PubMed Health⌕ Search

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 1,189 records · Page 66Linked to original sources
PubMedDate not supplied

Statistical shape model based segmentation of medical images.

This paper reviews several kinds of 2D shape representations by a set of parameters based on labeled points, Fourier descriptors and wavelet descriptors. Seven shape models for axial slices of spinal vertebra are derived by a statistical analysis of parameters corresponding to a set of example shapes and are subsequently compared. Two of the developed models are incorporated into methods for interactive segmentation of 2D gray level images. The first method is founded on Fourier descriptors, the second one is based on normalized sets of labeled points. Both methods are based on a model guided shape exploration.

Data Interpretation, Statistical↗
PubMed1997

Binning clones by hybridization with complex probes: statistical refinement of an inner product mapping method.

Molecular methods that use long-range information to solve genomics problems (i.e., top-down strategies) efficiently have become increasingly prominent in the genomics literature. One such method, an implementation of inner product mapping (IPM), uses noisy, long-range radiation hybrid (RH)/YAC overlap data and relatively noise-free RH/STS overlap data to localize clones to specific chromosomal regions. Because the molecular data are rarely noise-free, statistical models tailored to the top-down molecular methods make the methods far more effective. We develop two statistical models for IPM (or any other top-down strategy of similar form), a parametric logit model and a nonparametric order-restricted model, and show how these models can be implemented within a hierarchical Bayes framework. Using these models, we refine the chromosome 11 map reported in M. Perlin et al. (1995, Genomics 28: 315-327). Our analyses improve the IPM map, both in terms of successful localization of clones and in terms of the confidence with which they are localized.

Chromosome Mapping↗
PubMed1989

Reliability and validity of screening scales: effect of reducing scale length.

Self-report measures are often useful as the first phase of a multiphase case identification procedure for estimating rates of untreated disorder, but such measures are susceptible to several sources of unreliability. The reliability of respondents' reports can be greatly improved by employing a well-established screening scale composed of multiple replicate items. A practical question that arises is whether interview time can be saved by using only a portion of the original scale. The cost of shortening established scales in terms of the sensitivity and specificity of the screen is modelled statistically, and the robustness of the model is assessed using mental health data obtained on samples of psychiatric patients and community controls. Both the statistical model and the empirical examples suggest that items from highly reliable measures can be dropped without much loss in sensitivity or specificity. Suggestions are made for selecting subsets of items when shortening a screening scale and these are illustrated with screening scales for depression.

Adolescent↗
PubMed2001

A joint physics-based statistical deformable model for multimodal brain image analysis.

A probabilistic deformable model for the representation of multiple brain structures is described. The statistically learned deformable model represents the relative location of different anatomical surfaces in brain magnetic resonance images (MRIs) and accommodates their significant variability across different individuals. The surfaces of each anatomical structure are parameterized by the amplitudes of the vibration modes of a deformable spherical mesh. For a given MRI in the training set, a vector containing the largest vibration modes describing the different deformable surfaces is created. This random vector is statistically constrained by retaining the most significant variation modes of its Karhunen-Loève expansion on the training population. By these means, the conjunction of surfaces are deformed according to the anatomical variability observed in the training set. Two applications of the joint probabilistic deformable model are presented: isolation of the brain from MRI using the probabilistic constraints embedded in the model and deformable model-based registration of three-dimensional multimodal (magnetic resonance/single photon emission computed tomography) brain images without removing nonbrain structures. The multi-object deformable model may be considered as a first step toward the development of a general purpose probabilistic anatomical atlas of the brain.

Anatomy, Cross-Sectional↗
PubMed1998

Analytic decision-making in patients with critical limb ischaemia.

Patients with critical limb ischaemia have a relative short life expectancy. However, if the limb can be preserved or amputation postponed for at least 1-2 years there can be justification for revascularisation if it is possible. Published Scandinavian data indicate that of those operated on about one-third are primarily amputated and two-third revascularised. At least one-third in the revascularisation group are subjected to secondary amputation within six months, and about 50% are alive with the limb intact. Hence, prediction of successful reconstruction should be amended to reduce costs and reduce burden to the patients. In order to achieve statistical models, which preoperatively could predict outcome with an acceptable diagnostic accuracy, national as well as international cooperation with standardisation of predictors and appropriate use of statistical models should be strived for. Prediction, based on a statistical model, is always associated with some uncertainty, and will never replace qualified clinical judgement.

