SEARCH · PubMed Health

Results for “Gaussian process”

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 1,081 records · Page 60Linked to original sources

PubMed2006

Sound source localization in real sound fields based on empirical statistics of interaural parameters.

The role of temporal fluctuations and systematic variations of interaural parameters in localization of sound sources in spatially distributed, nonstationary noise conditions was investigated. For this, Bayesian estimation was applied to interaural parameters calculated with physiologically plausible time and frequency resolution. Probability density functions (PDFs) of the interaural level differences (ILDs) and phase differences (IPDs) were estimated by measuring histograms for a directional sound source perturbed by several types of interfering noise at signal-to-noise ratios (SNRs) between -5 and +30 dB. A moment analysis of the PDFs reveals that the expected values shift and the standard deviations increase considerably with decreasing SNR, and that the PDFs have non-Gaussian shape at medium SNRs. A d' analysis of the PDFs indicates that elevation discrimination is possible even at low SNRs in the median plane by integrating information across frequency. Absolute sound localization was simulated by a Bayesian maximum a posteriori (MAP) procedure. The simulation is based on frequency integration of broadly tuned "detectors." Confusion patterns of real and estimated sound source directions are similar to those of human listeners. The results indicate that robust processing strategies are needed to exploit interaural parameters successfully in noise conditions due to their strong temporal fluctuations.

Acoustic Stimulation↗

PubMed2002

Direct generation of Bessel beams.

Two implementations are identified to create a Bessel beam directly, i.e. without the spatial filtering of an initially Gaussian beam. The first implementation is based on a resonator configuration whose lowest-loss transverse mode is a Bessel beam. Numerical simulation to corroborate the geometrical optical arguments is presented. The second implementation is based on the theorem that the angular-plane wave spectrum of a Bessel beam is composed of a cone of wave vectors. This cone is also generated through a phase-matching condition in a four-wave mixing process. This leads to the conclusion that anti-Stokes radiation generated in a nonlinear material will leave the substrate under the form of a Bessel beam.

Journal Article↗

PubMed2003

Concerning the reaction between singlet nitrenium ions and water: a computational investigation on competitive reaction paths.

The reaction between singlet nitrenium ions XNH(+) (X = F and Cl) and H(2)O has been investigated by high-level of theory ab initio calculations. The geometries of the involved intermediates, transition structures, and dissociation products have been optimized at the MP2(full)/6-31G(d) level of theory, and accurate total energies have been obtained using the Gaussian-3 (G3) procedure. The reaction commences by the exothermic formation of the F-NH-OH(2) (+) and Cl-NH-OH(2) (+) intermediates, which are in turn able to undergo two distinct low-energy reaction paths, namely, the isomerization to the N-protonated isomers of the hydroxylamines F-NH-OH or Cl-NH-OH, and the eventual extrusion of HF or HCl. The competitive or alternative occurrence of these two processes strictly depends on the nature of the substituent X. In the reaction between FNH(+) and H(2)O, the energy gained in the formation of the complex F-NH-OH(2) (+) from the association between FNH(+) and H(2)O, 52.1 kcal mol(-1), is by far larger than the activation barrier for the loss of HF from F-NH-OH(2) (+), computed as 24.9 kcal mol(-1). In addition, the F-NH-OH(2) (+) intermediate requires 33.0 kcal mol(-1) to overcome the barrier for the isomerization to F-NH(2)-OH(+). Therefore, the reaction between FNH(+) and H(2)O is expected to occur practically exclusively by HF elimination with formation of the HN-OH(+) ionic product. On the other hand, for the reaction between ClNH(+) and H(2)O, it is not possible to get a definitive conclusion on the competitive or alternative occurrence of the two reaction paths. In fact, the transition structure involved in the elimination of HCl from Cl-NH-OH(2) (+) is only 3.4 kcal mol(-1) lower in energy than the transition structure for the isomerization of Cl-NH-OH(2) (+) to Cl-NH(2)-OH(+). In addition, the absolute values of the energy barriers of these two processes, 24.2 and 27.6 kcal mol(-1), respectively, are comparable with the energy gained in the formation of the complex Cl-NH-OH(2) (+) from the association between ClNH(+) and H(2)O, 24.0 kcal mol(-).1 Therefore, the ClNH(+) cation is predicted to react with water significantly slower than FNH(+).

