Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Parallel Algorithms”

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 1,711 records · Page 95Linked to original sources
PubMed1997

An Extension of the Back-Propagation Algorithm to Complex Numbers.

This paper presents a complex-valued version of the back-propagation algorithm (called 'Complex-BP'), which can be applied to multi-layered neural networks whose weights, threshold values, input and output signals are all complex numbers. Some inherent properties of this new algorithm are studied. The results may be summarized as follows. The updating rule of the Complex-BP is such that the probability for a "standstill in learning" is reduced. The average convergence speed is superior to that of the real-valued back-propagation, whereas the generalization performance remains unchanged. In addition, the number of weights and thresholds needed is only about the half of real-valued back-propagation, where a complex-valued parameter z=x+iy (where i=-1) is counted as two because it consists of a real part x and an imaginary part y. The Complex-BP can transform geometric figures, e.g. rotation, similarity transformation and parallel displacement of straight lines, circles, etc., whereas the real-valued back-propagation cannot. Mathematical analysis indicates that a Complex-BP network which has learned a transformation, has the ability to generalize that transformation with an error which is represented by the sine. It is interesting that the above characteristics appear only by extending neural networks to complex numbers.

Journal Article↗
PubMed2005

Fully 3D Monte Carlo reconstruction in SPECT: a feasibility study.

In single photon emission computed tomography (SPECT) with parallel hole collimation, image reconstruction is usually performed as a set of bidimensional (2D) analytical or iterative reconstructions. This approach ignores the tridimensional (3D) nature of scatter and detector response function that affects the detected signal. To deal with the 3D nature of the image formation process, iterative reconstruction can be used by considering a 3D projector modelling the 3D spread of photons. In this paper, we investigate the value of using accurate Monte Carlo simulations to determine the 3D projector used in a fully 3D Monte Carlo (F3DMC) reconstruction approach. Given the 3D projector modelling all physical effects affecting the imaging process, the reconstruction problem is solved using the maximum likelihood expectation maximization (MLEM) algorithm. To validate the concept, three data sets were simulated and F3DMC was compared with two other 3D reconstruction strategies using analytical corrections for attenuation, scatter and camera point spread function. Results suggest that F3DMC improves spatial resolution, relative and absolute quantitation and signal-to-noise ratio. The practical feasibility of the approach on real data sets is discussed.

Algorithms↗
PubMed2004

Characterization of spiculation on ultrasound lesions.

Spiculation is a stellate distortion caused by the intrusion of breast cancer into surrounding tissue. Its existence is an important clue to characterizing malignant tumors. Many successful mammographic methods have been proposed to detect tumors with spiculation. Traditional two-dimensional (2-D) ultrasound cannot easily find spiculations because spiculations normally appear parallel to the surface of the skin. Recently, three-dimensional (3-D) ultrasound has been gradually used in clinical applications and it has been proven to be useful in determining the architectural distortion or spiculation that surrounds a breast tumor. This paper aims to identify spiculation from 3-D ultrasonic volume data of a tumor found by a physician. In the proposed method, each coronal slice of volume data is successively extracted and then analyzed as a 2-D ultrasound image by the proposed spiculation detection method. First, in each horizontal slice, the modified rotating structuring element (ROSE) operation is used to find the central region in which spiculation lines converge. Second, the stick algorithm is used to estimate the direction of the edge of each pixel around the central region. A pixel whose edge points toward the central region is marked as a potential spiculation. Finally, the marked pixels are collected around the central region and their distribution is analyzed to determine whether spiculation is present. The 3-D test datasets were obtained using the Voluson 530 or 730, Kretztechnik, Austria. First, the proposed method was tested on 104 2-D typical coronal images (selected by an experienced physician) extracted from 52 3-D ultrasonic datasets. Finally, 225 3-D pathologically proven datasets were tested to evaluate the performance. Spiculations are more easily observed in the coronal view than in the other two views. That is, the 3-D ultrasound is a powerful tool for identifying spiculations. Furthermore, 16% (19/120) of benign cases and 90% (94/105) of malignant cases are detected as spiculations.

Algorithms↗
PubMed2005

De-immunization of therapeutic proteins by T-cell epitope modification.

