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 901 records · Page 50Linked to original sources

PubMed1993

Development of artificial neural filters for pattern recognition in protein sequences.

Four different artificial neural network architectures have been tested for their suitability to extract and predict sequence features. For optimization of the network weights an evolutionary computing method has been applied. The networks have feedforward architecture and provide adaptive neural filter systems for pattern recognition in primary structures and sequence classification. The recognition and prediction of signal peptidase cleavage sites of E. coli periplasmic protein precursors serves as an example for filter development. The primary structures are represented by seven physicochemical residue properties. This amino acid description provides the feature space for network optimization. The properties hydrophobicity, hydrophilicity, side-chain volume, and polarity allowed an accurate classification of the data. A three-layer network architecture reached a learning success of 100%; the highest prediction accuracy in an independent test set of sequences was 97%. This network architecture appears to be most suited for the analysis of E. coli signal peptidase cleavage sites. Further suggestions about the design and future applications of artificial neural networks for protein sequence analysis are made.

Bacterial Proteins↗

PubMed1998

MAVIS: an interactive visualization tool for computational chemistry calculations in a distributed networked environment.

This paper describes the software package, MAVIS (Molecular Animation and VI-sualization System), which was developed as a collaborative project between the University of Manchester Department of Chemistry and the Manchester Visualization Centre as a tool for the visualization of chemical systems and their properties and for the generation, submission, monitoring and study of results from calculations in a networked computing environment. MAVIS has been developed using AVS, a general visualization environment, as a suite of modules which can be extended with relative ease. Using the AVS system gives the application the power and flexibility of a general system with no requirement for the programmer to understand many of the complexities involved in modern image rendering systems and allowing them to concentrate upon the features of the functionality and the user interface. It also allows the code to be as portable between platforms as the AVS package itself and so it is available on a wide range of systems.

Chemistry, Organic↗

PubMed2002

Neural network simulations of the primate oculomotor system IV. A distributed bilateral stochastic model of the neural integrator of the vertical saccadic system.

The present report examines the performance of a distributed bi-directional neural network that simulates the vertical velocity to position integrator of the primate brain. Consistent with anatomy and physiology, its units receive stochastically weighted input from vertical medium-lead burst neurons. Also consistent with anatomy, units belonging to integrators with opposite on-directions (up or down) are interconnected via the posterior commissure (again in a stochastically weighted manner) and they can be excitatory or inhibitory. To demonstrate that integration can be a one-step process, the output of model units was routed directly to vertical motoneurons. Model units replicate the wide range of saccade-related discharge patterns encountered in the portion of the primate brain that is thought to house the vertical neural integrator (the interstitial nucleus of Cajal) while "lesions" of model units and/or their interconnections replicate the symptoms which follow insults to this brain area.

Animals↗

PubMed2006

Computational simulation of the input-output relationship in hippocampal pyramidal cells.

The precise mapping of how complex patterns of synaptic inputs are integrated into specific patterns of spiking output is an essential step in the characterization of the cellular basis of network dynamics and function. Relative to other principal neurons of the hippocampus, the electrophysiology of CA1 pyramidal cells has been extensively investigated. Yet, the precise input-output relationship is to date unknown even for this neuronal class. CA1 pyramidal neurons receive laminated excitatory inputs from three distinct pathways: recurrent CA1 collaterals on basal dendrites, CA3 Schaffer collaterals, mostly on oblique and proximal apical dendrites, and entorhinal perforant pathway on distal apical dendrites. We implemented detailed computer simulations of pyramidal cell electrophysiology based on three-dimensional anatomical reconstructions and compartmental models of available biophysical properties from the experimental literature. To investigate the effect of synaptic input on axosomatic firing, we stochastically distributed a realistic number of excitatory synapses in each of the three dendritic layers. We then recorded the spiking response to different stimulation patterns. For all dendritic layers, synchronous stimuli resulted in trains of spiking output and a linear relationship between input and output firing frequencies. In contrast, asynchronous stimuli evoked non-bursting spike patterns and the corresponding firing frequency input-output function was logarithmic. The regular/irregular nature of the input synaptic intervals was only reflected in the regularity of output inter-burst intervals in response to synchronous stimulation, and never affected firing frequency. Synaptic stimulations in the basal and proximal apical trees across individual neuronal morphologies yielded remarkably similar input-output relationships. Results were also robust with respect to the detailed distributions of dendritic and synaptic conductances within a plausible range constrained by experimental evidence. In contrast, the input-output relationship in response to distal apical stimuli showed dramatic differences from the other dendritic locations as well as among neurons, and was more sensible to the exact channel densities.

Action Potentials↗

PubMed2006

Output convergence analysis for a class of delayed recurrent neural networks with time-varying inputs.

