SEARCH · PubMed Health

Results for “parameter estimation”

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 991 records · Page 55Linked to original sources

PubMedDate not supplied

Selective genotyping for QTL detection using sib pair analysis in outbred populations with hierarchical structures.

A simulation study illustrates the effects of the inclusion of half-sib pairs as well as the effects of selective genotyping on the power of detection and the parameter estimates in a sib pair analysis of data from an outbred population. The power of QTL detection obtained from samples of sib pairs selected according to their within family variance or according to the mean within family variance within half sib family was compared and contrasted with the power obtained when only full sib pair analysis was used. There was an increase in power (4-16% ) and decrease in the bias of parameter estimates with the use of half-sib information. These improvements in power and parameter estimates depended on the number of the half sib pairs (half sib family size). Almost the same power as that obtained using all the available sib pairs could be achieved by selecting only 50-60% the animals. The most effective method was to select both full and half sib pairs on the basis of high within full sib family variance for the trait in question. The QTL position estimates were in general slightly biased towards the center of the chromosome and the QTL variance estimates were biased upwards, there being quite large differences in bias depending on the selection method.

Journal Article↗

PubMed2006

The reliability of the ankle-brachial index in the Atherosclerosis Risk in Communities (ARIC) study and the NHLBI Family Heart Study (FHS).

BACKGROUND: A low ankle-brachial index (ABI) is associated with increased risk of coronary heart disease, stroke, and death. Regression model parameter estimates may be biased due to measurement error when the ABI is included as a predictor in regression models, but may be corrected if the reliability coefficient, R, is known. The R for the ABI computed from DINAMAP readings of the ankle and brachial SBP is not known. METHODS: A total of 119 participants in both the Atherosclerosis Risk in Communities (ARIC) study and the NHLBI Family Heart Study (FHS) had repeat ABIs taken within 1 year, using a common protocol, automated oscillometric blood pressure measurement devices, and technician pool. RESULTS: The estimated reliability coefficient for the ankle systolic blood pressure (SBP) was 0.68 (95% CI: 0.57, 0.77) and for the brachial SBP was 0.74 (95% CI: 0.62, 0.83). The reliability for the ABI based on single ankle and arm SBPs was 0.61 (95% CI: 0.50, 0.70) and the reliability of the ABI computed as the ratio of the average of two ankle SBPs to two arm SBPs was estimated from simulated data as 0.70. CONCLUSION: These reliability estimates may be used to obtain unbiased parameter estimates if the ABI is included in regression models. Our results suggest the need for repeated measures of the ABI in clinical practice, preferably within visits and also over time, before diagnosing peripheral artery disease and before making therapeutic decisions.

Ankle↗

PubMed1994

[Parameters for estimating abnormal reticulocyte specimens from blood cell counts].

In order to estimate reticulocyte levels, low, normal or high, parameters for discrimination are calculated using hematology count data. Multivariate discriminant functions were used for detecting abnormal reticulocyte specimens for computerized laboratories. The method detects 60.2%, 60.3% and 15.7% of low, high and normal reticulocyte specimens respectively. Discrimination method with individual test data are used for uncomputerized laboratories. The tests used for detecting abnormal reticulocyte specimens are white blood cells, red blood cells, mean cell hemoglobin concentration and red blood cell distribution width. Detection efficiencies for low, high and normal reticulocyte specimens are 55.9%, 50.4% and 14.7%.

Blood Cell Count↗

PubMed2005

Reduction of selection bias in genomewide studies by resampling.

The accuracy of gene localization, the reliability of locus-specific effect estimates, and the ability to replicate initial claims of linkage and/or association have emerged as major methodological concerns in genomewide studies of complex diseases and quantitative traits. To address the issue of multiple comparisons inherent in genomewide studies, the use of stringent criteria for assessing statistical significance has been generally acknowledged as a strategy to control type I error. However, the application of genomewide significance criteria does not take account of the selection bias introduced into parameter estimates, e.g., estimates of locus-specific effect size of disease/trait loci. Some have argued that reliable locus-specific parameter estimates can only be obtained in an independent sample. In this report, we examine statistical resampling techniques, including cross-validation and the bootstrap, applied to the initial sample to improve the estimation of locus-specific effects. We compare them with the naive method in which all data are used for both hypothesis testing and parameter estimation, as well as with the split-sample approach in which part of the data are reserved for estimation. Upward bias of the naive estimator and inadequacy of the split-sample approach are derived analytically under a simple quantitative trait model. Simulation studies of the resampling methods are performed for both the simple model and a more realistic genomewide linkage analysis. Our results suggest that cross-validation and bootstrap methods can substantially reduce the estimation bias, especially when the effect size is small or there is no genetic effect.