Journal Article↗

PubMed2002

Estimation of the conduction velocity distribution of human sensory nerve fibers.

A new method for estimating the distribution of conduction velocities (DCV) of peripheral nerve fibers has been developed. It also enables estimation of single nerve fiber action potential (SFAP), which agrees with the physiological knowledge. Two compound nerve action potentials (CAPs) elicited by electrical stimulation of a nerve bundle were recorded at different conduction distances. The distances between the stimulation and recording electrodes were measured on the skin surface along the nerve bundle. Starting with an arbitrary SFAP, the first estimated DCV was calculated from a CAP by the regularized non-negative least squares method. The next SFAP was then calculated by deconvolution of the other CAP and the estimated DCV. A lowpass filter with an appropriate cutoff frequency was used to obtain better conversion. The process was iterated until the CAP error defined as /CAP(calculated)-CAP/(2) was small enough. The conduction distances contained errors in measurement, especially in the distal segment, that distorted the estimated results. The Fibonacci search, therefore, was adopted to optimize the distance according to the CAP error. The accuracy of this method was demonstrated by a simulation study performed with two CAPs calculated from an arbitrary bimodal DCV and a biphasic SFAP to which a Gaussian white noise was added. The reliability of this method was checked in normal subjects by recording a pair of CAPs elicited by stimulation of the median nerve at the wrist and the elbow.

Action Potentials↗

PubMed2005

Evolution of the electronic structure of Be clusters.

Using a modified symbiotic genetic algorithm approach and many-body interatomic potential derived from first principles, we have calculated equilibrium geometries and binding energies of the ground-state and low-lying isomers of Be clusters containing up to 41 atoms. Molecular-dynamics study was also carried out to study the frequency of occurrence of the various geometrical isomers as these clusters are annealed during the simulation process. For a selected group of these clusters, higher-energy isomers were more often found than their ground-state structures due to large catchment areas. The accuracy of the above ground-state geometries and their corresponding binding energies were verified by carrying out separate ab initio calculations based on molecular-orbital approach and density-functional theory with generalized gradient approximation for exchange and correlation. The atomic orbitals were represented by a Gaussian 6-311G** basis, and the geometry optimization was carried out using the GAUSSIAN 98 code without any symmetry constraint. While the ground-state geometries and their corresponding binding energies obtained from ab initio calculations do not differ much from those obtained using the molecular-dynamics approach, the relative stability of the clusters and the energy gap between the highest occupied and the lowest unoccupied molecular orbitals show significant differences. The energy gaps, calculated using the density-functional theory, show distinct shell closure effects, namely, sharp drops in their values for Be clusters containing 2, 8, 20, 34, and 40 electrons. While these features may suggest that small Be clusters behave free-electron-like and, hence, are metallic, the evolution of the structure, binding energies, coordination numbers, and nearest-neighbor distances do not show any sign of convergence towards the bulk value. We also conclude that molecular-dynamics simulation based on many-body interatomic potentials may not always give the correct picture of the evolution of the structure and energetics of clusters although they may serve as a useful tool for obtaining starting geometries by efficiently searching a large part of the phase space.

Journal Article↗

PubMed1991

Dynamics of adaptation for vision with a stabilized image.

The addition of a uniform increment of light to a high-contrast image that has been stabilized on the retina reveals marked perceptual nonlinearities. When the increment is small, the pattern appears in its original phase (OP), large increments produce an apparent phase reversal (APR), and intermediate increments may yield an apparently blank field or an oscillation of the apparent phase. In the present series of studies the threshold values used to produce a stable OP and APR were determined as a function of adaptation time before the application of the increment. The stabilized target had a luminance profile consisting of the difference of two Gaussians. A model of detection incorporating a multiplicative gain controlled by a filtered version of the stimulus was used to account for the occurrence of the OP and the APR and the transitory phenomena following the uniform increment. It is argued that the midpoint of the transition zone between the OP and APR, corresponding to blanking, enables us to estimate the shape of the step response function of the gain filter independently of the subsequent detection processes.

