Bayesian spatiotemporal modelling for identifying unusual and unstable trends in mammography utilisation.
Ontology highlight
ABSTRACT:
SUBMITTER: Duncan EW
PROVIDER: S-EPMC4885312 | biostudies-other | 2016
REPOSITORIES: biostudies-other
Similar Datasets
Project description:Spatiotemporal patterns and trends of COVID-19 at a local spatial scale using Bayesian approaches are hardly observed in literature. Also, studies rarely use satellite-derived long time-series data on the environment to predict COVID-19 risk at a spatial scale. In this study, we modelled the COVID-19 pandemic risk using a Bayesian hierarchical spatiotemporal model that incorporates satellite-derived remote sensing data on land surface temperature (LST) from January 2020 to October 2021 (89 weeks) and several socioeconomic covariates of the 140 neighbourhoods in Toronto. The spatial patterns of risk were heterogeneous in space with multiple high-risk neighbourhoods in Western and Southern Toronto. Higher risk was observed during Spring 2021. The spatiotemporal risk patterns identified 60% of neighbourhoods had a stable, 37% had an increasing, and 2% had a decreasing trend over the study period. LST was positively, and higher education was negatively associated with the COVID-19 incidence. We believe the use of Bayesian spatial modelling and the remote sensing technologies in this study provided a strong versatility and strengthened our analysis in identifying the spatial risk of COVID-19. The findings would help in prevention planning, and the framework of this study may be replicated in other highly transmissible infectious diseases.
| S-EPMC9172088 | biostudies-literature
Project description:BACKGROUND:As one of its Millennium Development Goals (MDGs), China has achieved a dramatic reduction in the maternal mortality ratio (MMR), although a distinct spatial heterogeneity still persists. Evidence of the quantitative effects of determinants on MMR in China is limited. A better understanding of the spatiotemporal heterogeneity and quantifying determinants of the MMR would support evidence-based policymaking to sustainably reduce the MMR in China and other developing areas worldwide. METHODS AND FINDINGS:We used data on MMR collected by the National Maternal and Child Health Surveillance System (NMCHSS) at the county level in China from 2010 to 2013. We employed a Bayesian space-time model to investigate the spatiotemporal trends in the MMR from 2010 to 2013. We used Bayesian multivariable regression and GeoDetector models to address 3 main ecological determinants of the MMR, including per capita income (PCI), the proportion of pregnant women who delivered in hospitals (PPWDH), and the proportion of pregnant women who had at least 5 check-ups (PPWFC). Among the 2,205 counties, there were 925 (42.0%) hotspot counties, located mostly in China's western and southwestern regions, with a higher MMR, and 764 (34.6%) coldspot counties with a lower MMR than the national level. China's westernmost regions, including Tibet and western Xinjiang, experienced a weak downward trend over the study period. Nationwide, medical intervention was the major determinant of the change in MMR. The MMR decreased by 1.787 (95% confidence interval [CI]: 1.424-2.142, p < 0.001) per 100,000 live births when PPWDH increased by 1% and decreased by 0.623 (95% CI 0.436-0.798, p < 0.001) per 100,000 live births when PPWFC increased by 1%. The major determinants for the MMR in China's western and southwestern regions were PCI and PPWFC, while that in China's eastern and southern coastlands was PCI. The MMR in western and southwestern regions decreased nonsignificantly by 1.111 (95% CI -1.485-3.655, p = 0.20) per 100,000 live births when PCI in these regions increased by 1,000 Chinese Yuan and decreased by 1.686 (95% CI 1.275-2.090, p < 0.001) when PPWFC increased by 1%. Additionally, the western and southwestern regions showed the strongest interactive effects between different factors, in which the corresponding explanatory power of any 2 interacting factors reached up to greater than 80.0% (p < 0.001) for the MMR. Limitations of this study include a relatively short study period and lack of full coverage of eastern coastlands with especially low MMR. CONCLUSIONS:Although China has accomplished a 75% reduction in the MMR, spatial heterogeneity still exists. In this study, we have identified 925 (hotspot) high-risk counties, mostly located in western and southwestern regions, and among which 332 counties are experiencing a slower pace of decrease than the national downward trend. Nationally, medical intervention is the major determinant. The major determinants for the MMR in western and southwestern regions, which are developing areas, are PCI and PPWFC, while that in China's developed areas is PCI. The interactive influence of any two of the three factors, PCI, PPWDH, and PPWFC, in western and southwestern regions was up to and in excess of 80% (p < 0.001).