Algorithms↗

PubMed2006

Improved hypothesis testing for coefficients in generalized estimating equations with small samples of clusters.

The sandwich standard error estimator is commonly used for making inferences about parameter estimates found as solutions to generalized estimating equations (GEE) for clustered data. The sandwich tends to underestimate the variability in the parameter estimates when the number of clusters is small, and reference distributions commonly used for hypothesis testing poorly approximate the distribution of Wald test statistics. Consequently, tests have greater than nominal type I error rates. We propose tests that use bias-reduced linearization, BRL, to adjust the sandwich estimator and Satterthwaite or saddlepoint approximations for the reference distribution of resulting Wald t-tests. We conducted a large simulation study of tests using a variety of estimators (traditional sandwich, BRL, Mancl and DeRouen's BC estimator, and a modification of an estimator proposed by Kott) and approximations to reference distributions under diverse settings that varied the distribution of the explanatory variables, the values of coefficients, and the degree of intra-cluster correlation (ICC). Our new method generally worked well, providing accurate estimates of the variability of fitted coefficients and tests with near-nominal type I error rates when the ICC is small. Our method works less well when the ICC is large, but it continues to out-perform the traditional sandwich and other alternatives.

Cluster Analysis↗

PubMed2006

Impact of correlation between pesticide parameters on estimates of environmental exposure.

Monte Carlo techniques are increasingly used in pesticide exposure modelling to evaluate the uncertainty in predictions arising from uncertainty in input parameters and to estimate the confidence that should be assigned to modelling results. The approach typically involves running a deterministic model repeatedly for a large number of input values sampled from statistical distributions. A key decision in setting up a probabilistic analysis is whether there is correlation between any of the inputs to the analysis. Pesticide properties are often the most sensitive in exposure assessment. Analysis of the literature demonstrated that there are examples of both positive and negative correlation between the sorption and degradation behaviour of a pesticide, but that general trends are not apparent at present. The inclusion of even weak correlation between sorption and degradation was found to greatly influence a probabilistic analysis of leaching through soil. Correlation will reduce the predicted extent of leaching for pesticides, and it is recommended to set the correlation to zero unless the experimental data support an alternative assumption (i.e. where the correlation is statistically significant (P <or= 0.05) and experimental artefacts can be excluded).

Environmental Exposure↗

PubMed2005

Sensor fault diagnosis for nonlinear processes with parametric uncertainties.

This paper addresses the problem of detecting, discriminating, and reconstructing sensor faults for nonlinear systems with known model structure but uncertainty in the parameters of the process. The convenience of the proposed technique lies in the fact that historical operational data and/or a priori fault information is not required to achieve accurate fault reconstruction except for fixed, short intervals. The overall fault diagnosis algorithm is composed of a series of nonlinear estimators, which estimates parameter and a fault isolation and identification filter. Parameter estimation and fault reconstruction cannot be performed accurately since faults and parametric uncertainty interact with each other. Therefore, these two tasks are performed at different time scales, where the fault diagnosis takes place at a more frequent rate than the parameter estimation. It is shown that the fault can be reconstructed under some realistic assumptions and the performance of the proposed methodology is evaluated on a simulated chemical process exhibiting nonlinear dynamic behavior.

Algorithms↗

PubMed2004

A/C(i) curve analysis across a range of woody plant species: influence of regression analysis parameters and mesophyll conductance.

