Regression Models Research Articles

Experimental analysis on fiber diameter of spunbond nonwoven fabrics through Plackett–Burman and Box–Behnken designs and its impact on mechanical properties

AbstractThe COVID‐19 pandemic sparked a surge in demand for nonwoven protective materials, prompting a significant increase in nonwoven fabric production. To advance understanding, particularly in the Spunbond process, we conducted experiments to analyze the effects of various input parameters on fiber diameter, and mechanical tests to study how fiber diameter influences the mechanical properties of spunbond nonwoven fabrics. Employing Plackett–Burman design and Box–Behnken design with Minitab 18, we have examined different process parameters and study the cause‐and‐effect relationships between input parameters and the response. A set of experiments were carried out with nine varying parameters, including polymer melt index, initial polymer temperature, and air velocity. By regression modeling, we assessed the effects and interactions of these factors on fiber diameter. The Box–Behnken approach revealed that only three factors significantly influenced fiber diameter. Our analysis unveiled valuable insights for optimizing process parameters to achieve a target fiber diameter of 31.0 μm, crucial for enhancing nonwoven fabric production. Moreover, the results of the tensile property tests show that fiber diameter influence mechanical properties of nonwoven fabrics. As the fiber diameters increase, the mechanical properties of the nonwoven fabric are increased.Highlights Market research of textile industry, particularly of nonwoven sector. Identification of the parameters of the Spunbond process. Selection of Placket–Burman and Box–Behnken designs to analyze, optimize, and predict fiber diameter. Evaluation of the designs results (analysis of variance, Pareto, Regression model …), determination of the optimal conditions and the factors that has most effect on fiber diameter. Investigation of the effect of fiber diameter on mechanical properties.

Differences in Use of a Patient Portal Across Sociodemographic Groups: Observational Study of the NHS App in England

Background The adoption of patient portals, such as the National Health Service (NHS) App in England, may improve patient engagement in health care. However, concerns remain regarding differences across sociodemographic groups in the uptake and use of various patient portal features, which have not been fully explored. Understanding the use of various functions across diverse populations is essential to ensure any benefits are equally distributed across the population. Objective This study aims to explore differences in the use of NHS App features across age, sex, deprivation, ethnicity, long-term health care needs, and general practice (GP) size categories. Methods We used weekly NHS App use data from the NHS App dashboard for 6386 GPs in England from March 2020 to June 2022. Negative binomial regression models explored variations in weekly rates of NHS App features used (registrations, log-ins, prescriptions ordered, medical record views, and appointments booked). Outcomes were measured as weekly rates per 1000 GP-registered patients, and we conducted separate models for each outcome. Regression models included all covariates mentioned above and produced incident rate ratios, which we present here as relative percentages for ease of interpretation. GP-level covariate data on sociodemographic variables were used as categorical variables in 5 groups for deprivation (Q1=least deprived practices and Q5=most deprived practices) and 4 groups for all other variables (Q1=least deprived practices and Q4=most deprived practices). Results We found variations in the use of different features overall and across sociodemographic categories. Fully adjusted regression models found lower use of features overall in more deprived practices (eg, Q5 vs Q1: registrations=–34%, log-ins=–34.9%, appointments booked=–39.7%, medical record views=–32.3%, and prescriptions ordered=–9.9%; P<.001). Practices with greater proportions of male patients also had lower levels of NHS App use (eg, Q4 vs Q1: registration=–7.1%, log-in=–10.4%, and appointments booked=–36.4%; P<.001). Larger practices had an overall higher use of some NHS App features (eg, Q4 vs Q1: registration=3.2%, log-ins=11.7%, appointments booked=73.4%, medical record views=23.9%, and prescriptions ordered=20.7%; P<.001), as well as those with greater proportions of White patients (eg, Q4 vs Q1: registration=1.9%, log-ins=9.1%, appointments booked=14.1%, medical record views=28.7%, and prescriptions ordered=130.4%; P<.001). Use patterns varied for practices with greater proportions of patients with long-term health care needs (eg, Q4 vs Q1: registrations=–3.6%, appointments booked=–20%, and medical record views=6%; P≤.001). Conclusions This study highlights that the use of the NHS App features varied across sociodemographic groups. In particular, it is used less by people living in more deprived areas. Tailored interventions and patient support are required to ensure that any benefits from the NHS App are spread equally throughout the population.