Adaptation, Ocular↗

PubMedDate not supplied

Fluid transport in glass beads phantoms: spatial velocity measurements and confirmation of the stochastic model.

A stochastic model of fluid flow in porous rocks has been previously developed to explain the measured distribution of local velocity. The theoretical predictions of this model agree well with experimental results obtained from the magnetic resonance imaging-based measurements of the spatial variation of velocity of water permeating through Bentheimer and Clashach sandstones. To further verify previous results, we have performed new velocity measurement experiments using an efficient velocity encoded pi-echo planar imaging sequence on glass bead phantoms that exhibit more regular pore size distribution than rocks. The results show that velocity distributions in glass bead phantoms also exhibit Gaussian profiles and the linear relationship between the velocity variance and the mean velocity (the Mansfield-Issa equation).

Diffusion↗

PubMed2006

Optimal spatial adaptation for patch-based image denoising.

A novel adaptive and patch-based approach is proposed for image denoising and representation. The method is based on a pointwise selection of small image patches of fixed size in the variable neighborhood of each pixel. Our contribution is to associate with each pixel the weighted sum of data points within an adaptive neighborhood, in a manner that it balances the accuracy of approximation and the stochastic error, at each spatial position. This method is general and can be applied under the assumption that there exists repetitive patterns in a local neighborhood of a point. By introducing spatial adaptivity, we extend the work earlier described by Buades et al. which can be considered as an extension of bilateral filtering to image patches. Finally, we propose a nearly parameter-free algorithm for image denoising. The method is applied to both artificially corrupted (white Gaussian noise) and real images and the performance is very close to, and in some cases even surpasses, that of the already published denoising methods.

Algorithms↗

PubMed2006

Waveform design for ultrasonic pulse-inversion fundamental imaging.

Pulse-inversion (PI) fundamental imaging exhibits significantly better contrast detection than linear and second-harmonic imaging. PI fundamental imaging involves two firings with inverted waveforms. When the returning echoes from the two firings are summed, the residual signal related to tissue is limited to even-order harmonics, whereas for microbubbles, the fundamental signal is not completely canceled due to the echo under compression differing from that under rarefaction. The efficacy of PI fundamental imaging has been reported previously. In this study, we investigated the performance of PI fundamental imaging using both simulations and in vitro experiments with various transmit waveforms, including coded excitation and asymmetrical waveforms (i.e., asymmetrical between compression and rarefaction). For coded excitation, a longer waveform was found to increase the similarity in the responses to positive and negative pulses, thus lowering the contrast between microbubbles and tissue. In addition, imperfect pulse compression also decreases the contrast because it increases the residue fundamental signal emanating from tissue. Using asymmetrical waveforms noticeably increased the residual microbubble signal in the fundamental band but the nonzero DC component that is inherent in such waveforms also increases the tissue fundamental signal. The combination of these two effects decreases the contrast. From these results, it is concluded that the use of coded excitation is undesirable in PI fundamental imaging and that the waveforms should contain no DC component. Furthermore, the transmit waveform needs to be appropriately windowed in order to reduce spectral leakage. Therefore, a Gaussian pulse with the pulse length determined by the signal-to-noise ratio of the imaging system is generally optimal for PI fundamental imaging.

Artifacts↗

PubMed1986

Non-Gaussian distribution of arterial pressure and heart rate during sleep and waking in hypertensive and normotensive subjects.

