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 1,675 records · Page 93Linked to original sources
PubMed1979

A neural network model for the development of direction selectivity in the visual cortex.

A neural network model is proposed to explain the development of direction selectivity of cortical cells. The model is constructed under the following three hypotheses that are very plausible from recent neurophysiological findings. (1) Direction selectivity is developed by modifiable inhibitory synapses. (2) It results not from the direct convergence of many excitatory inputs from LGN cells but from cortical neural networks. (3) Direction-selective mechanism is independent of orientation-selective mechanism.--The model was simulated on a computer for a few kinds of inhibitory connections and initial conditions. The results were consistent with neurophysiological facts not only for normal cats but for cats reared in an abnormal visual environment.

Animals↗
PubMed1995

Use of a neural network and a multiple regression model to predict histologic grade of astrocytoma from MRI appearances.

Several MRI features of supratentorial astrocytomas are associated with high histologic grade by statistically significant p values. We sought to apply this information prospectively to a group of astrocytomas in the prediction of tumor grade. We used 10 MRI features of fibrillary astrocytomas from 52 patient studies to develop neural network and multiple linear regression models for practical use in predicting tumor grade. The models were tested prospectively on MR images from 29 patients studies. The performance of the models was compared against that of a radiologist. Neural network accuracy was 61% in distinguishing between low and high grade tumors. Multiple linear regression achieved an accuracy of 59%. Assessment of the images by a radiologist yielded 57% accuracy. We conclude that while certain MRI parameters may be statistically related to astrocytoma histologic grade, neural network and linear regression models cannot reliably use them to predict tumor grade.

Astrocytoma↗
PubMed1994

Pattern recognition, chaos, and multiplicity in neural networks of excitable systems.

We study a neural network composed of excitable FitzHugh neurons that interact by diffusive type connections. Patterns of neural activity may be stored by a Hebbian rule. The stored patterns are recalled and given by the transient activity of the neurons after the network has been perturbed by related patterns and relaxes back to its steady state. Periodic perturbations of the network are repeated requests for computations and result in simple periodic, complex periodic, and chaotic responses and corresponding computational performances.

Models, Theoretical↗
PubMed2000

Artificial neural networks for the automated detection of trichloroethylene by passive Fourier transform infrared spectrometry

Artificial neural networks are applied to the automated classification of trichloroethylene (TCE) signatures from passive Fourier transform infrared remote sensing interferogram data. Through the use of three data collection methods, a combination of laboratory and field data is acquired that allows the methodology to be evaluated under a variety of infrared background conditions and in the presence of potentially interfering compounds such as sulfur hexafluoride, methyl ethyl ketone, acetone, carbon tetrachloride, and ammonia. To maximize the computational efficiency of the network optimization, experimental design techniques are employed to develop a training protocol for the network that takes into account the relationships among five variables that are related either to the network architecture or to the training process. This protocol is implemented for the case of a back-propagation neural network (BNN) and is used to develop an optimized network for the detection of TCE. The classification performance of the network is assessed by comparing both TCE detection capabilities and false detection rates to similar classification results obtained with the technique of piecewise linear discriminant analysis (PLDA). When applied to prediction data withheld from the optimization of both the BNN and PLDA algorithms, the BNN method is observed to outperform PLDA overall, with TCE detection rates in excess of 99% and false detection rates less than 0.5%.

Journal Article↗
PubMed1998

Motion deblurring in a neural network model of retino-cortical dynamics.

Simulations of a neural network model of retino-cortical dynamics (Oğmen H, Neural Netw 6 (1993) 245-273) are presented. The temporal-step response of the model to a single dot (spatial impulse) consists of three post-retinal phases: reset, feed-forward dominant and feedback dominant. In response to a single moving dot, the model predicts the perception of extensive blur. This extensive blur is proposed to be due to the relative spatial and temporal offsets between transient and sustained signals conveyed from retina to post-retinal levels. In response to a pair of horizontally separated dots moving in the horizontal direction, the model predicts extensive blur for the trailing dot irrespective of dot-to-dot separation. For the leading dot, the model predicts a decrease in perceived blur for long exposure durations when dot-to-dot separations are small. The reduction of perceived blur at long exposure durations for small dot-to-dot separations is proposed to stem from the spatio-temporal overlap between the transient activity generated by the trailing dot and the sustained activity generated by the leading dot. The model also predicts that targets moving at higher speeds generate more blur even when blur is normalized with respect to speed. The mechanism in the model generating this effect is a slow inhibition within the sustained channel. These predictions are compared with recent psychophysical data (Chen S, Bedell HE, Oğmen H, Vis Res 35 (1995) 2315-2328) and are found to be in excellent agreement. The model is used to offer a coherent explanation for several controversial findings published in the literature. This computational study shows that a model without any motion-compensation mechanism can give a good account of motion deblurring phenomenon and supplements our recent experimental study which provided evidence against motion-compensation type models in explaining the motion deblurring phenomenon (Chen S, Bedell HE, Oğmen H, Vis Res 35 (1995) 2315-2328).