Amputation, Surgical↗
PubMed2001

A statistical riverine litter propagation model.

A statistical riverine litter propagation (RLP) model based on importance sampling Monte Carlo (ISMC) simulation was developed in order to predict the frequency distribution of certain litter types in river reaches. The model was preliminarily calibrated for plastic sheeting by a pilot study conducted on the River Taff, Wales (UK). Litter movement was predominantly controlled by reach characteristics, such as vegetation overhang and water-course obstructions. These affects were modeled in the simulations, by utilizing geometric distributions of river reaches in the time domain. The proposed model satisfactorily simulated the dosing experiments performed at the River Taff. It was concluded from the preliminary calibrations that, the RLP model can be efficiently utilized to portray litter propagation at any arbitrarily selected river site, provided that the stream flows and reach characteristics are calibrated by representative probability distributions of similar sections. Therefore, the RLP model can be considered as a new statistical technique that can predict litter propagation in river sections.

Calibration↗
PubMed2000

Cumulative risk and early cognitive development: a comparison of statistical risk models.

Although it is generally accepted that social risk factors predict delays in early cognitive and language development, there is less agreement about how to represent such associations statistically. Using data collected prospectively on 87 African American children during their first 4 years, this study examined 3 analytic methods for describing a child's level of social risk: (a) individual risk variables, (b) factor scores derived from those risk variables, and (c) a risk index computed by tallying the number of risk conditions present. Comparisons indicated that the individual-risk-variables approach provides better overall prediction of developmental outcomes at a particular age but is less useful in predicting developmental patterns. The risk-factor approach provides good prediction of developmental trajectories when sample sizes are moderate to large. Finally, the risk-index approach is useful for relating social risk to developmental patterns when a large number of risk variables are assessed with a small sample or when other constructs are of primary interest.

Black or African American↗
PubMed2003

Singularities in mixture models and upper bounds of stochastic complexity.

A learning machine which is a mixture of several distributions, for example, a gaussian mixture or a mixture of experts, has a wide range of applications. However, such a machine is a non-identifiable statistical model with a lot of singularities in the parameter space, hence its generalization property is left unknown. Recently an algebraic geometrical method has been developed which enables us to treat such learning machines mathematically. Based on this method, this paper rigorously proves that a mixture learning machine has the smaller Bayesian stochastic complexity than regular statistical models. Since the generalization error of a learning machine is equal to the increase of the stochastic complexity, the result of this paper shows that the mixture model can attain the more precise prediction than regular statistical models if Bayesian estimation is applied in statistical inference.

Bayes Theorem↗
PubMed2001

On the problem of using 'optimal' cutpoints in the assessment of quantitative prognostic factors.

The identification and assessment of prognostic factors is an important task in clinical cancer research. Quantitative prognostic factors are often categorized by using one or several cutpoints to obtain an easier interpretation with respect to prognosis of the resulting patients' subgroups. Considering the selection of a data-driven 'optimal' cutpoint as a 'prototype' of statistical model building, we demonstrate that prognostic relevance of a single factor with no effect can solely be produced by the statistical model building process. Furthermore, we show how to overcome these problems by using corrected P-values and shrinkage methods. The problem and its solutions are illustrated by using the data of two breast cancer studies.

Adult↗
PubMed1998

A statistical mechanical model for beta-hairpin kinetics.

Understanding the mechanism of protein secondary structure formation is an essential part of the protein-folding puzzle. Here, we describe a simple statistical mechanical model for the formation of a beta-hairpin, the minimal structural element of the antiparallel beta-pleated sheet. The model accurately describes the thermodynamic and kinetic behavior of a 16-residue, beta-hairpin-forming peptide, successfully explaining its two-state behavior and apparent negative activation energy for folding. The model classifies structures according to their backbone conformation, defined by 15 pairs of dihedral angles, and is further simplified by considering only the 120 structures with contiguous stretches of native pairs of backbone dihedral angles. This single sequence approximation is tested by comparison with a more complete model that includes the 2(15) possible conformations and 15 x 2(15) possible kinetic transitions. Finally, we use the model to predict the equilibrium unfolding curves and kinetics for several variants of the beta-hairpin peptide.

