SEARCH · PubMed Health

Results for “Mathematical Computing”

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 307 records · Page 17Linked to original sources

PubMed2003

A multiple classifier system for early melanoma diagnosis.

Melanoma is the most dangerous skin cancer and early diagnosis is the key factor in its successful treatment. Well-trained dermatologists reach a diagnosis via visual inspection, and reach sensitivity and specificity levels of about 80%. Several computerised diagnostic systems were reported in the literature using different classification algorithms. In this paper, we will illustrate a novel approach by which a suitable combination of different classifiers is used in order to improve the diagnostic performances of single classifiers. We used three different kinds of classifiers, namely linear discriminant analysis (LDA), k-nearest neighbour (k-NN) and a decision tree, the inputs of which are 38 geometric and colorimetric features automatically extracted from digital images of skin lesions. Multiple classifiers were generated by combining the diagnostic outputs of single classifiers with appropriate voting schemata. This approach was evaluated on a set of 152 digital skin images. We compared the performances of multiple classifiers (2- and 3-classifier groups) between them and with respect to single ones (1-classifier group). We further compared the classifiers' performances with those of eight dermatologists. Classifiers' performances were measured in terms of distance from the ideal classifier. Compared with 1- and 2-classifier groups, performances of 3-classifier systems were significantly higher (P<0.0005 and P<0.001, respectively). No statistically significant differences were found between the 1- and 2-classifier groups (P=0.352). While the dermatologists group showed a level of performances significantly higher than the 1-classifier systems (P<0.020), no differences were found between the multiple classifier groups and the dermatologists groups, indicating comparable performances. This work suggests that a suitable combination of different kinds of classifiers can improve the performances of an automatic diagnostic system.

Algorithms↗

PubMed2001

Quantitative analysis of errors in fractionated stereotactic radiotherapy.

Fractionated stereotactic radiotherapy (FSRT) offers a technique to minimize the absorbed dose to normal tissues; therefore, quality assurance is essential for these procedures. In this study, quality assurance for FSRT of 58 cases, between August 1995 and August 1997 are described, and the errors for each step and overall accuracy were estimated. Some of the important items for FSRT procedures are: accuracy in CT localization, transferred image distortion, laser alignment, isocentric accuracy of linear accelerator, head frame movement, portal verification, and various human errors. A geometric phantom, that has known coordinates was used to estimate the accuracy of CT localization. A treatment planning computer was used for checking the transferred image distortion. The mechanical isocenter standard (MIS), rectilinear phantom pointer: (RLPP), and laser target localizer frame (LTLF) were used for laser alignment and target coordinates setting. Head-frame stability check was performed by a depth confirmation helmet (DCH). A film test was done to check isocentric accuracy and portal verification. All measured data for the 58 patients were recorded and analyzed for each item. 4-MV x-rays from a linear accelerator, were used for FSRT, along with homemade circular cones with diameters from 20 to 70 mm (interval: 5 mm). The accuracy in CT localization was 1.2+/-0.5 mm. The isocentric accuracy of the linear accelerator, including laser alignment, was 0.5+/-0.2 mm. The reproducibility of the head frame was 1.1+/-0.6 mm. The overall accuracy was 1.7+/-0.7 mm, excluding human errors.

Brain Neoplasms↗

PubMed1998

Numerical evaluation of the production of radionuclides in a nuclear reactor (Part II).

A computer program called LAURA has been developed to predict the production rates of any member of a nuclei network undergoing spontaneous decay and/or induced neutron transformation in a nuclear reactor. The theoretical bases for the development of LAURA were discussed in Part I. In particular, in Part I, we described how an expression based on the Rubinson (1949) approach is used to evaluate the depletion function. In this paper (Part II), we describe the full simulation of radionuclide production including the decomposition of a reaction network into independent linear chains, provisions for periodic reactor shutdown and restart, and implementation of an approximate solution given by Raykin and Shlyakhter (1989) to account for the effect of feedback due to alpha decay. Also included are some examples which demonstrate possible uses for LAURA.

Computer Simulation↗

PubMed1997

Compensation for displacement of the focal point in cone beam single photon emission computed tomography reconstruction.

