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 919 records · Page 51Linked to original sources

PubMed2005

Scale-freeness and biological networks.

The notion of scale-freeness and its prevalence in both natural and artificial networks have recently attracted much attention. The concept of scale-freeness is enthusiastically applied to almost any conceivable network, usually with affirmative conclusions. Well-known scale-free examples include the internet, electric lines among power plants, the co-starring of movie actors, the co-authorship of researchers, food webs, and neural, protein-protein interactional, genetic, and metabolic networks. The purpose of this review is to clarify the relationship between scale-freeness and power-law distribution, and to assess critically the previous related works, especially on biological networks. In addition, I will focus on the close relationship between power-law distribution and lognormal distribution to show that power-law distribution is not a special characteristic of natural selection.

Computational Biology↗

PubMed2001

Topological properties of citation and metabolic networks.

Topological properties of "scale-free" networks are investigated by determining their spectral dimensions d(S), which reflect a diffusion process in the corresponding graphs. Data bases for citation networks and metabolic networks together with simulation results from the growing network model [A.-L. Barabasi and R. Albert, Science 286, 509 (1999)] are probed. For completeness and comparisons lattice, random and small-world models are also investigated. We find that d(S) is around 3 for citation and metabolic networks, which is significantly different from the growing network model, for which d(S) is approximately 7.5. This signals a substantial difference in network topology despite the observed similarities in vertex-order distributions. In addition, the diffusion analysis indicates that the citation networks are treelike in structure, whereas the metabolic networks contain many loops.

Animals↗

PubMed2001

A novel large-memory neural network as an aid in medical diagnosis applications.

This paper describes the application of a large memory storage and retrieval (LAMSTAR) neural network to medical diagnosis and medical information retrieval problems. The network is based on Minsky's knowledge-lines (k-lines) theory of memory storage and retrieval in the central nervous system. It employs arrays of self-organized map modules, such that the k-lines are implemented via link weights (address correlation) that are being updated by learning. The network also employs features of forgetting and of interpolation and extrapolation, thus being able to handle incomplete data sets. It can deal equally well with exact and fuzzy information, thus being specifically applicable to medical diagnosis where the diagnosis is based on exact data, fuzzy patient interview information, patient history, observed images, and test records. Furthermore, the network can be operated in closed loop with Internet search engines to intelligently use data from the Internet in a higher hierarchy of learning. All of the above features are shown to make the LAMSTAR network suitable for medical diagnosis problems that concern large data sets of many categories that are often incomplete and fuzzy. Applications of the network to three specific medical diagnosis problems are described: two from nephrology and one related to an emergency-room drug identification problem. It is shown that the LAMSTAR network is hundreds and thousands times faster in its training than back-propagation-based networks when used for the same problem and with exactly the same information.

Computational Biology↗

PubMedDate not supplied

Discovering gene networks with a neural-genetic hybrid.

Recent advances in biology (namely, DNA arrays) allow an unprecedented view of the biochemical mechanisms contained within a cell. However, this technology raises new challenges for computer scientists and biologists alike, as the data created by these arrays is often highly complex. One of the challenges is the elucidation of the regulatory connections and interactions between genes, proteins and other gene products. In this paper, a novel method is described for determining gene interactions in temporal gene expression data using genetic algorithms combined with a neural network component. Experiments conducted on real-world temporal gene expression data sets confirm that the approach is capable of finding gene networks that fit the data. A further repeated approach shows that those genes significantly involved in interaction with other genes can be highlighted and hypothetical gene networks and circuits proposed for further laboratory testing.

Algorithms↗

PubMed2005

A hybrid neural network system for pattern classification tasks with missing features.

A hybrid neural network comprising Fuzzy ARTMAP and Fuzzy C-Means Clustering is proposed for pattern classification with incomplete training and test data. Two benchmark problems and a real medical pattern classification task are employed to evaluate the effectiveness of the hybrid network. The results are analyzed and compared with those from other methods.

Acute Disease↗

PubMed2006

Recent progresses in the application of machine learning approach for predicting protein functional class independent of sequence similarity.