Animals↗
PubMed1999

Modeling the occurrence of cardiac arrest as a poisson process.

STUDY OBJECTIVE: A statistical model for the occurrence of cardiac arrest has not been described in the literature. Independent events occurring along the time axis may constitute a Poisson process, described by the Poisson and exponential probability distributions. This statistical model defines the probability distribution of events occurring within time intervals and enables construction of confidence intervals for the mean rate. Moreover, the probability that 2 or more events will occur close in time can be estimated from knowledge of the mean rate. We investigated whether the occurrence of cardiac arrests constitutes a Poisson process. METHODS: Time and date for cardiac arrests requiring CPR out of hospital (county population, 155,000) or in hospital (850 beds) during 5 years were analyzed. Goodness of fit was assessed by comparing the observed weekly counts of cardiac arrests and the observed time intervals between such events with the values predicted from the model. RESULTS: The Poisson parameter estimates (mean weekly rates) for out-of-hospital and in-hospital cardiac arrest were 2.02 and 1.09 events per week, respectively. There was close agreement between observed and predicted values, indicating an adequate model fit. CONCLUSION: Occurrence of cardiac arrest along the time axis constitutes a Poisson process and may be adequately modeled by the Poisson and exponential distributions. The model provides information about the nature of these events and allows for probability calculations based on the mean rate of events. Examples of such calculations are given.

Cardiopulmonary Resuscitation↗
PubMed2003

The conceptual basis of mathematics in cardiology IV: statistics and model fitting.

This is the fourth in a series of four articles developed for the readers of Coronary Artery Disease. Without language ideas cannot be articulated. What may not be so immediately obvious is that they cannot be formulated either. One of the essential languages of cardiology is mathematics. Unfortunately, medical education does not emphasize, and in fact, often neglects empowering physicians to think mathematically. Reference to statistics, conditional probability, multicompartmental modeling, algebra, calculus and transforms is common but often without provision of genuine conceptual understanding. At the University of Vermont College of Medicine, Professor Bates developed a course designed to address these deficiencies. The course covered mathematical principles pertinent to clinical cardiovascular and pulmonary medicine and research. It focused on fundamental concepts to facilitate formulation and grasp of ideas. This series of four articles was developed to make the material available for a wider audience. The articles will be published sequentially in Coronary Artery Disease. Beginning with fundamental axioms and basic algebraic manipulations they address algebra, function and graph theory, real and complex numbers, calculus and differential equations, mathematical modeling, linear system theory and integral transforms and statistical theory. The principles and concepts they address provide the foundation needed for in-depth study of any of these topics. Perhaps of even more importance, they should empower cardiologists and cardiovascular researchers to utilize the language of mathematics in assessing the phenomena of immediate pertinence to diagnosis, pathophysiology and therapeutics. The presentations are interposed with queries (by Coronary Artery Disease abbreviated as CAD) simulating the nature of interactions that occurred during the course itself. Each article concludes with one or more examples illustrating application of the concepts covered to cardiovascular medicine and biology.

Cardiology↗
PubMed1999

On the utility of the Dirichlet distribution for meta-analysis of clinical studies.

Recently, there has been an increasing interest in combining efficacy (or safety) results from several clinical trials to draw an overall conclusion about the efficacy (or safety) of new investigational drugs. In a two-armed clinical trial, these efficacy outcomes are often expressed in terms of the two treatment effect means or the proportions of treatment success. For several clinical trials, efficacy assessment could be based on different types of clinical outcomes. These outcomes might be both categorical and noncategorical. Therefore, it is desirable to have a statistical model that is flexible enough for combining clinical trial data of both outcomes across several trials or studies. We discuss a simple statistical model with a potential for such flexibility. We illustrate the use of the model by providing a couple of numerical examples on the meta-analyses of efficacy data from prospective clinical trials/studies.

Bayes Theorem↗
PubMed1995

Case mix adjustment in comparative audit.

