SEARCH · PubMed Health

Results for “Uncertainty modelling”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 595 records · Page 33Linked to original sources

PubMed2003

An estimate of spatial uncertainty of mean concentrations predicted by Gaussian dispersion models.

Atmospheric dispersion models based on statistical theory predict the stochastic mean concentration at fixed points downwind. The uncertainty in the computed concentration, often expressed as a peak-to-mean ratio, is known to vary between two and five for Gaussian dispersion models. This paper shows that this can be correlated to uncertainty in the predicted location of a fixed concentration, and that in the centerline of the plume it is related to sigma(theta), the standard deviation of the observed horizontal wind direction. Practical formulae and methods are given that may be applied in the post-dispersion calculation phase of the risk assessment to ascertain the likely hazard zones.

Air Pollution, Radioactive↗

PubMed1998

Uncertainty and variability analysis in multiplicative risk models.

Currently, there is a trend away from the use of single (often conservative) estimates of risk to summarize the results of risk analyses in favor of stochastic methods which provide a more complete characterization of risk. The use of such stochastic methods leads to a distribution of possible values of risk, taking into account both uncertainty and variability in all of the factors affecting risk. In this article, we propose a general framework for the analysis of uncertainty and variability for use in the commonly encountered case of multireplicative risk models, in which risk may be expressed as a product of two or more risk factors. Our analytical methods facilitate the evaluation of overall uncertainty and variability in risk assessment, as well as the contributions of individual risk factors to both uncertainty and variability which is cumbersome using Monte Carlo methods. The use of these methods is illustrated in the analysis of potential cancer risks due to the ingestion of radon in drinking water.

Analysis of Variance↗

PubMed2006

Effect of uncertainty on depression in patients with newly diagnosed atrial fibrillation.

This cross-sectional survey study aimed to determine the effects of uncertainty on depression through the appraisal within the "uncertainty in illness" model. In 81 subjects with atrial fibrillation, individuals with greater symptom severity perceived more uncertainty. Uncertainty was appraised as a danger rather than an opportunity. While the appraisal of opportunity had a negative relationship with depression, the appraisal of danger was positively associated with depression. The indirect effect of uncertainty on depression through appraisal was supported by the empirical data on patients with atrial fibrillation. Helping patients to correctly interpret symptoms and to understand the effectiveness of treatment may modify and reduce their uncertainty so that uncertainty leading to worse health outcomes may be reduced.

Adaptation, Psychological↗

PubMed2006

Diazepam pharamacokinetics from preclinical to phase I using a Bayesian population physiologically based pharmacokinetic model with informative prior distributions in WinBUGS.

Modelling is an important applied tool in drug discovery and development for the prediction and interpretation of drug pharmacokinetics. Preclinical information is used to decide whether a compound will be taken forwards and its pharmacokinetics investigated in human. After proceeding to human little to no use is made of these often very rich data. We suggest a method where the preclinical data are integrated into a whole body physiologically based pharmacokinetic (WBPBPK) model and this model is then used for estimating population PK parameters in human. This approach offers a continuous flow of information from preclinical to clinical studies without the need for different models or model reduction. Additionally, predictions are based upon single parameter values, but making realistic predictions involves incorporating the various sources of variability and uncertainty. Currently, WBPBPK modelling is undertaken as a two-stage process: (i) estimation (optimisation) of drug-dependent parameters by either least squares regression or maximum likelihood and (ii) accounting for the existing parameter variability and uncertainty by stochastic simulation. To address these issues a general Bayesian approach using WinBUGS for estimation of drug-dependent parameters in WBPBPK models is described. Initially applied to data in rat, this approach is further adopted for extrapolation to human, which allows retention of some parameters and updating others with the available human data. While the issues surrounding the incorporation of uncertainty and variability within prediction have been explored within WBPBPK modeling methodology they have equal application to other areas of pharmacokinetics, as well as to pharmacodynamics.

Algorithms↗

PubMed1991

Magnetic resonance diffusion/perfusion phantom experiments.

