SEARCH · PubMed Health

Results for “Model performance”

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 361 records · Page 20Linked to original sources

PubMed2004

Basic principles of metabolic modeling of NMR (13)C isotopic turnover to determine rates of brain metabolism in vivo.

Metabolic modeling is a necessary part of the analysis of isotopic labeling data that is being obtained in the brain and other organs. Here are explained the basic principles of metabolic modeling of isotopic labeling studies, particularly with regard to (13)C isotopic measurements performed in vivo. The basic elements needed to simulate isotopic flows are described, and how to combine them to perform modeling analyses is explained. Procedures to introduce and evaluate model constraints and simplifications are discussed. The basic principle of isotopomer analysis is explained, as are mechanics of least-squares fitting of simulations to data. Closely related to the fitting is the effect of data scatter, which is discussed in the context of the non-normal distributions of uncertainty that are often seen with (13)C labeling measurements in vivo. This article is meant to provide a general background for investigators to begin to apply metabolic modeling analysis to (13)C isotopic labeling studies performed in vivo.

Brain↗

PubMed2004

[Comparison of various spatial interpolation methods for non-stationary regional soil mercury content].

Accurate delineating of the spatial distribution of soil heavy metal content is essential for pollution assessment and remediation. The objective of this paper is to evaluate various spatial interpolation methods, including ordinary Kriging (OK), simple Kriging (SK), lognormal Kriging (LNK), universal Kriging (UK), disjunctive Kriging (DK) and inverse distance weighting interpolation (IDW) for estimating soil surface Hg content with lognormal distribution, the linear and second-order polynomial trend, and to determine the optimal interpolation method. The predicted errors, statistical feature values and prediction maps obtained by different interpolation methods were compared. The result indicated that first-order trend OK method performed better than both zero and second-order OK methods. Within the method of first-order trend OK, Gaussian semi-variogram model performed better than both the spherical and exponential models. The method using transformed data performed worse than the methods without data transformation because of the 'distortion' effect arising from log transformation. Those with trend effect were better than those without trend effect. First-order trend UK method is the best method among the six methods studied, while the IDW method is the least.

Environmental Monitoring↗

PubMed2004

Removal of persistent polar pollutants through improved treatment of wastewater effluents (P-THREE).

The EU-project P-THREE started with the establishment of analytical methods for persistent polar pollutants (P3) and quality assurance, followed by screening of P3 in influents and effluents of known wastewater (WW) treatment plants (TP), receiving waters and tap water produced thereof in several European countries. A final selection of analytes for further studies has been performed. Model MBR reactors have been constructed and an optimisation on synthetic wastewater spiked with P3 (linear alkylbenzene sulfonates (LAS), naphthalene sulfonates) has been performed. An initial dynamic modelling of treatment processes has also started. Advanced oxidation process (AOP) treatment has been done in groundwater with isoproturon and nonylphenol ethoxylates (NPEO). The analysis of individual P3 and potential degradation products was also performed. An integrated systeme analysis of the WW treatment processes has also been initiated.

Alkanesulfonic Acids↗

PubMed2004

Statistical evaluation of mathematical models for microbial growth.

The aim of this study was to evaluate the suitability of several mathematical functions for describing microbial growth curves. The nonlinear functions used were: three-phase linear, logistic, Gompertz, Von Bertalanffy, Richards, Morgan, Weibull, France and Baranyi. Two data sets were used, one comprising 21 growth curves of different bacterial and fungal species in which growth was expressed as optical density units, and one comprising 34 curves of colony forming units counted on plates of Yersinia enterocolitica grown under different conditions of pH, temperature and CO(2) (time-constant conditions for each culture). For both sets, curves were selected to provide a wide variety of shapes with different growth rates and lag times. Statistical criteria used to evaluate model performance were analysis of residuals (residual distribution, bias factor and serial correlation) and goodness-of-fit (residual mean square, accuracy factor, extra residual variance F-test, and Akaike's information criterion). The models showing the best overall performance were the Baranyi, three-phase linear, Richards and Weibull models. The goodness-of-fit attained with other models can be considered acceptable, but not as good as that reached with the best four models. Overall, the Baranyi model showed the best behaviour for the growth curves studied according to a variety of criteria. The Richards model was the best-fitting optical density data, whereas the three-phase linear showed some limitations when fitting these curves, despite its consistent performance when fitting plate counts. Our results indicate that the common use of the Gompertz model to describe microbial growth should be reconsidered critically, as the Baranyi, three-phase linear, Richards and Weibull models showed a significantly superior ability to fit experimental data than the extensively used Gompertz.

