SEARCH · PubMed Health

Results for “Model interpretability”

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 1,333 records · Page 74Linked to original sources

PubMed2005

Nurses as imperfect role models for health promotion.

The purpose of this qualitative study was to discover ways in which nurses describe themselves as health-promoting role models. Focus groups and individual interviews were conducted with nurses working in a variety of settings. Transcribed interviews were analyzed thematically. Nurses defined themselves as role models of health promotion according to the meaning they gave the term, their perceptions of societal expectations, and their self-constructed personal and professional domains. The term role model evoked diverse interpretations ranging from negative perceptions of the idealized image to a humanized, authentic representation. Nurses perceived that society expected them as role models to be informational resources and to practice what they preached. Nurses defined themselves independently of societal expectations according to personal and professional domains. Valuing health, accepting imperfections, and self-reflecting were aspects of the personal domain, whereas gaining trust, caring, and partnering were facets of the professional domain.

Health Promotion↗

PubMed2003

Tumour grading from magnetic resonance spectroscopy: a comparison of feature extraction with variable selection.

Magnetic resonance spectroscopy (MRS) provides a non-invasive measurement of the biochemistry of living tissue. However, signal variation due to tissue heterogeneity causes considerable mixing between different disease categories, making accurate class assignments difficult. This paper compares a systematic methodology for classifier design using multivariate bayesian variable selection (MBVS), with one based on feature extraction using independent component analysis (ICA). We illustrate the methodology and assess the classification performance using a data set comprising 41 magnetic resonance spectra acquired in vivo from two grades of brain tumour, namely low- and medium-grade astrocytic tumours, labelled astrocytomas (AST), and high-grade gliomas and glioblastomas labelled glioblastomas (GL). The aim of this study is threefold. First, to describe the application of the alternative methodologies to MRS, then to benchmark their classification performance, and finally to interpret the classification models in terms of biologically relevant signals derived from the spectra. The classification performance is assessed using the bootstrap method and by application to a test sample in a retrospective study.

Astrocytoma↗

PubMed2003

Determination of multicanonical weight based on a stochastic model of sampling dynamics.

Based on the stochastic interpretation of the sampling process modeled by a Langevin equation, we present an effective iteration scheme to determine the weight in multicanonical molecular dynamics. Our method enables an automatic determination of the weight producing a uniform energy sampling via an iterative cancellation of the deterministic force in a Langevin equation. The deterministic force has been calculated from the energy trajectory by identifying the moments of the transition probability of a Fokker-Planck equation associated with a Langevin equation. The intimate relationship between the sampling process and the stochastic dynamics has been verified by applying the iteration scheme to a helix-coil transition of the 8-polyalanine system in a gas phase.

Journal Article↗

PubMed2000

Detection of compound spatial patterns: further evidence for different channel interactions.

It is shown that contrast interrelation functions for small compound gratings and for small compound edge-type patterns have different shapes: the former are lozenges, the latter ellipses in the normalised contrast space. These findings can be described by a simple p-norm model, comprising one channel which is most sensitive to the first pattern, another channel which is most sensitive to the second pattern and a non-linear summation of the channel outputs. Since the value of the summation exponent p is constant within but varies between pattern classes, an interpretation of the model parameters is limited to pattern class. Channel models based on the p-norm are discussed, taking into account the interpretation of the first derivative of the contrast interrelation function and the summation exponent, p.

Computer Simulation↗

PubMed2002

Calculation of dielectric spectra of suspensions of rod-shaped cells using boundary element method.

The boundary element method (BEM) has been applied to the calculation of dielectric spectra of suspensions of rod-shaped cells using two kinds of models: model-R consisting of a cylinder and two hemispheres and model-PU of prolate spheroid shape. Both models have an insulating shell phase of a uniform thickness. The calculations were compared with those using a conventional spheroidal model with a confocal shell (model-PC) and previous observations on rod-shaped yeast cells. The differences among the three models were not considerable and all the models succeeded in interpreting the observed data on yeast cells.

Models, Biological↗

PubMed2004

Mechanisms of mammalian otoacoustic emission and their implications for the clinical utility of otoacoustic emissions.

We review recent progress in understanding the physical and physiological mechanisms that generate otoacoustic emissions (OAEs). Until recently, the conceptual model underlying the interpretation of OAEs has been an integrated view that regards all OAEs as manifestations of cochlear nonlinearity. However, OAEs appear to arise by at least two fundamentally different mechanisms within the cochlea: nonlinear distortion and linear reflection. These differences in mechanism have be used to construct a new taxonomy for OAEs that identifies OAEs based on their mechanisms of generation rather than the details of their measurement. The mechanism-based taxonomy provides a useful conceptual framework for understanding and interpreting otoacoustic responses. As commonly measured in the clinic, distortion-product and other evoked OAEs comprise a mixtures of emissions produced by both mechanisms. This mixing precludes any fixed correspondence with the conventional, measurement-based nomenclature. We discuss consequences of the taxonomy for the clinical measurement and interpretation of OAEs.

