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 1,621 records · Page 90Linked to original sources
PubMed2003

Role of ventilation scintigraphy in diagnosis of acute pulmonary embolism: an evaluation using artificial neural networks.

The purpose of this study was to assess the value of the ventilation study in the diagnosis of acute pulmonary embolism using a new automated method. Either perfusion scintigrams alone or two different combinations of ventilation/perfusion scintigrams were used as the only source of information regarding pulmonary embolism. A completely automated method based on computerised image processing and artificial neural networks was used for the interpretation. Three artificial neural networks were trained for the diagnosis of pulmonary embolism. Each network was trained with 18 automatically obtained features. Three different sets of features originating from three sets of scintigrams were used. One network was trained using features obtained from each set of perfusion scintigrams, including six projections. The second network was trained using features from each set of (joint) ventilation and perfusion studies in six projections. A third network was trained using features from the perfusion study in six projections combined with a single ventilation image from the posterior view. A total of 1,087 scintigrams from patients with suspected pulmonary embolism were used for network training. The test group consisted of 102 patients who had undergone both scintigraphy and pulmonary angiography. Performances in the test group were measured as area under the receiver operation characteristic curve. The performance of the neural network in interpreting perfusion scintigrams alone was 0.79 (95% confidence limits 0.71-0.86). When one ventilation image (posterior view) was added to the perfusion study, the performance was 0.84 (0.77-0.90). This increase was statistically significant ( P=0.022). The performance increased to 0.87 (0.81-0.93) when all perfusion and ventilation images were used, and the increase in performance from 0.79 to 0.87 was also statistically significant ( P=0.016). The automated method presented here for the interpretation of lung scintigrams shows a significant increase in performance when one or all ventilation images are added to the six perfusion images. Thus, the ventilation study has a significant role in the diagnosis of acute lung embolism.

Adolescent↗
PubMedDate not supplied

An efficient learning algorithm for improving generalization performance of radial basis function neural networks.

This paper presents an efficient recursive learning algorithm for improving generalization performance of radial basis function (RBF) neural networks. The approach combines the rival penalized competitive learning (PRCL) [Xu, L., Kizyzak, A. & Oja, E. (1993). Rival penalized competitive learning for clustering analysis, RBF net and curve detection, IEEE Transactions on Neural Networks, 4, 636-649] and the regularized least squares (RLS) to provide an efficient and powerful procedure for constructing a minimal RBF network that generalizes very well. The RPCL selects the number of hidden units of network and adjusts centers, while the RLS constructs the parsimonious network and estimates the connection weights. In the RLS we derived a simple recursive algorithm, which needs no matrix calculation, and so largely reduces the computational cost. This combined algorithm significantly enhances the generalization performance and the real-time capability of the RBF networks. Simulation results of three different problems demonstrate much better generalization performance of the present algorithm over other existing similar algorithms.

Algorithms↗
PubMed2003

Application of artificial neural network algorithm to detection of parathyroid adenoma.