Recently, several models for determining microcirculatory parameters using magnetic resonance imaging have been proposed. These include the intravoxel incoherent motion (IVIM) model, the intravoxel coherent motion (IVCM) model, and various tracer models. In order to evaluate these models before extension into physiological systems, phantom studies were used to assess model assumptions, measurement uncertainties, and sensitivity to changes in perfusion. Emphasis is placed on the IVIM model, but the techniques discussed could be extended to evaluation of other models as well. An overview of considerations in pulse sequence development, phantom design, and data interpretation is presented for a variety of phantoms ranging in complexity from stationary volumes of fluid and mechanically pumped phantoms to isolated animal kidneys and finally to an in vivo animal model.

Animals↗

PubMedDate not supplied

A neural network approach to approximating MAP in belief networks.

Bayesian belief networks (BBN) are a widely studied graphical model for representing uncertainty and probabilistic interdependence among variables. One of the factors that restricts the model's wide acceptance in practical applications is that the general inference with BBN is NP-hard. This is also true for the maximum a posteriori probability (MAP) problem, which is to find the most probable joint value assignment to all uninstantiated variables, given instantiation of some variables in a BBN. To circumvent the difficulty caused by MAP's computational complexity, we suggest in this paper a neural network approximation approach. With this approach, a BBN is treated as a neural network without any change or transformation of the network structure, and the node activation functions are derived based on an energy function defined over a given BBN. Three methods are developed. They are the hill-climbing style discrete method, the simulated annealing method, and the continuous method based on the mean field theory. All three methods are for BBN of general structures, with the restriction that nodes of BBN are binary variables. In addition, rules for applying these methods to noisy-or networks are also developed, which may lead to more efficient computation in some cases. These methods' convergence is analyzed, and their validity tested through a series of computer experiments with two BBN of moderate size and complexity. Although additional theoretical and empirical work is needed, the analysis and experiments suggest that this approach may lead to effective and accurate approximation for MAP problems.

Algorithms↗

PubMed2001

Influences of parameter uncertainties within the ICRP 66 respiratory tract model: particle deposition.

Risk assessment associated with the inhalation of radioactive aerosols requires as an initial step the determination of particle deposition within the various anatomic regions of the respiratory tract. The model outlined in ICRP Publication 66 represents to date one of the most complete overall descriptions of not only particle deposition, but of particle clearance and local radiation dosimetry of lung tissues. In this study, a systematic review of the deposition component within the ICRP 66 respiratory tract model was conducted in which probability density functions were assigned to all input parameters. These distributions were subsequently incorporated within a computer code LUDUC (LUng Dose Uncertainty Code) in which Latin hypercube sampling techniques are used to generate multiple (e.g., 1,000) sets of input vectors (i.e., trials) for all of the model parameters needed to assess particle deposition within the extrathoracic (anterior and posterior), bronchial, bronchiolar, and alveolar-interstitial regions of the ICRP 66 respiratory tract model. Particle deposition values for the various trial simulations were shown to be well described by lognormal probability distributions. Geometric mean deposition fractions from LUDUC were found to be within approximately +/- 10% of the single-value estimates from the LUDEP computer code for each anatomic region and for particle diameters ranging from 0.001 to 50 microm. In all regions of the respiratory tract, LUDUC simulations for an adult male at light exertion show that uncertainties in particle deposition fractions are distributed only over a range of about a factor of approximately 2-4 for particle sizes between 0.005 to 0.2 microm. Below 0.005 microm, uncertainties increase only for deposition within the alveolar region. At particle sizes exceeding 1 microm, uncertainties in the deposition fraction within the extrathoracic regions are relatively small, but approach a factor of 20 for deposition in the bronchial region. Deposition fractions for particles above 1 microm become very uncertain within the deeper regions of the lungs (bronchiolar and alveolar-interstitial).

Body Height↗

PubMed1992

Use of historical assessment for evaluation of process-based model projections of future environmental change: Lake acidification in the Adirondack mountains, New York, USA.

Because of the considerable uncertainties associated with modeling complex ecosystem processes, it is essential that every effort be made to test model performance prior to relying on model projections for assessment of future surface water chemical response to environmental perturbation. Unfortunately, long-term chemical data with which to validate model performance are seldom available. The authors present here an evaluation of historical acidification of lake waters in the northeastern United States, and compare historical changes in a set of lakes to hindcasts from the same watershed model (MAGIC) used to estimate future changes in response to acidic deposition. The historical analyses and comparisons with MAGIC model hindcasts and forecasts of acid-base response demonstrate that the acidic and low-ANC lakes in this region are responsive to strong acid inputs. However, the model estimates suggest lakewater chemistry is more responsive to atmospheric inputs of sulfur than do the estimates based on paleolimnological historical analyses. A 'weight-of-evidence approach' that incorporates all available sources of information regarding acid-base response provides a more reasonable estimate of future change than an approach based on model projections alone. The results of these analyses have important implications for predicting future surface water chemical change in response to acidic deposition, establishing critical loads of atmospheric pollutants, and other environmental assessment activities where natural variation often exceeds the trends under investigation (high noise-to-signal ratio). Under these conditions, it is particularly important to evaluate future model projections in light of historical trends data.

