PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Neural Networks, Computer”

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 235 records · Page 13Linked to original sources
PubMed1994

Automated transformation of probabilistic knowledge for a medical diagnostic system.

Iliad is a large medical diagnostic system that covers more than 2000 diagnoses and 9000 findings. Due to the size and the complexity of this system, a robust knowledge representation is essential to consistently and efficiently model the medical knowledge involved. In this paper, we describe the knowledge representation currently used in Iliad and a probabilistic representation based on the Bayesian network formalism which can be derived using the information that the Iliad knowledge base contains.

Algorithms↗
PubMed1992

Vestibuloocular reflex arc analysis using an experimentally constrained neural network.

The primary function of the vestibuloocular reflex (VOR) is to maintain the stability of retinal images during head movements. This function is expressed through a complex array of dynamic and adaptive characteristics whose essential physiological basis is a disynaptic arc. We present a model of normal VOR function using a simple neural network architecture constrained by the physiological and anatomical characteristics of this disynaptic reflex arc. When tuned using a method of global optimization, this network is capable of exhibiting the broadband response characteristics observed in behavioral tests of VOR function. Examination of the internal units in the network show that this performance is achieved by rediscovering the solution to VOR processing first proposed by Skavenski and Robinson (1973). Type I units at the intermediate level of the network possess activation characteristics associated with either pure position or pure velocity. When the network is made more complex either through adding more pairs of internal units or an additional level of units, the characteristic division of unit activation properties into position and velocity types remains unchanged. Although simple in nature, the results of our simulations reinforce the validity of bottom-up approaches to modeling of neutral function. In addition, the architecture of the network is consistent with current ideas on the characteristics and site of adaptation of the reflex and should be compatible with current theories regarding learning rules for synaptic modification during VOR adaptation.

Computer Simulation↗
PubMed1996

Probabilistic interpretation of feedforward network outputs, with relationships to statistical prediction of ordinal quantities.

Several problems require the estimation of discrete random variables whose values can be put in a one-to-one ordered correspondence with a finite subset of the natural numbers. This happens whenever quantities are involved that represent integer items, or have been quantized on a fixed number of levels, or correspond to "graded" linguistic values. Here we propose a correct probabilistic approach to such kind of problems that fully exploits all the available prior knowledge about their own structure. In spite of the very stringent constraints induced in output space, the method can be directly applied to standard feed-forward networks while keeping local computation of both outputs and error signals. According to these guidelines, we devised a neural implementation of a complex image pre-processing algorithm by using very poor resolution on the computing elements in the network.

Algorithms↗
PubMed1992

Introduction of a neuronal network as a tool for diagnostic analysis and classification based on experimental pathologic data.

A neuronal network, as well as uni- and multivariate statistics and a discriminant analysis were applied to a morphometric database of 58 cases with thyroid neoplasms and normal thyroid tissue. The ability to classify cases correctly according to their diagnosis was compared between the neuronal network and discriminant analysis. For all pairwise comparisons, classification by neuronal network was as least as good as classification by discriminant analysis. For some comparisons, the neuronal network provided more correct diagnoses than discriminant analysis. On the contrary, in a comparison between tumors which are not significantly different according to multivariate statistics, the network reclassifies only half of the cases correctly, whereas discriminant analysis falsely suggests the possibility of classifying cases with either diagnosis. Our results confirm a higher sensitivity of the neuronal network to the diagnostic information contained in the present morphometric database, and we will therefore use this concept for analysis and diagnostic classification in further morphometric studies.

Diagnosis, Computer-Assisted↗
PubMed1995

A neural network approach for the determination of interhospital transport mode.

