Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Neural network”

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 145 records · Page 8Linked to original sources
PubMed1993

Acoustical detection of coronary occlusions using neural networks.

A nonlinear neural network classifier was applied to noninvasive acoustic detection of coronary artery disease; the classifier included a feature vector, derived from diastolic heart sounds, and a multi-layered network trained by the backpropagation. The feature vector is based on the linear prediction coefficients of the autoregressive method after an adaptive line enhancement method was used as the input pattern to the neural network. One hundred and twelve recordings (70 abnormal, 42 normal) were studied and the network was trained on a randomly chosen set of six abnormal and six normal patients. It was tested on a database consisting of 100 recordings to which it had not been exposed. The network correctly identified 50 of the 64 patients with coronary artery disease and 32 of the 36 patients without any coronary artery occlusions. These results showed that this neural network is capable of distinguishing normal patients from abnormal patients. In addition, the diagnostic capability of this approach is much better than any other available noninvasive approach.

Coronary Disease↗
PubMedDate not supplied

Modeling of cyclin-dependent kinase inhibition by 1H-pyrazolo[3,4-d]pyrimidine derivatives using artificial neural network ensembles.

Artificial neural network ensembles were used for modeling the cyclin-dependent kinase inhibition of 1H-pyrazolo[3,4-d]pyrimidine derivatives. The structural characteristics of these inhibitors were encoded in relevant 3D-spatial descriptors extracted by genetic algorithm feature selection. Bayesian-regularized multilayer neural networks, trained by the back-propagation algorithm, were developed using these variables as inputs. The predictive power of the model was tested by leave-one-out cross validation. In addition, for a more rigorous measure of the predictive capacity, multiple validation sets were randomly generated as members of neural network ensembles, which makes doing averaged predictions feasible. In this way, the predictive power was analyzed accounting for the averaged test set R values and test set mean-square errors. Otherwise, Kohonen self-organizing maps were used as an additional tool for the same modeling. The location of the inhibitors in a map facilitates the analysis of the connection between compounds and serves as a useful tool for qualitative predictions.

Algorithms↗
PubMed1993

An architecture for fully integrated large scale neural networks.

To apply neural networks to many engineering applications, large networks will be required. Such networks are difficult to build using standard crystalline silicon technology due to limitations in both the fabrication and packaging processes. An architecture is proposed where amorphous silicon photoresistors are used to store the synaptic weights. A single plate of amorphous silicon is able to contain up to 100 million photoresistors, exploiting readily available fabrication technology. Using an external light source, each photoresistor can be individually adjusted allowing them to be configured as programmable fixed-value resistors. The processing compatibility of polysilicon and amorphous silicon allows the same glass substrate to be used for large-area integration of the photosensors, the analogue neural network and the neurons. The integration of the photosensors and the rest of the network may be used to alleviate the interface problem at the inputs resulting in a design with a very simple architecture that is both elegant and simple to fabricate. This paper describes such a design in which amorphous silicon technology is applied to neural network hardware.

Algorithms↗
PubMed2001

Detection of immunologically significant factors for chronic fatigue syndrome using neural-network classifiers.

Neural-network classifiers were used to detect immunological differences in groups of chronic fatigue syndrome (CFS) patients that heretofore had not shown significant differences from controls. In the past linear methods were unable to detect differences between CFS groups and non-CFS control groups in the nonveteran population. An examination of the cluster structure for 29 immunological factors revealed a complex, nonlinear decision surface. Multilayer neural networks showed an over 16% improvement in an n-fold resampling generalization test on unseen data. A sensitivity analysis of the network found differences between groups that are consistent with the hypothesis that CFS symptoms are a consequence of immune system dysregulation. Corresponding decreases in the CD19(+) B-cell compartment and the CD34(+) hematopoietic progenitor subpopulation were also detected by the neural network, consistent with the T-cell expansion. Of significant interest was the fact that, of all the cytokines evaluated, the only one to be in the final model was interleukin-4 (IL-4). Seeing an increase in IL-4 suggests a shift to a type 2 cytokine pattern. Such a shift has been hypothesized, but until now convincing evidence to support that hypothesis has been lacking.