Impact of Credit in Rural Settlements of Minas Gerais, Brazil

Objective: The objective of this study is to analyze the impact of credit from the National Agrarian Reform Program (PNRA) on the income of family plots allocated by the National Institute of Agrarian Reform (INCRA) in the state of Minas Gerais, aiming to identify the correlation between the credit provided and the income generated. Theoretical Framework: Credit programs aimed at family farming enable the development of their activities in a sustainable, regulated manner and in compliance with environmental preservation, facilitating the construction and improvement of production systems according to each producer’s needs. Method: The study uses data collected from the RADIS/UFMT Project, in agrarian reform settlements in Minas Gerais state, Brazil, were processed and systematized using the Propensity Score Matching (PSM) regression model, with the use of Stata 16® software. Results and Discussion: The most accessed credit line by the beneficiaries was the Semi-arid line. The hypothesis that credit granted by the National Agrarian Reform Program (PNRA) is positively correlated with the income of plots allocated to families settled by INCRA is confirmed, showing an approximate 19% increase in the income of plots compared to those without access to credit. That is, for every R$1.00 of credit provided to the plot, there is an approximate increase of R$0.19 in income.

α-Synuclein Oligomers in Skin Biopsies Predict the Worsening of Cognitive Functions in Parkinson’s Disease: A Single-Center Longitudinal Cohort Study

α-synuclein oligomers within synaptic terminals of autonomic fibers of the skin reliably discriminate Parkinson’s disease (PD) patients from healthy controls. Nonetheless, the prognostic role of oligomers for disease progression is unknown. We explored whether α-synuclein oligomers evaluated as proximity ligation assay (PLA) score may predict the worsening of cognitive functions in patients with Parkinson’s disease. Thirty-four patients with PD and thirty-four healthy controls (HC), matched 1:1 for age and sex, were enrolled. Patients with PD underwent baseline skin biopsy and an assessment of cognitive domains including Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), Clock Drawing Test, and Frontal Assessment Battery. At the last follow-up visit available, patients were either cognitively stable (PD-CS) or cognitively deteriorated (PD-CD). α-synuclein oligomers were quantified as PLA scores. Differences between groups were assessed, controlling for potential confounders. The relationship between skin biopsy measures and cognitive changes was explored using correlation and multivariable regression analyses. The discrimination power of the PLA score was assessed via ROC curve. To elucidate the relationship between skin biopsy and longitudinal cognitive measures, we conducted multivariable regression analyses using delta scores of cognitive tests (Δ) as dependent variables. We found that PD-CD had higher baseline PLA scores than PD-CS (p = 0.0003), and they were correctly identified in the ROC curve analysis (AUC = 0.872, p = 0.0003). Furthermore, ANCOVA analysis with Bonferroni correction, considering all groups (PD-CS, PD-CD, and HC), showed significant differences between PD-CS and PD-CD (p = 0.003), PD-CS and HC (p = 0.002), and PD-CD and HC (p < 0.001). In the regression model using ΔMMSE as the dependent variable, the PLA score was found to be a significant predictor (β = −0.441, p = 0.016). Similar results were observed when evaluating the model with ΔMoCA (β = −0.378, p = 0.042). In conclusion, patients with Parkinson’s disease with higher α-synuclein burden in the peripheral nervous system may be more susceptible to cognitive decline.

The spillover effect of global uncertainty through input–output linkage

AbstractThis study examines the cross‐country spillover effects of global uncertainty on domestic value‐added by introducing indices for input‐ and output‐linked uncertainty. Employing a panel fixed‐effect regression model, the analysis covers 70 countries from 1995 to 2019, using the World Uncertainty Index and the OECD Inter‐Country Input–Output Table. Our findings reveal that input‐linked uncertainty from procuring foreign intermediate goods has the most significant negative impact on domestic value‐added. The effects of domestic uncertainty and output‐linked uncertainty from foreign product sales follow this. Additionally, these negative effects exhibit a dynamic cumulative impact over time, rather than being confined to a specific period. Considering both the magnitude and duration of these impacts, the primary transmission pathway of global uncertainty on a country's domestic production activity is through its input linkage. Furthermore, the study identifies country‐specific variations in the ripple effects of global uncertainty. Analysing specific cases such as Korea and China affirms the empirical relevance of our findings to their economic contexts.