We report on the construction of neural networks for determining whether pediatric patients requiring transport to a tertiary care center should be moved by air or by ground. The networks were based on the functional-link net architecture. In two experiments, feedforward supervised-learning neural nets were trained with examples of an expert's decisions and then were used in a consulting mode to provide advice on cases not previously encountered. Training and validation were performed by a combination of the k-fold cross-validation and leaving-one-out sampling methods. Use of the functional-link net rather than the customary backpropagation net enabled us to carry out the training with fairly large amounts of data in realistically short time periods. In the first experiment, capillary refill, skin color, and stridor were consistently the input variables that were most strongly associated with the decision output. In both experiments, the networks were validated by comparing their performance retrospectively against the determination of an expert pediatric transport physician. The network was trained based on the expert's opinion about the correct mode of transport for each case with error rates of less than 10(-5).

Artificial Intelligence↗
PubMed1993

Regional cerebral blood flow estimation by neural network-based parametric regression analysis.

An artificial neural network (ANN) model was proposed for real-time estimation of regional cerebral blood flow (rCBF), by given head and expired air curves obtained through 133Xe inhalation. The network was constructed according to a regression model described by a linear differential equation. Experimental results compare well with those obtained by conventional curve fitting strategies, but the parameter estimation process is much simplified. A systematic procedure in developing ANN for parametric regression analysis was introduced; networks are constructed according to the selected regression model so that the obtained weights of a trained network directly represent parameters of the regression model which best fits the observed data set. Such a design-oriented methodology extends the classification-based applications of ANN to parametric regression analysis, and therefore may have more generalized applications besides rCBF estimation.

Algorithms↗
PubMedDate not supplied

Diagnostic decision support by inference networks.

Inference networks permit the combining of diagnostic evidence in such a fashion that the mutual dependence structure of different pieces of evidence is considered, and that a probabilistic measure of the uncertainty of the final diagnostic decision is provided. Operated in an automatic reasoning mode, an inference network allows a decoupling of the false negative rate from the false positive rate in diagnostic procedures involving rare event detection such as the prescreening for cervical cancer.

Diagnosis, Computer-Assisted↗
PubMed1995

Nonlinear gated experts for time series: discovering regimes and avoiding overfitting.

In the analysis and prediction of real-world systems, two of the key problems are nonstationarity (often in the form of switching between regimes) and overfitting (particularly serious for noisy processes). This article addresses these problems using gated experts, consisting of a (nonlinear) gating network, and several (also nonlinear) competing experts. Each expert learns to predict the conditional mean, and each expert adapts its width to match the noise level in its regime. The gating network learns to predict the probability of each expert, given the input. This article focuses on the case where the gating network bases its decision on information from the inputs. This can be contrasted to hidden Markov models where the decision is based on the previous state(s) (i.e. on the output of the gating network at the previous time step), as well as to averaging over several predictors. In contrast, gated experts soft-partition the input space, only learning to model their region. This article discusses the underlying statistical assumptions, derives the weight update rules, and compares the performance of gated experts to standard methods on three time series: (1) a computer-generated series, obtained by randomly switching between two nonlinear processes; (2) a time series from the Santa Fe Time Series Competition (the light intensity of a laser in chaotic state); and (3) the daily electricity demand of France, a real-world multivariate problem with structure on several time scales. The main results are: (1) the gating network correctly discovers the different regimes of the process; (2) the widths associated with each expert are important for the segmentation task (and they can be used to characterize the sub-processes); and (3) there is less overfitting compared to single networks (homogeneous multilayer perceptrons), since the experts learn to match their variances to the (local) noise levels. This can be viewed as matching the local complexity of the model to the local complexity of the data.

Computers↗
PubMed1995

Artificial neural networks for recognition of electrocardiographic lead reversal.

