SEARCH · PubMed Health

Results for “Models, Statistical”

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 1,351 records · Page 75Linked to original sources

PubMedDate not supplied

Modelling the community as a determinant of health.

Widespread recognition is given to the idea that an individual's health is influenced by the ways an individual works, rests and plays as well as genetic endowments. More recently, some researchers have suggested that the relationships between individual 'characteristics' and health are influenced by the context or community of the individual. Although the notion of the community as a determinant of health is not new, the incorporation of the community in empirical research on the determinants of health has been based on simple statistical models that fail to reflect the complex nature of the individual-community interface. In this paper we use the methods developed in educational research to show how separate statistical models for variations in health between communities and between individuals can be combined to provide a multi-level model for the determinants of health of populations.

Community Networks↗

PubMed2005

Data mining of tree-based models to analyze freeway accident frequency.

INTRODUCTION: Statistical models, such as Poisson or negative binomial regression models, have been employed to analyze vehicle accident frequency for many years. However, these models have their own model assumptions and pre-defined underlying relationship between dependent and independent variables. If these assumptions are violated, the model could lead to erroneous estimation of accident likelihood. Classification and Regression Tree (CART), one of the most widely applied data mining techniques, has been commonly employed in business administration, industry, and engineering. CART does not require any pre-defined underlying relationship between target (dependent) variable and predictors (independent variables) and has been shown to be a powerful tool, particularly for dealing with prediction and classification problems. METHOD: This study collected the 2001-2002 accident data of National Freeway 1 in Taiwan. A CART model and a negative binomial regression model were developed to establish the empirical relationship between traffic accidents and highway geometric variables, traffic characteristics, and environmental factors. RESULTS: The CART findings indicated that the average daily traffic volume and precipitation variables were the key determinants for freeway accident frequencies. By comparing the prediction performance between the CART and the negative binomial regression models, this study demonstrates that CART is a good alternative method for analyzing freeway accident frequencies. IMPACT ON INDUSTRY: By comparing the prediction performance between the CART and the negative binomial regression models, this study demonstrates that CART is a good alternative method for analyzing freeway accident frequencies.

Accidents, Traffic↗

PubMed2006

A Bayesian model for joint segmentation and registration.

A statistical model is presented that combines the registration of an atlas with the segmentation of magnetic resonance images. We use an Expectation Maximization-based algorithm to find a solution within the model, which simultaneously estimates image artifacts, anatomical labelmaps, and a structure-dependent hierarchical mapping from the atlas to the image space. The algorithm produces segmentations for brain tissues as well as their substructures. We demonstrate the approach on a set of 22 magnetic resonance images. On this set of images, the new approach performs significantly better than similar methods which sequentially apply registration and segmentation.

Artifacts↗

PubMed1992

Predictive power of various models for longitudinal attachment level change.

Several statistical models that have been suggested in the periodontal literature for describing longitudinal attachment level changes, such as the gradual loss, single-burst, multiple-burst, and random walk models as well as other models introduced in this paper are compared by their power to predict future attachment loss. The data used in this analysis is from 1061 sites of 8 subjects, with moderate to severe periodontal disease, monitored monthly for about a year. This study found that none of the suggested models could significantly outperform the naïve mean predictor, which predicts the future attachment level from the past mean. It was also found that no single model, such as the burst, gradual, or random walk, together with measurement error can fully explain the variation in the data. These results indicate that in the course of one year, the attachment level change may not follow the same model. Consequently, a model that fits well to past data cannot be accurately extended to the future.

Disease Susceptibility↗

PubMed1994

[Dynamic modelling of the pathology-performance relationship].

Statistical modelling is insufficient to formalize the dynamic process of biological phenomena. This kind of model is also poor at explaining such processes. This theoretical position leads us to search for modelling techniques that allow the outcome to be transformed into (ideal) objects, substituting the intuition that underlies the observed signs with a specific and explicit mathematical construction. In this article, we present one example of such a model in ecopathology. This model is constructed with bibliographic data and permits us to measure the distance between a physical model based on maximization of a biomass (fattening of pigs), and a semantic model based on automatization of a system (porcine respiratory pathology).

Animals↗

PubMed2004

Improved spike-sorting by modeling firing statistics and burst-dependent spike amplitude attenuation: a Markov chain Monte Carlo approach.