Acoustic Stimulation↗

PubMed2004

Obsessive-compulsive disorder subtypes: an attempted replication and extension of a symptom-based taxonomy.

A symptom-based subgroup taxonomy for obsessive-compulsive disorder (OCD) was evaluated and refined. The Yale-Brown Obsessive-Compulsive Scale symptom checklist was scored and cluster analysis was conducted with a sample of OCD patients (N = 114). Results were compared to Calamari et al.'s (Behaviour Research and Therapy 37 (1999) 113) five subgroup model. Rules for determining the number of subgroups supported a more complex model. In between sample comparisons, a stable contamination subgroup was found in both a five and seven subgroup taxonomy. Between sample stability was not as strong for Harming, Obsessionals, Symmetry, and Certainty subgroups. Hoarding, as a distinctive subgroup, was unstable in separate samples. When the Calamari et al. sample and the present sample were combined (N = 220), we found a reliable Hoarding subgroup. More interpretable and stable models emerged with the combined samples suggesting that large clinical samples are needed to identify OCD subgroups. Greater support was found for a seven subgroup taxonomy based subgroup interpretability and validation measure differences. The potential utility of symptom-based subgroup models of OCD and alternative approaches are discussed. Identification of reliable and valid OCD subtypes may advance theory and treatment.

Cluster Analysis↗

PubMed1993

Neural network analysis of ventilation-perfusion lung scans.

A neural network model was constructed to interpret ventilation-perfusion (V/Q) lung scans. This model was trained with data from 100 consecutive V/Q scans with pulmonary angiographic correlation. The network was constructed from 28 input parameters that described various standard V/Q findings, which were fed into a single hidden layer that contained 10-20 nodes. The network output indicated the percentage probability of pulmonary embolism for each set of findings on V/Q scans. This network was then used to classify 28 new scans; the resultant classifications were compared with the rankings of an experienced observer who read the scans without knowledge of the correlative angiographic data. The network with 15 hidden nodes outperformed the experienced observer in prediction of the likelihood of pulmonary embolism in the 28-case test set (P = .039). The neural network has several advantages over current algorithms for interpretation of V/Q scans, including the ability to synthesize many variables into a single conclusion and to learn, or modify itself, at exposure to additional data.

Adult↗

PubMed1997

The Methyl-Rotor Electron-Spin Dynamics in the Smoluchowski Drift-Diffusional Framework

The EPR behavior of the internal motion of the methyl fragment in radicals is simulated using the Smoluchowski drift diffusion model. EPR stochastic Liouville lineshape calculations using this model are presented, allowing the exploration of the whole span of conditions between the discrete-site-exchange and the stepless-free-diffusion limits and offering a unification of these two approximate theories. An accurate value of the correlation time for the isotropic hyperfine interaction is calculated for the full ranges of the parameters describing the system and is compared to those given by the approximate models in the two limiting cases, allowing for a discussion of their quality. Long and short correlation time limit spectra are reported and interpreted using analytic models. For the isotropic hyperfine interaction studied here, no broadening effects are observed under Redfield (strong narrowing) conditions. This suggests absence of lifetime broadening effects in agreement with site-model results. The addition of a cos varphi term in the hyperfine interaction is essential for pyramidal radical centres, as is demonstrated by simulation of experimental reports in the literature.

Journal Article↗

PubMed2006

Assessment of basic indirect pharmacodynamic response models with physiological limits.

Many physiological factors are regulated by homeostatic mechanisms to maintain normal body function. Empirical lower Rl (Model I and IV) or upper Rh limits (Model II and III) were included in current basic indirect response (IDR) models to account for the additional role of physiological limits (IDRPL). Various characteristics of these models were evaluated with simulations and explicit equations. The simulations reveal that the expanded models exhibit most properties of basic indirect response models, such as slow response initiation, lag time between the kinetic and dynamic peaks, a large dose plateau, and shift in Tmax with dose. The proposed models always produce lesser net responses than predicted by basic IDR models. Simulations demonstrate that addition of a parameter limit which is close to the baseline has a great influence on the overall and maximum responses and fitted model parameters. Only stimulatory IDRPL Models III and IV allow resolution of all model parameters in the absence of clear indications or predetermined values of the lower or upper limits. However, all four models are able to resolve model parameters when subgroups with different baselines are simultaneously fitted. These models create new interpretations of the determinants of baseline conditions which can be important in assessing inter-subject variability in responses. The applicability of IDRPL models is demonstrated using several examples from the published literature. Indirect response models with physiological limits will be useful in characterizing drug responses for turnover systems which are maintained within a certain range.