| S-EPMC7228041 | biostudies-literature
Project description:This paper describes a Bayesian hierarchical approach to predict short-term concentrations of particle pollution in an urban environment, with application to inhalable particulate matter (PM10) in Greater London. We developed and compared several spatiotemporal models that differently accounted for factors affecting the spatiotemporal properties of particle concentrations. We considered two main source contributions to ambient measurements: (i) the long-range transport of the secondary fraction of particles, which temporal variability was described by a latent variable derived from rural concentrations; and (ii) the local primary component of particles (traffic- and non-traffic-related) captured by the output of the dispersion model ADMS-Urban, which site-specific effect was described by a Bayesian kriging. We also assessed the effect of spatiotemporal covariates, including type of site, daily temperature to describe the seasonal changes in chemical processes affecting local PM10 concentrations that are not considered in local-scale dispersion models and day of the week to account for time-varying emission rates not available in emissions inventories. The evaluation of the predictive ability of the models, obtained via a cross-validation approach, revealed that concentration estimates in urban areas benefit from combining the city-scale particle component and the long-range transport component with covariates that account for the residual spatiotemporal variation in the pollution process.
| S-EPMC3994509 | biostudies-other
Project description:Monitoring and control of West Nile virus (WNV) presents a challenge to state and local vector control managers. Models of mosquito presence and viral incidence have revealed that variations in mosquito autecology and land use patterns introduce unique dynamics of disease at the scale of a county or city, and that effective prediction requires locally parameterized models. We applied Bayesian spatiotemporal modeling to West Nile surveillance data from 49 mosquito trap sites in Nassau County, New York, from 2001 to 2015 and evaluated environmental and sociological predictors of West Nile virus incidence in Culex pipiens-restuans. A Bayesian spike-and-slab variable selection algorithm was used to help select influential independent variables. This method can be used to identify locally-important predictors. The best model predicted West Nile positives well, with an Area Under Curve (AUC) of 0.83 on holdout data. The temporal trend was nonlinear and increased throughout the year. The spatial component identified increased West Nile incidence odds in the northwestern portion of the county, with lower odds in wetlands on the south shore of Long Island. High Normalized Difference Vegetation Index (NDVI) areas, wetlands, and areas of high urban development had negative associations with WNV incidence. In this study we demonstrate a method for improving spatiotemporal models of West Nile virus incidence for decision making at the county and community scale, which empowers disease and vector control organizations to prioritize and evaluate prevention efforts.
| S-EPMC7676626 | biostudies-literature
Project description:BackgroundAlcohol use has widespread effects on health and contributes to over 200 detrimental conditions. Although the pattern of heavy episodic drinking independently increases the risk for injuries and transmission of some infectious diseases, long-term average consumption is the fundamental predictor of risk for most conditions. Population surveys, which are the main source of data on alcohol exposure, suffer from bias and uncertainty. This article proposes a novel triangulation method to reduce bias by rescaling consumption estimates by sex and age to match country-level consumption from administrative data.MethodsWe used data from 17 population surveys to estimate age- and sex-specific trends in alcohol consumption in the adult population of South Africa between 1998 and 2016. Independently for each survey, we calculated sex- and age-specific estimates of the prevalence of drinkers and the distribution of individuals across consumption categories. We used these aggregated results, together with data on alcohol production, sales and import/export, as inputs of a Bayesian model and generated yearly estimates of the prevalence of drinkers in the population and the parameters that characterise the distribution of the average consumption among drinkers.ResultsAmong males, the prevalence of drinkers decreased between 1998 and 2009, from 56.2% (95% CI 53.7%; 58.7%) to 50.6% (49.3%; 52.0%), and increased afterwards to 53.9% (51.5%; 56.2%) in 2016. The average consumption from 52.1 g/day (49.1; 55.6) in 1998 to 42.8 g/day (40.0; 45.7) in 2016. Among females the prevalence of current drinkers rose from 19.0% (17.2%; 20.8%) in 1998 to 20.0% (18.3%; 21.7%) in 2016 while average consumption decreased from 32.7 g/day (30.2; 35.0) to 26.4 g/day (23.8; 28.9).ConclusionsThe methodology provides a viable alternative to current approaches to reconcile survey estimates of individual alcohol consumption patterns with aggregate administrative data. It provides sex- and age-specific estimates of prevalence of drinkers and distribution of average daily consumption among drinkers in populations. Reliance on locally sourced data instead of global and regional trend estimates better reflects local nuances and is adaptable to the inclusion of additional data. This provides a powerful tool to monitor consumption, develop burden of disease estimates and inform and evaluate public health interventions.