Spike-sorting techniques attempt to classify a series of noisy electrical waveforms according to the identity of the neurons that generated them. Existing techniques perform this classification ignoring several properties of actual neurons that can ultimately improve classification performance. In this study, we propose a more realistic spike train generation model. It incorporates both a description of "nontrivial" (i.e., non-Poisson) neuronal discharge statistics and a description of spike waveform dynamics (e.g., the events amplitude decays for short interspike intervals). We show that this spike train generation model is analogous to a one-dimensional Potts spin-glass model. We can therefore tailor to our particular case the computational methods that have been developed in fields where Potts models are extensively used, including statistical physics and image restoration. These methods are based on the construction of a Markov chain in the space of model parameters and spike train configurations, where a configuration is defined by specifying a neuron of origin for each spike. This Markov chain is built such that its unique stationary density is the posterior density of model parameters and configurations given the observed data. A Monte Carlo simulation of the Markov chain is then used to estimate the posterior density. We illustrate the way to build the transition matrix of the Markov chain with a simple, but realistic, model for data generation. We use simulated data to illustrate the performance of the method and to show that this approach can easily cope with neurons firing doublets of spikes and/or generating spikes with highly dynamic waveforms. The method cannot automatically find the "correct" number of neurons in the data. User input is required for this important problem and we illustrate how this can be done. We finally discuss further developments of the method.

Action Potentials↗

PubMed1976

Quantitative studies of single-cell properties in monkey striate cortex. V. Multivariate statistical analyses and models.

1. Several statistical analyses were performed on 205 S-type and CX-type cells which had been completely analyzed on 12 response variables: orientation tuning, end stopping, spontaneous activity, response variability, direction selectivity, contrast selectivity for flashed or moving stimuli, selectivity for interaction of contrast and direction of stimulus movement, spatial-frequency selectivity, spatial separation of subfields responding to light increment of light decrement, sustained/transient response to flash, receptive-field size, and ocular dominance. 2. Correlation of these variables showed that within any cell group, these response variables vary independently. 3. A multivariate discriminant analysis showed that orientation specificity, receptive-field size, interaction of direction and contrast specificity ocular dominance, and spontaneous activity, taken together can adequately assign cells into the S-type or CX-type subgroups. 4. Various models of visual cortex are examined in view of the findings reported here and in the previous papers of this series, which suggest that a) orientation and direction selectivities are produced by separate neural mechanisms, b) there may be hierarchy among simple (S type) cells, and c) complex (CX-type) cells appear to receive a prominent S-type cell input.

Animals↗

PubMed1996

Modeling for intergroup comparisons of imaging data.

Intergroup comparisons pose unique challenges in the analysis of functional imaging data. Imperfections in intersubject stereotaxis can give rise to artifactual results and make it particularly important to allow for intersubject differences in task-related changes when formulating statistical models. Because intergroup comparisons generally involve inferences about the populations from which the subjects were drawn rather than inferences about the particular subjects themselves, subjects must be treated as random rather than fixed effects in the statistical model. These requirements, when combined with the need to adjust for multiple spatial comparisons, result in low statistical power when the number of subjects in each group is small. Functional imaging studies to identify differences between groups generally require many more subjects than other types of functional imaging studies and require careful advance planning to maximize the likelihood of reaching meaningful conclusions.

Brain Mapping↗

PubMed2001

Natural image statistics and neural representation.

It has long been assumed that sensory neurons are adapted, through both evolutionary and developmental processes, to the statistical properties of the signals to which they are exposed. Attneave (1954)Barlow (1961) proposed that information theory could provide a link between environmental statistics and neural responses through the concept of coding efficiency. Recent developments in statistical modeling, along with powerful computational tools, have enabled researchers to study more sophisticated statistical models for visual images, to validate these models empirically against large sets of data, and to begin experimentally testing the efficient coding hypothesis for both individual neurons and populations of neurons.

Animals↗

PubMed2005

Functional impairment as a predictor of spine loading.

STUDY DESIGN: Spine loadings during a variety of lifting exertions were compared with individual torso kinematic abilities. Relationships were evaluated between these measures. OBJECTIVE: To determine if trunk kinematic status (functional impairment) is indicative of spine loading increases in patients with low back pain (LBP) compared to asymptomatic individuals. SUMMARY OF BACKGROUND DATA: Recurrent LBP is a common and costly problem that may be related to increased spine loads in those individuals with LBP. Previous studies suggest that patients with LBP had greater loading than their asymptomatic counterparts when performing work. However, we know little about how to identify when a patient with LBP can resume lifting tasks without having exaggerated spine loading. METHODS: Sixty-two patients with LBP and 61 who were asymptomatic were evaluated for signs of kinematic compromise (i.e., inability to generate normal trunk kinematic patterns) during a prelift test. All subjects were then asked to perform a variety of lifting exertions that varied in lift origin (region), lift asymmetry position, and weight lifted. An electromyography-assisted model was used to evaluate spine loading in each subject during the lifting exertions. Statistical models were used to assess the relationship between kinematic compromise and spine loading. RESULTS: Patients with LBP had greater spine loading as well as greater kinematic compromise. The degree of kinematic compromise was related to the degree of spine loading increases in those individuals with LBP. A statistical model was developed that was able to describe 87% of the variability in compression, 61% in anteroposterior shear, and 65% in lateral shear. CONCLUSIONS: Those patients with greater kinematic compromise used higher levels of antagonistic muscle coactivation that not only reduced trunk motion but also resulted in increases in spine loading. Given the degree of kinematic compromise and the lifting task conditions, a method has been devised to predict the increase in spine loading above and beyond that of an asymptomatic individual when performing typical materials handling tasks.

