Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “probabilistic modelling”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,765 records · Page 98Linked to original sources
PubMed2006

Bayesian analysis of polyphonic western tonal music.

This paper deals with the computational analysis of musical audio from recorded audio waveforms. This general problem includes, as subtasks, music transcription, extraction of musical pitch, dynamics, timbre, instrument identity, and source separation. Analysis of real musical signals is a highly ill-posed task which is made complicated by the presence of transient sounds, background interference, or the complex structure of musical pitches in the time-frequency domain. This paper focuses on models and algorithms for computer transcription of multiple musical pitches in audio, elaborated from previous work by two of the authors. The audio data are supposedly presegmented into fixed pitch regimes such as individual chords. The models presented apply to pitched (tonal) music and are formulated via a Gabor representation of nonstationary signals. A Bayesian probabilistic structure is employed for representation of prior information about the parameters of the notes. This paper introduces a numerical Bayesian inference strategy for estimation of the pitches and other parameters of the waveform. The improved algorithm is much quicker and makes the approach feasible in realistic situations. Results are presented for estimation of a known number of notes present in randomly generated note clusters from a real musical instrument database.

Journal Article↗
PubMed2004

Phylogenetic motif detection by expectation-maximization on evolutionary mixtures.

The preferential conservation of transcription factor binding sites implies that non-coding sequence data from related species will prove a powerful asset to motif discovery. We present a unified probabilistic framework for motif discovery that incorporates evolutionary information. We treat aligned DNA sequence as a mixture of evolutionary models, for motif and background, and, following the example of the MEME program, provide an algorithm to estimate the parameters by Expectation-Maximization. We examine a variety of evolutionary models and show that our approach can take advantage of phylogenic information to avoid false positives and discover motifs upstream of groups of characterized target genes. We compare our method to traditional motif finding on only conserved regions. An implementation will be made available at http://rana.lbl.gov.

Algorithms↗
PubMed2006

[Teenagers and condom use: choices by young Brazilians from three Brazilian State capitals in their first and last sexual intercourse].

Condom use has increased among Brazilian youth, although condoms are not used in all sexual relations; in addition, their use varies over the course of an individual's affective and sexual history. This study focused on the prevalence and factors associated with condom use during sexual initiation and in the most recent sexual relations in young men and women (18-24 years of age). The data are from the GRAVAD Research Project, a cross-sectional study with a stratified probabilistic sample using household interviews in three Brazilian State capitals. The analysis used multinomial logistic regression with a hierarchical model. Prevalence of condom use in first sexual intercourse among individuals who used any contraceptive method was 80.7% for females and 88.6% for males. This proportion dropped to 38.8% and 56%, respectively, for the most recent intercourse. In both events, and for both genders, condom use was associated with social belonging and age at initiation. Condom use at sexual initiation was correlated with use in the most recent intercourse (OR = 2.42 for males and 1.89 for females). Increased condom use among youth does not mean continuing use. Women used condoms less than men in the events studied here.

Adolescent↗
PubMed1999

Probabilistic approach to ranking sources of uncertainty in ELF magnetic field exposure limits.

There is a need to improve the biological data, dosimetry, and risk assessment methodology used for setting guidelines for occupational exposures to extreme-low-frequency magnetic fields. This paper illustrates how a probabilistic approach can be used to determine priorities for future research based upon the analysis of biological and dosimetric variables that affect stimulation of the heart by magnetically-induced currents. A model was constructed to predict a level of whole-body magnetic-field exposure below which no cardiac stimulation is expected (Bncs). For each iteration of the model, a value was selected from cardiac stimulation threshold, shape factor, and conductivity distributions by Latin Hypercube sampling, and a value for Bncs was computed. The very wide range of simulated Bncs magnetic-field values obtained indicates that there is considerable uncertainty about what constitutes a "safe" level of exposure. The results show that the major occupational-exposure guidelines are very conservative with respect to risks of cardiac stimulation. The minimum Bncs value computed (0.01 T) by either a circular-loop model or an ellipsoid dosimetric model is ten times the highest recommended workday exposure value in a guideline. The lowest 5% Bncs value calculated for a circular-loop model is about 50-times the ICNIRP occupational exposure limit; the lowest 5% Bncs value calculated for an ellipsoid model is more than 100 times the ACGIH occupational exposure limit. A new finding is that the method specified by several guidelines for determining guideline compliance when exposures occur at multiple frequencies (additive weighting of harmonics) leads to substantially lower Bncs estimates relative to a probably more valid rms-weighting method. Probabilistic sensitivity analyses indicate that lack of knowledge of the threshold for cardiac stimulation is the greatest source of uncertainty as to what is a "safe" level of exposure to extreme-low-frequency magnetic fields. Additional research to address this source of uncertainty may be expected to have the greatest potential impact to reduce overall uncertainty about safe magnetic field exposures.

