Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Bayesian networks”

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 433 records · Page 24Linked to original sources
PubMed1995

Effective access to distributed heterogeneous medical text databases.

INQUERY is an advanced text information retrieval system developed by the Information Retrieval Laboratory of the University of Massachusetts in Amherst. It is based on Bayesian inference networks, which are probabilistic models for reasoning with multiple sources of uncertain evidence. The evidence, in this case, is the presence or absence of words and/or phrases in a document. Evidence is combined into belief that a document is relevant. The INQUERY retrieval engine has been developed with the support of ARPA, NSF, and industrial funding. It has a number of unique features and has achieved excellent results in the TIPSTER and TREC evaluations. Informatics research and application development using INQUERY has recently begun in the medical domain, including a new ARPA initiative concerned with clinical text. The features that we will focus on in this demonstration are: Automatic processing of natural language queries, including the extraction of phrases and specific medical concepts such as drug doses; Document selection through automatic relevance feedback and routing techniques, including the construction of complex queries using the INQUERY query language; The integration of conventional database techniques with text analysis and retrieval; Automatic thesaurus generation and query expansion using the PhraseFinder system; Distributed database access, including automatic database selection and merging of local searches; this will be demonstrated using a collection of medical databases; Retrieval based on passages, rather than whole documents.

Bayes Theorem↗
PubMed1996

An interactive decision support system for breast fine needle aspiration cytology.

OBJECTIVE: To develop a computerized system to assist in the diagnosis of malignancy in breast fine needle aspiration cytology. STUDY DESIGN: A Bayesian belief network was designed to control uncertainty and allow a diagnostic decision to be reached based on the sequential collection of cytologic information. Ten cytologic features were defined as clues that contribute to the diagnostic discrimination of benign and malignant aspirates. The impact of each feature on the diagnostic decision was quantified by a conditional probability matrix. RESULTS: For the assessment of a new case, the computer guides the user through the diagnosis, prompting him or her for information on each of the diagnostic features in turn. For each feature, the user is presented with a series of digitally stored color microscopic images that have been selected to represent good examples of the different feature grades-e.g., pleomorphism: none, mild, moderate and severe. Each image is mapped to an overlapping curve, and by positioning a line on the spectrum where the user feels the case lies, a membership function vector is calculated and entered as evidence into the network. This results in an update in the belief in the diagnostic alternatives. After all the clues have been assessed, a final diagnostic probability is reported. In addition, a cumulative belief curve can be drawn that maps the change in the diagnostic probabilities after each piece of evidence has been submitted, providing unique insight into the diagnostic process. CONCLUSION: Systems like this represent an important step forward in the use of descriptive classifiers. They impose consistency in terminology, improve reproducibility in the grading of cellular abnormalities and remove subjectivity in interpreting the significance of pvisual clues to diagnosis. As such, they represent a necessary tool in pathologic decision making.

Biopsy, Needle↗
PubMed1996

Human causal discovery from observational data.

Utilizing Bayesian belief networks as a model of causality, we examined medical students' ability to discover causal relationships from observational data. Nine sets of patient cases were generated from relatively simple causal belief networks by stochastic simulation. Twenty participants examined the data sets and attempted to discover the underlying causal relationships. Performance was poor in general, except at discovering the absence of a causal relationship. This work supports the potential for combining human and computer methods for causal discovery.

Bayes Theorem↗
PubMed1996

Evaluation of prostatic intraepithelial neoplasia after treatment with a 5-alpha-reductase inhibitor (finasteride). A methodologic approach.