5-alpha Reductase Inhibitors↗

PubMed2006

A bilinear stress-strain relationship for arteries.

A comprehensive understanding of the mechanical properties of blood vessels is essential for vascular physiology, pathophysiology and tissue engineering. A well-known approach to study the elasticity of blood vessels is to postulate a strain energy function such as the exponential or polynomial forms. It is typically difficult to fit experimental data to derive material parameters for blood vessels, however, due to the highly nonlinear nature of the stress-strain relation. In this work, we generalize the strain definition to absorb the elastic nonlinearity and then propose a two-dimensional bilinear stress-strain relation between second Piola-Kirchhoff stress and the new strain measure. The model is found to represent the Fung's exponential model very well. The novel linearized constitutive relation simplifies the determination of material constants by reducing the nonlinearity and provides a clearer physical interpretation of the model parameters. The limitations of the constitutive model and its implications for vascular mechanics are discussed.

Animals↗

PubMed2005

Structure-based prediction of bZIP partnering specificity.

Predicting protein interaction specificity from sequence is an important goal in computational biology. We present a model for predicting the interaction preferences of coiled-coil peptides derived from bZIP transcription factors that performs very well when tested against experimental protein microarray data. We used only sequence information to build atomic-resolution structures for 1711 dimeric complexes, and evaluated these with a variety of functions based on physics, learned empirical weights or experimental coupling energies. A purely physical model, similar to those used for protein design studies, gave reasonable performance. The results were improved significantly when helix propensities were used in place of a structurally explicit model to represent the unfolded reference state. Further improvement resulted upon accounting for residue-residue interactions in competing states in a generic way. Purely physical structure-based methods had difficulty capturing core interactions accurately, especially those involving polar residues such as asparagine. When these terms were replaced with weights from a machine-learning approach, the resulting model was able to correctly order the stabilities of over 6000 pairs of complexes with greater than 90% accuracy. The final model is physically interpretable, and suggests specific pairs of residues that are important for bZIP interaction specificity. Our results illustrate the power and potential of structural modeling as a method for predicting protein interactions and highlight obstacles that must be overcome to reach quantitative accuracy using a de novo approach. Our method shows unprecedented performance in predicting protein-protein interaction specificity accurately using structural modeling and suggests that predicting coiled-coil interactions generally may be within reach.

Basic-Leucine Zipper Transcription Factors↗

PubMed2005

Self-organized patchiness in asthma as a prelude to catastrophic shifts.

Asthma is a common disease affecting an increasing number of children throughout the world. In asthma, pulmonary airways narrow in response to contraction of surrounding smooth muscle. The precise nature of functional changes during an acute asthma attack is unclear. The tree structure of the pulmonary airways has been linked to complex behaviour in sudden airway narrowing and avalanche-like reopening. Here we present experimental evidence that bronchoconstriction leads to patchiness in lung ventilation, as well as a computational model that provides interpretation of the experimental data. Using positron emission tomography, we observe that bronchoconstricted asthmatics develop regions of poorly ventilated lung. Using the computational model we show that, even for uniform smooth muscle activation of a symmetric bronchial tree, the presence of minimal heterogeneity breaks the symmetry and leads to large clusters of poorly ventilated lung units. These clusters are generated by interaction of short- and long-range feedback mechanisms, which lead to catastrophic shifts similar to those linked to self-organized patchiness in nature. This work might have implications for the treatment of asthma, and might provide a model for studying diseases of other distributed organs.

Asthma↗

PubMed1989

Applications of therapy in animal models to bacterial infection in human disease.

Animal models have proven to be invaluable in bridging the gap between in vitro susceptibility testing of an antibiotic and anticipating results obtained in clinical studies. Variables such as antibiotic concentration, inoculum of organism, and pharmacokinetic parameters of the drug can be carefully controlled to provide information about the principles of treating infectious diseases as well as an evaluation of specific antimicrobial agents. End points of treatment can be precisely defined (that is, CSF sterility in meningitis, vegetation counts of bacteria in endocarditis) to allow a quantitative evaluation of a new antibiotic. However, it is important to realize that there may be differences in disease pathogenesis and antibiotic pharmacokinetics between experimental infections in animal models and infections in humans. Therefore, results in animal models should be interpreted with caution and compared with results obtained with antimicrobial regimens in clinical studies. Perhaps one of the most useful features of animal models is suggesting which antimicrobials would not be expected to be of therapeutic benefit in man.