Many therapeutic proteins in clinical use have been shown to elicit antibody responses which in some cases have been linked to adverse events. Conventional animal models, although convenient, have rarely been predictive of immunogenicity in humans. New methods for predicting the potential immunogenicity of therapeutic proteins are needed. This treatise proposes a new approach which pairs in silico T-cell epitope analysis with in vitro studies. T-cell epitope mapping algorithms such as EpiMatrix can be used to evaluate a candidate therapeutic protein for T-helper epitopes, followed by confirmation of the T-helper epitopes using in vitro methods such as MHC binding assays and T-cell assays. Once these are identified, substitution of key amino acids in the T-cell epitopes may attenuate the immunogenicity of the protein, since modification of the amino acids in anchor position(s) can abrogate binding to human class II MHC molecules and presentation of the peptides, in the context of MHC, to T-helper cells. Following substitution of the key amino acids, immunogenicity of the modified protein can be evaluated in vitro. In parallel, the potential effect of the modifications on the structure of the protein can be evaluated using in silico modeling methods. This multi-step process has been termed DeFT for de-immunization of functional therapeutics. In this article we review the rationale for the approach, provide several retrospective examples that prove the approach in principle, and describe potential applications to therapeutic protein design. The demand for pre-clinical means of evaluating therapeutic proteins is expected to increase with the number of therapeutic proteins and monoclonal antibodies entering the pre-clinical pipeline. Examples provided offer some preliminary proof that the de-immunization approach may improve clinical outcomes.

Animals↗
PubMed2006

Intrinsic uniformity requirements for pinhole SPECT.

UNLABELLED: Pinhole SPECT is a fully 3-dimensional tomography technique. Uniformity requirements for gamma-cameras have been studied for 2-dimensional SPECT performed with parallel-hole collimators. This study investigated reconstruction artifacts in pinhole SPECT arising from intrinsic uniformity defects of the gamma-camera in the case of a pinhole aperture (5 mm) and a rotation radius (10 cm) suitable for human studies. METHODS: A cylindric phantom was filled with water and 99mTc. The count density in the pinhole SPECT projections largely exceeded the density that would be expected in human studies. Two-dimensional gaussian-shaped positive uniformity defects of various heights and various full widths at half maximum were simulated at 5 locations on the 64 projections. All these sets of projections were reconstructed using an iterative 3-dimensional ordered-subset expectation maximization (OSEM) algorithm tailored to the pinhole geometry. The influence of the number of OSEM iterations (2, 5, or 8 for 8 subsets) was also investigated. The height of the ring artifacts generated by the uniformity defects was measured on the reconstructed transverse slices and was compared with the noise in the noncorrupted slices. The uniformity defects were also generated on a 30-megacount flood image from the camera. These images were inspected visually, and the National Electrical Manufacturers Association (NEMA) differential and integral uniformities were calculated. RESULTS: Defects centered on the line corresponding to the orthogonal projection of the camera axis of rotation onto the detector plane generated artifacts whose magnitude-to-noise ratio exceeded 1% when the defect height was larger than 5%-10%, depending on the defect full width at half maximum. These defects were clearly visible on the slices and the flood images. Defects located elsewhere on the camera detector generated ring artifacts of magnitude-to-noise ratio smaller than 1%. They were visually observable both on the reconstructed slices and on the flood images for defect heights larger than 5%. The increase in the number of OSEM iterations conducted to a decrease in the artifact magnitude-to-noise ratio. CONCLUSION: Although hardly visible on the flood images and only slightly increasing the NEMA differential or integral uniformity, a detector uniformity defect of 3% height is able to generate a visible artifact on the reconstructed transverse slices. The study was conducted on a relatively high-count pinhole SPECT acquisition. Considering that far fewer counts would accumulate in clinical practice, any camera that fulfils these uniformity requirements should not lead to the presence of visible uniformity artifacts in the reconstructed slices.

Artifacts↗
PubMed2005

Local coherence oscillations in the EEG during development in the fetal baboon.

OBJECTIVE: To test the hypothesis that local coherence in extra-dural EEG recordings, a direct measure of synchronous cortical network activity, oscillates with the same periodicity as EEG power cycling and follows a similar developmental trajectory. METHODS: Local coherence was derived from continuous EEG recordings from closely spaced (1 cm) chronically implanted extra-dural electrodes over the right hemisphere in five fetal baboons (Papio sp.). A ratio of high to low frequency EEG band power (14-18 Hz to 4-7 Hz), was used to characterize EEG power cycling. RESULTS: Data were obtained within a developmental window (137-151 days; term approximately 175 days) during which fetal EEG power cycling becomes increasingly well-organized. During this period ultradian oscillations in local coherence with periods of approximately 1h developed in parallel with EEG power cycling of approximately the same periods. However, the cross-correlation and phase relationship between local coherence and power oscillations were variable. CONCLUSIONS: These findings suggest that cycles in cortical synchrony develop in parallel with fetal power cycles, but are not tightly coupled to them. SIGNIFICANCE: Coherence provides a direct measure of cortical network dynamics not possible with univariate EEG measures such as spectral power. Reported here are the first measurements of local coherence in the developing fetal brain. Local coherence is studied with the long-term goal of monitoring the development of cortical network activity.