BACKGROUND: The most common radionuclide procedures for parathyroid imaging are (99m)Tc-MIBI/ (99m)Tc pertechnetate subtraction scintigraphy and (99m)Tc-MIBI double-phase imaging, with estimation of MIBI wash-out rate. Those two methods are by some authors regarded as complementary techniques, yielding the best evaluation of parathyroid gland if performed conjointly. By such an approach it seems reasonable to substitute the visual assessment of neck scintigrams and semiquantitative evaluation of MIBI wash-out rate with a single, common procedure. The aim of this study was application of the Artificial Neural Network (ANN) simulated by the computer program to detection and localisation of pathological parathyroid tissue in the planar neck scintigrams. MATERIAL AND METHODS: The applied algorithm was based on simultaneous data processing in sets of 3 single pixels, each of them belonging to one of the three consecutive neck scintigrams generated 20 min. after (99m)TcO(4 )- administration, 10 min. after (99m)Tc-MIBI injection and 120 min. after (99m)Tc-MIBI injection, respectively. Those scintigrams were aligned which each other according to the same vertical and horizontal co-ordinates. The training patterns were obtained from 25 patients by searching for maximum count numbers within small ROIs drawn in selected scintigraphic areas, arbitrarily classified and coded in a numerical scale. In 10 pts the results of ANN simulation were compared with those obtained by common conventional assessment of two radionuclide parathyroid examinations: subtraction method and (99m)Tc-MIBI double-phase imaging. RESULTS: The training patterns processed by the neural network showed a close relationship with the results of visual assessment of original neck scintigrams, with R square coefficient R(2) = 0.717, and standard error equal to 0.243. Similar comparison between original data and results of multidimensional regression analysis yielded weaker relationship, with R(2) = 0.543 and standard error 0.567. Parametric images obtained by the neural network presented regions with homogeneously distributed, relatively high activity, greater than or equal to 750 cts/pixel, visualized in areas of confirmed abnormal parathyroid location. In all 10 patients with suspected parathyroid adenoma results obtained by ANN simulation agreed with those by conventional methods. In five of these cases no parathyroid abnormalities were found. In the remaining 5 subjects results of both approaches were positive but the abnormalities were depicted more distinctly and visualised more clearly in parametric images received by ANN than in original scans. CONCLUSIONS: Application of trained ANN enables objective and quantitative detection and localisation of parathyroid adenoma and is a good alternative for conventional radionuclide imaging procedures used in diagnosing parathyroid abnormality. Including in neural network simulation not only scintigraphic data, but also clinical symptoms and/or some other indicators of parathyroid abnormality, parathormone level first of all, should be a next step in developing a procedure for assessing parathyroid abnormality, of high diagnostic accuracy.

Adenoma↗
PubMed2000

Artificial neural networks in liquid chromatography: efficient and improved quantitative structure-retention relationship models.

The application of the principal neural network architecture, namely the multilayer perceptron (MLP), has been developed for obtaining sufficient quantitative structure-retention relationships (QSRR) with high accuracy. The present study is an extension to the excellent study of Cserhati et al. [LC-GC Int., 11 (1998) 240] for the retention behavior of solutes based on their structure. To this end, a dataset of 25 substances as solutes to two different stationary phases (polyethylene-silica and polyethylene-alumina) were analyzed to their structural descriptors and related to their retention behavior as expressed by the logarithms of their capacity factors (log k'). The results were compared to those of Cserhati et al. who studied the same problem using as many as ten different equations based on multiple regression analysis. In the present study a series of new and improved algorithms other than the 'old-fashioned' and problematic steepest descent were examined for training the MLP networks. The proposed methods led to substantial gain in both the prediction ability and the computation speed of the resulting models. For the development and evaluation of the artificial neural network (ANN) systems the same (eight) descriptors proposed by Cserhati were used also in this study. Furthermore, the results were compared to those produced from classical linear multivariate regression such as partial least squares regression (PLS). Some of the proposed ANN models diminished the number of outliers, during their implementation to unseen data (solutes), to zero.

Algorithms↗
PubMed2006

A Hopfield neural network for image change detection.

This paper outlines an optimization relaxation approach based on the analog Hopfield neural network (HNN) for solving the image change detection problem between two images. A difference image is obtained by subtracting pixel by pixel both images. The network topology is built so that each pixel in the difference image is a node in the network. Each node is characterized by its state, which determines if a pixel has changed. An energy function is derived, so that the network converges to stable states. The analog Hopfield's model allows each node to take on analog state values. Unlike most widely used approaches, where binary labels (changed/unchanged) are assigned to each pixel, the analog property provides the strength of the change. The main contribution of this paper is reflected in the customization of the analog Hopfield neural network to derive an automatic image change detection approach. When a pixel is being processed, some existing image change detection procedures consider only interpixel relations on its neighborhood. The main drawback of such approaches is the labeling of this pixel as changed or unchanged according to the information supplied by its neighbors, where its own information is ignored. The Hopfield model overcomes this drawback and for each pixel allows a tradeoff between the influence of its neighborhood and its own criterion. This is mapped under the energy function to be minimized. The performance of the proposed method is illustrated by comparative analysis against some existing image change detection methods.