Bacteria↗

PubMed2000

Laboratory performance in HTLV-I/II analysis.

BACKGROUND: Since 1989, the CDC's Model Performance Evaluation Program has shipped samples to voluntary participant laboratories that test for HTLV antibodies. Each laboratory tests the well-characterized samples, reports the results, and provides information about its testing practices. The data from 15 performance survey periods are reported here. STUDY DESIGN AND METHODS: Multiple logistic regression was used to analyze all data from 15 survey periods from 1989 through 1996. RESULTS: The mean analytic sensitivity for EIA was 99.2 percent per survey period (range, 96-100%), the mean analytic specificity was 97.8 percent (75.6-100%), and the overall accuracy was 88.8 percent (63.8-100%). The mean analytic sensitivity for Western blot was 88.8 percent (75.6-100%); the mean analytic specificity was 95.7 percent (86.7-100%), and the overall accuracy was 91.1 percent (78.1-100%). CONCLUSIONS: Statistical analyses suggested associations between performance and both the retroviral serologic status of the sample and the analytical testing method. Western blot accuracy was associated with weekly testing volume. In early survey periods, performance problems were noted in the analysis of samples from donors with concomitant HTLV and HIV infections and those from donors who were positive for HTLV-II. Technological developments in test methods, such as the addition of recombinant antigens, appeared to have improved the laboratory performance of specific testing methods.

Blotting, Western↗

PubMed2002

Proficiency testing performance: a case study with modeling.

OBJECTIVES: Previous literature has approached proficiency testing (PT) performance by defining the minimum levels, and combinations of imprecision and bias, necessary to meet PT requirements. In this case report, current PT performance was assessed and modeling performed to prioritize our quality improvement efforts. METHODS: A total of 1,006 chemistry challenge results from Ontario's Laboratory Proficiency Testing Program (LPTP, now QMPLS) performed on 69 tests during 1999 and 2000 were used for this retrospective analysis. Peer group means, all method means and results from reference labs were used for comparison. QMPLS flagging and recommended performance criteria were compiled, and modeling performed to predict different levels of performance. RESULTS: Our internal imprecision is <5% for 72% of our 69 tests; however, only 20% of our tests had a CV/PT <25%. Of the 1,006 challenges performed, 136 (13.5%) results were outside PT limits, 55 (5.5%) results were flagged, and 12 requests were received from QMPLS seeking clarification on 24 (2.4%) results. Follow-up identified 9 (38%) nonanalytical errors, 8 (33%) method bias errors, 4 (17%) random errors, 2 poor methods, and one with no error identified. Modeling predicted flagging rates of 2.4% using QMPLS recommended precision performance, 1.6% using our current internal imprecision, 2.2% or 7.0% if we included an overall 20% or 50% relative bias rate with our current imprecision levels, or 15.0% when an estimate of our actual bias for each analyte was considered along with our current imprecision levels. CONCLUSIONS: If imprecision were the only cause of PT errors, our flagging rate for this study period would be 1.6%, and we would need to formally investigate 8 results a year. In practice, strict application of the QMPLS PT criteria would result in 68 investigations annually; however, judicial review of the results before request for clarification significantly reduced this number to 12 investigations (of which 38% were nonanalytical errors). At the present time bias is a significant cause of poor PT performance in a variety of assays. Individual laboratories need to address the problem of bias, and ultimately so do manufacturers. It would be helpful if PT programs also acknowledged this necessary evolution in both their criteria and processes.