There is increasing pressure to measure and publicize the quality of health care, with, for example, hospital 'league tables' appearing regularly in the lay press. Equally, there is increasing recognition that such simplistic presentations of data can be seriously misleading, as differences in case mix mean that they do not compare like with like. We discuss what is required of a statistical model if it si to be used to make comparisons of clinical performance which are adjusted for case mix. Outcome measures must be well defined and relevant, and the risk factors used in the model must reflect the patients' potential prognosis rather than aspects of his or her clinical care. In addition, the calibration of the statistical model is crucial. We review a number of established clinical scoring systems and demonstrate that they fail to meet all of the required criteria. Of particular concern is the fact that these scoring systems are susceptible to biases which could actually mask differences in performance. We highlight areas where further research is required, with the aim of stimulating progress in this field.

Bias↗
PubMed2007

How can we identify low- and high-risk patients among unselected patients with possible acute coronary syndrome?

OBJECTIVE: Prognosis among patients admitted with possible acute coronary syndrome (ACS) may differ from that of patients with definite ACS. The aim of this study was to identify risk factors for mortality among unselected patients and to use the statistical model to identify patients at low or high mortality risk. METHODS: From April 1, 2000, to March 31, 2002, we identified all consecutive patients aged 30 to 69 years admitted to the 2 coronary care units covering the municipality of Aarhus, Denmark (population, 138,290). ACS was considered a possible diagnosis if the physician at admission (1) had noted the presence or absence of chest pain, (2) performed a 12-lead electrocardiogram, and (3) measured markers of myocardial necrosis. In 1576 consecutive patients these criteria were fulfilled. RESULTS: By logistic regression, predictors of mortality were age 60 and older, ST elevation, right bundle-branch block, arrhythmia, elevated markers of myocardial necrosis, and the diagnosis of ACS. The predictive validity of the model, as indicated by receiver operating characteristic curve area, was 85.7%, 87.8%, and 80.1% for 7-, 30-, and 365-day mortality, respectively. CONCLUSIONS: Mortality may be predicted with high precision based on a statistical model. Identification of survivors by the use of a statistical model was superior as compared to simply ruling out the clinical diagnosis of ACS.

Adult↗
PubMed1999

Optimal structure for automatic processing of DNA sequences.

The faithful recovery of the base sequence in automatic DeoxyriboNucleic Acid (DNA) sequencing fundamentally depends on the underlying statistics of the DNA electrophoresis time series. Current DNA sequencing algorithms are heuristic in nature and modest in their use of statistical information. In this paper, a formal statistical model of the DNA time series is presented and then used to construct the optimal maximum-likelihood (ML) processor. The DNA-ML algorithm that is derived in this paper features Kalman prediction of peak locations, peak parameter estimation, whitened waveform comparison and multiple hypothesis processing using the M-algorithm. Properties of the algorithm are examined using both simulated and real data. Model parameters of critical importance and their impact on different types of error mechanisms, such as insertions and deletions, are pointed out. The statistical model of the DNA time-series and the structure of the DNA-ML algorithm provides a basis for future investigation and refinement of DNA sequencing techniques.

Algorithms↗
PubMed2000

Mathematical modeling for the prediction and optimization of laser hair removal.

BACKGROUND AND OBJECTIVE: The study of hair removal is a slow, tedious process. Efficacy evaluations require test-site observation for at least one complete hair cycle, a minimum of 6-8 months. In addition, tracking and counting individual hairs is extremely labor intensive. The objective of this study was to develop and evaluate a mathematical model for hair removal that could significantly speed the entire process. STUDY DESIGN/MATERIALS AND METHODS: Generally accepted kinetic and statistical modeling methods were used to develop a mathematical description of hair growth. The anagen and telogen percentages and decay times were the variables used to predict the kinetics of untreated hair. In the case that the follicles were treated, it was necessary to additionally consider the possible outcomes after treatment, making the calculations much too complicated for simple mathematical formulations. Therefore, a computerized statistical model was developed that considered the probabilities of no, partial, or complete follicular damage in addition to the untreated model variables. These models were then evaluated by comparing them to data derived from the literature and a study center. RESULTS: Values derived from the mathematical model were capable of closely approximating the experimental results of untreated (shaving) and treated (plucking, electrolysis, ruby laser, Q-switched Nd:YAG laser) hair growth kinetics. The model was also shown to be useful for optimizing the number and interval of Q-switched Nd:YAG laser treatments. CONCLUSIONS: A mathematical model can be used to reliably predict results from a variety of hair removal techniques. It also appears to be useful for optimizing a particular treatment protocol. In addition, the development of new hair removal products may be aided by using this method.

Computer Simulation↗