B-Lymphocytes↗
PubMed2006

Risk factor identification and mortality prediction in cardiac surgery using artificial neural networks.

OBJECTIVE: The artificial neural network model is a nonlinear technology useful for complex pattern recognition problems. This study aimed to develop a method to select risk variables and predict mortality after cardiac surgery by using artificial neural networks. METHODS: Prospectively collected data from 18,362 patients undergoing cardiac surgery at 128 European institutions in 1995 (the European System for Cardiac Operative Risk Evaluation database) were used. Models to predict the operative mortality were constructed using artificial neural networks. For calibration a sixfold cross-validation technique was used, and for testing a fourfold cross-testing was performed. Risk variables were ranked and minimized in number by calibrated artificial neural networks. Mortality prediction with 95% confidence limits for each patient was obtained by the bootstrap technique. The area under the receiver operating characteristics curve was used as a quantitative measure of the ability to distinguish between survivors and nonsurvivors. Subgroup analysis of surgical operation categories was performed. The results were compared with those from logistic European System for Cardiac Operative Risk Evaluation analysis. RESULTS: The operative mortality was 4.9%. Artificial neural networks selected 34 of the total 72 risk variables as relevant for mortality prediction. The receiver operating characteristics area for artificial neural networks (0.81) was larger than the logistic European System for Cardiac Operative Risk Evaluation model (0.79; P = .0001). For different surgical operation categories, there were no differences in the discriminatory power for the artificial neural networks (P = .15) but significant differences were found for the logistic European System for Cardiac Operative Risk Evaluation (P = .0072). CONCLUSIONS: Risk factors in a ranked order contributing to the mortality prediction were identified. A minimal set of risk variables achieving a superior mortality prediction was defined. The artificial neural network model is applicable independent of the cardiac surgical procedure.

Adolescent↗
PubMed1990

Molecular networks as a sub-neural factor of neural networks.

We describe a new approach in the research of neural networks. This research is based on molecular networks in the neuron. If we use molecular networks as a sub-neuron factor of neural networks, it is a more realistic approach than today's concepts in this new computer technology field, because the artificial neural activity profile is similar to the profile of the action potential in the natural neuron. The molecular networks approach can be used in three technologies: neurocomputer, neurochip and molecular chip. This means that molecular networks open new fields of science and engineering called molecular-like machines and molecular machines.

Cytoskeleton↗
PubMed2006

Dynamic output feedback stabilization for nonlinear systems based on standard neural network models.

A neural-model-based control design for some nonlinear systems is addressed. The design approach is to approximate the nonlinear systems with neural networks of which the activation functions satisfy the sector conditions. A novel neural network model termed standard neural network model (SNNM) is advanced for describing this class of approximating neural networks. Full-order dynamic output feedback control laws are then designed for the SNNMs with inputs and outputs to stabilize the closed-loop systems. The control design equations are shown to be a set of linear matrix inequalities (LMIs) which can be easily solved by various convex optimization algorithms to determine the control signals. It is shown that most neural-network-based nonlinear systems can be transformed into input-output SNNMs to be stabilization synthesized in a unified way. Finally, some application examples are presented to illustrate the control design procedures.

Algorithms↗
PubMed2001

Prediction of programmed-temperature retention values of naphthas by wavelet neural networks.

The wavelet neural network (WNN) was used to predict the programmed-temperature retention values of naphthas. In WNN, a Morlet mother wavelet was used as a transfer function, and the convergence speed was faster than other neural networks. Sixty-four compounds (selected randomly from 94) were used as a training set, and the 30 remaining compounds were used as a test set. A very satisfactory result was obtained only after about 8000 training epochs. The other two methods, the artificial neural network (ANN) and the Simpson integral method, were also used for this study. The comparison of results obtained from three methods showed that the WNN is the most suitable tool in predicting programmed-temperature retention values of naphthas, consequently this method can be used to provide reliable data for the petrochemical industry.

Journal Article↗
PubMed2003

Prediction of n-octanol/water partition coefficients for polychlorinated dibenzo-p-dioxins using a general regression neural network.