This study examined the effects of focal point displacement on image quality in cone beam single photon emission computed tomography (SPECT). A new image reconstruction algorithm that accounts for the focal point shift was derived and three shift geometries were investigated. The geometries included a lateral shift with a fixed focal length but off-center focusing, a linear axial shift with a variable focal length that depends linearly on the distance between a bin of the detector and the center of the detector, and a random axial shift with a randomly varying focal length. Computer simulation was conducted to evaluate the shift effects with a phantom that was composed of 118 small spherical sources. The results demonstrated that the lateral shift of the focal point was more critical to image quality than was the axial shift. With a 0.64 cm (1 pixel) lateral shift, noticeable artifacts was observed, while an axial shift resulted in minimal changes in image quality until it reached 8 cm (12.5 pixels). The derived reconstruction algorithm eliminated most of the artifacts caused by a fixed lateral shift or a linear axial shift of the focal point, but failed to do so for a random axial shift since the linear distribution assumed in image reconstruction did not match the random shift occurred in acquisition of the data.

Algorithms↗

PubMed1993

Statistical clustering techniques for the analysis of long molecular dynamics trajectories: analysis of 2.2-ns trajectories of YPGDV.

The microscopic interactions and mechanisms leading to nascent protein folding events are generally unknown. While such short time-scale events are difficult to study experimentally, molecular dynamics simulations of peptides can provide a useful model for studying events related to protein folding initiation. Recently, two extremely long molecular dynamics simulations (2.2 ns each) were carried out on the pentapeptide Tyr-Pro-Gly-Asp-Val [Tobias, D. J., Mertz, J. E., & Brooks, C. L., III (1991) Biochemistry 30, 6054-6058] that forms stable reverse turns in solution. Tobias et al. examined folding events in this large system (approximately 30,000 conformations) using traditional methods of trajectory analysis. The shear magnitude of this problem prompted us to develop an automated approach, based on self-organizing neural nets, to extract the key features of the molecular dynamics trajectory. The neural net is used to perform conformational clustering, which reduces the complexity of a system while minimizing the loss of information. The conformations were grouped together using distances in dihedral angle space as a measure of conformational similarity. The resulting clusters represent "conformational states", and transitions between these states were examined to identify mechanisms of conformational change. Many conformational changes involved the rotation of only a single dihedral angle, but concerted angle changes were also found. Most of the conformational information in the 30,000 samples from the full trajectories was retained in the relatively few resultant clusters, providing a powerful tool for analysis of an expanding base of large molecular simulations.

Algorithms↗

PubMed2001

Mechanism of unfolding of a model helical peptide.

Synthetic model helical peptides, Acetyl-W(EAAAR)(5)A-amide with (13)C=O specifically labeled alanine segments in repeats n = 1,2 or 4,5 were studied in aqueous D(2)O solution as a function of temperature using Fourier transform infrared spectroscopy and two-dimensional correlation analysis. The (13)C==O provided a probe which was sensitive to the carbonyl stretch in the peptide bonds of the alanine residues at the amino terminal end in one peptide as compared to the probe in the carboxy terminal end of the other peptide during thermal perturbation. The relative stability of each terminal end was examined; the more stable terminal was determined to be the amino terminal end. Also studied were the glutamate and arginine side-chain modes involved in the salt bridging interaction. Two-dimensional correlation analysis enabled enhanced resolution in the spectral region of 1520--1700 cm(-1), and thus, the order in which these vibrational modes were perturbed as a function of increasing temperature were established.

Amino Acid Sequence↗

PubMed1997

Oligomerization of the amide sensor protein AmiC by x-ray and neutron scattering and molecular modeling.