| S-EPMC8565040 | biostudies-literature
Project description:To develop criteria to identify thresholds for the minimally acceptable performance of physicians interpreting diagnostic mammography studies.In an institutional review board-approved HIPAA-compliant study, an Angoff approach was used to set criteria for identifying minimally acceptable interpretive performance for both workup after abnormal screening examinations and workup of a breast lump. Normative data from the Breast Cancer Surveillance Consortium (BCSC) was used to help the expert radiologist identify the impact of cut points. Simulations, also using data from the BCSC, were used to estimate the expected clinical impact from the recommended performance thresholds.Final cut points for workup of abnormal screening examinations were as follows: sensitivity, less than 80%; specificity, less than 80% or greater than 95%; abnormal interpretation rate, less than 8% or greater than 25%; positive predictive value (PPV) of biopsy recommendation (PPV2), less than 15% or greater than 40%; PPV of biopsy performed (PPV3), less than 20% or greater than 45%; and cancer diagnosis rate, less than 20 per 1000 interpretations. Final cut points for workup of a breast lump were as follows: sensitivity, less than 85%; specificity, less than 83% or greater than 95%; abnormal interpretation rate, less than 10% or greater than 25%; PPV2, less than 25% or greater than 50%; PPV3, less than 30% or greater than 55%; and cancer diagnosis rate, less than 40 per 1000 interpretations. If underperforming physicians moved into the acceptable range after remedial training, the expected result would be (a) diagnosis of an additional 86 cancers per 100,000 women undergoing workup after screening examinations, with a reduction in the number of false-positive examinations by 1067 per 100,000 women undergoing this workup, and (b) diagnosis of an additional 335 cancers per 100,000 women undergoing workup of a breast lump, with a reduction in the number of false-positive examinations by 634 per 100,000 women undergoing this workup.Interpreting physicians who fall outside one or more of the identified cut points should be reviewed in the context of an overall assessment of all their performance measures and their specific practice setting to determine if remedial training is indicated.
| S-EPMC3632803 | biostudies-other
Project description:Identifying children who are at-risk for developmental delay, so that these children can have access to interventions as early as possible, is an important and challenging problem in developmental research. This research aimed to identify latent subgroups of children with developmental delay, by modelling and clustering developmental milestones. The main objectives were to (a) create a developmental profile for each child by modelling milestone achievements, from birth to three years of age, across multiple domains of development, and (b) cluster the profiles to identify groups of children who show similar deviations from typical development. The ensemble methodology used in this research consisted of three components: (1) Bayesian sequential updating was used to model the achievement of milestones, which allows for updated predictions of development to be made in real time; (2) a measure was created that indicated how far away each child deviated from typical development for each functional domain, by calculating the area between each child's obtained sequence of posterior means and a sequence of posterior means representing typical development; and (3) Dirichlet process mixture modelling was used to cluster the obtained areas. The data used were 348 binary developmental milestone measurements, collected from birth to three years of age, from a small community sample of young children (N = 79). The model identified nine latent groups of children with similar features, ranging from no delays in all functional domains, to large delays in all domains. The performance of the Dirichlet process mixture model was validated with two simulation studies.