This paper studies the output convergence of a class of recurrent neural networks with time-varying inputs. The model of the studied neural networks has different dynamic structure from that in the well known Hopfield model, it does not contain linear terms. Since different structures of differential equations usually result in quite different dynamic behaviors, the convergence of this model is quite different from that of Hopfield model. This class of neural networks has been found many successful applications in solving some optimization problems. Some sufficient conditions to guarantee output convergence of the networks are derived.

Algorithms↗

PubMed2004

Congestion control for ATM multiplexers using neural networks: multiple sources/single buffer scenario.

A new neural network based method for solving the problem of congestion control arising at the user network interface (UNI) of ATM networks is proposed in this paper. Unlike the previous methods where the coding rate for all traffic sources as controller output signals is tuned in a body, the proposed method adjusts the coding rate for only a part of the traffic sources while the remainder sources send the cells in the previous coding rate in case of occurrence of congestion. The controller output signals include the source coding rate and the percentage of the sources that send cells at the corresponding coding rate. The control methods not only minimize the cell loss rate but also guarantee the quality of information (such as voice sources) fed into the multiplexer buffer. Simulations with 150 ADPCM voice sources fed into the multiplexer buffer showed that the proposed methods have advantage over the previous methods in the aspect of the performance indices such as cell loss rate (CLR) and voice quality.

Algorithms↗

PubMed1993

The mind and Donald O. Hebb.

By rooting behavior in ideas, and ideas in the brain, Hebb laid the groundwork for modern neuroscience. His theory prefigured computer models of neural networks.

Animals↗

PubMed2003

Analysis of hospitalised patient flows using data-mining.

UNLABELLED: Face to the development of hospital information system in the "Hôpital Européen George Pompidou" (HEGP), computerized patients records made medical data easier to analyse than before. We use data-mining technology to analyse intra-hospital patients' paths with one year of PMSI data (a French medical information system similar to Diagnosis Related Group). METHODS: 1. "sequential patterns mining" was used to analyse the most frequent patients' paths, 2. an integrated framework of "association rules mining" and "classification rule mining" was used to build prediction rules of patients' paths. RESULT: We construct a rule based prediction model, which gives the tendency of the patient's paths between the different medical units.

Computer Communication Networks↗

PubMed1998

MoldaNet: a network distributed molecular graphics and modelling program that integrates secure signed applet and Java 3D technologies.

MoldaNet is a molecular graphics and modelling program that integrates several new Java technologies, including authentication as a Secure Signed Applet, and implementation of Java 3D classes to enable access to hardware graphics acceleration. It is the first example of a novel class of Internet-based distributed computational chemistry tool designed to eliminate the need for user pre-installation of software on their client computer other than a standard Internet browser. The creation of a properly authenticated tool using a signed digital X.509 certificate permits the user to employ MoldaNet to read and write the files to a local file store; actions that are normally disallowed in Java applets. The modularity of the Java language also allows straightforward inclusion of Java3D and Chemical Markup Language classes in MoldaNet to permit the user to filter their model into 3D model descriptors such as VRML97 or CML for saving on local disk. The implications for both distance-based training environments and chemical commerce are noted.

Computer Graphics↗

PubMed1997

Application of a neural network for gentamicin concentration prediction in a general hospital population.

Neural network (NN) computation is computer modeling based in part on simulation of the structure and function of the brain. These modeling techniques have been found useful as pattern recognition tools. In the present study, data including age, sex, height, weight, serum creatinine concentration, dose, dosing interval, and time of measurement were collected from 240 patients with various diseases being treated with gentamicin in a general hospital setting. The patient records were randomly divided into two sets: a training set of 220 patients used to develop relationships between input and output variables (peak and trough plasma concentrations) and a testing set (blinded from the NN) of 20 to test the NN. The network model was the back-propagation, feed-forward model. Various networks were tested, and the most accurate networks for peak and trough (calculated as mean percent error, root mean squared error of the testing group, and tau value between observed and predicted values) were reported. The results indicate that NNs can predict gentamicin serum concentrations accurately from various input data over a range of patient ages and renal function and may offer advantages over traditional dose prediction methods for gentamicin.

Adolescent↗

PubMed1996

Reproducibility of Bayesian belief network assessment of breast fine needle aspirates.