Chemistry, Clinical↗

PubMedDate not supplied

Lumped model of terminal aortic impedance in the dog.

The aim of this study was the formulation of a minimal lumped model of the aortic impedance as seen in the abdominal aorta just downstream of the origin of renal arteries. At this location simultaneous measurements of pressure and flow were taken in four anesthetized and open-chest dogs (weight, 30.9 +/- 5.8 kg) under basal, vasodilated (sodium nitroprusside) and vasoconstricted (methoxamine) conditions. Using these measurements we identified and compared three lumped models, A, B, and C, with decreasing complexity from A to C. The frequency response of these models was given the general form of peripheral resistance, Rp, multiplied by the ratio between (a) two zeros and two poles (model A); (b) two zeros and one pole (model B); and (c) one zero and one pole (model C). Rp was calculated as the ratio of mean pressure to mean flow. The other model parameters (time constants, damping factors, and natural frequencies) were estimated by minimizing the sum of squared differences between experimental and model generated pulsatile flows. After parameter estimation, the F-test was applied to compare the goodness of data fit obtained from the three models. Results of this test and the analysis of parameter estimation errors indicated that model B was preferable with respect to models A and C. The analysis of general model performance was followed by a consideration of alternative specific model structures that are physically realizable. With the aid of a determined model structure we evaluated the overall compliance of terminal aortic circulation under a variety of vascular states induced by injection of vasoactive agents.

Animals↗

PubMed1989

Risk evaluation for bovine embryo transfer services using computer simulation and economic decision theory.

A mathematical model was used to evaluate the dynamics and risks in bovine embryo transfer. Variables included embryo collection, fertilization, and transfer rates, plus overall pregnancy rates. Decision analysis was applied to three sets of 500 simulated flushes to test three strategies: 1) nonsuperovulation, 2) superovulation using follicle stimulating hormone (FSH) and 3) superovulation using pregnant mare serum gonadotropin (PMSG). Model validation involved comparing model performance against standards derived from 39 published studies. The model's repeatability was +/- 0.10 against standards in 93% of cases, and never exceeded +/- 0.13 in all 1500 simulations. Model outcomes were accurate to +/- 3% using average results. No pregnancies occurred in 22% of nonsuperovulated donors, 8% of FSH superovulated donors, and 6% of PMSG superovulated donors. PMSG treatment averaged more pregnancies per flush than FSH treatment (4.4 vs 3.9) but showed greater variation in response (64 vs 51%). Decision analysis suggests that a PMSG-induced flush would net $105 more than an FSH-induced flush, and that either superovulation strategy would yield approximately 10 times the net income of a nonsuperovulated flush. Pricing by response (PMSG) or by transfer (FSH) is optimal for the provider of embryo transfer services.

Journal Article↗

PubMed2006

"Blind" testing of models for predicting the 90Sr activity concentration in river systems using post-Chernobyl monitoring data.

Two different models for predicting the time-dependent mobility of (90)Sr in river systems have been evaluated using post-Chernobyl monitoring data for five large Belarusian rivers (Dnieper, Pripyat, Sozh, Besed and Iput) in the period between 1990 and 2004. The results of model predictions are shown to be in good agreement (within a factor of 5) with the measurements of (90)Sr activity concentration in river waters over a long period of time after the accident. This verifies the relatively good accuracy of the generalised input parameters of these models which were derived primarily from measurements of (90)Sr deposited after atmospheric nuclear weapons testing (NWT). For the cases studied here, the simpler AQUASCOPE model performed just as well as the more complex "Global" model which used GIS-based catchment data as an input. The reasons for this are discussed. Exponential decay equations were also curve-fitted to the data for each river to help assess the uncertainties in the predictive models.

Chernobyl Nuclear Accident↗

PubMed2000

A regression-based method for mapping traffic-related air pollution: application and testing in four contrasting urban environments.

