SEARCH · PubMed Health

Results for “Algorithms”

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 685 records · Page 38Linked to original sources

PubMed2004

Fast-searching algorithm for vector quantization using projection and triangular inequality.

In this paper, a new and fast-searching algorithm for vector quantization is presented. Two inequalities, one used for terminating the searching process and the other used to delete impossible codewords, are presented to reduce the distortion computations. Our algorithm makes use of a vector's features (mean value, edge strength, and texture strength) to reject many unlikely codewords that cannot be rejected by other available approaches. Experimental results show that our algorithm is superior to other algorithms in terms of computing time and the number of distortion calculations. Compared with available approaches, our method can reduce the computing time and the number of distortion computations significantly. Compared with the best method of reducing distortion computation, our algorithm can further reduce the number of distortion calculations by 29% to 58.4%. Compared with the best encoding algorithm for vector quantization, our approach also further reduces the computing time by 8% to 47.7%.

Algorithms↗

PubMed2005

Feature-based wavelet shrinkage algorithm for image denoising.

A selective wavelet shrinkage algorithm for digital image denoising is presented. The performance of this method is an improvement upon other methods proposed in the literature and is algorithmically simple for large computational savings. The improved performance and computational speed of the proposed wavelet shrinkage algorithm is presented and experimentally compared with established methods. The denoising method incorporated in the proposed algorithm involves a two-threshold validation process for real-time selection of wavelet coefficients. The two-threshold criteria selects wavelet coefficients based on their absolute value, spatial regularity, and regularity across multiresolution scales. The proposed algorithm takes image features into consideration in the selection process. Statistically, most images have regular features resulting in connected subband coefficients. Therefore, the resulting subbands of wavelet transformed images in large part do not contain isolated coefficients. In the proposed algorithm, coefficients are selected due to their magnitude, and only a subset of those selected coefficients which exhibit a spatially regular behavior remain for image reconstruction. Therefore, two thresholds are used in the coefficient selection process. The first threshold is used to distinguish coefficients of large magnitude and the second is used to distinguish coefficients of spatial regularity. The performance of the proposed wavelet denoising technique is an improvement upon several other established wavelet denoising techniques, as well as being computationally efficient to facilitate real-time image-processing applications.

Algorithms↗

PubMed2006

A new algorithm for distorted fingerprints matching based on normalized fuzzy similarity measure.

Coping with nonlinear distortions in fingerprint matching is a challenging task. This paper proposes a novel algorithm, normalized fuzzy similarity measure (NFSM), to deal with the nonlinear distortions. The proposed algorithm has two main steps. First, the template and input fingerprints were aligned. In this process, the local topological structure matching was introduced to improve the robustness of global alignment. Second, the method NFSM was introduced to compute the similarity between the template and input fingerprints. The proposed algorithm was evaluated on fingerprints databases of FVC2004. Experimental results confirm that NFSM is a reliable and effective algorithm for fingerprint matching with nonliner distortions. The algorithm gives considerably higher matching scores compared to conventional matching algorithms for the deformed fingerprints.

Algorithms↗

PubMed2005

A general tool for the evaluation of spiral CT interpolation algorithms: revisiting the effect of pitch in multislice CT.

While multislice spiral computed tomography (CT) scanners are provided by all major manufacturers, their specific interpolation algorithms have been rarely evaluated. Because the results published so far relate to distinct particular cases and differ significantly, there are contradictory recommendations about the choice of pitch in clinical practice. In this paper, we present a new tool for the evaluation of multislice spiral CT z-interpolation algorithms, and apply it to the four-slice case. Our software is based on the computation of a "Weighted Radiation Profile" (WRP), and compares WRP to an expected ideal profile in terms of widening and heterogeneity. It provides a unique scheme for analyzing a large variety of spiral CT acquisition procedures. Freely chosen parameters include: number of detector rows, detector collimation, nominal slice width, helical pitch, and interpolation algorithm with any filter shape and width. Moreover, it is possible to study any longitudinal and off-isocenter positions. Theoretical and experimental results show that WRP, more than Slice Sensitivity Profile (SSP), provides a comprehensive characterization of interpolation algorithms. WRP analysis demonstrates that commonly "preferred helical pitches" are actually nonoptimal regarding the formerly distinguished z-sampling gap reduction criterion. It is also shown that "narrow filter" interpolation algorithms do not enable a general preferred pitch discussion, since they present poor properties with large longitudinal and off-center variations. In the more stable case of "wide filter" interpolation algorithms, SSP width or WRP widening are shown to be almost constant. Therefore, optimal properties should no longer be sought in terms of these criteria. On the contrary, WRP heterogeneity is related to variable artifact phenomena and can pertinently characterize optimal pitches. In particular, the exemplary interpolation properties of pitch = 1 "wide filter" mode are demonstrated.