Feedback↗
PubMed1998

Universal approximation in p-mean by neural networks.

A feedforward neural net with d input neurons and with a single hidden layer of n neurons is given byg(x(1), em leader,x(d))= summation operator j=1na(j)sigma,where a(j), theta(j), w(ji) in R. In this paper we study the approximation of arbitrary functions F:R(d)-->R by a neural net in an L(p)(&mgr;) norm for some finite measure &mgr; on R(d). We prove that under natural moment conditions, a neural net with non-polynomial function can approximate any given function.

Journal Article↗
PubMed2004

Adaptive probabilistic neural networks for pattern classification in time-varying environment.

In this paper, we propose a new class of probabilistic neural networks (PNNs) working in nonstationary environment. The novelty is summarized as follows: 1) We formulate the problem of pattern classification in nonstationary environment as the prediction problem and design a probabilistic neural network to classify patterns having time-varying probability distributions. We note that the problem of pattern classification in the nonstationary case is closely connected with the problem of prediction because on the basis of a learning sequence of the length n, a pattern in the moment n + k, k > or = 1 should be classified. 2) We present, for the first time in literature, definitions of optimality of PNNs in time-varying environment. Moreover, we prove that our PNNs asymptotically approach the Bayes-optimal (time-varying) decision surface. 3) We investigate the speed of convergence of constructed PNNs. 4) We design in detail PNNs based on Parzen kernels and multivariate Hermite series.

Algorithms↗
PubMed1995

Artificial neural networks for screening patients needing emergency cranial computed tomography scans in emergency departments.

RATIONALE AND OBJECTIVES: We evaluated the potential for a neural network to screen candidates for emergency cranial computed tomography (CT) scans in an emergency department setting. METHODS: Data were collected from 1625 patients undergoing emergency cranial CT scanning in two different emergency departments (EDs). Singular value decomposition (SVD) was used to remap input data for network training. Data were randomly divided into six subsets, and one was reserved as a test set to analyze network performance. Five networks were then trained on data from the five remaining sets using fivefold cross-validation. Each trained network was allowed an independent vote on need for CT scanning in each case from the test set. The majority vote was used as the final prediction. A similar analysis was done on data from each individual ED. Results are compared with prior statistical studies of the same data. RESULTS: The network performed well when predicting clinical variable patterns that consistently produced negative CT scans and on patterns that were ambiguous in terms of the CT scan results. It performed poorly, however, on patterns that consistently predicted positive scans. This last finding appears to have resulted from inadequate training material. The two populations from which data were taken were shown to be distinct, but a network trained on the combined data performed as well as the networks from the individual EDs in predicting patients requiring CT scanning. Variables with the greatest contribution to the networks' prediction were consistent with those in prior statistical studies. CONCLUSION: Although preliminary in nature, neural networks show promise as a screening device for selecting patients for emergent cranial CT scanning.

Analysis of Variance↗
PubMed2006

Short-term memory for serial order: a recurrent neural network model.

Despite a century of research, the mechanisms underlying short-term or working memory for serial order remain uncertain. Recent theoretical models have converged on a particular account, based on transient associations between independent item and context representations. In the present article, the authors present an alternative model, according to which sequence information is encoded through sustained patterns of activation within a recurrent neural network architecture. As demonstrated through a series of computer simulations, the model provides a parsimonious account for numerous benchmark characteristics of immediate serial recall, including data that have been considered to preclude the application of recurrent neural networks in this domain. Unlike most competing accounts, the model deals naturally with findings concerning the role of background knowledge in serial recall and makes contact with relevant neuroscientific data. Furthermore, the model gives rise to numerous testable predictions that differentiate it from competing theories. Taken together, the results presented indicate that recurrent neural networks may offer a useful framework for understanding short-term memory for serial order.