A comparison of statistical and machine learning models for spatio-temporal prediction of ambient air pollutant concentrations in Scotland

AbstractThe spatio-temporal prediction of air pollutant concentrations is vital for assessing regulatory compliance and for producing exposure estimates in epidemiological studies. Numerous approaches have been utilised for making such predictions, including land use regression models, additive models, spatio-temporal smoothing models and machine learning prediction algorithms. However, relatively few studies have compared the predictive performance of these models thoroughly, which is one of the novel contributions of this paper. For the specific challenge of predicting monthly average concentrations of NO$$_{2}$$ 2 , PM$$_{10}$$ 10 and $$\hbox {PM}_{2.5}$$ PM 2.5 in Scotland, we find that random forests typically outperform (or are as good as) more traditional statistical prediction approaches. Additionally, we utilise the best performing model to provide a new data resource, namely, predictions of monthly average concentrations (with uncertainty quantification) of the above pollutants on a regular 1 km$$^{2}$$ 2 grid for all of Scotland between 2016 and 2020.

Individualized post‐operative prediction of cochlear implantation outcomes in children with prelingual deafness using functional near‐infrared spectroscopy

AbstractObjectiveThe goal of this study was to develop an objective measure and predictor of cochlear implantation (CI) outcomes using functional near‐infrared spectroscopy (fNIRS) for young children with prelingual deafness.MethodsSound‐evoked hemodynamic responses were recorded from auditory and language‐related cortical regions of 47 child CI recipients (35.47 ± 17.24 months of age) using fNIRS shortly after CI activation (0.26 ± 0.30 months). There were four sound conditions (natural speech, instrumental music, multi‐speaker babble noise, and speech‐in‐noise). Post‐CI auditory and verbal communication performance was evaluated using clinical questionnaires with caretakers. Both classification and individualized regression models were constructed to predict post‐CI behavioral improvement from fNIRS data using support vector machine (SVM) learning algorithms.ResultsAuditory cortical responses shortly after CI hearing onset yielded highly accurate prediction of behavioral development in young CI children. For classification models, optimal prediction was achieved using cortical responses to two or more sound conditions, with the highest accuracy of 98.20% (precision = 98.17%, sensitivity = 98.96%, area under the curve of the receiver operating characteristic curve = 99.61%) obtained with the combination of speech, noise, and music stimuli. Similarly, for regression models, best prediction of individual development was achieved using three (highest r = 0.919) or four (r = 0.966) sound conditions. The predictability of cortical responses far outperformed (Cohen's d: 18.56) that of the collection of audiological and demographic parameters (classification accuracy: 0.62) under the same SVM algorithms and could not benefit from the inclusion of the latter.ConclusionMachine learning models using auditory cortical hemodynamic responses shortly after CI activation were able to predict individualized post‐CI behavioral improvement in children with prelingual deafness.Level of EvidenceLevel 5.

Local residents’ attitudes toward and contact with international students: a perspective from Montreal, Quebec

As migrants holding temporary, foreign-resident status in their host communities, international students often experience prejudice and have little meaningful contact with locals. To date, a comprehensive account of international students’ experience is lacking, and existing conceptualizations exclude linguistic threat as a potential source of increased prejudice and diminished contact. Therefore, our goal in this study (set in Quebec, Canada) was to explore local residents’ attitudes toward and contact with international students in relation to five potential threats experienced by local residents, including cultural differences, competition over resources, intergroup anxiety, stereotypes, with linguistic threat added as a new, previously unexplored variable. We recruited 59 student and non-student local francophone residents as participants and examined their attitudes toward, perceptions of threat from, and the quality and quantity of contact with international students attending English-medium universities. Both student and non-student participants demonstrated positive attitudes toward and low levels of perceived threat from international students, except for linguistic threat. Compared to student participants, non-student participants reported significantly greater linguistic threat from international students and described contact with them that was both less frequent and lower in quality. Regression models accounted for 50–67% of variance in participants’ attitudes, with symbolic threat to social values and belief systems emerging as the common significant predictor of attitudes for both groups. Adding linguistic threat did not improve regression models. Finally, only contact quality showed significant relationships with attitudes and perceived threat, where greater contact quality was associated with more favorable attitudes and reduced threat. We discuss implications of intergroup attitudes and contact for language planning and use in multilingual contexts.