Misplacement of electrodes during the recording of an electrocardiogram (ECG) can cause an incorrect interpretation, misdiagnosis, and subsequent lack of proper treatment. The purpose of this study was twofold: (1) to develop artificial neural networks that yield peak sensitivity for the recognition of right/left arm lead reversal at a very high specificity; and (2) to compare the performances of the networks with those of 2 widely used rule-based interpretation programs. The study was based on 11,009 ECGs recorded in patients at an emergency department using computerized electrocardiographs. Each of the ECGs was used to computationally generate an ECG with right/left arm lead reversal. Neural networks were trained to detect ECGs with right/left arm lead reversal. Different networks and rule-based criteria were used depending on the presence or absence of P waves. The networks and the criteria all showed a very high specificity (99.87% to 100%). The neural networks performed better than the rule-based criteria, both when P waves were present (sensitivity 99.1%) or absent (sensitivity 94.5%). The corresponding sensitivities for the best criteria were 93.9% and 39.3%, respectively. An estimated 300 million ECGs are recorded annually in the world. The majority of these recordings are performed using computerized electrocardiographs, which include algorithms for detection of right/left arm lead reversals. In this study, neural networks performed better than conventional algorithms and the differences in sensitivity could result in 100,000 to 400,000 right/left arm lead reversals being detected by networks but not by conventional interpretation programs.

Algorithms↗
PubMed1994

Application of a bi-directional associative memory (BAM) network in computer assisted learning in chemistry.

A computer assisted learning software based on a bi-directional associative memory (BAM) network was developed. The software was implemented to assist students in associating the names of the elements in the periodic table with their chemical symbols. The use of the BAM facilitates the analysis and interpretation of students' responses. The software package can be modified easily as an educational tool for other disciplines.

Algorithms↗
PubMed1997

Evaluation of simplified compartmental models of reconstructed neocortical neurons for use in large-scale simulations of biological neural networks.

The electrotonic properties of the complex arborizations of neurons can be simulated by creating compartmental models based on the morphology of real neurons. These models can be very detailed with thousands of individual compartments and active channels. Large numbers of these models can be linked together into biologically realistic, large-scale neural networks with which to obtain a better understanding of the interactions among real neurons. However, the use of detailed compartmental models in such large networks is hindered by long computation times. Methods exist to reduce the complex morphology of detailed compartmental models to simpler reconstructions that retain many of the electrotonic properties of the original model yet are computationally efficient. However, little work exists that evaluates the limitations and performance of such reduced models with realistic active conductances modeled in both the soma and the dendrites to ensure that they are appropriate for use in biologically realistic network models. We have created detailed and reduced models of reconstructed dye-filled neurons from rat somatosensory neocortex and evaluated the ability of the reduced models to faithfully reproduce the input-output functions of the more detailed models. We find that the reduced models are not capable of perfectly reproducing the exact output of the detailed models using identical parameters. However, if the parameters are adjusted the reduced models are certainly capable of providing input-output patterns that are well within an acceptable range of known neural activity. The limitations and the benefits of such models are discussed.

Animals↗
PubMed1998

Ensembles of radial basis function networks for spectroscopic detection of cervical precancer.

The mortality related to cervical cancer can be substantially reduced through early detection and treatment. However, current detection techniques, such as Pap smear and colposcopy, fail to achieve a concurrently high sensitivity and specificity. In vivo fluorescence spectroscopy is a technique which quickly, noninvasively and quantitatively probes the biochemical and morphological changes that occur in precancerous tissue. A multivariate statistical algorithm was used to extract clinically useful information from tissue spectra acquired from 361 cervical sites from 95 patients at 337-, 380-, and 460-nm excitation wavelengths. The multivariate statistical analysis was also employed to reduce the number of fluorescence excitation-emission wavelength pairs required to discriminate healthy tissue samples from precancerous tissue samples. The use of connectionist methods such as multilayered perceptrons, radial basis function (RBF) networks, and ensembles of such networks was investigated. RBF ensemble algorithms based on fluorescence spectra potentially provide automated and near real-time implementation of precancer detection in the hands of nonexperts. The results are more reliable, direct, and accurate than those achieved by either human experts or multivariate statistical algorithms.

Algorithms↗
PubMedDate not supplied

Reliable disparity estimation through selective integration.