Algorithms↗
PubMed1995

Utilization of neural network in the elaboration of an evaluation scale for pain in cerebral palsy.

An interesting aspect of neural networks is shown in the elaboration of an evaluation scale for pain in cerebral palsy with severe mental retardation. Because of the diversity of cases, the number of items had to be limited in the final step of statistical validation. Classical analysis on prior data did not allow to decide whether the variability in results is more likely due to the type of disability (i.e., the possibility of pain expression) than to the actual presence of pain. A neural network was used to find implicit relations between the data, with the advantage of having total control on the variables' status by applying variations in the network architecture. This allowed for the rapid identification more significant item combinations as a function of degree of relationship to pain in cerebral palsy.

Adolescent↗
PubMed2005

Analysis of motor control and behavior in multi agent systems by means of artificial neural networks.

This article gives a short introduction to Self-Organizing Maps, a particular form of Artificial Neural Networks and shows by some examples, how these approaches can be used in order to analyze and visualize time series data originating from complex systems. The methods shown in this article have originally been developed for the analysis of RoboCup robot soccer games, a special kind of so-called Multi Agent Systems. Although this application has no direct connection to biomechanics, the examples shown here may give an impression of the abilities of Neural Networks in the field of Time Series Analysis in general. Because of the abstractness of the methods, it appears to be very likely that they can easily be adapted to Time Series Analysis problems within the biomechanics context.

Algorithms↗
PubMed1999

Discontinuities in recurrent neural networks.

This article studies the computational power of various discontinuous real computational models that are based on the classical analog recurrent neural network (ARNN). This ARNN consists of finite number of neurons; each neuron computes a polynomial net function and a sigmoid-like continuous activation function. We introduce arithmetic networks as ARNN augmented with a few simple discontinuous (e.g., threshold or zero test) neurons. We argue that even with weights restricted to polynomial time computable reals, arithmetic networks are able to compute arbitrarily complex recursive functions. We identify many types of neural networks that are at least as powerful as arithmetic nets, some of which are not in fact discontinuous, but they boost other arithmetic operations in the net function (e.g., neurons that can use divisions and polynomial net functions inside sigmoid-like continuous activation functions). These arithmetic networks are equivalent to the Blum-Shub-Smale model, when the latter is restricted to a bounded number of registers. With respect to implementation on digital computers, we show that arithmetic networks with rational weights can be simulated with exponential precision, but even with polynomial-time computable real weights, arithmetic networks are not subject to any fixed precision bounds. This is in contrast with the ARNN that are known to demand precision that is linear in the computation time. When nontrivial periodic functions (e.g., fractional part, sine, tangent) are added to arithmetic networks, the resulting networks are computationally equivalent to a massively parallel machine. Thus, these highly discontinuous networks can solve the presumably intractable class of PSPACE-complete problems in polynomial time.

Algorithms↗
PubMed2006

From gene networks underlying sex determination and gonadal differentiation to the development of neural networks regulating sociosexual behavior.