Activity Cycles↗
PubMed2006

MRI detects white matter reorganization after neural progenitor cell treatment of stroke.

We evaluated the effects of neural progenitor cell treatment of stroke on white matter reorganization using MRI. Male Wistar rats (n = 26) were subjected to 3 h of middle cerebral artery occlusion and were treated with neural progenitor cells (n = 17) or without treatment (n = 9) and were sacrificed at 5-7 weeks thereafter. MRI measurements revealed that grafted neural progenitor cells selectively migrated towards the ischemic boundary regions. White matter reorganization, confirmed histologically, was coincident with increases of fractional anisotropy (FA, P < 0.01) after stroke in the ischemic recovery regions compared to that in the ischemic core region in both treated and control groups. Immunoreactive staining showed axonal projections emanating from neurons and extruding from the corpus callosum into the ipsilateral striatum bounding the lesion areas after stroke. Fiber tracking (FT) maps derived from diffusion tensor imaging revealed similar orientation patterns to the immunohistological results. Complementary measurements in stroke patients indicated that FT maps exhibit an overall orientation parallel to the lesion boundary. Our data demonstrate that FA and FT identify and characterize cerebral tissue undergoing white matter reorganization after stroke and treatment with neural progenitor cells.

Algorithms↗
PubMed2005

The SWISS-MODEL workspace: a web-based environment for protein structure homology modelling.

MOTIVATION: Homology models of proteins are of great interest for planning and analysing biological experiments when no experimental three-dimensional structures are available. Building homology models requires specialized programs and up-to-date sequence and structural databases. Integrating all required tools, programs and databases into a single web-based workspace facilitates access to homology modelling from a computer with web connection without the need of downloading and installing large program packages and databases. RESULTS: SWISS-MODEL workspace is a web-based integrated service dedicated to protein structure homology modelling. It assists and guides the user in building protein homology models at different levels of complexity. A personal working environment is provided for each user where several modelling projects can be carried out in parallel. Protein sequence and structure databases necessary for modelling are accessible from the workspace and are updated in regular intervals. Tools for template selection, model building and structure quality evaluation can be invoked from within the workspace. Workflow and usage of the workspace are illustrated by modelling human Cyclin A1 and human Transmembrane Protease 3. AVAILABILITY: The SWISS-MODEL workspace can be accessed freely at http://swissmodel.expasy.org/workspace/

Algorithms↗
PubMed2004

Multisensor optimal information fusion input white noise deconvolution estimators.

The unified multisensor optimal information fusion criterion weighted by matrices is rederived in the linear minimum variance sense, where the assumption of normal distribution is avoided. Based on this fusion criterion, the optimal information fusion input white noise deconvolution estimators are presented for discrete time-varying linear stochastic control system with multiple sensors and correlated noises, which can be applied to seismic data processing in oil exploration. A three-layer fusion structure with fault tolerant property and reliability is given. The first fusion layer and the second fusion layer both have netted parallel structures to determine the first-step prediction error cross-covariance for the state and the estimation error cross-covariance for the input white noise between any two sensors at each time step, respectively. The third fusion layer is the fusion center to determine the optimal matrix weights and obtain the optimal fusion input white noise estimators. The simulation results for Bernoulli-Gaussian input white noise deconvolution estimators show the effectiveness.

Algorithms↗
PubMedDate not supplied

Interactive visual analysis of families of function graphs.

The analysis and exploration of multidimensional and multivariate data is still one of the most challenging areas in the field of visualization. In this paper, we describe an approach to visual analysis of an especially challenging set of problems that exhibit a complex internal data structure. We describe the interactive visual exploration and analysis of data that includes several (usually large) families of function graphs fi (x, t). We describe analysis procedures and practical aspects of the interactive visual analysis specific to this type of data (with emphasis on the function graph characteristic of the data). We adopted the well-proven approach of multiple, linked views with advanced interactive brushing to assess the data. Standard views such as histograms, scatterplots, and parallel coordinates are used to jointly visualize data. We support iterative visual analysis by providing means to create complex, composite brushes that span multiple views and that are constructed using different combination schemes. We demonstrate that engineering applications represent a challenging but very applicable area for visual analytics. As a case study, we describe the optimization of a fuel injection system in diesel engines of passenger cars.

Algorithms↗
PubMed1996

Criticality and parallelism in combinatorial optimization.