For statistical analysis of 24-hour recordings of arterial pressure (AP) and heart rate (HR), it is necessary to establish a theoretical probability density function of the distributions. In the present study, 24-hour recordings of direct AP and HR were performed in 15 normotensives (NT) and 39 patients with essential hypertension (EH) by means of a new portable device with a digital memory for analyzing frequency histograms. In both NT and EH, frequency histograms of AP (systolic and diastolic AP) and HR during a 24-hour period showed bi-modal curves, whereas those made during sleep and waking produced asymmetrical patterns resembling Gamma distribution. From the AP and HR histograms made for each subject during sleep and waking, such parameters as mean (M), standard deviation, mode (Mo), skewness (Sk), kurtosis (K) and minimum values (Mi) were calculated. The Sk and M minus Mo were positive, and K was greater than 3 in both HR and AP histograms; the AP and HR histograms during sleep can be more correctly analyzed with Gamma distribution (mean parameter errors were less than 7.3%) than with Gaussian distribution (in this: S = O, M minus Mo = O, K = 3). The Mi would accord with the location parameter of the Gamma distribution.

Adult↗

PubMed2004

Surgical classification of variations in the anatomy of the sigmoid sinus.

OBJECTIVE: Analyze the variability of anatomic relations between the sigmoid sinus and other surgical landmarks within the mastoid cavity to establish a classification of sinus location to facilitate surgical planning. STUDY DESIGN: A descriptive study of surgical landmarks using 96 temporal bones from the Temporal Bone Laboratory at the Department of Otorhinolaryngology, Valle University at Cali, Colombia. METHODS: A dissection simulating a canal wall-up mastoidectomy extended to include exposure of the facial recess, epitympanum, and sigmoid sinus was performed on 96 adult disease-free temporal bones. Seven standard measurements were taken on each bone between the sigmoid sinus and other structures in the mastoid cavity. All data were analyzed using Epi-Info 2000. RESULTS: This parametric study was consistent with three classes of anatomic variations in the location of the sigmoid sinus (type 1, type 2, and type 3) based on a Gaussian distribution of the seven variables. CONCLUSIONS: Variability in the anatomic location of the sigmoid sinus in the mastoid cavity is common and has surgical relevance. SIGNIFICANCE: This classification will help to rapidly identify the sigmoid sinus relationship within the mastoid cavity and to use the sigmoid sinus as a landmark during ear surgery in those cases in which traditional landmarks, such as the posterior auditory canal wall or the semicircular canals, are destroyed by a pathologic process.

Humans↗

PubMed1999

Cooperative fluctuations and subunit communication in tryptophan synthase.

Tryptophan synthase (TRPS), with linearly arrayed subunits alphabetabetaalpha, catalyzes the last two reactions in the biosynthesis of L-tryptophan. The two reactions take place in the respective alpha- and beta-subunits of the enzyme, and the intermediate product, indole, is transferred from the alpha- to the beta-site through a 25 A long hydrophobic tunnel. The occurrence of a unique ligand-mediated long-range cooperativity for substrate channeling, and a quest to understand the mechanism of allosteric control and coordination in metabolic cycles, have motivated many experimental studies on the structure and catalytic activity of the TRPS alpha2beta2 complex and its mutants. The dynamics of these complexes are analyzed here using a simple but rigorous theoretical approach, the Gaussian network model. Both wild-type and mutant structures, in the unliganded and various liganded forms, are considered. The substrate binding site in the beta-subunit is found to be closely coupled to a group of hinge residues (beta77-beta89 and beta376-beta379) near the beta-beta interface. These residues simultaneously control the anticorrelated motion of the two beta-subunits, and the opening or closing of the hydrophobic tunnel. The latter process is achieved by the large amplitude fluctuations of the so-called COMM domain in the same subunit. Intersubunit communications are strengthened in the presence of external aldimines bound to the beta-site. The motions of the COMM core residues are coordinated with those of the alpha-beta hinge residues beta174-beta179 on the interfacial helix betaH6 at the entrance of the hydrophobic tunnel. And the motions of betaH6 are coupled, via helix betaH1 and alphaL6, to those of the loop alphaL2 that includes the alpha-subunit catalytically active residue Asp60. Overall, our analysis sheds light on the molecular machinery underlying subunit communication, and identifies the residues playing a key role in the cooperative transmission of conformational motions across the two reaction sites.