Genes are not expressed in isolation any more than social behavior has meaning outside of society. Both are in dynamic flux with the immediate environment that the gene/individual finds itself, which in turn establishes the timing, pattern, and conditions of expression. This means that complex behaviors and their genetic underpinnings should be viewed as a cumulative process, or as the result of experiences up to that point in time and, at the same time, as setting the stage for what will follow. The evidence indicates that as experiences accumulate throughout life, early experiences shape how genes/individuals will respond to later experiences, whereas later experiences modify the effects of these earlier experiences. A method of graphically representing and analyzing change in gene and neural networks is presented. Results from several animal model systems will be described to illustrate these methods. First, we will consider the phenomenon of temperature-dependent sex determination in reptiles. We will illustrate how the experience of a particular temperature during a sensitive period of embryogenesis sculpts not only the patterns of expression of genes involved in sex determination and gonadal differentiation but also the morphological, physiological, neuroendocrine, and behavioral traits of the adult phenotype. The second model system concerns the effects of the sex ratio in the litter in rats, and the genotype ratio in the litter of transgenic mice, on the nature and frequency of maternal care and how this in turn influences the patterns of activation of identified neural circuits subserving the offspring's sociosexual behavior when it is an adult.

Animals↗
PubMed2000

Novel approach to acoustical voice analysis using artificial neural networks.

Perceptual rating scales are widely used for the assessment of voice quality. These ratings may be influenced by the individual experience of the listener. Thus, researchers have turned to acoustical measures which may eventually correlate with voice quality. In this study we tested whether multivariate statistics, combined with artificial neural networks, could identify patterns of acoustic voice parameters corresponding to a widely used perceptual rating scale. In a multicenter study with 31 raters, voice samples of 117 individuals with or without voice disorders were perceptually rated. The RBH index, consisting of a 4-point scale of roughness, breathiness, and hoarseness, was used. Voice samples were then analyzed with an acoustical feature extraction and classified using amultivariate regression tree analysis with the perceptual ratings as a priori information. Artificial neural networks were trained to selected acoustic parameters having high "relative importance" in the regression trees. Mean classification accuracies were around 30% with topographic feature maps (trained with Learning Vector Quantization algorithm) and 65-85% with feedforward networks (trained with RProp algorithm). Based on the best-fitting results with feedforward networks, a classification system (computer program) consisting of 50 simultaneous working networks was developed. Using this program, the classification matched 40% of the a prori values in both R and B domains. In 65% they matched at least in one domain. These accuracies are within the range reported by other authors using artificial neural networks in biology and clinical medicine. Thus, the results encourage further research of feedforward networks for acoustic voice analysis.

Humans↗
PubMed2005

Comparative effectiveness of total population versus disease-specific neural network models in predicting medical costs.

The objective of this research was to compare the accuracy of two types of neural networks in identifying individuals at risk for high medical costs for three chronic conditions. Two neural network models-a population model and three disease-specific models-were compared regarding effectiveness predicting high costs. Subjects included 33,908 health plan members with diabetes, 19,264 with asthma, and 2,605 with cardiac conditions. For model development/ testing, only members with 24 months of continuous enrollment were included. Models were developed to predict probability of high costs in 2000 (top 15% of distribution) based on 1999 claims factors. After validation, models were applied to 2000 claims factors to predict probability of high 2001 costs. Each member received two scores-population model score applied to cohort and disease model score. Receiver Operating Characteristic (ROC) curves compared sensitivity, specificity, and total performance of population model and three disease models. Diabetes-specific model accuracy, C = 0.786 (95%CI = 0.779-0.794), was greater than that of population model applied to diabetic cohort, C = 0.767 (0.759-0.775). Asthma-specific model accuracy, C = 0.835 (0.825-0.844), was no different from that of population model applied to asthma cohort, C = 0.844 (0.835-0.853). Cardiac-specific model accuracy, C = 0.651 (0.620-0.683), was lower than that of population model applied to cardiac cohort, C = 0.726 (0.697-0.756). The population model predictive power, compared to the disease model predictive power, varied by disease; in general, the larger the cohort, the greater the advantage in predictive power of the disease model compared to the population model. Given these findings, disease management program staff should test multiple approaches before implementing predictive models.

Age Factors↗
PubMed2003

Identifying multiple abdominal organs from CT image series using a multimodule contextual neural network and spatial fuzzy rules.