Accurate, high-resolution maps of traffic-related air pollution are needed both as a basis for assessing exposures as part of epidemiological studies, and to inform urban air-quality policy and traffic management. This paper assesses the use of a GIS-based, regression mapping technique to model spatial patterns of traffic-related air pollution. The model--developed using data from 80 passive sampler sites in Huddersfield, as part of the SAVIAH (Small Area Variations in Air Quality and Health) project--uses data on traffic flows and land cover in the 300-m buffer zone around each site, and altitude of the site, as predictors of NO2 concentrations. It was tested here by application in four urban areas in the UK: Huddersfield (for the year following that used for initial model development), Sheffield, Northampton, and part of London. In each case, a GIS was built in ArcInfo, integrating relevant data on road traffic, urban land use and topography. Monitoring of NO2 was undertaken using replicate passive samplers (in London, data were obtained from surveys carried out as part of the London network). In Huddersfield, Sheffield and Northampton, the model was first calibrated by comparing modelled results with monitored NO2 concentrations at 10 randomly selected sites; the calibrated model was then validated against data from a further 10-28 sites. In London, where data for only 11 sites were available, validation was not undertaken. Results showed that the model performed well in all cases. After local calibration, the model gave estimates of mean annual NO2 concentrations within a factor of 1.5 of the actual mean (approx. 70-90%) of the time and within a factor of 2 between 70 and 100% of the time. r2 values between modelled and observed concentrations are in the range of 0.58-0.76. These results are comparable to those achieved by more sophisticated dispersion models. The model also has several advantages over dispersion modelling. It is able, for example, to provide high-resolution maps across a whole urban area without the need to interpolate between receptor points. It also offers substantially reduced costs and processing times compared to formal dispersion modelling. It is concluded that the model might thus be used as a means of mapping long-term air pollution concentrations either in support of local authority air-quality management strategies, or in epidemiological studies.

Air Pollutants↗

PubMed2003

Estimation of primary and secondary particulate matter intake fractions for power plants in Georgia.

Air pollution benefit-cost analyses depend on dispersion models to predict population exposures to pollutants, but it is difficult to determine the reasonableness of the model estimates. This is in part because validation with field measurements is not feasible for marginal concentration changes and because few models can capture the necessary spatial and temporal domains with adequate sophistication. In this study, we use the concept of an intake fraction (the fraction of a pollutant or its precursor emitted that is eventually inhaled) to provide insight about population exposures and model performance. We apply CALPUFF, a regional-scale dispersion model common in health benefits assessments, to seven power plants in northern Georgia, considering both direct emissions of fine particulate matter (PM2.5) and secondarily formed ammonium sulfate and ammonium nitrate particles over a domain within 500 km of Atlanta. We estimate emission-weighted average intake fractions of 6 x 10(-7) for primary PM2.5, 2 x 10(-7) for ammonium sulfate from SO2, and 6 x 10(-8) for ammonium nitrate from NOx, with no effect of SO2 on ammonium nitrate. To provide insight about model strengths and limitations, we compare our findings with those from a frequently applied source-receptor (S-R) matrix. Using S-R matrix over an identical domain, the corresponding intake fractions are 5 x 10(-7), 2 x 10(-7), 3 x 10(-8), and -2 x 10(-8), respectively, with the values approximately doubling if the domain is expanded to cover the continental United States. Evaluation of model assumptions and comparison of past intake fraction estimates using these two models illustrates the importance of assumptions about the relative concentrations of ammonia, sulfate, and nitrate, which significantly influences ammonium nitrate intake fractions. These findings provide a framework for improved understanding of the factors that influence population exposures to particulate matter.

Air Pollutants↗

PubMed2005

A density-functional model of the dispersion interaction.

We have recently introduced [J. Chem. Phys. 122, 154104 (2005)] a simple parameter-free model of the dispersion interaction based on the instantaneous in space, dipole moment of the exchange hole. The model generates remarkably accurate interatomic and intermolecular C6 dispersion coefficients, and geometries and binding energies of intermolecular complexes. The model involves, in its original form, occupied Hartree-Fock or Kohn-Sham orbitals. Here we present a density-functional reformulation depending only on total density, the gradient and Laplacian of the density, and the kinetic-energy density. This density-functional model performs as well as the explicitly orbital-dependent model, yet offers obvious computational advantages.