The analysis and interpretation of A/C(i) curves (net CO(2) assimilation rate, A, versus calculated substomatal CO(2) concentration, C(i)) is dependent upon a number of underlying assumptions. The influence of the C(i) value at which the A/C(i) curve switches between the Rubisco- and electron transport-limited portions of the curve was examined on A/C(i) curve parameter estimates, as well as the effect of mesophyll CO(2) conductance (g(m)) values on estimates of the maximum rate of Rubisco-mediated carboxylation (V(cmax)). Based on an analysis using 19 woody species from the Pacific Northwest, significant variation occurred in the C(i) value where the Rubisco- and electron transport-limited portions of the curve intersect (C(i_t)), ranging from 20 Pa to 152 Pa and averaging c. 71 Pa and 37 Pa for conifer and broadleaf species, respectively. Significant effects on estimated A/C(i) parameters (e.g. V(cmax)) may arise when preliminary estimates of C(i_t), necessary for the multiple regression analyses, are set either too high or too low. However, when the appropriate threshold is used, a significant relationship between A/C(i) and chlorophyll fluorescence estimates of carboxylation is achieved. The use of the V(cmax) parameter to describe accurately the Rubisco activity from the A/C(i) curve analysis is also dependent upon the assumption that C(i) is approximately equal to chloroplast CO(2) concentrations (C(c)). If leaf mesophyll conductance is low, C(c) will be much lower than C(i) and will result in an underestimation of V(cmax) from A/C(i) curves. A large range of mesophyll conductance (g(m)) values was observed across the 19 species (0.005+/-0.002 to 0.189+/-0.011 mol m(-2) s(-1) for Tsuga heterophylla and Quercus garryana, respectively) and, on average, g(m) was 1.9 times lower for the conifer species (0.058+/-0.017 mol m(-2) s(-1) for conifers versus 0.112+/-0.020 mol m(-2) s(-1) for broadleaves). When this mesophyll limitation was accounted for in V(cmax) estimates, considerable variation still existed between species, but the difference in V(cmax) between conifer and broadleaf species was reduced from c. 11 micromol m(-2) s(-1) to 4 micromol m(-2) s(-1). For example, A/C(i) curve estimates of V(cmax) were 31.2+/-6.2 and 42.2+/-4.4 micromol m(-2) s(-1), and A/C(c) curve estimates were 41.2+/-7.1 micromol m(-2) s(-1) and 45.0+/-4.8 micromol m(-2) s(-1), for the conifer and broadleaf species, respectively.

Biological Transport↗

PubMed2005

Macromolecular transport in the arterial wall: alternative models for estimating barriers.

Early atherosclerosis, or atherogenesis, is characterized by the abnormal accumulation of plasma-borne macromolecules (e.g., LDL) in the arterial intima. The change of barrier characteristics of tissue in the arterial wall requires evaluation of macromolecular transport across the endothelial cell layer (ECL) and internal elastic lamina (IEL), the luminal and abluminal boundaries of the arterial intima, respectively. In this study, alternative mathematical models are derived from dynamic mass balances to describe macromolecular transport across the arterial wall. One model considers each medial layer as a spatially lumped compartment, whereas another model consists of a spatially lumped intima and spatially distributed media. Model simulations of a tracer concentration distribution in the arterial wall are compared with concentration distributions of horseradish peroxidase (HRP) after i.v. injection in mice. For each model, optimal parameter values are obtained that yield model outputs matching the data well for two different HRP circulation times. The model parameter estimates show that the ECL is the major barrier for macromolecular transport across the normal arterial wall. Sensitivity analysis indicates that the parameter estimates of the transport coefficients of the ECL and IEL are well determined. Optimal circulation times are determined and expected to yield improved precision of parameter estimates in future experiments to reflect disease progression.

Animals↗

PubMed1990

Use of telephone interviewing in health care research.

Increasingly, telephone interviewing has become the data collection procedure of choice in large-scale health services research surveys. Previous analyses indicate that excluding nontelephone households does not seriously affect most national parameter estimates, such as yearly estimates of number of ambulatory visits, mainly because the proportion of households without telephones is small. Moreover, if the exclusion of nontelephone households simply underestimates the proportions in the population with certain characteristics, such as age and ethnicity, and the "true" proportions are known, it is possible to appropriately weight the study group in order to mitigate the telephone-exclusion bias. However, regression analyses undertaken on three years of national Health Interview Surveys indicate, at least on some key measures such as having health insurance, that persons living in households with and without telephones represent different populations, and parameter estimates are distorted by excluding nontelephone households. Under these circumstances, it is not possible to adjust parameter estimates to take into account nontelephone households.

Ambulatory Care↗

PubMed1988

D-optimal design applied to binding saturation curves of an enkephalin analog in rat brain.