OBJECTIVE: To assess the consistency of diagnosis of fine needle aspiration biopsies of breast lesions by three experienced and five less experienced pathologists using conventional means and applying a Bayesian belief network (BBN) to 10 diagnostic features to support diagnostic decision making. STUDY DESIGN: Forty fine needle aspiration biopsies, previously assessed by one of the experienced pathologists both conventionally and using a BBN, were assessed by two further experienced pathologists and five less experienced pathologists. RESULTS: Using the BBN, the experienced pathologists arrived at diagnoses in agreement with an established consensus at a slightly lower rate than by conventional means. The less experienced pathologists arrived at the correct diagnoses no more frequently with the help of the BBN than conventionally. CONCLUSION: As used in this study, the BBN did not help less experienced pathologists to interpret their observations but did not enable less experienced pathologists to identify how their observations differed and affected their diagnoses. The prototype system used in this study has since been upgraded by providing computer graphic displays of the features to be observed so that a more uniform mental image can be held by the participating pathologists. This will be tested with the same study design.

Biopsy, Needle↗

PubMed2004

Whole-genome annotation by using evidence integration in functional-linkage networks.

The advent of high-throughput biology has catalyzed a remarkable improvement in our ability to identify new genes. A large fraction of newly discovered genes have an unknown functional role, particularly when they are specific to a particular lineage or organism. These genes, currently labeled "hypothetical," might support important biological cell functions and could potentially serve as targets for medical, diagnostic, or pharmacogenomic studies. An important challenge to the scientific community is to associate these newly predicted genes with a biological function that can be validated by experimental screens. In the absence of sequence or structural homology to known genes, we must rely on advanced biotechnological methods, such as DNA chips and protein-protein interaction screens as well as computational techniques to assign putative functions to these genes. In this article, we propose an effective methodology for combining biological evidence obtained in several high-throughput experimental screens and integrating this evidence in a way that provides consistent functional assignments to hypothetical genes. We use the visualization method of propagation diagrams to illustrate the flow of functional evidence that supports the functional assignments produced by the algorithm. Our results contain a number of predictions and furnish strong evidence that integration of functional information is indeed a promising direction for improving the accuracy and robustness of functional genomics.

Chromosome Mapping↗

PubMed1995

Mapping of phase response properties of monopolar ECG voltages.

In the present investigation an attempt has been made to study the phase response properties of monopolar chest lead ECG voltages. Using a generator model of the heart an equivalent circuit of ECG network has been developed. The equivalent impedance between WCT and probe electrode has been determined by reduction techniques. From this equivalent impedance the phaser characteristics of monopolar ECG voltages have been analysed for change in probe electrode locations. The source of the generated voltage, i.e., the heart, will develop a different voltage for its different condition. There will also be a change in impedances. Thus for the normal subject the distribution of the phaser of the ECG voltages will be different from that of the abnormal one. A software tool has been developed to evaluate the relative phase response of ECG voltages. The data acquisition of monopolar ECG records of chest leads V1 to V6 from chart recorder has been done with the help of AutoCAD application package. The harmonic constituents of ECG voltages have been evaluated at each harmonic plane and the phase characteristics have been studied in polar coordinate for normal subjects as well as for a typical case. An interesting result has been observed in typical cases which are indicated in the paper.

Algorithms↗

PubMed2002

[Data mining in diagnostic knowledge acquisition from patients with brain glioma].

In order to correctly predict the malignant degree of brain glioma, three data mining algorithms: multi-layer perceptron network(MLP), decision tree, and rule induction are adopted to acquire diagnostic knowledge from patients with brain glioma cases. Totally 280 cases are collected, and some of them contain missing values. Preprocessing is taken to make them applicable to all three algorithms. Performance comparisons are carried out with a 10-fold cross validation test. Although the result of MLP is hard to be understood and cannot be applied directly, its reliability and accuracy are the highest when only a few hidden nodes are involved. Unlike MLP, both decision tree and rule induction use attribute-value pairs to represent diagnostic knowledge derived from treated cases. These could improve both the understandability and applicability of their results. When compared with rule induction, the inherent restriction in structure makes decision tree more efficient in decision-making but meanwhile hurts its simplicity, accuracy, and reliability. For testing samples, results of all these algorithms can achieve accuracy rate over 80%, which satisfies the basic requirement of neuroradiologists. If diagnostic accuracy rate is the main factor to be considered, MLP with only a few hidden nodes is the best. If the result is expected to be further checked or evaluated, rule induction will be the best algorithm. This work proves that data mining techniques can be used to obtain valid diagnostic knowledge from brain glioma cases and make computer aided diagnosis system in this field feasible.

Algorithms↗

PubMed2006

Artificial neural network classifier for the diagnosis of Parkinson's disease using [99mTc]TRODAT-1 and SPECT.