A general regression neural network was used for the first time to study quantitative structure and property relationships of organic pollutants to correlate and predict n-octanol/water partition coefficients of polychlorinated dibenzo- p -dioxins from their topological molecular descriptors. In total, 42 polychlorinated dibenzo- p -dioxins and dibenzo- p -dioxins were available for this study-42 polychlorinated dibenzo- p -dioxins and dibenzo- p -dioxins in the training data set and 41 polychlorinated dibenzo- p -dioxins in the test data set. Partial least squares regression, back propagation network and general regression neural network models were trained using the training data set, and the accuracy of the models obtained were examined by the use of leave-one-out cross-validation. For prediction of the n-octanol/water partition coefficient, the best method is the general regression neural network. With the test data set, the correlation coefficient, root mean square error and mean absolute relative error for the general regression neural network model are 0.9276, 0.22 and 2.79%, respectively. For describing the structure of polychlorinated dibenzo- p -dioxins, the topological molecular descriptors outperform the mobile order and disorder thermodynamic method.

Algorithms↗
PubMed2006

Use of unsupervised and supervised artificial neural networks for the identification of lactic acid bacteria on the basis of SDS-PAGE patterns of whole cell proteins.

Conventional multivariate statistical techniques (hierarchical cluster analysis, linear discriminant analysis) and unsupervised (Kohonen Self Organizing Map) and supervised (Bayesian network) artificial neural networks were compared for as tools for the classification and identification of 352 SDS-PAGE patterns of whole cell proteins of lactic acid bacteria belonging to 22 species of the genera Lactobacillus, Leuconostoc, Enterococcus, Lactococcus and Streptococcus including 47 reference strains. Electrophoretic data were pre-treated using the logistic weighting function described by Piraino et al. [Piraino, P., Ricciardi, A., Lanorte, M. T., Malkhazova, I., Parente, E., 2002. A new procedure for data reduction in electrophoretic fingerprints of whole-cell proteins. Biotechnol. Lett. 24, 1477-1482]. Hierarchical cluster analysis provided a satisfactory classification of the patterns but was unable to discriminate some species (Leuconostoc, Lb. sakei/Lb. curvatus, Lb. acidophilus/Lb. helveticus, Lb. plantarum/Lb. paraplantarum, Lc. lactis/Lc. raffinolactis). A 7x7 Kohonen self-organizing map (KSOM), trained with the patterns of the reference strains, provided a satisfactory classification of the patterns and was able to discriminate more species than hierarchical cluster analysis. The map was used in predictive mode to identify unknown strains and provided results which in 85.5% of cases matched the classification obtained by hierarchical cluster analysis. Two supervised tools, linear discriminant analysis and a 23:5:2 Bayesian network were proven to be highly effective in the discrimination of SDS-PAGE patterns of Lc. lactis from those of other species. We conclude that data reduction by logistic weighting coupled to traditional multivariate statistical analysis or artificial neural networks provide an effective tool for the classification and identification of lactic acid bacteria on the basis of SDS-PAGE patterns of whole cell proteins.

Bacterial Proteins↗
PubMed2003

Prediction of beta-turns in proteins from multiple alignment using neural network.

A neural network-based method has been developed for the prediction of beta-turns in proteins by using multiple sequence alignment. Two feed-forward back-propagation networks with a single hidden layer are used where the first-sequence structure network is trained with the multiple sequence alignment in the form of PSI-BLAST-generated position-specific scoring matrices. The initial predictions from the first network and PSIPRED-predicted secondary structure are used as input to the second structure-structure network to refine the predictions obtained from the first net. A significant improvement in prediction accuracy has been achieved by using evolutionary information contained in the multiple sequence alignment. The final network yields an overall prediction accuracy of 75.5% when tested by sevenfold cross-validation on a set of 426 nonhomologous protein chains. The corresponding Q(pred), Q(obs), and Matthews correlation coefficient values are 49.8%, 72.3%, and 0.43, respectively, and are the best among all the previously published beta-turn prediction methods. The Web server BetaTPred2 (http://www.imtech.res.in/raghava/betatpred2/) has been developed based on this approach.

Algorithms↗
PubMed1993

Evolutionary artificial neural networks.