Animals↗
PubMed1996

[Paradigmatic shifts in clinical practice in the the last generation].

Physicians have always used theoretical models (paradigms) to interpret clinical reality, and have changed the prevailing model only when it could no longer satisfy clinical needs. The purpose of this essay is to review some of the paradigmatic changes in clinical reasoning that have occurred since my undergraduate medial education. My training in the 50's was along the bio-medical model that reduced all diseases to structural or biochemical dysfunctions. Within this framework, causes were perceived as leading inevitably rather than probabilistically to their consequences, and chance and ambiguity had a very small role in explication of pathophysiologic mechanisms and in diagnostic reasoning. The doctor-patient relationship was paternalistic and the orientation to extending survival rejected notions of quality of life and involved parsimonious utilization of health care resources. Today however, clinical reasoning has shifted from deductive and deterministic to inductive (evidence-based) and probabilistic. Disease is believed to result from multiple factors rather than from single causes, and there is increasing acceptance of psycho-social factors of disease. Awareness of the confounding effects of false-positive and false-negative tests has changed the attitude to diagnostic evaluation. Terms, such as risk indicators of disease, predictive value of tests and risk-benefit ratio are increasingly used in discussing clinical decisions. We respect the patient's autonomy more than we did in the past, and consider his/her preferences and quality of life in clinical decision-making. Fair distribution of medical resources is considered as an ethical principle. Finally, clinical guidelines are no longer viewed as counter-intuitive, but rather as effective means to reduce the disturbingly high rates of medical error.

Clinical Medicine↗
PubMed2004

Comparing categorization models.

Four experiments are presented that competitively test rule- and exemplar-based models of human categorization behavior. Participants classified stimuli that varied on a unidimensional axis into 2 categories. The stimuli did not consistently belong to a category; instead, they were probabilistically assigned. By manipulating these assignment probabilities, it was possible to produce stimuli for which exemplar- and rule-based explanations made qualitatively different predictions. F. G. Ashby and J. T. Townsend's (1986) rule-based general recognition theory provided a better account of the data than R. M. Nosofsky's (1986) exemplar-based generalized context model in conditions in which the to-be-classified stimuli were relatively confusable. However, generalized context model provided a better account when the stimuli were relatively few and distinct. These findings are consistent with multiple process accounts of categorization and demonstrate that stimulus confusion is a determining factor as 10 which process mediates categorization.

Decision Making↗
PubMed2006

Hierarchical stochastic image grammars for classification and segmentation.

We develop a new class of hierarchical stochastic image models called spatial random trees (SRTs) which admit polynomial-complexity exact inference algorithms. Our framework of multitree dictionaries is the starting point for this construction. SRTs are stochastic hidden tree models whose leaves are associated with image data. The states at the tree nodes are random variables, and, in addition, the structure of the tree is random and is generated by a probabilistic grammar. We describe an efficient recursive algorithm for obtaining the maximum a posteriori estimate of both the tree structure and the tree states given an image. We also develop an efficient procedure for performing one iteration of the expectation-maximization algorithm and use it to estimate the model parameters from a set of training images. We address other inference problems arising in applications such as maximization of posterior marginals and hypothesis testing. Our models and algorithms are illustrated through several image classification and segmentation experiments, ranging from the segmentation of synthetic images to the classification of natural photographs and the segmentation of scanned documents. In each case, we show that our method substantially improves accuracy over a variety of existing methods.

Algorithms↗
PubMed2005

Extension of Lander-Waterman theory for sequencing filtered DNA libraries.