OBJECTIVE: To develop a methodology applicable to the morphologic study of the efficacy of finasteride on prostatic intraepithelial neoplasia (PIN), a putative precursor of prostate cancer. STUDY DESIGN: Three PIN foci were reviewed in two simple prostatectomy specimens from patients with clinical diagnoses of benign prostatic hyperplasia and treated with finasteride for six months. The feasibility of PIN diagnosis and grading based on "diagnostic distance" was investigated. It is a measure of the "extent" to which the observed features are different from those of the untreated prototypes representing the following diagnostic categories: normal prostate, low and high grade PIN and prostatic adenocarcinoma with a cribriform or large acinar pattern. Uncertainty in the PIN diagnosis and grading was dealt with by means of a Bayesian belief network (BBN). RESULTS: The distance measure values of the three PIN foci from the prototype of untreated, nonneoplastic prostate were 9, 7 and 8, respectively, in relative, arbitrary units. Their distance from the two prostate cancer patterns (large acinar and cribriform) was as high as 8-10. The distance of these foci from either low or high grade PIN were as low as 5, 3 and 2, and 3, 5 and 4, respectively. BBN produced the highest belief values for PIN, thus confirming the morphology-based and diagnostic distance-supported diagnosis; however, the belief values were low for both grades. CONCLUSION: The results provided by BBN analyses and diagnostic distance measures support the conclusion that this methodology is applicable to assessing the efficacy of finasteride treatment of PIN.

5-alpha Reductase Inhibitors↗
PubMed1998

Case diagnosis as positive identification in prostatic neoplasia.

OBJECTIVE: To apply a distance measure and Bayesian belief network-based methodology to the positive identification of case diagnosis in prostatic neoplasia. STUDY DESIGN: Eight morphologic and cellular features were analyzed in 20 cases of normal prostate, 20 of low grade prostatic intraepithelial neoplasia (PIN), 20 of high grade PIN, 20 of prostatic adenocarcinoma with a cribriform pattern and 20 of prostatic adenocarcinoma with an acinar pattern. The diagnostic distance was evaluated to measure the "extent" to which the feature outcomes of the individual cases differed from the expected profile of outcomes in typical cases of normal prostate, low and high grade PIN, and cribriform and large acinar adenocarcinoma. Belief values were evaluated with a Bayesian belief network (BBN). RESULTS: A bivariate representation of the cumulative absolute diagnostic distances of all the cases from the prototypes of normal prostate and cribriform adenocarcinoma was made. Three separate groups of cases were observed, corresponding to normal prostate, low grade PIN and cribriform adenocarcinoma. An additional group was formed by the cases of high grade PIN and acinar adenocarcinoma--i.e., there was complete overlap between the diagnostic distance values of cases belonging to these two categories. However, these cases showed differences in clue outcomes. To explore the contribution of such observations to case identification, a bivariate representation of the diagnostic distances from high grade PIN and acinar adenocarcinoma was made. The cases then formed five separate groups corresponding to the five diagnostic categories. When the individual cases were considered, their shortest distance was from the prototype of the category into which they were originally diagnosed. The BBN gave these diagnostic categories the highest belief values. CONCLUSION: The combined evaluation of diagnostic distance and belief represents an identification procedure. The numeric value of certainty characterizes individual cases according to the level of progression from PIN toward cancer.

Bayes Theorem↗
PubMed1992

Predicting protein secondary structure using neural net and statistical methods.

A comparison of neural network methods and Bayesian statistical methods is presented for prediction of the secondary structure of proteins given their primary sequence. The Bayesian method makes the unphysical assumption that the probability of an amino acid occurring in each position in the protein is independent of the amino acids occurring elsewhere. However, we find the predictive accuracy of the Bayesian method to be only minimally less than the accuracy of the most sophisticated methods used to date. We present the relationship of neural network methods to Bayesian statistical methods and show that, in principle, neural methods offer considerable power, although apparently they are not particularly useful for this problem. In the process, we derive a neural formalism in which the output neurons directly represent the conditional probabilities of structure class. The probabilistic formalism allows introduction of a new objective function, the mutual information, which translates the notion of correlation as a measure of predictive accuracy into a useful training measure. Although a similar accuracy to other approaches (utilizing a mean-square error) is achieved using this new measure, the accuracy on the training set is significantly and tantalizingly higher, even though the number of adjustable parameters remains the same. The mutual information measure predicts a greater fraction of helix and sheet structures correctly than the mean-square error measure, at the expense of coil accuracy, precisely as it was designed to do. By combining the two objective functions, we obtain a marginally improved accuracy of 64.4%, with Matthews coefficients C alpha, C beta and Ccoil of 0.40, 0.32 and 0.42, respectively. However, since all methods to date perform only slightly better than the Bayes algorithm, which entails the drastic assumption of independence of amino acids, one is forced to conclude that little progress has been made on this problem, despite the application of a variety of sophisticated algorithms such as neural networks, and that further advances will require a better understanding of the relevant biophysics.