| S-EPMC7266333 | biostudies-literature
Project description:We introduce a probabilistic (Bayesian) framework and associated software toolbox for mapping population receptive fields (pRFs) based on fMRI data. This generic approach is intended to work with stimuli of any dimension and is demonstrated and validated in the context of 2D retinotopic mapping. The framework enables the experimenter to specify generative (encoding) models of fMRI timeseries, in which experimental stimuli enter a pRF model of neural activity, which in turns drives a nonlinear model of neurovascular coupling and Blood Oxygenation Level Dependent (BOLD) response. The neuronal and haemodynamic parameters are estimated together on a voxel-by-voxel or region-of-interest basis using a Bayesian estimation algorithm (variational Laplace). This offers several novel contributions to receptive field modelling. The variance/covariance of parameters are estimated, enabling receptive fields to be plotted while properly representing uncertainty about pRF size and location. Variability in the haemodynamic response across the brain is accounted for. Furthermore, the framework introduces formal hypothesis testing to pRF analysis, enabling competing models to be evaluated based on their log model evidence (approximated by the variational free energy), which represents the optimal tradeoff between accuracy and complexity. Using simulations and empirical data, we found that parameters typically used to represent pRF size and neuronal scaling are strongly correlated, which is taken into account by the Bayesian methods we describe when making inferences. We used the framework to compare the evidence for six variants of pRF model using 7 T functional MRI data and we found a circular Difference of Gaussians (DoG) model to be the best explanation for our data overall. We hope this framework will prove useful for mapping stimulus spaces with any number of dimensions onto the anatomy of the brain.
| S-EPMC7417811 | biostudies-literature
Project description:BACKGROUND:Opioid use has more than doubled over recent decades, and Denmark occupies fifth place in the global ranking. These increases have been partly attributed to the ageing population. OBJECTIVE:Our objective was to assess the impact of age over time on utilisation of the most commonly used opioids in Denmark. METHODS:We retrieved nationwide sales data on opioid sales in Denmark from 1999 to 2017. We investigated utilisation trends in age groups for the four opioids with the highest use. We used three volume-based metrics (defined daily doses/1000/day, oral morphine equivalents/1000/day, and packages dispensed/year) and one person-based metric (users/1000/year). RESULTS:The four opioids selected according to users/1000/year were tramadol (46.1), codeine and combination products (12.4 for codeine, 3.7 for codeine and acetylsalicylic acid, and 4.2 for codeine and paracetamol), morphine (17.0), and oxycodone (12.1). Overall utilisation according to volume and person metrics increased for all except codeine and combination products. Tramadol doses or strength increased, albeit less with increasing age. Oxycodone doses or strength decreased for all age groups but were nearly unchanged for the age group ??80 years. CONCLUSION:Tramadol is the most utilised opioid in Denmark and was prescribed at increasing doses or strengths over the study period, particularly in the younger (<?80 years) age groups. Overall, oxycodone was prescribed at decreasing doses or strengths over time but remained unchanged for the age group ??80 years. There is a need to address the pharmacological treatment of pain in terms of age, with tramadol and oxycodone being possible targets for regulatory efforts.
| S-EPMC6879688 | biostudies-literature
Project description:Recently, evidence has emerged that humans approach learning using Bayesian updating rather than (model-free) reinforcement algorithms in a six-arm restless bandit problem. Here, we investigate what this implies for human appreciation of uncertainty. In our task, a Bayesian learner distinguishes three equally salient levels of uncertainty. First, the Bayesian perceives irreducible uncertainty or risk: even knowing the payoff probabilities of a given arm, the outcome remains uncertain. Second, there is (parameter) estimation uncertainty or ambiguity: payoff probabilities are unknown and need to be estimated. Third, the outcome probabilities of the arms change: the sudden jumps are referred to as unexpected uncertainty. We document how the three levels of uncertainty evolved during the course of our experiment and how it affected the learning rate. We then zoom in on estimation uncertainty, which has been suggested to be a driving force in exploration, in spite of evidence of widespread aversion to ambiguity. Our data corroborate the latter. We discuss neural evidence that foreshadowed the ability of humans to distinguish between the three levels of uncertainty. Finally, we investigate the boundaries of human capacity to implement Bayesian learning. We repeat the experiment with different instructions, reflecting varying levels of structural uncertainty. Under this fourth notion of uncertainty, choices were no better explained by Bayesian updating than by (model-free) reinforcement learning. Exit questionnaires revealed that participants remained unaware of the presence of unexpected uncertainty and failed to acquire the right model with which to implement Bayesian updating.
| S-EPMC3024253 | biostudies-literature