BACKGROUND: The degree to which conventional DNA sequencing techniques will be successful for highly repetitive genomes is unclear. Investigators are therefore considering various filtering methods to select against high-copy sequence in DNA clone libraries. The standard model for random sequencing, Lander-Waterman theory, does not account for two important issues in such libraries, discontinuities and position-based sampling biases (the so-called "edge effect"). We report an extension of the theory for analyzing such configurations. RESULTS: The edge effect cannot be neglected in most cases. Specifically, rates of coverage and gap reduction are appreciably lower than those for conventional libraries, as predicted by standard theory. Performance decreases as read length increases relative to island size. Although opposite of what happens in a conventional library, this apparent paradox is readily explained in terms of the edge effect. The model agrees well with prototype gene-tagging experiments for Zea mays and Sorghum bicolor. Moreover, the associated density function suggests well-defined probabilistic milestones for the number of reads necessary to capture a given fraction of the gene space. An exception for applying standard theory arises if sequence redundancy is less than about 1-fold. Here, evolution of the random quantities is independent of library gaps and edge effects. This observation effectively validates the practice of using standard theory to estimate the genic enrichment of a library based on light shotgun sequencing. CONCLUSION: Coverage performance using a filtered library is significantly lower than that for an equivalent-sized conventional library, suggesting that directed methods may be more critical for the former. The proposed model should be useful for analyzing future projects.

Chromosome Mapping↗
PubMed1992

Transition play in team performance of volleyball: a log-linear analysis.

The purpose of this study was to develop and test a method to analyze and evaluate sequential skill performances in a team sport. An on-line computerized system was developed to record and summarize the sequential skill performances in volleyball. Seventy-two sample games from the third Federation of International Volleyball Cup men's competition were videotaped and grouped into two categories according to the final team standing and game outcome. Log-linear procedures were used to investigate the nature and degree of the relationship in the first-order (pass-to-set, set-to-spike) and second-order (pass-to-spike) transition plays. Results showed that there was a significant dependency in both the first-order and second-order transition plays, indicating that the outcome of a skill performance is highly influenced by the quality of a preceding skill performance. In addition, the pattern of the transition plays was stable and consistent, regardless of the classification status: Game Outcome, Team Standing, or Transition Process. The methodology and subsequent results provide valuable aids for a thorough understanding of the characteristics of transition plays in volleyball. In addition, the concept of sequential performance analysis may serve as an example for sport scientists in investigating probabilistic patterns of motor performance.

Humans↗
PubMed2002

Growth of a virtual tumour using probabilistic methods of cell generation.

A study into treatment enhancement in combined chemo-radiotherapy for unresectable head and neck cancer has initiated the development of a computer model of tumour growth. The model is based on biological parameters, and characterises tumour growth prior to chemo-radiotherapy. Tumour growth starting from a single stem cell is modelled using the Monte Carlo method. The type of the cell function, their relative proportions on mitosis, their proliferative capacity, the duration of the four phases of the cell cycle, the mean cell cycle time, and the cell loss due to natural causes are the main parameters of the basic model. A Gaussian distribution function operates in establishing the cell cycle time, with a mean value of 33 hours, while the cell type is sampled from a uniform distribution. With the established model, the sensitivity of the developed tumour's cell population to the stem, proliferative and nonproliferative ratio at mitosis was assessed. The present model accurately reflects the exponential distribution of cells along the cell cycle (70% cells in GI phase, 15% in S, 10% in G2, 5% in M) of a developed tumour as described in the literature. The proportion of stem, finitely proliferating and resting cells during tumour growth is maintained within their biological limits (2% stem, 13% finitely proliferating, 85% nonproliferating cells). The ratio (R = 3) between the time necessary to develop a clinically detectable tumour (10(9) cells) and the further time to grow to its lethal size (10(12) cells) is in accordance with the biological data when tumour volume is compared for the two periods (30 doublings and 10 doublings respectively). In conclusion, computer simulation can illustrate the biological growth of a tumour and the cell distribution along the cell cycle. These distributions may then be used in the assessment of tumour response to radiotherapy and to specific chemotherapeutic agents.

Cell Count↗
PubMed2001

Probabilistic prediction of unknown metabolic and signal-transduction networks.