The D-optimal design, a minimal sample design that minimizes the volume of the joint confidence region for the parameters, was used to evaluate binding parameters in a saturation curve with a view to reducing the number of experimental points without loosing accuracy in binding parameter estimates. Binding saturation experiments were performed in rat brain crude membrane preparations with the opioid mu-selective ligand [3H]-[D-Ala2,MePhe4,Gly-ol5]enkephalin (DAGO), using a sequential procedure. The first experiment consisted of a wide-range saturation curve, which confirmed that [3H]-DAGO binds only one class of specific sites and non-specific sites, and gave information on the experimental range and a first estimate of binding affinity (Ka), capacity (Bmax) and non-specific constant (k). On this basis the D-optimal design was computed and sequential experiments were performed each covering a wide-range traditional saturation curve, the D-optimal design and a splitting of the D-optimal design with the addition of 2 points (+/- 15% of the central point). No appreciable differences were obtained with these designs in parameter estimates and their accuracy. Thus sequential experiments based on D-optimal design seem a valid method for accurate determination of binding parameters, using far fewer points with no loss in parameter estimation accuracy.

Animals↗

PubMed1988

Estimating the parameters of a Polya-Aeppli distribution.

"The paper provides expressions for the 'maximum likelihood equations' and the asymptotic variances and co-variance of the maximum likelihood estimators of the parameters in a Polya-Aeppli distribution provided that we have a knowledge of the number of groups with zero objects...the number of groups with one object...and the total number of groups.... The results are illustrated using an example involving demographic data [for India]." The focus is on the number and types of households.

Asia↗

PubMed1994

Bayesian forecasting of serum vancomycin concentrations with non-steady-state sampling strategies.

The application of three non-steady-state sampling strategies and the fitting of either three or five pharmacokinetic parameter estimates by a two-compartment Bayesian forecasting program was evaluated retrospectively in 27 adult patients with stable renal function. Sampling strategies included a single midpoint concentration, a set of peak and trough concentrations, and three serial vancomycin concentrations. The most precise and least-bias predictions of steady-state peak vancomycin concentrations were observed by using population-based parameter estimates [mean prediction error (ME) = -0.40 and mean absolute error = 5.77]. The addition of non-steady-state feedback concentration(s) did not provide additional information for predictions of future steady-state peak concentrations. The least-bias prediction of steady-state trough vancomycin concentrations was seen when a single midpoint non-steady-state concentration was used (ME = 0.92 and -0.17 for five and three fitted parameter estimates, respectively). The MEs of serial and peak and trough feedback strategies were similar in magnitude to those obtained using population parameters, but in opposite directions (underprediction vs. overprediction, respectively). The fitting of only three parameters produced results similar to those using five parameters. The results from this study confirm our previous evaluation that non-steady-state concentrations provide very minimal information to Bayesian forecasting of future steady-state concentrations.

Adult↗

PubMed1989

Empirical comparisons of proportional hazards and logistic regression models.

We compare parameter estimates from the proportional hazards model, the cumulative logistic model and a new modified logistic model (referred to as the person-time logistic model), with the use of simulated data sets and with the following quantities varied: disease incidence, risk factor strength, length of follow-up, the proportion censored, non-proportional hazards, and sample size. Parameter estimates from the person-time logistic regression model closely approximated those from the Cox model when the survival time distribution was close to exponential, but could differ substantially in other situations. We found parameter estimates from the cumulative logistic model similar to those from the Cox and person-time logistic models when the disease was rare, the risk factor moderate, and censoring rates similar across the covariates. We also compare the models with analysis of a real data set that involves the relationship of age, race, sex, blood pressure, and smoking to subsequent mortality. In this example, the length of follow-up among survivors varied from 5 to 14 years and the Cox and person-time logistic approaches gave nearly identical results. The cumulative logistic results had somewhat larger p-values but were substantively similar for all but one coefficient (the age-race interaction). The latter difference reflects differential censoring rates by age, race and sex.

Adolescent↗

PubMed1987

The clinical pharmacology of vancomycin in seriously ill preterm infants.