L-serovar rectal Chlamydia trachomatis in patients who were male-assigned at birth attending two sexual health clinics, Baltimore, Maryland 2009-2016.

ABSTRACT Introduction Lymphogranuloma venereum (LGV) caused by L-serovars of Chlamydia trachomatis (CT) is a sexually transmitted infection (STI) causing asymptomatic infection, and anogenital disease. Epidemiologically, LGV has been strongly associated with HIV infection. Methods Using a retrospective longitudinal study design we first described unadjusted and adjusted associations between demographic and sociobehavioral characteristics and rectal (r) CT (rCT) and rLGV in sexual health clinic attendees in Baltimore, MD between 2009-2016 at baseline visits. Rectal samples that tested positive for CT with the Aptima AC2 assay underwent DNA extraction and in-house L-serovar PCR analysis. Then, patients were followed over 24 months and incident STIs were described using multivariable Cox-proportional hazard regression models. Results In 292/391 (73.4%) patients who were male-assigned at birth (MAB), LGV prevalence was 20/292 (6.8%, 95%CI:4.2%-10.4%). In the final multivariable regression model, LGV was significantly associated with age > 25 years, transgender identity, and a history of syphilis. Baseline LGV was not associated with incident HIV or gonorrhea, but after adjusting for age, race, HIV status, and history of syphilis, LGV was marginally associated with a new syphilis diagnosis [HR 2.35 (95%CI:0.96, 5.73, p = 0.06) over 24 months follow up. Conclusion LGV accounted for almost 7% of rCT in patients MAB attending sexual health clinics in Baltimore between 2009-2016. LGV was not associated with prevalent or incident HIV as in other studies. FDA-cleared tests and better surveillance measures are required to understand changes in the epidemiology of LGV in the U.S.

Research on Forecasting Sales of Pure Electric Vehicles in China Based on the Seasonal Autoregressive Integrated Moving Average–Gray Relational Analysis–Support Vector Regression Model

Aiming to address the complexity and challenges of predicting pure electric vehicle (EV) sales, this paper integrates a time series model, support vector machine and combined model to forecast EV sales in China. Firstly, a seasonal autoregressive integrated moving average (SARIMA) model was constructed using historical EV sales data, and the model was trained on sales statistics to obtain forecasting results. Secondly, variables that were highly correlated with sales were analyzed using gray relational analysis (GRA) and utilized as input parameters for the support vector regression (SVR) model, which was constructed to optimize sales predictions for EVs. Finally, a combined model incorporating different algorithms was verified against market sales data to explore the optimal sales prediction approach. The results indicate that the SARIMA-GRA-SVR model with the squared prediction error and inverse method achieved the best predictive performance, with MAPE, MAE and RMSE values of 12%, 1.45 and 2.08, respectively. This empirical study validates the effectiveness and superiority of the SARIMA-GRA-SVR model in forecasting EV sales.