Animals↗

PubMedDate not supplied

Electric field plethysmography signals of the human thorax as determined by a 2-D FE-model.

Electric field plethysmography (EFPG) signals of the human thorax consist of a baseline component, a respiratory component and a cardiac component. This paper investigates to what extent a two-dimensional finite-element model of the human thorax can serve as a tool for signal prediction allowing for both optimisation of electrode position and effective signal interpretation. In the model inspiration is taken into account by decreasing lung conductivity. Modelling of cardiac activity involves heart geometry changes and lung conductivity increase. Comparison with experiments shows that the model is more effective for strip than spot electrodes for current injection. Modelling results suggest that the thorax behaves approximately as a homogeneous structure with an inspiratory conductivity increase of about 5%. Two electrode pairs are preferable for monitoring respiration in general. As cardiac activity involves two counter-acting signal sources in the thorax, optimum electrode position is obtained using four separated electrodes. However, recommendable positions for two electrode pairs are on the left sternal and right parasternal line. The investigation shows that in providing a proper modelling of respiration and cardiac activity, a 2-D model provides a practical tool for modelling EFPG signals with the cost of a few structural and systematical restrictions.

Cardiography, Impedance↗

PubMedDate not supplied

[Application of micronucleus test for the demonstration of antimutagenic properties of natural substances].

Authors have used a reliable and reproductible original model of controlled production of free radicals for the cytokinesis-block micronucleus assay. This mutagenicity test is of easy realization and quick interpretation. An experimental model is proposed to test mutagenic activity of free radical generators and antioxydant properties of many substances, such as ascorbic acid. Its addition to the culture medium showed a significant decrease in the rate of micronucleated cells exposed to the free radical generator. The study of other antioxydant compounds (beta-caroten and 5-hydroxyquinolin) have confirmed results previously obtained with ascorbic acid. Thus, the free radical generator coupled to the cytokinesis-block micronucleus assay represents a reliable test to study and evaluate the antioxydant power of natural substances.

Analysis of Variance↗

PubMed2003

Logistic regression of family data from retrospective study designs.

We wish to study the effects of genetic and environmental factors on disease risk, using data from families ascertained because they contain multiple cases of the disease. To do so, we must account for the way participants were ascertained, and for within-family correlations in both disease occurrences and covariates. We model the joint probability distribution of the covariates of ascertained family members, given family disease occurrence and pedigree structure. We describe two such covariate models: the random effects model and the marginal model. Both models assume a logistic form for the distribution of one person's covariates that involves a vector beta of regression parameters. The components of beta in the two models have different interpretations, and they differ in magnitude when the covariates are correlated within families. We describe ascertainment assumptions needed to estimate consistently the parameters beta(RE) in the random effects model and the parameters beta(M) in the marginal model. Under the ascertainment assumptions for the random effects model, we show that conditional logistic regression (CLR) of matched family data gives a consistent estimate beta(RE) for beta(RE) and a consistent estimate for the covariance matrix of beta(RE). Under the ascertainment assumptions for the marginal model, we show that unconditional logistic regression (ULR) gives a consistent estimate for beta(M), and we give a consistent estimator for its covariance matrix. The random effects/CLR approach is simple to use and to interpret, but it can use data only from families containing both affected and unaffected members. The marginal/ULR approach uses data from all individuals, but its variance estimates require special computations. A C program to compute these variance estimates is available at http://www.stanford.edu/dept/HRP/epidemiology. We illustrate these pros and cons by application to data on the effects of parity on ovarian cancer risk in mother/daughter pairs, and use simulations to study the performance of the estimates.

Algorithms↗

PubMed1992

A model for variations in single and repeated egg counts in Schistosoma mansoni infections.

Faecal egg counts are often used to measure Schistosoma mansoni infection, but the considerable variation between successive counts complicates their interpretation. The stochastic model described in this paper gives a description of observed egg counts in a population and can be used as a tool to gain an insight into the underlying worm load distribution. The model distinguishes between two sources of variation in egg counts: (1) variation caused by the difference in worm load between individuals, and (2) the variability of egg counts for an individual with a given worm load. Empirical data, single and repeated measurements, from surveys in five villages in Burundi and Zaire have been used to fit and validate the model. We have discussed possible mechanisms that explain the differences in estimated values between the villages. The model indicates that the expected number of eggs in a stool sample per S. mansoni worm pair is lower than suggested by autopsy data and that, possibly as a consequence of immunity, the inter-individual variation in worm loads decreases with age.

Adolescent↗