The first dose and steady state pharmacokinetics of vancomycin were studied in 16 seriously ill preterm infants (less than or equal to 34 wk gestational age) with documented Staphylococcus epidermidis infections. One infant was dropped from the study due to peripheral flushing occurring during administration of the first dose. Individual vancomycin doses ranged from 9.8 to 17.8 mg/kg and were infused intravenously over 15-37 min. Fifteen infants were studied after the first dose of vancomycin, whereas only 12 of these 15 were able to be studied under steady state conditions. Vancomycin half-life, steady-state volume of distribution, and body clearance averaged 6.0 h, 0.53 liter/kg, and 1.22 ml/min after the first dose and only slight differences were observed in these parameter estimates under steady state conditions. However, substantial accumulation of vancomycin in serum was observed with multiple dosing. Complete 8-h urine collections were possible in 12 of 15 premature infants after the first dose of vancomycin. Overall, 44.6% of the dose was recovered in the urine with a corresponding vancomycin renal ClR averaging 0.88 ml/min. Vancomycin body Cl correlated directly with renal ClR (r = 0.88, p less than 0.001) and body weight (r = 0.8, p less than 0.001). Vancomycin pharmacokinetic parameter estimates Vdss and Cl correlated directly with body weight, surface area, and postconceptional age. No significant relationships were observed between these parameter estimates and gestational age or postnatal age. Fourteen of 15 infants were treated successfully for their underlying infectious process. These data support the use of lower doses of vancomycin than previously recommended for the treatment of preterm infants.

Female↗

PubMed2006

A comparison of methods to estimate cross-environment genetic correlations.

Advanced techniques for quantitative genetic parameter estimation may not always be necessary to answer broad genetic questions. However, simpler methods are often biased, and the extent of this determines their usefulness. In this study we compare family mean correlations to least squares and restricted error maximum likelihood (REML) variance component approaches to estimating cross-environment genetic correlations. We analysed empirical data from studies where both types of estimates were made, and from studies in our own laboratories. We found that the agreement between estimates was better when full-sib rather than half-sib estimates of cross-environment genetic correlations were used and when mean family size increased. We also note biases in REML estimation that may be especially important when testing to see if correlations differ from 0 or 1. We conclude that correlations calculated from family means can be used to test for the presence of genetic correlations across environments, which is sufficient for some research questions. Variance component approaches should be used when parameter estimation is the objective, or if the goal is anything other than determining broad patterns.

Analysis of Variance↗

PubMed1983

Model testing in radioligand/receptor interaction by Monte Carlo simulation.

Monte Carlo simulations have been applied for evaluating the reliability of parameter estimates as well as for testing models in radioligand saturation binding experiments. Scatchard analysis was compared to the nonlinear least-square curve fitting method for one-site saturation binding curves. It was found that linear regression analysis from the transformed data in the Scatchard plot yielded generally less accurate parameter estimates than nonlinear regression analysis of untransformed data. The advantage of the nonlinear least-squares curve fitting method was especially pronounced in cases where the scatter and number of data points, as well as the radioligand concentration range, were chosen similar to less optimal experimental conditions. Under such circumstances, several KD and Bmax values derived by Scatchard analysis led to physically impossible negative values whereas the same data analyzed by nonlinear regression yielded reasonable parameter estimates. Furthermore, it was found that for both means of analysis, KD and Bmax correlated positively. In another set of Monte Carlo experiments, saturation binding curves involving two receptor sites were generated and subsequently analyzed according to both a one-site and a two-site model. The confidence with which one is able to distinguish the two-site model from nonlinear least-squares curve fitting was then estimated for optimal, as well as for, less ideal experimental conditions.

Models, Biological↗

PubMed2003

On bias in the estimation of autocorrelations for fMRI voxel time-series analysis.

For fMRI time-series analysis to be statistically valid, it is important to deal correctly with temporal autocorrelation in the noise. Most of the approaches in the literature adopt a two-stage approach in which the autocorrelation structure is estimated using the residuals of an initial model fit. This estimate is then used to "prewhiten" the data and the model before the model is refit to obtain final activation parameter estimates. An assumption implicit in this scheme is that the residuals from the initial model fit represent a realization of the "true" noise process. In general this assumption will not be correct as certain components of the noise will be removed by the model fit. In this paper we examine (i) the form of the bias induced by the initial model fit, (ii) methods of correcting for the bias, and (iii) the impact of bias correction on the model parameter estimates. We find that while bias correction does result in more accurate estimates of the correlation structure, this does not translate into improved estimates of the model parameters. In fact estimates of the model parameters and their standard errors are seen to be so accurate that we conclude that bias correction is unnecessary.

Artifacts↗