Association of life’s essential 8 with chronic obstructive pulmonary disease: a population-based analysis of NHANES 2007–2018

Abstract Background Chronic obstructive pulmonary disease (COPD) is closely linked to cardiovascular diseases. We aimed to investigate the association between Life’s Essential 8 (LE8), the newly established measurement to assess cardiovascular health (CVH), and COPD among U.S. general adults. Methods This study extracted the National Health and Nutrition Examination Survey (2007–2018) data. Multivariate logistic regression models were used to examine the associations between LE8 and COPD. A restricted cubic spline regression model was used to explore the dose-response relationships between LE8 scores and COPD. In addition, subgroup and sensitivity analyses were performed to assess the robustness of our results. Results Our study included 19,774 participants representing 145.2 million non-institutionalized U.S. population aged ≥ 20 years. The overall age-adjusted prevalence of COPD was 4.5%. After adjusting for the potential covariates, LE8 was inversely associated with COPD [adjusted odds ratio (AOR) = 0.169, 95% CI: 0.115, 0.249], exhibiting a nonlinear dose-response relationship (P for nonlinearity < 0.05). Similar trends in the associations of health behavior score (AOR = 0.300, 95% CI: 0.223, 0.404) and health factor score (AOR = 0.603, 95% CI: 0.426, 0.852) with COPD were also identified. Furthermore, higher LE8 metric scores of nicotine exposure and sleep health were associated with a lower prevalence of COPD. Conclusion LE8 was inversely associated with spirometric/self-reported COPD in a nonlinear trend, primarily driven by the nicotine exposure metric of LE8. Adhering to LE8 guidelines, especially smoking cessation, to sustain optimal CVH levels may be beneficial to alleviate the burden of COPD.

Phase angle and body composition in long-term type 1 diabetes in adults: a comparative study in a Brazilian public reference outpatient clinic

Abstract Introduction Type 1 Diabetes Mellitus (DM1) affects a small percentage of the population. Nevertheless, its prevalence is currently growing with alarming data on uncontrolled cases. The importance of body composition and Phase Angle (PA), assessed by Bioelectrical Impedance (BIA), in long- term DM1 patients lies in the fact that alterations in cellular integrity and body compartments may affect risk profiles and metabolic control. The objective of this study was to compare PA and body composition parameters between adults with DM1 and healthy controls. Methods A comparative study was carried out in a public university outpatient clinic including a cohort of adult patients of both sexes diagnosed with DM1 and healthy controls matched by age and sex in a 2:1 ratio. Anthropometric measurements included weight, height and BMI. Using the raw BIA data of Resistance and Reactance, fat-free mass (FFM), fat mass (FM), fat-free mass index (FFMI), fat mass index (FMI), PA and standardized PA (SPA) were calculated. Means or medians were compared between the groups. Regression models were used to identify distinguishing characteristics of the groups and associations within the DM1 group (i.e. glycated hemoglobin (HbA1c), disease duration, presence of microvascular complications, capillary blood glucose, BMI and FMI). Results 88 patients with DM1and 46 healthy controls were evaluated. PA (6.05 vs. 6.85, p = 0.000) and SPA (-1.47 vs. -0,37, p = 0.000) were lower in patients with DM1 compared to healthy controls. People with DM1 displayed higher adiposity (%FM = 29.6 vs. 27.6, p = 0.016; FMI = 7.00 vs. 6.33, p = 0.016) and lower %FFM compared to healthy controls. Most of the differences were maintained after sex stratification; however, men with DM1 showed a lower FFMI than male controls (18.2 vs. 20.16, p = 0.029). Conclusion Patients with DM1 present lower PA than healthy controls, which may be related to worse cell membrane integrity. Significant body composition differences between the groups and between sexes were identified, with data showing greater adiposity in women with DM1 and men displaying lower muscle mass. These findings suggest the importance of including PA and body composition evaluations in the follow-up of patients with DM1. The ultimate goal is to obtain a better metabolic control and, consequently, a better prognosis.

Analyzing the adoption of nutrition-sensitive carp-SIS polyculture technology: evidence from a case study in Bangladesh