Protein sequence contains clues to its function. Functional prediction from sequence presents a challenge particularly for proteins that have low or no sequence similarity to proteins of known function. Recently, machine learning methods have been explored for predicting functional class of proteins from sequence-derived properties independent of sequence similarity, which showed promising potential for low- and non-homologous proteins. These methods can thus be explored as potential tools to complement alignment- and clustering-based methods for predicting protein function. This article reviews the strategies, current progresses, and underlying difficulties in using machine learning methods for predicting the functional class of proteins. The relevant software and web-servers are described. The reported prediction performances in the application of these methods are also presented, which need to be interpreted with caution as they are dependent on such factors as datasets used and choice of parameters.

Algorithms↗

PubMed1999

Automatic segmentation of lung fields on chest radiographic images.

In this work we have implemented a system for the automatic segmentation of lung fields in chest radiographic images. The image analysis process is carried out in three levels. In the first one we perform operations on the image that are independent from domain knowledge. This knowledge is implicitly and not very elaborately used in the intermediate level and used in an explicit manner in the high level block, globally corresponding to the idea of progressive segmentation. The representation of knowledge in the high level block is in the form of production rules. The control structure is in general bottom-up but there are certain hybrid control stages, in which the control is driven by the region model (main organs) we are seeking. We have applied the global system to a set of 45 posteroanterior (PA) chest radiographs, obtaining a mean degree of overlap with contours drawn by radiologists of 87%.

Algorithms↗

PubMed2006

Computational model of response maps in the dorsal cochlear nucleus.

The neurons in the mammalian (gerbil, cat) dorsal cochlear nucleus (DCN) have responses to tones and noise that have been used to classify them into unit types. These types (I-V) are based on excitatory and inhibitory responses to tones organized into plots called response maps (RMs). Type I units show purely excitatory responses, while type V units are primarily inhibited. A computational model of the neural circuitry of the mammalian DCN, based on the MacGregor neuromime, was used to investigate RMs of the principal cells (P-cells) that represent the fusiform and giant cells. In gerbils, fusiform cells have been shown to have primarily type III unit response properties; however, fusiform cells in the cat DCN are thought to have type IV unit response properties. The DCN model is based on a previous computational model of the cat (Hancock and Voigt Ann Biomed Eng 27: 73-87, 1999) and gerbil (Zheng and Voigt Ann Biomed Eng 34: 697-708, 2006) DCN. The basic model for both species is architecturally the same, and to get either type III unit RMs or type IV unit RMs, connection parameters were adjusted. Interestingly, regardless of the RM type, these units in gerbils and cats show spectral notch sensitivity and are thought to play a role in sound localization in the median plane. In this study, further parameter adjustments were made to systematically explore their effect on P-cell RMs. Significantly, type I, type III, type III-i, type IV, type IV-T and type V unit RMs can be created for the modeled P-cells. Thus major RMs observed in the cat and gerbil DCN are recreated by the model. These results suggest that RMs of individual DCN projection neurons are the result of specific assortment of excitatory and inhibitory inputs to that neuron and that subtle differences in the complement of inputs can result in different RM types. Modulation of the efficacy of certain synapses suggests that RM type may change dynamically.

Animals↗

PubMed2001

Converting neural signals from place codes to rate codes.

The nervous system uses two basic types of formats for encoding information. The parameters of many sensory (and some premotor) signals are represented by the pattern of activity among an array of neurons each of which is optimally responsive to a different parameter value. This type of code is commonly referred to as a place code. Motor commands, in contrast, use rate coding: the desired force of a muscle is specified as a monotonic function of the aggregate rate of discharge across all of its motor neurons. Generating movements based on sensory information often requires converting signals from a place code to a rate code. In this paper I discuss three possible models for how the brain does this.

Animals↗

PubMed2002

Computer-aided detection and diagnosis at the start of the third millennium.