Imaging the dopaminergic neurotransmitter system with positron emission tomography (PET) or single photon emission tomography (SPECT) is a powerful tool for the diagnosis of Parkinson's disease (PD). Previous studies have indicated that human observers have a diagnostic accuracy similar to conventional ROI analysis of SPECT imaging data. Consequently, it has been hypothesized that an artificial neural network (ANN), which can mimic the pattern recognition skills of human observers, may provide similar results. A set of patients with PD, and normal healthy control subjects, were studied using the dopamine transporter tracer [(99m)Tc]TRODAT-1 and SPECT. The sample was comprised of 81 patients (mean age +/- SD: 63.4 +/- 10.4 years; age range: 39.0-84.2 years) and 94 healthy controls (mean age +/- SD: 61.8 +/- 11.0 years; age range: 40.9-83.3 years). The images were processed to extract the striatum and the striatal pixel values were used as inputs to a three-layer ANN. The same set of data was used to both train and test the ANN, in a 'leave one out' procedure. The diagnostic accuracy of the ANN was higher than any previous analysis method applied to the same data (94.4% total accuracy, 97.5% specificity and 91.4% sensitivity). However, it should be stressed that, as with all applications of an ANN, it was difficult to interpret precisely what triggers in the images were being detected by the network.

Adult↗

PubMed2000

A model of invariant object recognition in the visual system: learning rules, activation functions, lateral inhibition, and information-based performance measures.

VisNet2 is a model to investigate some aspects of invariant visual object recognition in the primate visual system. It is a four-layer feedforward network with convergence to each part of a layer from a small region of the preceding layer, with competition between the neurons within a layer and with a trace learning rule to help it learn transform invariance. The trace rule is a modified Hebbian rule, which modifies synaptic weights according to both the current firing rates and the firing rates to recently seen stimuli. This enables neurons to learn to respond similarly to the gradually transforming inputs it receives, which over the short term are likely to be about the same object, given the statistics of normal visual inputs. First, we introduce for VisNet2 both single-neuron and multiple-neuron information-theoretic measures of its ability to respond to transformed stimuli. Second, using these measures, we show that quantitatively resetting the trace between stimuli is not necessary for good performance. Third, it is shown that the sigmoid activation functions used in VisNet2, which allow the sparseness of the representation to be controlled, allow good performance when using sparse distributed representations. Fourth, it is shown that VisNet2 operates well with medium-range lateral inhibition with a radius in the same order of size as the region of the preceding layer from which neurons receive inputs. Fifth, in an investigation of different learning rules for learning transform invariance, it is shown that VisNet2 operates better with a trace rule that incorporates in the trace only activity from the preceding presentations of a given stimulus, with no contribution to the trace from the current presentation, and that this is related to temporal difference learning.

Algorithms↗

PubMed2006

Real time computation: zooming in on population codes.

Information processing in nervous systems intricately combines computation at the neuronal and network levels. Many computations may be envisioned as sequences of signal processing steps along some pathway. How can information encoded by single cells be mapped onto network population codes, and how do different modules or layers in the computation synchronize their communication and computation? These fundamental questions are particularly severe when dealing with real time streams of inputs. Here we study this problem within the context of a minimal signal perception task. In particular, we encode neuronal information by externally applying a space- and time-localized stimulus to individual neurons within a network. We show that a pulse-coupled recurrent neural network can successfully handle this task in real time, and obeys three key requirements: (i) stimulus dependence, (ii) initial-conditions independence, and (iii) accessibility by a readout mechanism. In particular, we suggest that the network's overall level of activity can be used as a temporal cue for a robust readout mechanism. Within this framework, the network can rapidly map a local stimulus onto a population code that can then be reliably read out during some narrow but well defined window of time.

Computer Simulation↗

PubMed2002

On the capabilities of neural networks using limited precision weights.

This paper analyzes some aspects of the computational power of neural networks using integer weights in a very restricted range. Using limited range integer values opens the road for efficient VLSI implementations because: (i) a limited range for the weights can be translated into reduced storage requirements and (ii) integer computation can be implemented in a more efficient way than the floating point one. The paper concentrates on classification problems and shows that, if the weights are restricted in a drastic way (both range and precision), the existence of a solution is not to be taken for granted anymore. The paper presents an existence result which relates the difficulty of the problem as characterized by the minimum distance between patterns of different classes to the weight range necessary to ensure that a solution exists. This result allows us to calculate a weight range for a given category of problems and be confident that the network has the capability to solve the given problems with integer weights in that range. Worst-case lower bounds are given for the number of entropy bits and weights necessary to solve a given problem. Various practical issues such as the relationship between the information entropy bits and storage bits are also discussed. The approach presented here uses a worst-case analysis. Therefore, the approach tends to overestimate the values obtained for the weight range, the number of bits and the number of weights. The paper also presents some statistical considerations that can be used to give up the absolute confidence of a successful training in exchange for values more appropriate for practical use. The approach presented is also discussed in the context of the VC-complexity.

Neural Networks, Computer↗