IntroductionSmall indigenous species (SIS) of fish are rich in micronutrients that are essential to combat the existing malnutrition in Bangladesh. However, their availability is constantly decreasing due to gradual environmental degradation making their availability irregular and hence expensive in the market. Integrating SIS with carps in homestead ponds is being promoted as a form of nutrition-sensitive aquaculture to enhance both production and consumption of these nutrient-rich species. Various improved pond management techniques (IPMTs) are suggested to boost the nutrition-sensitive carp-SIS polyculture.ObjectivesThis study examines the trends and factors influencing the adoption of IPMTs using the sustainable livelihood framework (SLF) approach. We address the following three key questions: (i) What are the trends in production and consumption of fish, in particular SIS from homestead ponds? (ii) Have farm households adopted or disadopted IPMTs over time? (iii) What are the determinants of adoption and disadoption of IPMTs?MethodsBased on primary data of 234 households from Barishal district from 2014 and 2022, we perform significance tests to compare project and non-project households in 2014 versus 2022 and apply fixed effects Poisson regression and fixed effects negative binomial regression models to identify household decisions to adopt the IPMTs.Results and conclusionQuantitative survey results indicate a significant increase in the production and consumption of SIS over time. In 2014, the project households exhibited adoption rates of 60% or more for various IPMTs such as stocking of fast-growing species, pre-stocking liming, pond dike construction, fertilizer application, stocking of high quality fish seeds, aquatic weed control, and turbidity management. However, by 2022, some of these IPMTs were disadopted by project households and only a few techniques, including providing sunlight exposure to ponds, pre-stocking liming, and supplementary feeding, were more widely used. Additionally, many non-project households also increased their adoption rates of IPMTs significantly. The fixed effects regression model shows that adoption is positively influenced by the number of household members participating in aquaculture, size of the pond, sole ownership of the pond and the number of years household has been involved in aquaculture. Follow-up activities such as monitoring and training of the project beneficiaries and government support are recommended to support long-term adoption of the IPMTs.

Prediction of Floor Failure Depth in Coal Mines: A Case Study of Xutuan Mine, China

Accurately predicting the failure depth of coal seam floors is crucial for preventing water damage, ensuring the safe and efficient mining of coal seams, and protecting the ecological environment of mining areas. In order to improve the prediction accuracy of the coal seam floor failure depth, an improved support vector regression (SVR) model is proposed to predict the floor failure depth by taking the 3234 working face in Xutuan Mine as an example. This improved model incorporates principal component analysis (PCA) and slime mould algorithm (SMA) optimization techniques. First, based on the measured data of seam floor failure depth in several mining areas, a prediction index system of floor failure depth was constructed. Subsequently, the PCA method was used to reduce the dimension of the measured data of the coal seam floor failure depth, and the input structure of the SVR model was optimized. Then, the SMA was used to optimize the key parameters, namely the penalty factor (C) and kernel function parameter (g), in the SVR model, achieving automatic parameter selection and obtaining the optimal parameter combination. This process led to the establishment of a coal seam floor failure depth prediction model based on PCA-SMA-SVR. The predictive performance of the PCA-SMA-SVR model, SMA-SVR model, and SVR model was quantitatively evaluated and compared using four quantitative indicators, and the results showed that the PCA-SMA-SVR model had the smallest MAE, RMSE, MRE, and TIC values, which were 1.0470 m, 1.2928 m, 0.0628, and 0.0374, respectively. Finally, the PCA-SMA-SVR model was used to predict that the floor failure depth of the 3234 working face in Xutun Mine was 17.09 m, and the predicted result was compared and analyzed with the results of four commonly used empirical formulas (16.03–21.74 m). The results show that the model is close to the results of four commonly used empirical formulas, indicating that the model has high predictive performance and good practicality. This study is of great significance for the safety, green mining, and ecological environment protection of coal mines.

Strategies for feeding tilapia in smallholder aquaculture systems: a study of aquaculture feed access and availability in two districts of Northern Province, Zambia