Local search methods constitute one of the most successful approaches to solving large-scale combinatorial optimization problems. As these methods are increasingly parallelized, optimization performance initially improves but then abruptly degrades to no better than that of random search beyond a certain point. The existence of this transition is demonstrated for a family of generalized spin-glass models and the traveling salesman problem. Finite-size scaling is used to characterize size-dependent effects near the transition, and analytical insight is obtained through a mean-field approximation.

Algorithms↗
PubMed1999

Influence of upper airway shunt on total respiratory impedance in infants.

When input impedance is determined by means of the forced oscillation technique, part of the oscillatory flow measured at the mouth is lost in the motion of the upper airway wall acting as a shunt. This is avoided by applying the oscillations around the subject's head (head generator) rather than at the mouth (conventional technique). In seven wheezing infants, we compared both techniques to estimate the importance of the upper airway wall shunt impedance (Zuaw) for the interpretation of the conventional technique results. Computation of Zuaw required, in addition, estimation of nasal impedance values, which were drawn from previous measurements (K. N. Desager, M. Willemen, H. P. Van Bever, W. De Backer, and P. A. Vermeire. Pediatr. Pulmonol. 11: 1-7, 1991). Upper airway resistance and reactance at 12 Hz ranged from 40 to 120 and from 0 to -150 hPa. l(-1). s, respectively. Varying nasal impedance within the range observed in infants did not result in major changes in the estimates of Zuaw or lung impedance (ZL), the impedance of the respiratory system in parallel with Zuaw. The conventional technique underestimated ZL, depending on the value of Zuaw. The head generator technique slightly overestimated ZL, probably because the pressure gradient across the upper airway was not completely suppressed. Because of the need to enclose the head in a box (which is not required with the conventional technique), the head generator technique is difficult to perform in infants.

Airway Resistance↗
PubMed2004

Comments on "a parallel mixture of SVMs for very large scale problems".

Collobert, Bengio, and Bengio (2002) recently introduced a novel approach to using a neural network to provide a class prediction from an ensemble of support vector machines (SVMs). This approach has the advantage that the required computation scales well to very large data sets. Experiments on the Forest Cover data set show that this parallel mixture is more accurate than a single SVM, with 90.72% accuracy reported on an independent test set. Although this accuracy is impressive, their article does not consider alternative types of classifiers. We show that a simple ensemble of decision trees results in a higher accuracy, 94.75%, and is computationally efficient. This result is somewhat surprising and illustrates the general value of experimental comparisons using different types of classifiers.

Algorithms↗
PubMed2004

The use of an aSi-based EPID for routine absolute dosimetric pre-treatment verification of dynamic IMRT fields.

BACKGROUND AND PURPOSE: In parallel with the increased use of intensity modulated radiation treatment (IMRT) fields in radiation therapy, flat panel amorphous silicon (aSi) detectors are becoming the standard for online portal imaging at the linear accelerator. In order to minimise the workload related to the quality assurance of the IMRT fields, we have explored the possibility of using a commercially available aSi portal imager for absolute dosimetric verification of the delivery of dynamic IMRT fields. PATIENTS AND METHODS: We investigated the basic dosimetric characteristics of an aSi portal imager (aS500, Varian Medical Systems), using an acquisition mode especially developed for portal dose (PD) integration during delivery of a-static or dynamic-radiation field. Secondly, the dose calculation algorithm of a commercially available treatment planning system (Cadplan, Varian Medical Systems) was modified to allow prediction of the PD image, i.e. to compare the intended fluence distribution with the fluence distribution as actually delivered by the dynamic multileaf collimator. Absolute rather than relative dose prediction was applied. The PD image prediction was compared to the corresponding acquisition for several clinical IMRT fields by means of the gamma evaluation method. RESULTS AND CONCLUSIONS: The acquisition mode is accurate in integrating all PD over a wide range of monitor units, provided detector saturation is avoided. Although the dose deposition behaviour in the portal image detector is not equivalent to the dose to water measurements, it is reproducible and self-consistent, lending itself to quality assurance measurements. Gamma evaluations of the predicted versus measured PD distribution were within the pre-defined acceptance criteria for all clinical IMRT fields, i.e. allowing a dose difference of 3% of the local field dose in combination with a distance to agreement of 3 mm.

Algorithms↗
PubMed2006

N log N method for hydrodynamic interactions of confined polymer systems: Brownian dynamics.