Bayes Theorem↗
PubMed2006

A novel radial basis function neural network for discriminant analysis.

A novel radial basis function neural network for discriminant analysis is presented in this paper. In contrast to many other researches, this work focuses on the exploitation of the weight structure of radial basis function neural networks using the Bayesian method. It is expected that the performance of a radial basis function neural network with a well-explored weight structure can be improved. As the weight structure of a radial basis function neural network is commonly unknown, the Bayesian method is, therefore, used in this paper to study this a priori structure. Two weight structures are investigated in this study, i.e., a single-Gaussian structure and a two-Gaussian structure. An expectation-maximization learning algorithm is used to estimate the weights. The simulation results showed that the proposed radial basis function neural network with a weight structure of two Gaussians outperformed the other algorithms.

Algorithms↗
PubMed2005

Protein structure and fold prediction using Tree-Augmented naïve Bayesian classifier.

Due to the large volume of protein sequence data, computational methods to determine the structure class and the fold class of a protein sequence have become essential. Several techniques based on sequence similarity, Neural Networks, Support Vector Machines (SVMs), etc. have been applied. Since most of these classifiers use binary classifiers for multi-classification, there may be (N) c2 classifiers required. This paper presents a framework using the Tree-Augmented Bayesian Networks (TAN) which performs multi-classification based on the theory of learning Bayesian Networks and using improved feature vector representation of (Ding et al., 2001). In order to enhance TAN's performance, pre-processing of data is done by feature discretization and post-processing is done by using Mean Probability Voting (MPV) scheme. The advantage of using Bayesian approach over other learning methods is that the network structure is intuitive. In addition, one can read off the TAN structure probabilities to determine the significance of each feature (say, hydrophobicity) for each class, which helps to further understand the complexity in protein structure. The experiments on the datasets used in three prominent recent works show that our approach is more accurate than other discriminative methods. The framework is implemented on the BAYESPROT web server and it is available at http://www-appn.comp.nus.edu.sg/~bioinfo/bayesprot/Default.htm. More detailed results are also available on the above website.

Algorithms↗
PubMed1994

Removing the assumption of conditional independence from Bayesian decision models by using artificial neural networks: some practical techniques and a case study.

The article describes how artificial neural networks with special designs can be applied to approximate a subjective Bayesian decision model without the assumption of conditional independence. New techniques are proposed to resolve some of the practical difficulties during the processes of problem structuring, knowledge elicitation, quantitative modeling, and model interpretation. A Bayesian model considering the conditional dependencies to predict a teenager's marijuana use was constructed by experts using these techniques, and compared to another conventional Bayesian model which assumed conditional independence. The new approach without the assumption of conditional independence had predictive power (r = 0.7) in the test of linearity compared to the conventional approach (r = 0.58) on a data set (n = 129). Its receiver operating characteristic curve dominated the alternative approach within the range (true positive fraction > 0.7) that we were interested in. The interpretations of the possible conditional dependencies provided by the artificial neural network after the training process were consistent with the expert's descriptions.

Adolescent↗
PubMed2005

An analysis of thyroid function diagnosis using Bayesian-type and SOM-type neural networks.

Thyroid function diagnosis is an important classification problem, and we made reanalysis of the human thyroid data, which had been analyzed by the multivariate analysis, by the two notable neural networks. One is the self-organizing map approach which clusters the patients and displays visually a characteristic of the distribution according to laboratory tests. We found that self-organizing map (SOM) consists of three well separated clusters corresponding to hyperthyroid, hypothyroid and normal, and more detailed information for patients is obtained from the position in the map. Besides, the missing value SOM which we had introduced to investigate QSAR problem turned out to be also useful in treating such classification problem. We estimated the classification rates of thyroid disease using Bayesian regularized neural network (BRNN) and found that its prediction accuracy is better than multivariate analysis. Automatic relevance determination (ARD) method of BRNN was surely verified to be effective by the direct calculation of classification rates using BRNN without ARD for all possible combinations of laboratory tests.