Artificial Intelligence↗
PubMed1994

Automatic diagnosis of neuro-muscular diseases using neural network.

An automatic diagnostic tool for neuromuscular diseases, based on the feature extraction and classification of myoelectric patterns using neural network is described. Electromyogram (EMG) signals are extracted from the patients during maximal contraction using needle electrodes. This EMG signal is digitized at a rate of 1000 samples/second. The myoelectric signal is divided into many time segments. Five time domain features are extracted from each of these segments and are averaged over the segments to obtain one feature set. This is applied to the neural network for classification. Results are presented for the diagnosis of polymyositis.

Electromyography↗
PubMed2004

Neural network analysis for evaluating cancer risk in thyroid nodules with an indeterminate diagnosis at aspiration cytology: identification of a low-risk subgroup.

Thyroid nodules with a predominant follicular structure are often diagnosed as indeterminate at fine-needle aspiration biopsy (FNAB). We studied 453 patients with a thyroid nodule diagnosed as indeterminate at FNAB by using a feed-forward artificial neural network (ANN) analysis to integrate cytologic and clinical data, with the goal of subgrouping patients into a high-risk and in a low-risk category. Three hundred seventy-one patients were used to train the network and 82 patients were used to validate the model. The cytologic smears were blindly reviewed and classified in a high-risk and a low-risk subgroup on the basis of standard criteria. Neural network analysis subdivided the 371 lesions of the first series into a high-risk group (cancer rate of approximately 33% at histology) and a low-risk group (cancer rate of 3%). Only cytologic parameters contributed to this classification. Analysis of the receiver operating characteristic (ROC) curves demonstrated that the ANN model discriminated with higher sensitivity and specificity between benign and malignant nodules compared to standard cytologic criteria (p < 0.001). This value did not show degradation when ANN predictions were applied to the validation series of 82 nodules. In conclusion, neural network analysis of cytologic data may be a useful tool to refine the risk of cancer in patients with lesions diagnosed as indeterminate by FNAB.

Adult↗
PubMed1993

Peptides secondary structure prediction with neural networks: a criterion for building appropriate learning sets.

Artificial neural networks have been recently applied with success for protein secondary structure prediction. So far, one of the two main aspects on which neural net performance depends, the topology of the net, has been considered. The present work addresses the other main aspect, the building up of the learning set. We present a criterion to build up suitable learning sets based on the alpha-helix percentage. Starting from a set of several well known proteins, we formed 7 groups of proteins with similar helix percentages and we used them for the learning of the same neural net. We found that the best secondary structure prediction for each of the tested proteins (not belonging to the initial set) was the one obtained using the learning set whose helix percentage was closest to that of the tested protein. The accuracy of correct prediction of our method on three types of secondary structure (alpha-helix, beta-sheet and coil), has been compared with the accuracy of other secondary structure prediction methods.

Algorithms↗
PubMed2007

Functional organization within a neural network trained to update target representations across 3-D saccades.

The goal of this study was to understand how neural networks solve the 3-D aspects of updating in the double-saccade task, where subjects make sequential saccades to the remembered locations of two targets. We trained a 3-layer, feed-forward neural network, using back-propagation, to calculate the 3-D motor error the second saccade. Network inputs were a 2-D topographic map of the direction of the second target in retinal coordinates, and 3-D vector representations of initial eye orientation and motor error of the first saccade in head-fixed coordinates. The network learned to account for all 3-D aspects of updating. Hidden-layer units (HLUs) showed retinal-coordinate visual receptive fields that were remapped across the first saccade. Two classes of HLUs emerged from the training, one class primarily implementing the linear aspects of updating using vector subtraction, the second class implementing the eye-orientation-dependent, non-linear aspects of updating. These mechanisms interacted at the unit level through gain-field-like input summations, and through the parallel "tweaking" of optimally-tuned HLU contributions to the output that shifted the overall population output vector to the correct second-saccade motor error. These observations may provide clues for the biological implementation of updating.

Animals↗
PubMed1993

Implementation of the three-dimensional-pattern search problem on Hopfield-like neural networks.