Journal Article↗

PubMed2006

Uncertainties in aerosol deposition within the respiratory tract using the icrp 66 model: a study in workers.

This study estimates uncertainties in aerosol deposition within the main regions of the human respiratory tract calculated using the ICRP 66 model. Uniform, triangular, normal, or lognormal distributions were assigned to the model parameters, which involve physical properties of aerosols, their inhalability, their thermo- and aerodynamic deposition efficiencies, and the anatomy, physiology, and exertion level of the individuals. Calculations were performed over a range of aerosol sizes from 0.01 to 50 mum. Monodispersed aerosols were characterized by their aerodynamic diameter (dae). Polydispersed aerosols were characterized by their activity median aerodynamic diameters (AMADs) and the geometric standard deviation (GSD) in diameter. Lognormal distributions of particle deposition were generally observed with low GSD (< 2). The highest uncertainties were observed within the deep lung for the smallest and the largest aerosol sizes, which were mainly due either to particle density or to aerodynamic deposition efficiencies and anatomical and physiological variability, respectively. In the case of diameters larger than 5 mum, uncertainties in the deep lung deposition were much more important for monodispersed than for polydispersed aerosols. This was explained both by the size distribution of the deposited aerosol, the median of which corresponded to a maximal dae value of about 7 and 5 in bronchioles and alveoli, respectively, and by the absence of deposition, which occurs for dae equal to or larger than 50 mum, depending on the exertion level. Thus, in the range of AMADs considered, for the four default workers proposed by ICRP 66, uncertainties in aerosol deposition remain low, with GSD smaller than 3.

Aerosols↗

PubMed2005

The cost-effectiveness of basiliximab induction in "old-to-old" kidney transplant programs: Bayesian estimation, simulation, and uncertainty analysis.

INTRODUCTION: Markov models are employed in economic analyses to evaluate all possible expectations in a dilemna. The introduction of a new clinical protocol (Basiliximab induction with calcineurin-sparing protocols) for a group of kidney transplant recipients receiving organs from marginal donors was validated with a Markov simulation model, demonstrating the usefulness of combining simulation with Bayesian estimation methods for analysis of cost-effectiveness data collected alongside a clinical trial. We sought to determine whether calcineurin-sparing protocols using anti-interleukin-2/antibody induction (Simulect) would show a beneficial effect on initial kidney function and reduce transplantation costs upon admission, clinical incidences, graft function, and complications during the first month after transplant. PATIENTS AND METHODS: A Markov Chain Monte Carlo (MCMC) was used to estimate a system of generalized linear models relating costs and outcomes to a kidney transplant process affected by treatment under alternative therapies. The Markov simulation model was established following three chains: a calcineurin-free regimen with Basiliximab induction (chain A); a calcineurin-sparing protocol with Basiliximab induction (chain B); and a conventional immunosuppressive regimen (chain C). The MCMC draws were used as parameters in simulations that yielded inferences about the relative cost-effectiveness of the novel therapy under a variety of scenarios. After designing the Markov chain and cohorts, 31 patients from the "old-to-old" program were assigned; eight to chain A; eight to chain B; and 15 to chain C. A year after transplantation a cost-benefit study was performed guided by the three branches of the Markov model. RESULTS: The Markov model showed a benefit of induction therapies in elderly patients. A cost-benefit model showed that after a year, there was a clear benefit from calcineurin-free plus Basiliximab induction therapies, with a slight benefit from calcineurin-sparing protocols. CONCLUSIONS: Markov models are extremely useful when introducing new clinical therapies. The approach allows flexibility in assessing treatment using various premises and quantifies the global effect of parametric uncertainty on a decision maker's confidence to adopt one therapy over another. In our transplant program, a cost-effective analysis of outcomes in old patients using the Markov model showed a clear benefit of calcineurin-sparing protocols with Basixilimab induction.