Bayes Theorem↗
PubMed2006

Bayesian sequential inference for stochastic kinetic biochemical network models.

As postgenomic biology becomes more predictive, the ability to infer rate parameters of genetic and biochemical networks will become increasingly important. In this paper, we explore the Bayesian estimation of stochastic kinetic rate constants governing dynamic models of intracellular processes. The underlying model is replaced by a diffusion approximation where a noise term represents intrinsic stochastic behavior and the model is identified using discrete-time (and often incomplete) data that is subject to measurement error. Sequential MCMC methods are then used to sample the model parameters on-line in several data-poor contexts. The methodology is illustrated by applying it to the estimation of parameters in a simple prokaryotic auto-regulatory gene network.

Kinetics↗
PubMed2006

Bayesian learning of sparse gene regulatory networks.

Differential equations (DEs) have been the most widespread formalism for gene regulatory network (GRN) modeling, as they offer natural interpretation of biological processes, easy elucidation of gene relationships, and the capability of using efficient parameter estimation methods. However, an important limitation of DEs is their requirement of O(d(2)) parameters where d is the number of genes modeled, which often causes over-parameterization for large d, leading to the over-fitting of data and dense parameter sets that are hard to interpret. This paper presents the first effort to address the over-parameterization problem by applying the sparse Bayesian learning (SBL) method to sparsify the GRN model of DEs. SBL operates on the parsimony principle, with the objective to reduce the number of effective parameters by driving the redundant parameters to zero. The resulting sparse parameter set offers three important advantages for GRN inference: first, the inferred GRNs are more plausible, since the biological counterparts are known to be sparse; second, gene relationships can be more easily elucidated from sparse sets than from dense sets; and third, the solutions become more optimal and consistent, due to the reduction in the volume of solution space. Experiments are conducted on the yeast Saccharomyces cerevisiae time-series gene expression data, in which known regulatory events related to the cell cycle G1/S phase are reliably reproduced.

Bayes Theorem↗
PubMed2006

Assisting the diagnosis of thyroid diseases with Bayesian-type and SOM-type neural networks making use of routine test data.

Patients with hyperthyroidism sometimes take much time to receive the final diagnosis. To improve patient QOL, simple screening for hyperthyroidism by thyroid non-specialists at the physical check-up is highly expected. Therefore, we applied both Bayesian-type and SOM-type neural networks since we assured the approach useful in analysing thyroid function diagnosis in the previous work. Routine test (14 parameters) data from 66 subjects with a known diagnosis (18 patients with hyperthyroidism and 48 healthy volunteers) were adopted as learning data, and then 142 individuals who also received the same routine tests at the Tohoku University Hospital were screened to predict patients with hyperthyroidism. Both neural networks using 14 parameters predicted several patients as having hyperthyroidism with high probability, including all three hyperthyroid patients diagnosed later by the physician. Further detailed analysis of the routine test parameters that were important for classification found that screening with a set of three parameters (alkaline phosphatase, serum creatinine and total cholesterol) or plus aspartate aminotransferase allowed for quite accurate screening. These results showed that the same neural networks as previous work allows simple screening of patients for hyperthyroidism on the basis of routine test data, and that physicians not specializing in the thyroid can rapidly identify individuals suspected of having hyperthyroidism, to permit a rapid referral for examination and treatment by thyroid specialists.

Bayes Theorem↗
PubMed2005

Ontology Driven Construction of a Knowledgebase for Bayesian Decision Models Based on UMLS.

All decision models use some form of language to describe domain elements and their interactions. The terminology is often specific and even unique to the algorithm and is a choice of designers. Nevertheless the domain elements and concepts of any decision problem are almost never unique and are used and reused in many other decision problems. The same is true about the information about those elements in the context of different decision problems. Put together, the information about any given element forms our knowledge about the element and if stored properly in a knowledgebase, can be used and reused as necessary without the need for duplication.In this paper we discuss creation of an ontology using UMLS vocabulary and semantic network that provides an abstract understanding of elements (or objects) in the problem domain. Based on this ontology, a knowledgebase will be constructed that provides further information about the object in relation to another object or objects as described in the semantic links.A knowledgebase structured as such will have the benefit of problem-independence. It can be expanded as needed to include other objects that are used in a different series of problems and therefore, will have a one to many mapping between knowledgebase and decision models. Updating the knowledgebase will update the decision models seamlessly and maintenance will be less of an issue across decision models and within the knowledgebase. We are using this approach in building Bayesian decision models using Bayesian networks; however, this approach is not limited to Bayesian networks and has been and can be used for other decision making purposes.