Regulatory networks provide control over complex cell behavior in all kingdoms of life. Here we describe a statistical model, based on representing proteins as collections of domains or motifs, which predicts unknown molecular interactions within these biological networks. Using known protein-protein interactions of Saccharomyces cerevisiae as training data, we were able to predict the links within this network with only 7% false-negative and 10% false-positive error rates. We also use Markov chain Monte Carlo simulation for the prediction of networks with maximum probability under our model. This model can be applied across species, where interaction data from one (or several) species can be used to infer interactions in another. In addition, the model is extensible and can be analogously applied to other molecular data (e.g., DNA sequences).

Amino Acid Motifs↗
PubMed2001

Assessing acute and chronic copper risks to freshwater aquatic life using species sensitivity distributions for different taxonomic groups.

Using copper as an example, we present a method for assessing chemical risks to an aquatic community using species sensitivity distributions (SSDs) for different taxonomic groups. This method fits probability models to chemical exposure and effects data to estimate the percentage of aquatic species potentially at risk and expands on existing probabilistic risk assessment methodologies. Due to a paucity of chronic toxicity data for many chemicals, this methodology typically uses an acute-chronic ratio (ACR) to estimate the chronic effects distribution from the acute effects distribution. We expanded on existing methods in two ways. First, copper SSDs were developed for different organism groups (e.g., insects, fish) that share similar sensitivities or ecological functions. Integration of exposure and effects distributions provides an estimate of which organism groups may be at risk. These results were then compared with a site-specific food web, allowing an estimation of whether key food web components are potentially at risk and whether the overall aquatic community may be at risk from the perspective of ecosystem function. Second, chronic SSDs were estimated using the relationship between copper ACRs and acute toxicity (i.e., the less acutely sensitive a species, the larger the ACR). This correction in the ACR removes concerns previously identified with use of the ACR and allows evaluation of a significantly expanded chronic data set with the same approach as that for assessing acute risks.

Animals↗
PubMed2002

Generalized gap model for bacterial artificial chromosome clone fingerprint mapping and shotgun sequencing.

We develop an extension to the Lander-Waterman theory for characterizing gaps in bacterial artificial chromosome fingerprint mapping and shotgun sequencing projects. It supports a larger set of descriptive statistics and is applicable to a wider range of project parameters. We show that previous assertions regarding inconsistency of the Lander-Waterman theory at higher coverages are incorrect and that another well-known but ostensibly different model is in fact the same. The apparent paradox of infinite island lengths is resolved. Several applications are shown, including evolution of the probability density function, calculation of closure probabilities, and development of a probabilistic method for computing stopping points in bacterial artificial chromosome shotgun sequencing.

Chromosome Mapping↗
PubMed2006

Bayesian inference of hemodynamic changes in functional arterial spin labeling data.

The study of brain function using MRI relies on acquisition techniques that are sensitive to different aspects of the hemodynamic response contiguous to areas of neuronal activity. For this purpose different contrasts such as arterial spin labeling (ASL) and blood oxygenation level dependent (BOLD) functional MRI techniques have been developed to investigate cerebral blood flow (CBF) and blood oxygenation, respectively. Analysis of such data typically proceeds by separate, linear modeling of the appropriate CBF or BOLD time courses. In this work an approach is developed that provides simultaneous inference on hemodynamic changes via a nonlinear physiological model of ASL data acquired at multiple echo times. Importantly, this includes a significant contribution by changes in the static magnetization, M, to the ASL signal. Inference is carried out in a Bayesian framework. This is able to extract, from dual-echo ASL data, probabilistic estimates of percentage changes of CBF, R(2) (*), and the static magnetization, M. This approach provides increased sensitivity in inferring CBF changes and reduced contamination in inferring BOLD changes when compared with general linear model approaches on single-echo ASL data. We also consider how the static magnetization, M, might be related to changes in CBV by assuming the same mechanism for water exchange as in vascular space occupancy.

Bayes Theorem↗
PubMed2006

Context-sensitive autoassociative memories as expert systems in medical diagnosis.