Age Factors↗

PubMed2006

On the structure of simple molecular liquids SbCl5 and WCl6.

Neutron diffraction measurements on liquid antimony pentachloride and tungsten hexachloride have been carried out using the Studsvik liquids and amorphous diffractometer (SLAD) at the Studsvik Neutron Research Laboratory. The corrected structure factors have been interpreted by means of reverse Monte Carlo modeling which provides large structural models, containing thousands of atoms, that are consistent with the experimental data within their uncertainties. From these models, partial structure factors and pair correlation functions can be calculated. It is demonstrated that the intramolecular structure can be determined on the basis of data extending up to 10 Angstrom (-1). SbCl(5) is found to have a trigonal bipyramidal shape in the liquid, while liquid WCl(6) consists of octahedral molecules. The intermolecular structure of liquid SbCl(5) and WCl(6) seems to be determined largely by steric effects (excluded volume and molecular shape).

Journal Article↗

PubMed1979

Comparison of WKB and finite difference calculations for a two-dimensional cochlear model.

There are many points of uncertainty in the subject of cochlear models. In this paper only the question of efficient computing methods is addressed. For the cochlear model with a one-dimensional approximation for the fluid motion, Zweig, Lipes, and Pierce [J. Acoust. Soc. Am. 59, 975-982 (1976)] have shown that the WKB method agrees well with a direct numerical integration. For the two-dimensional fluid model, Neely [E.D. thesis, California Institute of Technology, Pasadena, CA (1977)] has shown that a direct finite difference solution is an order of magnitude faster than the integral equation approach used by Allen [J. Acoust. Soc. Am 61, 110-119 (1977)]. In the present work, a formal WKB solution is derived following Whitham [Linear and Nonlinear Waves (Wiley, New York, 1974)]. The advantage of this formulation is simplicity, but the disadvantage is that no error estimate is available. We find that the numerical results from the WKB solution agree well with those of Neely (1977), while the computer time is reduced by another order of magnitude. Thus, the WKB method seems to offer the satisfactory accuracy, efficiency, and flexibility for treating the more realistic cochlear models.

Cochlea↗

PubMed2005

Use of Bayesian modelling in risk assessment: application to growth of Listeria monocytogenes and food flora in cold-smoked salmon.

An attempt to use a Bayesian approach to model variability and uncertainty separately in microbial growth in a risk assessment is presented. It was conducted within the framework of a French project aiming at assessing the exposure to Listeria monocytogenes in cold-smoked salmon. The chosen model describes the effect of time and temperature on bacterial growth. A Bayesian approach close to the one proposed by Pouillot et al. [Int. J. Food Microbiol. 81 (2003) 87] is used to estimate the variability and uncertainty of growth parameters from both literature data and data experimentally acquired during the project. Variability between strains and between products is taken into account. The growth of the food flora of cold-smoked salmon is also modelled by the same method. The results obtained for both models are used to predict the simultaneous growth of L. monocytogenes and food flora in cold-smoked salmon with a competitive model, expressing variability and uncertainty through a second-order Monte Carlo simulation.

Animals↗

PubMed2004

Evaluation of a TG-43 compliant analytical dosimetry model in clinical 192Ir HDR brachytherapy treatment planning and assessment of the significance of source position and catheter reconstruction uncertainties.

A simple, time efficient, analytical model incorporating heterogeneities and body dimensions around a point 192Ir source is generalized for accurate dosimetry around commercially available 192Ir brachytherapy sources. The generalized model was verified in dosimetry of a clinical 192Ir high dose rate prostate monotherapy application, involving 16 catheters and 83 source dwell positions, through comparison with corresponding treatment planning system data. The computational time efficiency and accuracy of the proposed model allowed the assessment of the impact that uncertainties in source dwell positions and catheter reconstruction may have on dose distributions, and how these could potentially affect the clinical outcome. Results revealed that a 0.1 cm catheter reconstruction uncertainty and a 0.15 cm source position uncertainty along the catheter lead to a dose uncertainty of less than 2% for doses lower than 200% of the prescribed dose, reaching up to 5% for points lying in close proximity to the catheters. These uncertainties were found to have no impact (less than 1%) on dose volume histogram results of both the planning target volume and the urethra. A catheter reconstruction uncertainty as high as 0.2 cm results in a dose uncertainty greater than 2%, reaching up to 9%, only for points inside the 150% contour. However, even in this case, the impact on dose volume histogram calculations is less than 3%.