AmiC is the negative regulator of the amidase operon which is involved in amide metabolism in the cytosol of Pseudomonas aeruginosa. Crystal structures show that AmiC contains two large domains that are very similar to the periplasmic leucine-isoleucine-valine binding protein (LivJ) of Escherichia coli. Synchrotron X-ray and neutron (in 100% 2H2O buffer) scattering data were obtained for AmiC in the presence of its substrate acetamide and its anti-inducer butyramide which binds more weakly to AmiC than acetamide. Guinier analyses to obtain radius of gyration RG and molecular weight Mr values showed that AmiC formed trimers whose formation was favored in the presence of acetamide and which exhibited concentration-dependent properties at concentrations between 0.4 and 2 mg/mL. Above 2 mg/mL, where trimers predominated, the RG data were identical within 0.05 nm for AmiC-acetamide and AmiC-butyramide with mean X-ray and neutron RG values of 3.35 and 3. 28 nm, respectively. Scattering curve fits constrained by the crystal structure of AmiC-acetamide were evaluated in order to describe a model for trimeric AmiC. A translational search of parallel alignments of three monomers to form a symmetric AmiC homotrimer gave a good X-ray curve fit. Combinations of calculated curves for monomeric, dimeric, trimeric, and tetrameric AmiC as seen in the crystal structure of AmiC gave reasonable but weaker X-ray curve fits which did not favor the existence of tetrameric AmiC. It is concluded that AmiC exhibits novel ligand-dependent oligomerization properties in solution when these are compared to other members of the periplasmic binding protein superfamily, where AmiC exists in monomeric and trimeric forms, the proportions of which depend on the presence of acetamide or butyramide.

Bacterial Proteins↗

PubMed1991

Mean residence times and distribution volumes for drugs undergoing linear reversible metabolism and tissue distribution and linear or nonlinear elimination from the central compartments.

Equations for the mean residence times in the body (MRT) and in the central compartment (MRTc) are derived for bolus central dosing of a drug and its metabolite which undergo linear tissue distribution and linear reversible metabolism but are eliminated either linearly or nonlinearly (Michaelis-Menten kinetics) from the central compartments. In addition, a new approach to calculate the steady-state volumes of distribution for nonlinear systems (reversible or nonreversible) is proposed based on disposition decomposition analysis. The application of these equations to a dual reversible two-compartment model is illustrated by computer simulations.

Computer Simulation↗

PubMed1996

Influence of surface free energies and cohesion parameters on pharmaceutical material interaction parameters-theoretical simulations.

PURPOSE: The aim of this study was to perform simulations of the influence of surface free energies and cohesion parameters on various interaction parameters within binary systems. METHODS: Using predictive equations derived from surface free energies and cohesion parameters originally proposed by Wu (2, 3) and by Rowe (4), values of interfacial tension, spreading and reduced spreading coefficients, interaction parameter and strength of interaction were simulated by means of a data processor. The influence of polar and disperse fractions of the two interacting materials was also examined. RESULTS: From the simulations, boundary conditions could be drawn: minimum interfacial tension, positive spreading coefficient, reduced spreading coefficient superior to unity, maximum value of the interaction parameter or of the strength of interaction. CONCLUSIONS: Simulations of the various parameters will help the formulator to select proper materials, eg. an agent that will efficiently bind some powdered substrate, a film-forming agent that will properly coat given cores or a material that will enhibit high interaction with a substrate.

Biocompatible Materials↗

PubMed1997

Bayesian population pharmacokinetic and pharmacodynamic analyses using mixture models.

Population studies of the pharmacokinetics or pharmacodynamics of drugs help us, learn about the variability in drug disposition and effects, information that can be used to treat future patients at safe and effective doses. We present a new approach to population modeling based on a weighted mixture of normal distributions having random weights and means. This method allows estimation of underlying continuous population distributions without prespecifying the parametric form or shape of these probability distributions. Additionally, this method can carry out nonparametric regression of pharmacokinetic or dynamic parameters on patient covariates while estimating the underlying distributions. Two examples illustrate the method and its flexibility.

Bayes Theorem↗

PubMed2003

[Prospective analysis of quantification of contrast media enhanced power Doppler sonography of equivocal breast lesions].