Bayes Theorem↗
PubMed2000

Consistency of posterior distributions for neural networks.

In this paper we show that the posterior distribution for feedforward neural networks is asymptotically consistent. This paper extends earlier results on universal approximation properties of neural networks to the Bayesian setting. The proof of consistency embeds the problem in a density estimation problem, then uses bounds on the bracketing entropy to show that the posterior is consistent over Hellinger neighborhoods. It then relates this result back to the regression setting. We show consistency in both the setting of the number of hidden nodes growing with the sample size, and in the case where the number of hidden nodes is treated as a parameter. Thus we provide a theoretical justification for using neural networks for nonparametric regression in a Bayesian framework.

Bayes Theorem↗
PubMed1996

Dynamical computational properties of local cortical networks for visual and motor processing: a bayesian framework.

A major unsolved question concerns the interaction between the coding of information in the cortex and the collective neural operations (such as perceptual grouping, mental rotation) that can be performed on this information. A key property of the local networks in the cerebral cortex is to combine thalamocortical or feedforward information with horizontal cortico-cortical connections. Among different types of neural networks compatible with the known functional and architectural properties of the cortex, we show that there exist interesting bayesian solutions resulting in an optimal collective decision made by the neuronal population. We suggest that thalamo-cortical and cortico-cortical synaptic plasticity can be differentially modulated to optimize this collective bayesian decision process. We take two examples of cortical dynamics, one for perceptual grouping in MT, and the other one for mental rotation in M1. We show that a neural implementation of the bayesian principle is both computationally efficient to perform these tasks and consistent with the experimental data on the related neuronal activities. A major implication is that a similar collective decision mechanism should exist in different cortical regions due to the similarity of the cortical functional architecture.

Algorithms↗
PubMed2002

Appropriate medical data categorization for data mining classification techniques.

Some data mining (DM) methods, or software tools, require normalized data, others rely on categorized data, and some can accommodate multiple data scales. Each DM technique has a specific background theory; therefore, different results are expected when applying multiple methods. The purpose of this study is to find the data format appropriate for each DM classification technique for wider applications, and efficiently to obtain trustworthy results. Considering the nature of medical data, categorical variables are sometimes useful for making decisions and can make it easier to extrapolate knowledge. In this study, three mathematical data categorization methods (Fusinter, minimum description length principle [MDLPC] and Chi-merge) were applied to accommodate five data mining classification techniques (statistics discriminant analysis, supervised classification with Neural Networks, Decision trees, Genetic supervised clustering and Bayesian classification [probability neural networks; PNN]) using a heart disease database with four types of data (continuous data, binary data, nominal data, and ordinal data). Compared with original or normalized data, data categorized by the MDLPC categorization method was found to perform better in most of the DM classification techniques used in this study. Categorical data is good for most DM classification techniques (e.g. classification of disease and non-disease groups) and is relatively easy to use for extracting medical knowledge.

Bayes Theorem↗
PubMed1997

Decision support for psychiatric diagnosis based on a simple questionnaire.

This paper compares two classifiers: Pseudo Bayesian and Neural Network for assisting in making diagnoses of psychiatric patients based on a simple yes/no questionnaire which is provided at the outpatient's first visit to the hospital. The classifiers categorize patients into three most commonly seen ICD classes, i.e. schizophrenic, emotional and neurotic disorders. One hundred completed questionnaires were utilized for constructing and evaluating the classifiers. Average correct decision rates were 73.3% for the Pseudo Bayesian Classifier and 77.3% for the Neural Network classifier. These rates were higher than the rate which an experienced psychiatrist achieved based on the same restricted data as the classifiers utilized. These classifiers may be effectively utilized for assisting psychiatrists in making their final diagnoses.

Affective Symptoms↗