Computer-aided diagnosis has been under development for more than 3 decades. The rate of progress appears exponential, with either recent approval or pending approval for devices focusing on mammography, chest radiographs, and chest CT. Related technologies improve diagnosis for many other types of medical images including virtual colonography, vascular imaging, as well as automated quantitation of image-derived metrics. A variety of techniques are currently employed with success, likely reflecting the variety of imagery used, as well as the variety of tasks. Most areas of medical imaging have had efforts at computer assistance, and some have even received FDA approval and can be reimbursed. We anticipate that the rapid advance of these technologies will continue, and that application will broaden to cover much of medical imaging. Acceptance of, and integration of computer-aided diagnosis technology with the electronic radiology practice is a current challenge. These challenges will be overcome, and we expect that computer-aided diagnosis will be routinely applied to medical images.

Colonography, Computed Tomographic↗

PubMed2004

Avoiding overfitting in multilayer perceptrons with feeling-of-knowing using self-organizing maps.

Overfitting in multilayer perceptron (MLP) training is a serious problem. The purpose of this study is to avoid overfitting in on-line learning. To overcome the overfitting problem, we have investigated feeling-of-knowing (FOK) using self-organizing maps (SOMs). We propose MLPs with FOK using the SOMs method to overcome the overfitting problem. In this method, the learning process advances according to the degree of FOK calculated using SOMs. The mean square error obtained for the test set using the proposed method is significantly less than that in a conventional MLP method. Consequently, the proposed method avoids overfitting.

Algorithms↗

PubMed2004

A [17F]-fluoromethane PET/TMS study of effective connectivity.

We used transcranial magnetic stimulation (TMS) in combination with positron emission tomography (PET) to investigate the effective connectivity of four cortical regions within the same study. By employing [17F]-[CH3F] ([17F]-fluoromethane) as a radiotracer of blood-flow, we were able to obtain increased sensitivity compared to [15O]-H2O for both cortical and subcortical structures. The brain areas investigated were left primary motor cortex, right primary visual cortex, and left and right prefrontal areas. We found that each site of stimulation yielded a different pattern of activation/deactivation consistent with its anatomical connectivity. Moreover, we found that TMS of prefrontal and motor cortical areas gave rise to trans-synaptic activation of subcortical circuits.

Adult↗

PubMed2006

Origin, structure, and role of background EEG activity. Part 4: Neural frame simulation.

OBJECTIVE: To develop a method for simulating background EEG based on the premise that the self-organized activity from synaptic interaction among populations of neurons creates sustained fluctuations that can be modeled with the filtered output of a random number generator. METHODS: The logarithm of the amplitude of activity was weighted in accordance with 1/f, the log frequency in both temporal (PSD(T)) and spatial (PSD(X)) power spectral densities. The activity was spatially smoothed by volume conduction. Further deviation from full randomness was by sustained spatial coherence averaging 25% of total power. The departure from the background state to an active state, as seen in the awake EEG, was simulated by adding segments that were 90% correlated while attenuating by 50% the uncorrelated background activity in those segments. Spatial amplitude modulation was imposed on the correlated noise to create signals that simulated AM patterns. RESULTS: The statistical properties of the EEG that were replicated (Freeman, 2004a,b, 2005) included the PSD(T), PSD(X), point spread function (PSF), partitioning of the variance with PCA, and the percentages of correct classification of AM patterns. CONCLUSIONS: The origin of background EEG was traced to self-sustaining mutual excitation among pyramidal cells creating stable noise that was filtered by self-organized criticality to give 1/f(2) PSD, by inhibitory feedback to give oscillations in the classic clinical bands, and by volume conduction to give smoothing. The essential change that identified a frame in EEG was transient synchrony by phase transition among cortical populations in beta and gamma bands of the PSD(T). SIGNIFICANCE: This simulation can provide test data with which to optimize techniques for noninvasively extracting information from the EEG for diagnosis and treatment evaluation of neuropsychiatric disorders and for operation by paraplegics of prosthetic devices.

Brain↗

PubMedDate not supplied

Anterior osteophyte discrimination in lumbar vertebrae using size-invariant features.