Journal Article↗

PubMed2000

A tiered approach to deterministic models for indoor air exposures.

There are a number of deterministic mathematical approaches available for modeling indoor air pollution concentrations. These models range in complexity from simple saturation vapor pressure models to models using computational fluid dynamics, with many in between these extremes. This range reflects the variety of ways pollutant generation, transport, and mixing are treated in the different models. In selecting which model to use, a tiered approach is useful. The tiered approach considers the goal of the modeling, the availability of model inputs, and the degree of uncertainty that is acceptable. The simpler models are easy to use and the inputs are often readily available. However, they usually have an inherently high degree of uncertainty which requires conservative assumptions that may result in overestimates of concentrations. The more complex models are more difficult to use and require more precise inputs. However, they have the potential to allow the modeler to reduce or quantify uncertainty. This ability to address uncertainty may be limited by the understanding of and the availability of the model inputs. An application example of the tiered approach using a completely mixed space model, a two-zone model, and a turbulent diffusion model illustrates the selection criteria for model use and the model performances. Ultimately, an understanding of the principles behind the models and their inherent strengths and weaknesses is required for their appropriate application to exposure assessment.

Air Movements↗

PubMed1999

Use of a mathematical model of rodent in vitro benzene metabolism to predict human in vitro metabolism data.

Benzene, a ubiquitous environmental pollutant, is known to cause leukemia and aplastic anemia in humans and hematotoxicity and myelotoxicity in rodents. Toxicity is thought to be exerted through oxidative metabolites formed in the liver, primarily via pathways mediated by cytochrome P450 2E1 (CYP2E1). Phenol, hydroquinone and trans-trans-muconaldehyde have all been hypothesized to be involved in benzene-induced toxicity. Recent reports indicate that benzene oxide is produced in vitro and in vivo and may be sufficiently stable to reach the bone marrow. Our goal was to improve existing mathematical models of microsomal benzene metabolism by including time course data for benzene oxide, by obtaining better parameter estimates and by determining if enzymes other than CYP2E1 are involved. Microsomes from male B6C3F1 mice and F344 rats were incubated with [(14)C]benzene (14 microM), [(14)C]phenol (303 microM) and [(14)C]hydroquinone (8 microM). Benzene and phenol were also incubated with mouse microsomes in the presence of trans-dichloroethylene, a CYP2E1 inhibitor, and benzene was incubated with trichloropropene oxide, an epoxide hydrolase inhibitor. These experiments did not indicate significant contributions of enzymes other than CYP2E1. Mathematical model parameters were fitted to rodent data and the model was validated by predicting human data. Model simulations predicted the qualitative behavior of three human time course data sets and explained up to 81% of the total variation in data from incubations of benzene for 16 min with microsomes from nine human individuals. While model predictions did deviate systematically from the data for benzene oxide and trihydroxybenzene, overall model performance in predicting the human data was good. The model should be useful in quantifying human risk due to benzene exposure and explicitly accounts for interindividual variation in CYP2E1 activity.

Animals↗

PubMed1995

A modeling approach for quantifying tumor hypoxia with [F-18]fluoromisonidazole PET time-activity data.

[F-18]fluoromisonidazole (FMISO), a positron-emitting nitroimidazole, binds preferentially to hypoxic cells. It has been used to image hypoxia in human tumors with positron emission tomography (PET). In order to quantify tumor oxygenation status from these PET data, a kinetic model of FMISO cellular bioreduction has been developed to relate cellular oxygen concentration to the cellular FMISO reaction rate constant, kappa A. Also, a compartmental model of FMISO transport and metabolism has been developed to compute the volume average kappa A in tissue regions from [F-18]FMISO PET time-activity data. This compartmental model was characterized using Monte Carlo simulations and [F-18]FMISO PET time-activity data. The model performed well in Monte Carlo simulations; performance was enhanced by fixing three of the seven model parameters at physiologically reasonable values. The four parameters optimized were blood flow rate, kappa A for two partial volume/spillover correction factors. The model was able to accurately determine kappa A for a variety of computer-generated time-activity curv including those for hypothetical heterogeneous tissue regions and poorly perfused tissue regions. The model was also able to fit [H-3]FMISO time-activity data from 36B-10 rat tumors as well as [F-18]FMISO PET time-activity data from a human patient with a base of the tongue squamous cell carcinoma. The kappa A values in muscles ROIs were comparable to those in well-oxygenated cell monolayers while kappa A values in tumor ROIs were greater, suggesting the presence of hypoxic cells in the tumor.