ABSTRACT Aquaculture contributes to agricultural development, economic growth, and poverty reduction initiatives in Zambia. In northern and rural regions of the country, development requires improvements to the management of local and farm-generated inputs. Fish farmers have limited market access and purchasing power, which restricts access to commercial inputs, such as formulated feeds and chemical fertilizers. Instead, farmers adopt integrated-agriculture-aquaculture (IAA) practices, whereby on-farm resources supply essential inputs. Research is limited regarding local systems of resource management, and this hinders development of innovations for improving efficiency. This mixed-methods study explored local patterns of resource utilization. Smallholders (n = 63) participated in surveys and interviews regarding resource management. A series of quantitative analyses, including descriptive statistics, principal component analyses (PCA), and multinomial logistic regression (logit) models, identified relationships between these resources, household livelihood, and system productivity. Qualitative analyses explored farmer perceptions regarding seasonal resource availability. Results suggested that livestock manure, homemade feeds comprising maize and soybean, garden residues, and kitchen wastes were the most frequently selected inputs. Logit models of feeding intensity found that income, proximity to Luwingu Town, experience in aquaculture, and access to extension services were key household resources. High-intensity feeding strategies incorporated commercial feeds, chemical fertilizers, as well as home-made feeds containing a wide diversity of both high- and low-quality ingredients. Low-intensity feeding strategies incorporated only farm waste and termites. Potential innovations that may improve system performance in resource poor contexts are considered and recommendations for additional research are offered.

Dengue Prediction using Short-Term Memory

Dengue is a severe infectious disease on the rise in Malaysia, and there is a demand for artificial intelligence to support the health system. However, the application of deep learning, specifically Long Short-Term Memory (LSTM) time series forecasting, has not been explored by many in dengue prediction studies. However, considering the availability of daily weather data being collected, the ability of LSTM to capture long term dependencies can be leveraged in forecasting dengue cases. Therefore, this study investigates the performance and viability of LSTM time series forecasting on predicting dengue cases. An LSTM model is developed and evaluated to be compared to a Support Vector Regression (SVR) model by utilising the availability of a dengue dataset consisting of weather and climate data. The results indicated LSTM time series forecasting performed better than SVR, with R2 and MAE scoring 0.75 and 8.76. In short, LSTM has shown better performance and, in addition, capturing trends in the rise and fall of dengue cases. Altogether, this research could contribute to the fight against the increase of dengue cases without relying on forecasted weather data but instead, history.

Abstract A028: Precise monitoring of osteosarcoma patients by noninvasive assessment of matrix metalloproteinases activities in plasma extracellular vesicles

Abstract Osteosarcoma (OS) is the most prevalent bone cancer in children, characterized by aggressive primary tumors and high metastatic potential. Current treatment protocols involve a combination of neoadjuvant chemotherapy, surgery, and adjuvant chemotherapy, tailored based on histological grade and metastatic risk. Identifying metastatic lesions through routine imaging significantly impacts treatment outcomes, highlighting the need for innovative diagnostic tools for timely monitoring. Noninvasive in vitro diagnostics, analyzing circulating tumor components like tumor-derived extracellular vesicles (EVs), would offer promise for monitoring treatment responses in pediatric sarcomas. Matrix metalloproteinases (MMPs) play a crucial role in tumor progression by degradation of the extracellular matrix; thus, recent research has focused on understanding their functional roles by profiling their proteolytic activities. Tumor EVs are crucial in transporting MMPs to facilitate invasion and metastasis, making them suitable surrogate for noninvasive profiling of MMP activities. Herein, we introduce an innovative OS EV MMP Activity Assay to assess MMP activities in OS patients noninvasively. This assay evaluates six different combinations of surface markers on OS EVs and OS-associated MMPs, generating a unique profile for each patient. Using a logistic regression model, an OS EV MMP Activity Score is generated from the most effective combinations. This Score has shown excellent performance in distinguishing between metastatic and localized OS. In a longitudinal study of six OS patients undergoing treatment, the Score strongly correlated with treatment response as measured by routine imaging. The assay represents a significant advancement in harnessing tumor EVs for diagnostics, providing a noninvasive, quantitative, and reliable approach to assess MMP activities in solid tumors. Citation Format: Junseok Lee, You-Ren Ji, Hyoyong Kim, Zhao Chen, Ryan Zhang, Diego Banuet, Yusef Mathkour, Kenny Vo, Steven Jonas, Yazhen Zhu, Hsian-Rong Tseng. Precise monitoring of osteosarcoma patients by noninvasive assessment of matrix metalloproteinases activities in plasma extracellular vesicles [abstract]. In: Proceedings of the AACR Special Conference: Liquid Biopsy: From Discovery to Clinical Implementation; 2024 Nov 13-16; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(21_Suppl):Abstract nr A028.