A network model of disparity estimation was developed based on disparity-selective neurons, such as those found in the early stages of processing in the visual cortex. The model accurately estimated multiple disparities in regions, which may be caused by transparency or occlusion. The selective integration of reliable local estimates enabled the network to generate accurate disparity estimates on normal and transparent random-dot stereograms. The model was consistent with human psychophysical results on the effects of spatial-frequency filtering on disparity sensitivity. The responses of neurons in macaque area V2 to random-dot stereograms are consistent with the prediction of the model that a subset of neurons responsible for disparity selection should be sensitive to disparity gradients.

Animals↗
PubMed1993

Prediction of protein secondary structure at better than 70% accuracy.

We have trained a two-layered feed-forward neural network on a non-redundant data base of 130 protein chains to predict the secondary structure of water-soluble proteins. A new key aspect is the use of evolutionary information in the form of multiple sequence alignments that are used as input in place of single sequences. The inclusion of protein family information in this form increases the prediction accuracy by six to eight percentage points. A combination of three levels of networks results in an overall three-state accuracy of 70.8% for globular proteins (sustained performance). If four membrane protein chains are included in the evaluation, the overall accuracy drops to 70.2%. The prediction is well balanced between alpha-helix, beta-strand and loop: 65% of the observed strand residues are predicted correctly. The accuracy in predicting the content of three secondary structure types is comparable to that of circular dichroism spectroscopy. The performance accuracy is verified by a sevenfold cross-validation test, and an additional test on 26 recently solved proteins. Of particular practical importance is the definition of a position-specific reliability index. For half of the residues predicted with a high level of reliability the overall accuracy increases to better than 82%. A further strength of the method is the more realistic prediction of segment length. The protein family prediction method is available for testing by academic researchers via an electronic mail server.

Mathematical Computing↗
PubMed1997

Analyses on the temporal patterns of spikes of auditory neurons of the macaque monkey by means of an artificial neural network and tree-based models.

The time scale over which information in the primary auditory cortex is processed was estimated. An artificial neural network was used to learn the temporal patterns of spikes. After learning, test patterns were input to the network. Comparison of the accuracy of the network with that of the maximum likelihood function computed from the spike count reveals that the temporal patterns of spikes are closely related to stimulus discrimination. Next, a tree-based model from a subset of the spike trains with a fixed time resolution was constructed and validated the model with another. By repeating this for different bin widths, it was found that there are no simple models for the time bin width larger than 50 ms. This indicates that the time scale in the auditory cortex is not larger than 50 ms.

Animals↗
PubMed1996

Web interface for the Heart Disease Program.

The task of making a large complex diagnostic program available to a broad audience of physicians has become more feasible with the ubiquitous accessibility of the client-server architecture of the World Wide Web. This paper describes the design and implementation of a Web interface for the Heart Disease Program (HDP). The client-server architecture imposes a number of requirements on the program. The graphical capabilities of the Web enable a number of enhancements to the program but also cause some limitations. Our initial experience with physicians using the HDP through the Web interface has been positive and we are now conducting an evaluation of the HDP using this form of access.

Computer Communication Networks↗
PubMed1994

Energy minimization method using automata network for sequence and side-chain conformation prediction from given backbone geometry.

Globular proteins have high packing densities as a result of residue side chains in the core achieving a tight, complementary packing. The internal packing is considered the main determinant of native protein structure. From that point of view, we present here a method of energy minimization using an automata network to predict a set of amino acid sequences and their side-chain conformations from a desired backbone geometry for de novo design of proteins. Using discrete side-chain conformations, that is, rotamers, the sequence generation problem from a given backbone geometry becomes one of combinatorial problems. We focused on the residues composing the interior core region and predicted a set of amino acid sequences and their side-chain conformations only from a given backbone geometry. The kinds of residues were restricted to six hydrophobic amino acids (Ala, Ile, Met, Leu, Phe, and Val) because the core regions are almost always composed of hydrophobic residues. The obtained sequences were well packed as was the native sequence. The method can be used for automated sequence generation in the de novo design of proteins.

Bacterial Proteins↗