PURPOSE: To evaluate the potential diagnostic value of quantification of contrast enhanced power Doppler sonography in breast lesions. MATERIALS AND METHODS: Before and after bolus application of 300 mg/ml of the contrast agent Levovist (Schering, Germany), 76 randomly selected equivocal breast lesions (37 benign, 39 malignant) underwent power Doppler sonography. The data were acquired with a 7.5 MHz linear array transducer (Sonoline Versa Pro and Sonoline Elegra Plus, Siemens, Germany). Following postprocessing digitalization, color pixel density (CPD) and dynamic contrast enhancement were calculated and compared using different regions of the tumor (entire lesion, center and periphery). RESULTS: A measurable pixel signal prior to the application of the contrast agent was observed in 12 of 37 benign and in 18 of 39 malignant lesions (32.4 % and 46.2 %, respectively). After administration of the contrast agent, the lesions with measurable signal increased to 20 of 37 (54 %) and 25 of 39 (64.1 %), respectively. No statistically significant difference of dignity-related signal intensity could be verified before or after administration of the contrast agent. Maximum CPD and relative increase in pixel density after contrast medium was appreciably higher in the center of malignant lesions than in corresponding regions of benign lesions. CONCLUSION: Quantification of color pixel density is rather nonspecific for the discrimination of equivocal breast lesions and not suitable for clinical routine. Signal detection and quantification of color coded pixels enable comparison of dynamic contrast enhancement.

Adult↗

PubMed2002

On the simulation of protein folding by short time scale molecular dynamics and distributed computing.

There are proposals to overcome the current incompatibilities between the time scales of protein folding and molecular dynamics simulation by using a large number of short simulations of only tens of nanoseconds (distributed computing). According to the principles of first-order kinetic processes, a sufficiently large number of short simulations will include, de facto, a small number of long time scale events that have proceeded to completion. But protein folding is not an elementary kinetic step: folding has a series of early conformational steps that lead to lag phases at the beginning of the kinetics. The presence of these lag phases can bias short simulations toward selecting minor pathways that have fewer or faster lag steps and so miss the major folding pathways. Attempts to circumvent the lags by using loosely coupled parallel simulations that search for first-order transitions are also problematic because of the difficulty of detecting transitions in molecular dynamics simulations. Nevertheless, the procedure of using parallel independent simulations is perfectly valid and quite feasible once the time scale of simulation proceeds past the lag phases into a single exponential region.

Computer Simulation↗

PubMed1996

Internal packing conditions and fluctuations of amino acid residues in globular proteins.

In order to investigate the environmental conditions of amino acid residues in protein molecules, four kinds of packing studies (atomic, geometric, hydrophobic and hydration) were formulated and tested on two proteins; bovine pancreatic trypsin inhibitor (BPTI) and bovine pancreatic ribonuclease S (RNase S). The inter-relationship of these packings on the fluctuations of amino acid residues was analysed by comparing the packing results with the dynamical studies, such as the root-mean-square-deviation values of atomic displacements obtained from the trajectories of molecular dynamics simulation, temperature factor information from crystal structures and residue fluctuations in proteins from continuum model. These analyses yield information about the most fluctuating and most stabilizing residue sites. Comparison of the results obtained by these methods indicate a good agreement, specifying an inverse correlation between the residue packing and fluctuations. This kind of study is helpful in identifying the specific residue sites such as nucleation, receptor binding and antigenic determining sites which in a way indirectly correlates with the functional residues in protein molecules.

Amino Acids↗

PubMed1997

Fast direct Fourier methods, based on one- and two-pass coordinate transformations, yield accurate reconstructions of x-ray CT clinical images.

The conversion from polar to Cartesian coordinates can be carried out with two-pass algorithms. The paper describes two different methods based on concentric square frames and octagonal frames and their results, obtained with accurate interpolations based on the "moving window Shannon reconstruction' (MWSR). The embedding of these algorithms in direct Fourier methods (DFMs) of tomographic reconstruction is discussed. With respect to one-pass methods and to the use of octagonal frames, the square frame method makes it possible to carry out the first pass, a radial resampling, in the direct space, before computing 1D Fourier transforms (FTs) of projections. Reconstructions of clinical images from the raw data of a third-generation x-ray tomograph are presented and compared with those obtained with one-pass DFMs and with the convolution back-projection method (CBPM) performed by the instrument. The simple algorithm using square frames yields results in complete agreement with other DFM protocols and the CBPM. On a general-purpose computer, the execution of DFM protocols based on one-pass and two-pass coordinate transformations is 35 to 55 times faster than the CBPM and make the algorithms attractive for modern instrumentation.