A Brownian dynamics simulation technique is presented where a Fourier-based N log N approach is used to calculate hydrodynamic interactions in confined flowing polymer systems between two parallel walls. A self-consistent coarse-grained Langevin description of the polymer dynamics is adopted in which the polymer beads are treated as point forces. Hydrodynamic interactions are therefore included in the diffusion tensor through a Green's function formalism. The calculation of Green's function is based on a generalization of a method developed for sedimenting particles by Mucha et al. [J. Fluid Mech. 501, 71 (2004)]. A Fourier series representation of the Stokeslet that satisfies no-slip boundary conditions at the walls is adopted; this representation is arranged in such a way that the total O(N2) contribution of bead-bead interactions is calculated in an O(N log N) algorithm. Brownian terms are calculated using the Chebyshev polynomial approximation proposed by Fixman [Macromolecules 19, 1195 (1986); 19, 1204 (1986)] for the square root of the diffusion tensor. The proposed Brownian dynamics simulation methodology scales as O(N1.25 log N). Results for infinitely dilute systems of dumbbells are presented to verify past predictions and to examine the performance and numerical consistency of the proposed method.

Algorithms↗
PubMed1997

Neural networks in neurotologic expert systems.

Artificial intelligence donates new possibilities to neurotologic research. Neural networks are a computer-based reasoning method which can be applied in expert systems created for clinical decision support. Neural networks have been used in medical imaging, in medical signal processing and to analyze both clinical and laboratory data. Principally, neural networks simulate the function of the brain. They have to be taught to make correct decisions from the input data. This learning process can be either supervised or unsupervised. The decision making is based on mathematical transformations and it occurs on a hidden level. Calculations are made on parallel manner and the decision making simulates pattern recognition method. Neural networks suit well in medical problems which cannot be defined in simple rules. A drawback of neural networks is that the decisions are irrational and cannot be motivated to the user. Another problem is neural networks' difficulty to handle incomplete input data, i.e., how to define some default or expected values for unknown input parameters. In a complex medical area, which would require multilayered neural networks, the neural networks require a large amount of solved cases for the learning process. In our experience neural networks seem not suitable for diagnosing vertigo and a better choice would be either case-based reasoning or possibly genetic algorithms or a combination of these.

Diagnosis, Computer-Assisted↗
PubMed2000

Kinetic model for 5-fluorouridine degradation. Catalytic effect of 5-fluorouracil.

The effects of 5-fluorouridine (5-FUR, CAS 316-46-1) degradation products, 5-fluorouracil (5-FU, CAS 51-21-8) and D-ribose (CAS 50-69-1), on its degradation rate was investigated following a 2(3) factorial design. The experimental data fitted to the proposed mathematical model which includes two parallel degradation mechanisms: the first one, a second order bimolecular reaction involving both 5-FUR and 5-FU, and the second, a first order one. Experimental data obtained show a high variability. Both graphic and statistical analysis of the experiments for which a full kinetic model was applied manifested that the degradation mechanism included an autocatalytic route and confirmed the role of the 5-FU on the hydrolysis of 5-FUR.

Algorithms↗
PubMed1997

Total urinary follicle stimulating hormone as a biomarker for detection of early pregnancy and periimplantation spontaneous abortion.

Total concentrations of follicle stimulating hormone (FSH) were evaluated in daily urine samples from conceptive and nonconceptive menstrual cycles by measurement of the FSH beta subunit following treatment of the samples to dissociate the FSH heterodimer. Samples were self-collected by normal subjects during cycles in which daily blood samples also were obtained. Daily blood and urine specimens were collected prospectively from 10 subject in conceptive cycles, which led to normal pregnancies, and from 10 subjects with bilateral tubal ligations to provide control samples form nonconceptive cycles. Mean serum and urinary FSH concentration profiles wer parallel in both groups following ovulation and during he first 9 days of the luteal phase. Mean values for both serum and urinary FSH rose significantly above the postovulatory baseline by 10-12 days following the midcycle luteinizing hormone (LH) peak in nonconceptive cycles, but did not rise at any time following ovulation during conceptive cycles. Following regression analysis of the changing FSH concentration between days 9-14 post-LH surge in conceptive cycles, a slope of </= 0.02 ng FHS/mg creatinine/day was selected as a cutoff point to identify conceptive cycles. There was a high concordance between the day of LH peak in serum and the day of FSH peak in urine. Therefore, in applying the algorithm, the day of FSH peak in urine was used to determine the days for which the FSH slope would be calculated, i.e., days 9-14 post-FSH peak in urine. The sensitivity and specificity of the change in urinary FSH concentrations to detect pregnancy in a different set of 55 cycles were found to 88.9% and 89.3%, respectively. All six cases of early fetal loss in the sample set were correctly identified. These results suggest that urinary FSH can be used as an additional biomarker for the verification of early pregnancy in prospective epidemiological studies in which early fetal loss is a suspected outcome.

Abortion, Spontaneous↗