Does Intellectual Capital Influence Firm's Financial Performance in Bangladesh?

This paper aimed to find the effects of intellectual capital and its components on firms’ financial performance in Bangladesh. A sample of 100 firms comprised of 48 manufacturing, 21 services, and 31 banking companies was studied for five years from 2017 to 2021, setting 500 firm-year observations. Data was collected purposively from secondary sources, such as annual reports of sampled companies. The robust fixed-effect regression model using STATA 14.2 software was applied to test the hypothesis due to autocorrelation and heteroscedasticity problems. The regression results documented that overall IC, human capital efficiency (HCE), and capital employed efficiency (CEE) positively and significantly enhanced the ROA and ROE of the firms. However, structural capital efficiency (SCE) and relational capital efficiency (RCE) negatively and insignificantly influenced the same. The study contributes to resource-based theory using econometric methods to extend the samples of both financial and non-financial companies. This study helps investors, managers, policy-makers, governments, and accounting regulatory bodies regarding the utilization pattern of invisible and tangible resources in Bangladeshi firms.

Incidence and predictors of clinical failure after early treatment for mild‐to‐moderate COVID‐19 in high‐risk individuals: A multicentric cohort study

AbstractObjectivesTo estimate the risk of COVID‐19‐related hospitalization and death (CovH/D), among high‐risk individuals early treated for COVID‐19 and to identify associated factors.Methods and resultsA multicenter cohort of 12,475 high‐risk outpatients (female 50.2%, median age 70 years [IQR 57–80], fully vaccinated 79.1%, immunocompromised 23.2%) treated with monoclonal antibodies or antivirals for mild‐to‐moderate COVID‐19 (March 2021–May 2023) in the Lazio region, Italy. The unadjusted risk of CovH/D by Day 30 was 3.08% (95% CI 2.7%–3.4%). By means of logistic regression models, which included a specific set of potential confounders for each exposure of interest, we observed a higher risk for the elderly, unvaccinated and immunocompromised participants. Using the “Delta period” as a reference, a decreased risk was observed for Omicron waves.ConclusionsDespite the administration of COVID‐19 early treatment and the decreasing risk of CovH/D across the calendar periods, the elderly, the unvaccinated and the immunocompromised people remain at high risk of clinical progression.

Identifying the geographic leading edge of Lyme disease in the United States with internet searches: A spatiotemporal analysis of Google Health Trends data

Background The geographic footprint of Lyme disease is expanding in the United States, which calls for novel methods to identify emerging endemic areas. The ubiquity of internet use coupled with the dominance of Google’s search engine makes Google user search data a compelling data source for epidemiological research. Objective We evaluated the potential of Google Health Trends to track spatiotemporal patterns in Lyme disease and identify the leading edge of disease risk in the United States. Materials and methods We analyzed internet search rates for Lyme disease-related queries at the designated market area (DMA) level (n = 206) for the 2011–2019 and 2020–2021 (COVID-19 pandemic) periods. We used maps and other exploratory methods to characterize changes in search behavior. To assess statistical correlation between searches and Lyme disease cases reported to Centers for Disease Control and Prevention (CDC) between 2011 and 2019, we performed a longitudinal ecological analysis with modified Poisson generalized estimating equation regression models. Results Mapping DMA-level changes in “Lyme disease” search rates revealed an expanding area of higher rates occurring along the edges of the northeastern focus of Lyme disease. Bivariate maps comparing search rates and CDC-reported incidence rates also showed a stronger than expected signal from Google Health Trends in some high-risk adjacent states such as Michigan, North Carolina, and Ohio, which may be further indication of a geographic leading edge of Lyme disease that is not fully apparent from routine surveillance. Searches for “Lyme disease” were a significant predictor of CDC-reported disease incidence. Each 100-unit increase in the search rate was significantly associated with a 10% increase in incidence rates (RR = 1.10, 95% CI: 1.07, 1.12) after adjusting for environmental covariates of Lyme disease identified in the literature. Conclusion Google Health Trends data may help track the expansion of Lyme disease and inform the public and health care providers about emerging risks in their areas.