Evolutionary artificial neural networks (EANNs) can be considered as a combination of artificial neural networks (ANNs) and evolutionary search procedures such as genetic algorithms (GAs). This paper distinguishes among three levels of evolution in EANNs, i.e. the evolution of connection weights, architectures and learning rules. It first reviews each kind of evolution in detail and then analyses major issues related to each kind of evolution. It is shown in the paper that although there is a lot of work on the evolution of connection weights and architectures, research on the evolution of learning rules is still in its early stages. Interactions among different levels of evolution are far from being understood. It is argued in the paper that the evolution of learning rules and its interactions with other levels of evolution play a vital role in EANNs.

Algorithms↗
PubMed2001

Less medical intervention after sharp demarcation of Grade 1-2 cervical intraepithelial neoplasia smears by neural network screening.

BACKGROUND: Neural network technology has been used for the daily screening of cervical smears in The Netherlands since 1992. The authors believe this method might have the potential to demarcate diagnoses of Grade 1-2 cervical intraepithelial neoplasia (CIN 1-2). METHODS: Of 133,196 women who were screened between 1992-1995, there were 2236 CIN 1-2 smears; 1128 of which were detected by means of neural network screening (NNS) (n = 83,404 women) and 1108 of which were diagnosed by conventional screening (n = 49,792 women). Cytologic and clinical outcomes (first cytologic or histologic follow-up diagnosis) were retrieved for all the women in the study population (n = 1920). Stratification based on clinical outcome resulted in the cases being grouped as overdiagnosed, concordant, or underdiagnosed. The smears were performed by general practitioners, whereas the biopsies were obtained by gynecologists. RESULTS: The prevalence rate for CIN 1-2 was 1.15% (95% confidence interval [95% CI], 1.08-1.23%) for NNS and 1.92% (95% CI, 1.80-2.04%) for conventional diagnosis (P < 0.001). Concordance with histology was significantly higher for NNS (53.9%; 95% CI, 50.7-57.0%) compared with conventional screening (29.2%; 95% CI, 26.4-32.2%). In addition, overdiagnosis was significantly lower for cases diagnosed by NNS (39.4%; 95% CI, 36.3-42.4%) compared with cases diagnosed by conventional screening (62.4%; 95% CI, 59.3-65.5%). CONCLUSIONS: Neural network-based screening can lead to fewer women being burdened unnecessarily with a cytologic diagnosis of CIN 1-2 by resulting in a sharp demarcation in these diagnoses and a corresponding reduction in unnecessary medical interventions. [See editorial on pages 171-172, this issue.]

Cytogenetic Analysis↗
PubMedDate not supplied

Recurrent neural networks of integrate-and-fire cells simulating short-term memory and wrist movement tasks derived from continuous dynamic networks.

Dynamic recurrent neural networks composed of units with continuous activation functions provide a powerful tool for simulating a wide range of behaviors, since the requisite interconnections can be readily derived by gradient descent methods. However, it is not clear whether more realistic integrate-and-fire cells with comparable connection weights would perform the same functions. We therefore investigated methods to convert dynamic recurrent neural networks of continuous units into networks with integrate-and-fire cells. The transforms were tested on two recurrent networks derived by backpropagation. The first simulates a short-term memory task with units that mimic neural activity observed in cortex of monkeys performing instructed delay tasks. The network utilizes recurrent connections to generate sustained activity that codes the remembered value of a transient cue. The second network simulates patterns of neural activity observed in monkeys performing a step-tracking task with flexion/extension wrist movements. This more complicated network provides a working model of the interactions between multiple spinal and supraspinal centers controlling motoneurons. Our conversion algorithm replaced each continuous unit with multiple integrate-and-fire cells that interact through delayed "synaptic potentials". Successful transformation depends on obtaining an appropriate fit between the activation function of the continuous units and the input-output relation of the spiking cells. This fit can be achieved by adapting the parameters of the synaptic potentials to replicate the input-output behavior of a standard sigmoidal activation function (shown for the short-term memory network). Alternatively, a customized activation function can be derived from the input-output relation of the spiking cells for a chosen set of parameters (demonstrated for the wrist flexion/extension network). In both cases the resulting networks of spiking cells exhibited activity that replicated the activity of corresponding continuous units. This confirms that the network solutions obtained through backpropagation apply to spiking networks and provides a useful method for deriving recurrent spiking networks performing a wide range of functions.

