 140 patients from the same centre and its performance was assessed with concordance statistic and calibration plots. The final model termed VENOMS from the term "Viper ENvenOming Mortality Score included 7 admission clinical parameters (recorded in the first 48 hours after bite): presence of overt bleeding manifestations, presence of capillary leak syndrome, haemoglobin <10 g/dL, bite to antivenom administration time > 6.5 h, systolic blood pressure < 100 mm Hg, urine output <20 mL/h in 24 h and female gender. The lowest possible VENOMS score of 0 predicted an in-hospital mortality risk of 0.06% while highest score of 12 predicted a mortality of 99.1%. The model had a concordance statistic of 0·86 (95% CI 0·79-0·94) in the validation cohort. Calibration plots indicated good agreement of predicted and observed outcomes. The VENOMS score is a good predictor of the mortality in viper envenoming in southern India where Russell's viper envenoming burden is high. The score may have potential applications in triaging patients and guiding management after further validation.

Machine Learning to Estimate Surface Roughness from Satellite Images

We apply the Support Vector Regression (SVR) machine learning model to estimate surface roughness on a large alluvial fan of the Kosi River in the Himalayan Foreland from satellite images. To train the model, we used input features such as radar backscatter values in Vertical–Vertical (VV) and Vertical–Horizontal (VH) polarisation, incidence angle from Sentinel-1, Normalised Difference Vegetation Index (NDVI) from Sentinel-2, and surface elevation from Shuttle Radar Topographic Mission (SRTM). We generated additional features (VH/VV and VH–VV) through a linear data fusion of the existing features. For the training and validation of our model, we conducted a field campaign during 11–20 December 2019. We measured surface roughness at 78 different locations over the entire fan surface using an in-house-developed mechanical pin-profiler. We used the regression tree ensemble approach to assess the relative importance of individual input feature to predict the surface soil roughness from SVR model. We eliminated the irrelevant input features using an iterative backward elimination approach. We then performed feature sensitivity to evaluate the riskiness of the selected features. Finally, we applied the dimension reduction and scaling to minimise the data redundancy and bring them to a similar level. Based on these, we proposed five SVR methods (PCA-NS-SVR, PCA-CM-SVR, PCA-ZM-SVR, PCA-MM-SVR, and PCA-S-SVR). We trained and evaluated the performance of all variants of SVR with a 60:40 ratio using the input features and the in-situ surface roughness. We compared the performance of SVR models with six different benchmark machine learning models (i.e., Gaussian Process Regression (GPR), Generalised Regression Neural Network (GRNN), Binary Decision Tree (BDT), Bragging Ensemble Learning, Boosting Ensemble Learning, and Automated Machine Learning (AutoML)). We observed that the PCA-MM-SVR perform better with a coefficient of correlation (R = 0.74), Root Mean Square Error (RMSE = 0.16 cm), and Mean Square Error (MSE = 0.025 cm2). To ensure a fair selection of the machine learning model, we evaluated the Akaike’s Information Criterion (AIC), corrected AIC (AICc), and Bayesian Information Criterion (BIC). We observed that SVR exhibits the lowest values of AIC, corrected AIC, and BIC of all the other methods; this indicates the best goodness-of-fit. Eventually, we also compared the result of PCA-MM-SVR with the surface roughness estimated from different empirical and semi-empirical radar backscatter models. The accuracy of the PCA-MM-SVR model is better than the backscatter models. This study provides a robust approach to measure surface roughness at high spatial and temporal resolutions solely from the satellite data.

A Predictive Model for Functional Outcome in Patients with Acute Ischemic Stroke Undergoing Endovascular Thrombectomy

IntroductionEndovascular thrombectomy (EVT) is a well-established treatment of acute ischemic stroke. Variability in outcomes among thrombectomy patients results in a need for patient centered approaches to recovery. Identifying key factors that are associated with outcomes can help prognosticate and direct resources for continued improvement post-treatment. Thus, we developed a comprehensive predictive model of short-term outcomes post-thrombectomy. MethodsThis is a retrospective chart review of adult patients who underwent EVT at our institution over the last four years. Primary outcome was dichotomized 90-day mRS (mRS 0–2 v mRS 3–6). Bivariate analyses were conducted, followed by logistic regression modelling via a backward-elimination approach to identify the best fit predictive model. Results326 thrombectomies were performed; 230 cases were included in the model. In the final predictive model, adjusting for age, gender, race, diabetes, and presenting NIHSS, pre-admission mRS = 0–2 (OR 18.1; 95% 3.44–95.48; p < 0.001) was the strongest predictor of a good outcome at 90-days. Other independent predictors of good outcomes included being a non-smoker (OR 5.4; 95% CI 1.53–19.00; p = 0.01) and having a post-thrombectomy NIHSS<10 (OR 9.7; 95% CI 3.90–24.27; p < 0.001). A decompressive hemicraniectomy (DHC) was predictive of a poor outcome at 90-days (OR 0.07; 95% CI 0.01–0.72; p = 0.03). This model had a Sensitivity of 79%, a Specificity of 89% and an AUC=0.89. ConclusionOur model identified low pre-admission mRS score, low post-thrombectomy NIHSS, non-smoker status and not requiring a DHC as predictors of good functional outcomes at 90-days. Future works include developing a prognostic scoring system.

On a Stepwise Hypotheses Testing Procedure and Information Criterion in Identifying Dynamic Relations between Time Series

This paper studies an effective stepwise hypotheses testing pro cedure in identifying dynamic relations between time series, and its close connection with popular information criteria such as AIC and BIC. This procedure, labeled M2, extends Chen and Lee’s (1990) procedure to cover both the strong and weak form dynamic relations; and to be used with a guided choice of significance levels which are adapting in nature. Intu itively, procedure M2 can be viewed as a backward-elimination approach that simplifies the all-possible pairwise comparisons approach implied by information criterion. New insights concerning identification of strong and weak form dynamic relations using these approaches are given. Extensive simulation experiments are conducted to illustrate the performance of the IC and M2 approach in different settings. For applications, we study the dynamic relations between price level and interest rate in US and UK, and the robustness of the model identified is also addressed.