BACKGROUND: The complexity of our contemporary medical practice has impelled the development of different decision-support aids based on artificial intelligence and neural networks. Distributed associative memories are neural network models that fit perfectly well to the vision of cognition emerging from current neurosciences. METHODS: We present the context-dependent autoassociative memory model. The sets of diseases and symptoms are mapped onto a pair of basis of orthogonal vectors. A matrix memory stores the associations between the signs and symptoms, and their corresponding diseases. A minimal numerical example is presented to show how to instruct the memory and how the system works. In order to provide a quick appreciation of the validity of the model and its potential clinical relevance we implemented an application with real data. A memory was trained with published data of neonates with suspected late-onset sepsis in a neonatal intensive care unit (NICU). A set of personal clinical observations was used as a test set to evaluate the capacity of the model to discriminate between septic and non-septic neonates on the basis of clinical and laboratory findings. RESULTS: We show here that matrix memory models with associations modulated by context can perform automatic medical diagnosis. The sequential availability of new information over time makes the system progress in a narrowing process that reduces the range of diagnostic possibilities. At each step the system provides a probabilistic map of the different possible diagnoses to that moment. The system can incorporate the clinical experience, building in that way a representative database of historical data that captures geo-demographical differences between patient populations. The trained model succeeds in diagnosing late-onset sepsis within the test set of infants in the NICU: sensitivity 100%; specificity 80%; percentage of true positives 91%; percentage of true negatives 100%; accuracy (true positives plus true negatives over the totality of patients) 93,3%; and Cohen's kappa index 0,84. CONCLUSION: Context-dependent associative memories can operate as medical expert systems. The model is presented in a simple and tutorial way to encourage straightforward implementations by medical groups. An application with real data, presented as a primary evaluation of the validity and potentiality of the model in medical diagnosis, shows that the model is a highly promising alternative in the development of accuracy diagnostic tools.

Decision Support Systems, Clinical↗
PubMed1991

Influence of gene flow and breeding tactics on gene diversity within populations.

Expressions describing the accumulation of gene correlations within and among lineages and individuals of a population are derived. The model permits different migration rates by males and females and accounts for various breeding tactics within lineages. The resultant equations enable calculation of the probabilistic quantities for the fixation indices, rates of loss of genetic variation, accumulation of inbreeding, and coefficients of relationship for the population at any generation. All fixation indices were found to attain asymptotic values rapidly despite the consistent loss of genetic variation and accumulation of inbreeding within the population. The time required to attain asymptotic values, however, was prolonged when gene flow among lineages was relatively low (less than 20%). The degree of genetic differentiation among breeding groups, inbreeding coefficients, and gene correlations within lineages were found to be primarily functions of breeding tactics within groups rather than gene flow among groups. Thus, the asymptotic value of S. Wright's island model is not appropriate for describing genetic differences among groups within populations. An alternative solution is provided that under limited conditions will reduce to the original island model. The evolution of polygynous breeding tactics appears to be more favorable for promoting intragroup gene correlations than modification of migration rates. Inbreeding and variance effective sizes are derived for populations that are structured by different migration and breeding tactics. Processes that reduce the inbreeding effective population size result in a concomitant increase in variance effective population size.

Animals↗
PubMed2006

A plea for new psychometric models in educational assessment.

OBJECTIVE: To describe the weaknesses of the current psychometric approach to assessment as a scientific model. DISCUSSION: The current psychometric model has played a major role in improving the quality of assessment of medical competence. It is becoming increasingly difficult, however, to apply this model to modern assessment methods. The central assumption in the current model is that medical competence can be subdivided into separate measurable stable and generic traits. This assumption has several far-reaching implications. Perhaps the most important is that it requires a numerical and reductionist approach, and that aspects such as fairness, defensibility and credibility are by necessity mainly translated into reliability and construct validity. These approaches are more and more difficult to align with modern assessment approaches such as mini-CEX, 360-degree feedback and portfolios. This paper describes some of the weaknesses of the psychometric model and aims to open a discussion on a conceptually different statistical approach to quality of assessment. FUTURE DIRECTIONS: We hope that the discussion opened by this paper will lead to the development of a conceptually different statistical approach to quality of assessment. A probabilistic or Bayesian approach would be worth exploring.

Bayes Theorem↗