Allosteric Regulation↗

PubMed2026

Making multi-axis Gaussian graphical models scalable to millions of cells.

MOTIVATION: Networks underlie the generation and interpretation of many biological datasets: gene networks shed light on the regulatory structure of the genome, and cell networks can capture structure of the tumor micro-environment. However, most methods that learn such networks make the faulty "independence assumption"; to learn the gene network, they assume that no cell network exists. "Multi-axis" methods, which do not make this assumption, fail to scale beyond a few thousand cells or genes. This limits their applicability to only the smallest datasets. RESULTS: We develop a multi-axis method, which learns conditional dependency networks, capable of processing million-cell datasets within minutes. This was previously impossible, and unlocks the use of such methods on modern scRNA-seq datasets, as well as more complex datasets. We apply the method to a new scRNA-seq dataset for neuronal cell development, and compare the result to an existing state of the art method, hdWGCNA. We demonstrate that the new method yields gene networks that have a more focused biological interpretation and that the simultaneously learned cell network has advantages over a conventional kNN-based clustering. Further, our method yields novel biological insights by identifying long non-coding RNAs that potentially have a role in neuronal development. AVAILABILITY AND IMPLEMENTATION: Our methodology is available as a Python package GmGM on PyPI (https://pypi.org/project/GmGM/0.5.3/). The code for all experiments performed in this article is available on GitHub (https://github.com/BaileyAndrew/GmGM-Bioinformatics) and Zenodo (10.5281/zenodo.20384566).

Gene Regulatory Networks↗

PubMed2003

Time-dependent entropy estimation of EEG rhythm changes following brain ischemia.

Our approach is motivated by the need to generate a rigorous measure of the degree of disorder (or complexity) of the EEG signal in brain injury. Entropy is a method to quantify the order/disorder of a time series. It is the first time that a time-dependent entropy (TDE) is used in the quantification of brain injury level. The TDE was sensitive enough to monitor the significant changes in the subject's brain rhythms during recovery from global ischemic brain injury. Among the different entropy measures, we used Tsallis entropy. This entropy is parametrized and is able to match with the particular properties of EEG, like long-range rhythms, spikes, and bursts. The method was tested in a signal composed of segments of synthetic signals (Gaussian and uniform distributions) and segments of real signals. The real signal segments were extracted from normal EEG, EEG recordings from early recovery, and normal EEG corrupted by simulated spikes and bursts. Adult Wistar rats were subjected to asphyxia-cardiac arrest for 3 and 5 min. The TDE detected the pattern of ischemia-induced EEG alterations and was able to discriminate the different injury levels. Two parameters seem to be good descriptors of the recovery process; the mean entropy and the variance of the estimate followed opposite trends, with the mean entropy decreasing and its variance increasing with injury.

Algorithms↗

PubMed1996

Methods in quantitative image analysis.