Biomechanical Phenomena↗

PubMed2005

Models of brain function in neuroimaging.

Inferences about brain function, using neuroimaging data, rest on models of how the data were caused. These models can be quite diverse, ranging from conceptual models of functional anatomy to nonlinear mathematical models of hemodynamics. However, they all have to be internally consistent because they model the same thing. This consistency encompasses many levels of description and places constraints on the statistical models, adopted for data analysis, and the experimental designs they embody. The aim of this review is to introduce the key models used in imaging neuroscience and how they relate to each other. We start with anatomical models of functional brain architectures, which motivate some of the fundaments of neuroimaging. We then turn to basic statistical models (e.g., the general linear model) used for making classical and Bayesian inferences about where neuronal responses are expressed. By incorporating biophysical constraints, these basic models can be finessed and, in a dynamic setting, rendered causal. This allows us to infer how interactions among brain regions are mediated.

Biophysical Phenomena↗

PubMed1992

Factors involved in burn mortality: a multivariate statistical approach based on discriminant analysis.

This article suggests an alternative statistical model for studying the mortality of burned patients: discriminant analysis. This model was applied to our population of 532 patients among whom 71 died. It is not the first time that this model has been applied to assess burn mortality, although it is not frequently used for that application. We found four factors that are statistically significant: age, TBSA, inhalation injury and sex (female). Discriminant analysis allowed us to demonstrate an impressive correlation between death, age and TBSA; inhalation injury by itself and sex, the two other significant factors in our study, seem to have a minor influence on the final outcome of the burned patients and their predictive value is virtually nil. The advantages of this statistical model are compared with logistic regression, the commonly chosen statistical method.

Adult↗

PubMed2004

A digital technique for art authentication.

We describe a computational technique for authenticating works of art, specifically paintings and drawings, from high-resolution digital scans of the original works. This approach builds a statistical model of an artist from the scans of a set of authenticated works against which new works then are compared. The statistical model consists of first- and higher-order wavelet statistics. We show preliminary results from our analysis of 13 drawings that at various times have been attributed to Pieter Bruegel the Elder; these results confirm expert authentications. We also apply these techniques to the problem of determining the number of artists that may have contributed to a painting attributed to Pietro Perugino and again achieve an analysis agreeing with expert opinion.

Journal Article↗

PubMed1994

Effects of clinical ketosis on test day milk yields in Finnish Ayrshire cattle.

A linear model for repeated measurements was used to estimate the effects of clinical ketosis on 722,198 test day milk yields collected from September 1, 1985 to January 31, 1988 on 60,851 Finnish Ayrshire cows of parity < 7. An index was created to differentiate among milk collected within 17 d following diagnosis of ketosis, milk collected before or > 17 d after diagnosis, and milk collected on nonketotic cows. For each parity separately, the statistical model included fixed effects (ketosis, calving season, year and season of milk sampling, and stage of lactation) and random effects (herd and permanent and temporary environments) on test day milk yields. The pattern underlying correlations between temporary environmental effects was accommodated in the statistical model. Compared with those for nonketotic cows, lactation curves of cows with ketosis showed a depression in early lactation; estimated milk loss was 44.3 kg for 17 d after diagnosis. The 305-d milk yield of cows diagnosed with ketosis was estimated to be 141.1 kg higher than that of cows free of ketosis. Although milk losses occurred after ketosis, ketotic cows yielded more milk over the entire lactation than did nonketotic cows; and yields would have been even higher if cows had not had ketosis.

Animals↗

PubMed2001

Carryover bias in visual assessment.

Carryover of stimuli in sequential judgments was studied for a visual assessment task involving estimation of the percentage cover of black circles on a white image. Seven image types with different levels of cover density were arranged in a sequentially balanced design in which each image type was preceded the same number of times by all image types. In the absence of carryover, when images were preceded by images with the same cover density, the response scores were well fitted by a power function of percentage cover with a mean exponent of 0.73 over subjects. Carryover took the form of an assimilation, so that the cover estimate for a target image was generally higher when preceded by an image with higher cover, and lower when preceded by an image with lower cover. However, the magnitude of the carryover effect showed little evidence of increasing with difference in cover between successive images. Nonparametric and parametric methods for testing for carryover are presented. The need for development of psychological models to explain the proposed statistical models is discussed.

Adolescent↗

PubMed2002

Bayesian infinite mixture model based clustering of gene expression profiles.