Identifying abdominal organs is one of the essential steps in visualizing organ structure to assist in teaching, clinical training, diagnosis, and medical image retrieval. However, due to partial volume effects, gray-level similarities of adjacent organs, contrast media affect, and the relatively high variations of organ position and shape, automatically identifying abdominal organs has always been a high challenging task. To conquer these difficulties, this paper proposes combining a multimodule contextual neural network and spatial fuzzy rules and fuzzy descriptors for automatically identifying abdominal organs from a series of CT image slices. The multimodule contextual neural network segments each image slice through a divide-and-conquer concept, embedded within multiple neural network modules, where the results obtained from each module are forwarded to other modules for integration, in which contextual constraints are enforced. With this approach, the difficulties arising from partial volume effects, gray-level similarities of adjacent organs, and contrast media affect can be reduced to the extreme. To address the issue of high variations in organ position and shape, spatial fuzzy rules and fuzzy descriptors are adopted, along with a contour modification scheme implementing consecutive organ region overlap constraints. This approach has been tested on 40 sets of abdominal CT images, where each set consists of about 40 image slices. We have found that 99% of the organ regions in the test images are correctly identified as its belonging organs, implying the high promise of the proposed method.

Algorithms↗
PubMed2006

Comparing the predictive value of neural network models to logistic regression models on the risk of death for small-cell lung cancer patients.

Cancer is one of the leading causes of mortality in the developed world, and prognostic assessment of cancer patients is indispensable in medical care. Medical researchers are accustomed to using regression models to predict patient outcomes. Neural networks have been proposed as an alternative with great potential. Nonetheless, empirical evidence remains lacking to support the application of this technique as the appropriate method to investigate cancer prognosis. Utilizing data on patients from two National Cancer Institute of Canada clinical trials, we compared predictive accuracy of neural network models and logistic regression models on risk of death of limited-stage small-cell lung cancer patients. Our results suggest that neural network and logistic regression models have similar predictive accuracy. The distributions of individual predicted probabilities are very similar. On occasion, however, the prediction pairs are quite different, suggesting that they do not always give the same interpretations of the same variables.

Adult↗
PubMed1995

Non-invasive identification of gastric contractions from surface electrogastrogram using back-propagation neural networks.

Gastric contractions play an important role in the digestive process of the stomach. The established method for the measurement of gastric contractions is invasive and involves the insertion through the nose of a manometric probe into the stomach. A non-invasive method is introduced in this paper for the identification of gastric contractions using the surface electrogastrogram. The electrogastrogram (EGG) was measured by placing surface electrodes on the abdominal skin over the stomach in ten subjects. Gastric contractions were simultaneously monitored using an intraluminal manometric probe. The back-propagation neural network was applied to identify gastric contractions from the EGG. The input of the neural network was composed of spectral data points of the EGG which was computed using the exponential distribution method. Experiments were conducted to optimize network structures and parameters. Using the EGG data in five subjects as the training set and the EGG data in another five subjects as the testing set, an overall accuracy of 92% was achieved in the identification of gastric contractions with an optimized three-layer back-propagation neural network (number of nodes for input:hidden:output layers being 64:10:2).

Adult↗
PubMed2005

Calibration of a Hall effect displacement measurement system for complex motion analysis using a neural network.

Biomechanics studies often require the analysis of position and orientation. Although a variety of transducer and camera systems can be utilized, a common inexpensive alternative is the Hall effect sensor. Hall effect sensors have been used extensively for one-dimensional position analysis but their non-linear behavior and cross-talk effects make them difficult to calibrate for effective and accurate two- and three-dimensional position and orientation analysis. The aim of this study was to develop and calibrate a displacement measurement system for a hydraulic-actuation joystick used for repetitive motion analysis of heavy equipment operators. The system utilizes an array of four Hall effect sensors that are all active during any joystick movement. This built-in redundancy allows the calibration to utilize fully connected feed forward neural networks in conjunction with a Microscribe 3D digitizer. A fully connected feed forward neural network with one hidden layer containing five neurons was developed. Results indicate that the ability of the neural network to accurately predict the x, y and z coordinates of the joystick handle was good with r(2) values of 0.98 and higher. The calibration technique was found to be equally as accurate when used on data collected 5 days after the initial calibration, indicating the system is robust and stable enough to not require calibration every time the joystick is used. This calibration system allowed an infinite number of joystick orientations and positions to be found within the range of joystick motion.