Animals↗
PubMed1985

An improved neural-network model for the neural integrator of the oculomotor system: more realistic neuron behavior.

The discharge rates of premotor, brain-stem neurons that create eye movements modulate in relation to eye velocity yet firing rates of extraocular motoneurons contain both eye-position and eye-velocity signals. The eye-position signal is derived from the eye-velocity command by means of a neural network which functions as a temporal integrator. We have previously proposed a network of lateral-inhibitory neurons that is capable of performing the required integration. That analysis centered on the temporal aspects of the signal processing for a limited class of idealized inputs. All of its cells were identical and carried only the integrated signal. Recordings in the brain stem, however, show that neurons in the region of the neural integrator have a variety of background firing rates, all carry some eye-velocity signal as well as the eye-position signal, and carry the former with different strengths depending on the type of eye movement being made. It was necessary to see if the proposed model could be modified to make its neurons more realistic. By modifying the spatial distribution of afferents to the network, we demonstrate that the same basic model functions properly in spite of afferents with nonuniform background firing rates. To introduce the eye-velocity signal a double-layer network, consisting of inhibitory and excitatory cells, was necessary. By presenting the velocity input to only local regions of this network it was shown that all cells in the network still carried the integrated signal and that its cells could carry different eye-velocity signals for different types of eye movements. Thus, this model stimulates quantitatively and qualitatively, the behavior of neurons seen in the region of the neural integrator.

Brain Stem↗
PubMed2003

Clustering gene expression pattern and extracting relationship in gene network based on artificial neural networks.

Massive datasets such as gene expression profiles are accumulating along with the development of DNA microarray technologies. In this paper, we focus on mining biological relevant information such as typical expression patterns and the interconnections of gene networks from massive datasets. At first, the algorithm of a self-organizing map (SOM) was used to cluster gene expression data. Then, for the typical patterns extracted by the SOM, a three-layer artificial neural network (ANN) model was used to extract the relationships between the expression patterns. In order to evaluate the clustering analysis based on the SOM, biological and statistical indices were introduced. To validate the efficiency of the scheme proposed for extracting the relationships between the expression patterns with the ANN, a test dataset was created and used for the test. Finally, the interconnections of a typical pattern of early G1, late G1, S, G2, and M phases in a yeast cell cycle were extracted and visualized.

Journal Article↗
PubMed2001

Prediction of the aroma quality and the threshold values of some pyrazines using artificial neural networks.

An artificial neural network is used to predict both the classification of aroma compounds and their flavor impression threshold values for a series of pyrazines. The classification set consists of 98 compounds (32 green, 43 bell-pepper, and 23 nutty smelling pyrazines), and the regression sets consist of 24 green and 37 bell-pepper odorous pyrazines. The best classification of the three aroma impressions (93.7%) is obtained by using a multilayer perceptron network architecture. To predict the threshold values of bell-pepper fragrance, a standard Pearson R correlation coefficient of 0.936 for the training set, 0.912 for the verification set, and 0.926 for the test set is received with two hidden layers consisting of two and one neurons. The network for the threshold prediction of the class of green-smelling pyrazines with one hidden layer containing three neurons turns out to be the best with a standard Pearson R correlation coefficient of 0.859 for the training, 0.918 for the verification, and 0.948 for the test set. These good correlations show that artificial neural networks are versatile tools for the classification of aroma compounds.

Flavoring Agents↗
PubMed1995

Escherichia coli O157:H7 restriction pattern recognition by artificial neural network.

An artificial neural network model for the recognition of Escherichia coli O157:H7 restriction patterns was designed. In the training phase, images of two classes of E. coli isolates (O157:H7 and non-O157:H7) were digitized and transmitted to the neural network. The system was then tested for recognition of images not included in the training set. Promising results were achieved with the designed network configuration, providing a basis for further study. This application of a new generation of computation technology serves as an example of its usefulness in microbiology.

Animals↗