The three-dimensional (3D)-pattern search problem can be summarized as finding, in a molecule, the subset of atoms that have the most similar spatial arrangement as those of a given 3D pattern. For this NP-complete combinatorial optimization problem we propose, by analogy to the travelling salesman problem, a new method taking advantage of the capability of Hopfield-like neural networks to carry out combinatorial optimization of an objective function. This objective function is built from the sum of the differences of interatomic distances in the pattern and the molecule. Here we present the implementation we have found of the 3D-pattern search problem on Hopfield-like neural networks. Initial tests indicate that this approach not only successfully retrieves a given pattern, but can also suggest partial solutions having one or two atoms less than the given pattern, an interesting feature in the case of local conformational flexibility of the molecule. The distributed representation of the problem on Hopfield-like neural networks offers a good perspective for parallel implementation.

Drug Design↗
PubMed2001

Prediction of postoperative prostatic cancer stage on the basis of systematic biopsies using two types of artificial neural networks.

OBJECTIVE: The choice of therapy for prostatic cancer should depend on a rational preoperative estimate of tumor stage. Artificial neural networks were used to predict postoperative staging of prostatic cancer from sextant biopsies and routinely available preoperative data. METHODS: In group I (97 cases), nonorgan confinement (tumor stage > or =pT3a) was predicted on the basis of age and six histopathological variables from sextant biopsies. In group II (77 cases), nonorgan confinement and extraprostatic organ infiltration (tumor classification > or =pT3b) were predicted from age, four histopathological variables, the preoperative PSA level, and the total prostate volume estimated by preoperative ultrasonography. Learning vector quantization (LVQ) networks were applied for this purpose and compared to multilayer perceptrons (MLP) and linear discriminant analysis (LDA). RESULTS: Nonorgan confinement could be predicted correctly in 90% of newly presented cases from sextant biopsy histopathology alone. A similar accuracy of predicting nonorgan confinement (83%) was obtained by combining preoperative biopsy histology with clinical data. Extraprostatic organ infiltration could be predicted correctly in 82%. The best results were obtained by LVQ networks, followed by MLP networks and LDA. CONCLUSION: The postoperative tumor stage of prostatic cancer can be estimated with high accuracy, sensitivity and specificity from preoperative routine parameters using artificial neural networks, especially LVQ networks. The results suggest that this methodology should be evaluated in a larger prospective study.

Aged↗
PubMed1990

[Phonographic diagnostic aid in heart defects using neural networks].

New approaches to pattern recognition and modeling using neural networks also provide powerful tools for the analysis of heart sounds in the diagnosis of heart failure. This paper describes how a neutral network can be used to estimate the duration of systolic and diastolic heart phases as well as to suggest a suitable diagnosis. A neural network-based auscultation system is capable of documenting and analysing a heart sound signal by applying methods similar to those used by physicians in the subjective interpretation of stethoscopic signals.

Diagnosis, Computer-Assisted↗
PubMed1993

Predicting behavior of an enzyme-linked immunoassay model by using commercially available neural network software.

Setting up new immunoassays can be a laborious and expensive task. A relatively new form of multivariate analysis known as neural networks can be applied to this problem with potential savings in reagents and technician time. Neural network software programs for personal computers are now available. We applied one such software package (Brainmaker) to a model ELISA system for measuring human serum albumin. Random combinations of four variable ELISA conditions (antigen concentration, primary and secondary antibody titers, and time for chromagen development) were used to train a three-layered feed-forward network. The trained network was then used to predict measured absorbances as a function of the four input variables in separate cross-validation sets. The network adequately predicted the effect of the input variables on the absorbance produced. With use of such methods, optimal conditions for the linear dependence of absorbance on antigen concentration can be evaluated on the computer rather than in the laboratory, with subsequent savings of time and money.

Antibodies↗
PubMed1997

A simple neural network exhibiting selective activation of neuronal ensembles: from winner-take-all to winners-share-all.

A neuroecological equation of the Lotka-Volterra type for mean firing rate is derived from the conventional membrane dynamics of a neural network with lateral inhibition and self-inhibition. Neural selection mechanisms employed by the competitive neural network receiving external inputs are studied with analytic and numerical calculations. A remarkable findings is that the strength of lateral inhibition relative to that of self-inhibition is crucial for determining the steady states of the network among three qualitatively different types of behavior. Equal strength of both types of inhibitory connections leads the network to the well-known winner-take-all behavior. If, however, the lateral inhibition is weaker than the self-inhibition, a certain number of neurons are activated in the steady states or the number of winners is in general more than one (the winners-share-all behavior). On the other hand, if the self-inhibition is weaker than the lateral one, only one neuron is activated, but the winner is not necessarily the neuron receiving the largest input. It is suggested that our simple network model provides a mathematical basis for understanding neural selection mechanisms.

Algorithms↗