Algorithms↗

PubMed2002

Estimating and evaluating the statistics of gapped local-alignment scores.

We present a novel maximum-likelihood-based algorithm for estimating the distribution of alignment scores from the scores of unrelated sequences in a database search. Using a new method for measuring the accuracy of p-values, we show that our maximum-likelihood-based algorithm is more accurate than existing regression-based and lookup table methods. We explore a more sophisticated way of modeling and estimating the score distributions (using a two-component mixture model and expectation maximization), but conclude that this does not improve significantly over simply ignoring scores with small E-values during estimation. Finally, we measure the classification accuracy of p-values estimated in different ways and observe that inaccurate p-values can, somewhat paradoxically, lead to higher classification accuracy. We explain this paradox and argue that statistical accuracy, not classification accuracy, should be the primary criterion in comparisons of similarity search methods that return p-values that adjust for target sequence length.

Algorithms↗

PubMed1988

The relationship of intralaboratory bias and imprecision on laboratories' ability to meet medical usefulness limits.

The previously described computer modeling technic empirically develops quantitative relationships between intralaboratory performance, as characterized by individual laboratories' coefficients of variation (CVs) and biases, and clinical "medical usefulness limits." These limits determine the magnitude of total analytic error, the combined effects of CV and bias that can be tolerated by the clinician. The computer model delineates all combinations of CV and bias compatible with specified medical usefulness limits. Both CV and bias are critical in determining a laboratory's ability to meet medical usefulness limits. For example, a laboratory with a 6% CV and zero bias will meet the +/- 10% or less total analytic error (medical usefulness limit) 90% of the time. If the medical usefulness limit is expanded to +/- 15%, a laboratory with a 6% CV can tolerate coexisting relative biases of up to 4% and still meet this limit 95% of the time. Plots of the limiting values for combinations of intralaboratory CV and bias are given that allow the laboratory's results to fall within medical usefulness limits of 2, 5, 10, 15, and 20%.

Clinical Laboratory Techniques↗

PubMed1994

Rapid numerical integration algorithm for finding the equilibrium state of a system of coupled binding reactions.

We have adapted a simple method of numerical integration to predict the equilibrium state of a population of components undergoing reversible association according to the Law of Mass Action. Its particular application is to populations of protein molecules in aqueous solution. The method is based on Euler integration but employs an adaptive step size: the time increment being reduced if it would make the concentration of any component negative and increased while the concentration of any component changes at greater than a specified rate. Parameters of the algorithm have been optimized empirically using a model set of binding equilibria with dissociation constants ranging from 10(-5) M to 10(-9) M. The method obtains the solution to a set of binding equilibria more rapidly than the conventional initial value methods (simple Euler, 4th order Runge-Kutta and variable-step Runge-Kutta methods were tested) for the same accuracy. A computer code in standard C is presented.

Algorithms↗

PubMed1999

Bayesian inference on biopolymer models.

MOTIVATION: Most existing bioinformatics methods are limited to making point estimates of one variable, e.g. the optimal alignment, with fixed input values for all other variables, e.g. gap penalties and scoring matrices. While the requirement to specify parameters remains one of the more vexing issues in bioinformatics, it is a reflection of a larger issue: the need to broaden the view on statistical inference in bioinformatics. RESULTS: The assignment of probabilities for all possible values of all unknown variables in a problem in the form of a posterior distribution is the goal of Bayesian inference. Here we show how this goal can be achieved for most bioinformatics methods that use dynamic programming. Specifically, a tutorial style description of a Bayesian inference procedure for segmentation of a sequence based on the heterogeneity in its composition is given. In addition, full Bayesian inference algorithms for sequence alignment are described. AVAILABILITY: Software and a set of transparencies for a tutorial describing these ideas are available at http://www.wadsworth.org/res&res/bioinfo/

Bayes Theorem↗