Radiologists often examine X-rays of cervical, thoracic and lumbar vertebrae for determining the presence of osteoarthritis and osteoporosis. For individual vertebra assessment, the boundary increasingly digresses from the general rectangular shape as the vertebra becomes less normal in appearance. For an abnormal vertebra, bony growths ('osteophytes') may appear at the vertebral comers, resulting in a change in the vertebra's shape. Image processing techniques are presented for computing size-invariant, convex hull-based features to highlight anterior osteophytes. Feature evaluation of 714 lumbar spine vertebrae using a multi-layer perceptron yielded normal and abnormal average correct discrimination of 90.5 and 86.6%, respectively.

Algorithms↗

PubMed2005

Vesicular roundness and compound release in PC-12 cells.

The principal goals of this study were to establish a quantitative morphological analysis of spatial and regional properties of dense core vesicles, and to use this analysis to assess whether homotypic fusion is prominent in chronically treated PC-12 cells at elevated release levels. Simple computerized image processing of electron-micrographs provided the binary images of vesicular dense cores, whilst the artificial intelligence methods were needed to determine the vesicular membranes. As in the past, the presence of large, highly irregular vesicles, provided the morphological evidence of fused vesicles, but the irregularity of vesicular shape was assessed quantitatively-from its roundness. Free space of each vesicle was determined from the distance to its nearest-neighbor, or from the size of its Voronoi polygon. Within a Voronoi polygon, each point is closer to that vesicle than to any other vesicle. Large vesicles were not less round and did not have larger free space, as expected if they result from fusion of several smaller vesicles. In conclusion, we present a novel and rigorous morphological analysis of spatial and regional properties of dense core vesicles. The results demonstrate that the homotypic fusion is not prominent in PC-12 cells, before or following a chronic treatment that enhances release.

Algorithms↗

PubMed2005

Brain-behaviour relationships in latent inhibition: a computational model.

In a series of studies, we applied a neural network to study the neural bases of latent inhibition. We first designed a model able to handle behavioral data and then we investigated whether structures and neural elements in the brain were able to carry out the operations described by network. We demonstrated that the network was able to describe many of the behavioral properties of LI, and elucidate the effects of several manipulations of the dopaminergic system, the hippocampus, and the nucleus accumbens on LI, as well as some of the positive symptoms of schizophrenia. The results support the idea that a 'conceptual nervous system' can be successfully used to relate brain and behavior.

Animals↗

PubMedDate not supplied

Roles of visual experience and intrinsic mechanism in the activity-dependent self-organization of orientation maps: theory and experiment.

It is widely accepted that functional maps in the mammalian visual cortex such as ocular dominance columns and orientation columns are formed depending on neural activity. There is still, however, controversy on how much visual experience contributes to the map formation during development. In the present study, we address this issue from mathematical modeling and experimental investigation. Using a model of activity-dependent self-organization of geniculo-cortical afferent inputs, we showed that spontaneous activity in the LGN can produce orientation maps, while the exposure to drifting gratings results in sharply segregated orientation maps as observed in cat visual cortex. The restricted exposure to a single orientation of the grating led to the over-representation of the exposed orientation, which was moderated by the contribution of learning based on the spontaneous activity. These theoretical results were confirmed by intrinsic optical recordings from area 18 of kittens reared under various visual conditions.

Animals↗

PubMedDate not supplied

SOM-based algorithms for qualitative variables.

It is well known that the SOM algorithm achieves a clustering of data which can be interpreted as an extension of Principal Component Analysis, because of its topology-preserving property. But the SOM algorithm can only process real-valued data. In previous papers, we have proposed several methods based on the SOM algorithm to analyze categorical data, which is the case in survey data. In this paper, we present these methods in a unified manner. The first one (Kohonen Multiple Correspondence Analysis, KMCA) deals only with the modalities, while the two others (Kohonen Multiple Correspondence Analysis with individuals, KMCA_ind, Kohonen algorithm on DISJonctive table, KDISJ) can take into account the individuals, and the modalities simultaneously.

Algorithms↗