The main steps of image analysis are image capturing, image storage (compression), correcting imaging defects (e.g. non-uniform illumination, electronic-noise, glare effect), image enhancement, segmentation of objects in the image and image measurements. Digitisation is made by a camera. The most modern types include a frame-grabber, converting the analog-to-digital signal into digital (numerical) information. The numerical information consists of the grey values describing the brightness of every point within the image, named a pixel. The information is stored in bits. Eight bits are summarised in one byte. Therefore, grey values can have a value between 0 and 256 (2(8)). The human eye seems to be quite content with a display of 5-bit images (corresponding to 64 different grey values). In a digitised image, the pixel grey values can vary within regions that are uniform in the original scene: the image is noisy. The noise is mainly manifested in the background of the image. For an optimal discrimination between different objects or features in an image, uniformity of illumination in the whole image is required. These defects can be minimised by shading correction [subtraction of a background (white) image from the original image, pixel per pixel, or division of the original image by the background image]. The brightness of an image represented by its grey values can be analysed for every single pixel or for a group of pixels. The most frequently used pixel-based image descriptors are optical density, integrated optical density, the histogram of the grey values, mean grey value and entropy. The distribution of the grey values existing within an image is one of the most important characteristics of the image. However, the histogram gives no information about the texture of the image. The simplest way to improve the contrast of an image is to expand the brightness scale by spreading the histogram out to the full available range. Rules for transforming the grey value histogram of an existing image (input image) into a new grey value histogram (output image) are most quickly handled by a look-up table (LUT). The histogram of an image can be influenced by gain, offset and gamma of the camera. Gain defines the voltage range, offset defines the reference voltage and gamma the slope of the regression line between the light intensity and the voltage of the camera. A very important descriptor of neighbourhood relations in an image is the co-occurrence matrix. The distance between the pixels (original pixel and its neighbouring pixel) can influence the various parameters calculated from the co-occurrence matrix. The main goals of image enhancement are elimination of surface roughness in an image (smoothing), correction of defects (e.g. noise), extraction of edges, identification of points, strengthening texture elements and improving contrast. In enhancement, two types of operations can be distinguished: pixel-based (point operations) and neighbourhood-based (matrix operations). The most important pixel-based operations are linear stretching of grey values, application of pre-stored LUTs and histogram equalisation. The neighbourhood-based operations work with so-called filters. These are organising elements with an original or initial point in their centre. Filters can be used to accentuate or to suppress specific structures within the image. Filters can work either in the spatial or in the frequency domain. The method used for analysing alterations of grey value intensities in the frequency domain is the Hartley transform. Filter operations in the spatial domain can be based on averaging or ranking the grey values occurring in the organising element. The most important filters, which are usually applied, are the Gaussian filter and the Laplace filter (both averaging filters), and the median filter, the top hat filter and the range operator (all ranking filters). Segmentation of objects is traditionally based on threshold grey values. (AB

Algorithms↗

PubMed1988

The use of difference of Gaussian image filtering to assess objectively the correlation between breast vascularity and breast cancer.

The potential of thermography as a screening technique for the early detection of breast cancer was assessed by using difference of Gaussian filtering to isolate and quantify the vascular contents of breast thermograms. Forty-five patients found to have breast cancer and 49 patients who developed breast cancer within 5 years of being screened were paired with normal variants and the thermograms of each group were assessed. No statistically significant separation was resolved between either of the two paired groups implying that the vascular content of isolated thermograms is unable to provide meaningful indications of breast cancer. The processing algorithm, although developed initially for thermograms, is equally valid for other diagnostic imaging techniques and could be of use where it is required to isolate vascularity or other fine detail from larger body structures.

Breast↗

PubMed2005

Exact probability function for bulk density and current in the asymmetric exclusion process.

We examine the asymmetric simple exclusion process with open boundaries, a paradigm of driven diffusive systems, having a nonequilibrium steady-state transition. We provide a full derivation and expanded discussion and digression on results previously reported briefly in M. Depken and R. Stinchcombe, Phys. Rev. Lett. 93, 040602 (2004). In particular we derive an exact form for the joint probability function for the bulk density and current, both for finite systems, and also in the thermodynamic limit. The resulting distribution is non-Gaussian, and while the fluctuations in the current are continuous at the continuous phase transitions, the density fluctuations are discontinuous. The derivations are done by using the standard operator algebraic techniques and by introducing a modified version of the original operator algebra. As a by-product of these considerations we also arrive at a very simple way of calculating the normalization constant appearing in the standard treatment with the operator algebra. Like the partition function in equilibrium systems, this normalization constant is shown to completely characterize the fluctuations, albeit in a very different manner.

Journal Article↗

PubMed1995

Multiple resolution Bayesian segmentation of ultrasound images.

We propose a novel method for obtaining the maximum a posteriori (MAP) probabilistic segmentation of speckle-laden ultrasound images. Our technique is multiple-resolution based, and relies on the conversion of speckle images with Rayleigh statistics to subsampled images with Gaussian statistics. This conversion reduces computation time, as well as allowing accurate parameter estimation for a probabilistic segmentation algorithm. Results appear to provide improvements over previous techniques in terms of low-contrast detail and accuracy.

Algorithms↗