Algorithms↗
PubMed2005

Artificial neural network modeling of peptide mobility and peptide mapping in capillary zone electrophoresis.

Recently, we have developed an artificial neural network model, which was able to predict accurately the electrophoretic mobilities of relatively small peptides. To examine the robustness of this methodology, a 3-3-1 back-propagation artificial neural network (BP-ANN) model was developed using the same inputs as the previous model, which were the Offord's charge over mass term (Q/M(2/3)), corrected steric substituent constant (E(s,c)) and molar refractivity (MR). The data set consisted of 102 peptides with a larger range of size than that of our earlier report - up to 42 amino acid residues as compared to 13 amino acids in the initial study - that also included highly charged and hydrophobic peptides. The entire data set was obtained from the published result by Janini and co-workers. The results of this model are compared with those obtained using multiple linear regressions (MLR) model developed in this work and the multi-variable model released by Janini et al. Better predictive ability of the BP-ANN model over the MLR indicates the non-linear characteristics of the electrophoretic mobility of peptides. The present model exhibits better robustness than the MLR models in predicting CZE mobilities of a diverse data set at different experimental conditions. To explore the utility of the ANN model in simulation of the CZE peptide maps, the profiles for the endoproteinase digests of melittin, glucagon and horse cytochrome C is studied in the present work.

Amino Acid Sequence↗
PubMed1999

Applications of neural network analyses to in vivo 1H magnetic resonance spectroscopy of epilepsy patients.

A total of 67 in vivo water-suppressed proton magnetic resonance spectra of the temporal lobes were recorded from 15 patients with long standing temporal lobe epilepsy and 13 healthy volunteers. Conventional data analysis indicated slightly lower N-acetyl aspartate levels in epileptic patients compared with controls. For further analysis of data, a spectral region (4.0-1.5 ppm) was used as input for artificial neural network analysis. Correct classification of spectra was obtained in 66 out of 67 cases, disregarding from which side of the brain the spectra were recorded. The ability of the trained network to recognize spectra recorded both contalaterally and ipsilaterally to the epileptic focus strongly indicates bilateral metabolic changes. Artificial neural networks could also be trained to recognize whether the spectra were recorded from the ipsilateral or contralateral side of the epileptic focus, indicating that neural network analysis of in vivo proton MR spectra can be used as an additional tool for pre-surgical lateralization of seizure foci.

Adolescent↗
PubMed2000

Artificial neural networks compared to factor analysis for low-dimensional classification of high-dimensional body fat topography data of healthy and diabetic subjects.

Subcutaneous adipose tissue thickness was measured in 590 healthy subjects at 15 specific body sites by means of the new optical device, lipometer, providing a high-dimensional and partly highly intercorrelated set of data, which had been analyzed by factor analysis previously. N-2-N back-propagation neural networks are able to perform low-dimensional display of high-dimensional data as a special application. We report about the performance of such a 15-2-15 network and compare its results with the output of factor analysis. As test data for verification, measurement values on women with proven diabetes mellitus type II (NIDDM) are used. Surprisingly our 15-2-15 neural network is able to reproduce the classification pattern resulting from factor analysis very precisely. After extracting the network weights the classification of new subjects is even more simple with the neural network as compared with factor analysis. In addition, the network weights are able to cluster highly correlated body sites nicely to different groups, corresponding to different regions of the human body. Thus, the analysis of these weights provides additional information about the structure of the data. Therefore, N-2-N networks seem to be a good alternative method for analyzing high-dimensional data with strong intercorrelation.

Adipose Tissue↗