Brachytherapy↗

PubMed2006

Success at first insemination in Australian Angus cattle: analysis of uncertain binary responses.

Field data from Australian Angus herds were used to investigate 2 methods of analyzing uncertain binary responses for success or failure at first insemination. A linear mixed model that included herd, year, and month of mating as fixed effects; unrelated service sire, additive animal, and residual as random effects; and linear and quadratic effects of age at mating as covariates was used to analyze binary data. An average gestation length (GL) derived from artificial insemination data was used to assign an insemination date to females mated to natural service sires. Females that deviated from this average GL led to uncertain binary responses. Two analyses were carried out: 1) a threshold model fitted to uncertain binary data, ignoring uncertainty (M1); and 2) a threshold model fitted to uncertain binary data, accounting for uncertainty via fuzzy logic classification (M2). There was practically no difference between point estimates obtained from M1 and M2 for service sire and herd variance; however, when uncertain binary data were analyzed ignoring uncertainty (M1), additive variance and heritability estimates were greater than with M2. Pearson correlations indicated that no major reranking would be expected for service sire effects and animal breeding values using M1 and M2. Given the results of the current study, a threshold model contemplating uncertainty is suggested for noisy binary data to avoid bias when estimating genetic parameters.

Animals↗

PubMed1989

Long-term change in the suitability of welsh streams for dippers Cinclus cinclus as a result of acidification and recovery: a modelling study.

Using a recognized and widely used hydrochemical model, MAGIC, long-term changes in acidity were simulated at 104 sites in the acid sensitive region of upland Wales. Conditions were modelled in the future (2010) under different reductions in sulphate deposition from 0 to 90% of 1984 values. Chemical output from the model was used to simulate change in the chemical suitability of streams for a species of river bird, the Dipper Cinclus cinclus, known to be affected by acidification According to simulations, only reductions in sulphate deposition by over 50% of 1984 levels prevented decline in the number of streams chemically suitable for Dippers. Greater reductions in deposition in the model permitted some recovery except where conifer forestry occupied acid sensitive catchments. There are several uncertainties with the models in their present form.

Journal Article↗

PubMed2003

Simulation results for on-line optimization of a batch bioreactor using nonlinear filtering and optimal control.

The computation of optimal control profiles for batch bioreactors is based on the use of simple and empirical dynamic models. Since these models present some level of uncertainty, the difference between the model dynamics and the reactor dynamics can have significant effects in the reliability of the calculated profile. To develop near optimal control trajectories considering this drawback, we propose to calculate successive control profiles on a moving time horizon using a mathematical model in which the kinetic parameters are estimated by an observer. The desired objective is to generate a near optimal control trajectory adapted to the "running" fermentation. This idea results in a nonlinear estimator plus an optimizer arrangement that so far has not been applied to batch fermentors. Numerical simulations are performed on xanthan-gum batch fermentations and reasonably good results are obtained.

Algorithms↗

PubMedDate not supplied

A methodology for validating numerical ground water models.

Ground water validation is one of the most challenging issues facing modelers and hydrogeologists. Increased complexity in ground water models has created a gap between model predictions and the ability to validate or build confidence in predictions. Specific procedures and tests that can be easily adapted and applied to determine the validity of site-specific ground water models do not exist. This is true for both deterministic and stochastic models, with stochastic models posing the more difficult validation problem. The objective of this paper is to propose a general validation approach that addresses important issues recognized in previous validation studies, conferences, and symposia. The proposed method links the processes for building, calibrating, evaluating, and validating models in an iterative loop. The approach focuses on using collected validation data to reduce uncertainty in the model and narrow the range of possible outcomes. This method is designed for stochastic numerical models utilizing Monte Carlo simulation approaches, but it can be easily adapted for deterministic models. The proposed methodology relies on the premise that absolute validity is not theoretically possible, nor is it a regulatory requirement. Rather, the proposed methodology highlights the importance of testing various aspects of the model and using diverse statistical tools for rigorous checking and confidence building in the model and its predictions. It is this confidence that will encourage regulators and the public to accept decisions based on the model predictions. This validation approach will be applied to a model, described in this paper, dealing with an underground nuclear test site in rural Nevada.

Environmental Monitoring↗