Animals↗

PubMed2000

Automating parallel implementation of neural learning algorithms.

Neural learning algorithms generally involve a number of identical processing units, which are fully or partially connected, and involve an update function, such as a ramp, a sigmoid or a Gaussian function for instance. Some variations also exist, where units can be heterogeneous, or where an alternative update technique is employed, such as a pulse stream generator. Associated with connections are numerical values that must be adjusted using a learning rule, and and dictated by parameters that are learning rule specific, such as momentum, a learning rate, a temperature, amongst others. Usually, neural learning algorithms involve local updates, and a global interaction between units is often discouraged, except in instances where units are fully connected, or involve synchronous updates. In all of these instances, concurrency within a neural algorithm cannot be fully exploited without a suitable implementation strategy. A design scheme is described for translating a neural learning algorithm from inception to implementation on a parallel machine using PVM or MPI libraries, or onto programmable logic such as FPGAs. A designer must first describe the algorithm using a specialised Neural Language, from which a Petri net (PN) model is constructed automatically for verification, and building a performance model. The PN model can be used to study issues such as synchronisation points, resource sharing and concurrency within a learning rule. Specialised constructs are provided to enable a designer to express various aspects of a learning rule, such as the number and connectivity of neural nodes, the interconnection strategies, and information flows required by the learning algorithm. A scheduling and mapping strategy is then used to translate this PN model onto a multiprocessor template. We demonstrate our technique using a Kohonen and backpropagation learning rules, implemented on a loosely coupled workstation cluster, and a dedicated parallel machine, with PVM libraries.

Algorithms↗

PubMed1994

Risk adjustment methods can affect perceptions of outcomes.

When comparing outcomes of medical care, it is essential to adjust for patient risk, including severity of illness. A variety of severity measures exist, but perceptions of outcomes may differ depending on how severity is defined. We used two severity-adjustment approaches to demonstrate that comparisons of outcomes across subgroups of patients can vary dramatically depending on how severity is assessed. We studied two approaches: model 1 was the admission MedisGroups score; model 2 was computed from age and 12 chronic conditions defined by diagnosis codes. Although common summary measures of model performance (R-squared and C) both suggested that model 1 is a better predictor of in-hospital death than model 2, the weaker model consistently produced more accurate expectations by payer class and age group. Using model 1 for severity adjustment suggested that Medicare patients did substantially worse than expected and Medicaid patients substantially better. In contrast, use of model 2 found Medicare patients doing as expected, but Medicaid patients faring poorly.

Adolescent↗

PubMed1998

The integrated exposure uptake biokinetic model for lead in children: independent validation and verification.

The U.S. Environmental Protection Agency employs a model, the integrated exposure biokinetic (IEUBK) model for lead in children, for the assessment of risks to children posed by environmental lead at hazardous waste sites. This paper describes results of an effort to verify the consistency of the documentation with the computer model and to test the computer code using a group that is independent from those involved in the model development. This review concluded that the IEUBK model correctly calculates the equations specified in the IEUBK model theory documentation. However, several issues were identified on model documentation, model performance, and the C++ programming language code (i.e., IEUBK model source code) documentation. These issues affect the ability of an independent reviewer to understand the workings of the IEUBK model but not the model's reliability. As a result of these findings, recommendations have been provided for updating documentation to the model as well as associated adjustments to the model documentation.

Child↗