MOTIVATION: The biologic significance of results obtained through cluster analyses of gene expression data generated in microarray experiments have been demonstrated in many studies. In this article we focus on the development of a clustering procedure based on the concept of Bayesian model-averaging and a precise statistical model of expression data. RESULTS: We developed a clustering procedure based on the Bayesian infinite mixture model and applied it to clustering gene expression profiles. Clusters of genes with similar expression patterns are identified from the posterior distribution of clusterings defined implicitly by the stochastic data-generation model. The posterior distribution of clusterings is estimated by a Gibbs sampler. We summarized the posterior distribution of clusterings by calculating posterior pairwise probabilities of co-expression and used the complete linkage principle to create clusters. This approach has several advantages over usual clustering procedures. The analysis allows for incorporation of a reasonable probabilistic model for generating data. The method does not require specifying the number of clusters and resulting optimal clustering is obtained by averaging over models with all possible numbers of clusters. Expression profiles that are not similar to any other profile are automatically detected, the method incorporates experimental replicates, and it can be extended to accommodate missing data. This approach represents a qualitative shift in the model-based cluster analysis of expression data because it allows for incorporation of uncertainties involved in the model selection in the final assessment of confidence in similarities of expression profiles. We also demonstrated the importance of incorporating the information on experimental variability into the clustering model. AVAILABILITY: The MS Windows(TM) based program implementing the Gibbs sampler and supplemental material is available at http://homepages.uc.edu/~medvedm/BioinformaticsSupplement.htm CONTACT: medvedm@email.uc.edu

Bayes Theorem↗

PubMedDate not supplied

A new coincidence model for single particle counters, Part II: Advances and applications.

Accuracy, acceptance limits and methods for U.S.P. (788) contaminating particle assays published in the XXII Revision are refined in U.S.P. XXIII. In both Revisions, although different numerical values and methods are employed, particle contamination limits remain constants for all S.V.I. container volumes. The effect of this quality standard is high particle concentration acceptance limits in the smallest S.V.I. container sizes. The effect of these high concentrations is to introduce both undercount errors and false counts into U.S.P. (788) SVI contaminating particle assays. There is general agreement that the count of high concentrations of particles by a single particle light extinction counter result in an increase of the average size of the distribution of particles reported and a decrease in their total number. The error mechanism is termed "signal coincidence." Understanding and control of both these problems is unified with the introduction of the count efficiency parameter. Part I of this paper makes available two core concepts with which evaluation and control of coincidence error in single particle counters can be accurately quantified. These two core concepts are the "Particle Triggered Poisson Model," a new more accurate statistical model of the particle counting process and a concentration measure that includes the effect of particle size on the counting capability of a detector. Use of these concepts make it possible to evaluate particle detector count efficiency capability from experimental data of the coincidence effect. This is an application paper. It combines the theory in the Part I paper with the replicability of particle counters into a simple test protocol. The test results can be used to calculate a contour of particle size and count within which both undercount errors and the introduction of false counts into U.S.P. (788) particle assays are controlled. From the data analyzed it can be seen that any single particle size test cannot effectively evaluate detector performance. The use of the theory and methodology described can help realize the intent of the U.S.P. (788) SVI particle contamination assay.

Calibration↗

PubMed1999

Significantly lower entropy estimates for natural DNA sequences.

If DNA were a random string over its alphabet {A, C, G, T}, an optimal code would assign two bits to each nucleotide. DNA may be imagined to be a highly ordered, purposeful molecule, and one might therefore reasonably expect statistical models of its string representation to produce much lower entropy estimates. Surprisingly, this has not been the case for many natural DNA sequences, including portions of the human genome. We introduce a new statistical model (compression algorithm), the strongest reported to date, for naturally occurring DNA sequences. Conventional techniques code a nucleotide using only slightly fewer bits (1.90) than one obtains by relying only on the frequency statistics of individual nucleotides (1.95). Our method in some cases increases this gap by more than fivefold (1.66) and may lead to better performance in microbiological pattern recognition applications. One of our main contributions, and the principle source of these improvements, is the formal inclusion of inexact match information in the model. The existence of matches at various distances forms a panel of experts which are then combined into a single prediction. The structure of this combination is novel and its parameters are learned using Expectation Maximization (EM). Experiments are reported using a wide variety of DNA sequences and compared whenever possible with earlier work. Four reasonable notions for the string distance function used to identify near matches, are implemented and experimentally compared. We also report lower entropy estimates for coding regions extracted from a large collection of nonredundant human genes. The conventional estimate is 1.92 bits. Our model produces only slightly better results (1.91 bits) when considering nucleotides, but achieves 1.84-1.87 bits when the prediction problem is divided into two stages: (i) predict the next amino acid-based on inexact polypeptide matches, and (ii) predict the particular codon. Our results suggest that matches at the amino acid level play some role, but a small one, in determining the statistical structure of nonredundant coding sequences.

Algorithms↗