Algorithms↗

PubMed2004

Efficient learning algorithms for three-layer regular feedforward fuzzy neural networks.

A key step of using gradient descend methods to develop learning algorithms of a regular feedforward fuzzy neural network (FNN) is to differentiate max--min functions, which contain max and min operations. The paper aims at several objectives. First, investigate further the differentiation of max--min functions. Second, employ general fuzzy numbers, which include triangular and trapezoidal fuzzy numbers as special cases to define a three-layer regular FNN. The general fuzzy numbers related can be approximately determined by their corresponding finite level sets. So, we can approximately represent the input-output (I/O) relationship of the regular FNN as functions of the endpoints of all finite level sets. Third, a fuzzy back-propagation algorithm is presented. And to speed up the convergence of the learning algorithm, a fuzzy conjugate gradient algorithm for fuzzy weights and biases is developed, furthermore, the convergence of the algorithm is analyzed, systematically. Finally, some real simulations demonstrate the efficiency of our learning algorithms. The regular FNN is applied to the approximate realization of fuzzy inference rules and fuzzy functions defined on given compact sets.

Algorithms↗

PubMed2005

Neural network learning algorithms for tracking minor subspace in high-dimensional data stream.

A novel random-gradient-based algorithm is developed for online tracking the minor component (MC) associated with the smallest eigenvalue of the autocorrelation matrix of the input vector sequence. The five available learning algorithms for tracking one MC are extended to those for tracking multiple MCs or the minor subspace (MS). In order to overcome the dynamical divergence properties of some available random-gradient-based algorithms, we propose a modification of the Oja-type algorithms, called OJAm, which can work satisfactorily. The averaging differential equation and the energy function associated with the OJAm are given. It is shown that the averaging differential equation will globally asymptotically converge to an invariance set. The corresponding energy or Lyapunov functions exhibit a unique global minimum attained if and only if its state matrices span the MS of the autocorrelation matrix of a vector data stream. The other stationary points are saddle (unstable) points. The globally convergence of OJAm is also studied. The OJAm provides an efficient online learning for tracking the MS. It can track an orthonormal basis of the MS while the other five available algorithms cannot track any orthonormal basis of the MS. The performances of the relative algorithms are shown via computer simulations.

Algorithms↗

PubMed2006

Efficient variant of algorithm FastICA for independent component analysis attaining the Cramér-Rao lower bound.

FastICA is one of the most popular algorithms for independent component analysis (ICA), demixing a set of statistically independent sources that have been mixed linearly. A key question is how accurate the method is for finite data samples. We propose an improved version of the FastICA algorithm which is asymptotically efficient, i.e., its accuracy given by the residual error variance attains the Cramér-Rao lower bound (CRB). The error is thus as small as possible. This result is rigorously proven under the assumption that the probability distribution of the independent signal components belongs to the class of generalized Gaussian (GG) distributions with parameter alpha, denoted GG(alpha) for alpha > 2. We name the algorithm efficient FastICA (EFICA). Computational complexity of a Matlab implementation of the algorithm is shown to be only slightly (about three times) higher than that of the standard symmetric FastICA. Simulations corroborate these claims and show superior performance of the algorithm compared with algorithm JADE of Cardoso and Souloumiac and nonparametric ICA of Boscolo et al. on separating sources with distribution GG (alpha) with arbitrary alpha, as well as on sources with bimodal distribution, and a good performance in separating linearly mixed speech signals.

Algorithms↗

PubMed2004

An efficient sequential learning algorithm for growing and pruning RBF (GAP-RBF) networks.

This paper presents a simple sequential growing and pruning algorithm for radial basis function (RBF) networks. The algorithm referred to as growing and pruning (GAP)-RBF uses the concept of "Significance" of a neuron and links it to the learning accuracy. "Significance" of a neuron is defined as its contribution to the network output averaged over all the input data received so far. Using a piecewise-linear approximation for the Gaussian function, a simple and efficient way of computing this significance has been derived for uniformly distributed input data. In the GAP-RBF algorithm, the growing and pruning are based on the significance of the "nearest" neuron. In this paper, the performance of the GAP-RBF learning algorithm is compared with other well-known sequential learning algorithms like RAN, RANEKF, and MRAN on an artificial problem with uniform input distribution and three real-world nonuniform, higher dimensional benchmark problems. The results indicate that the GAP-RBF algorithm can provide comparable generalization performance with a considerably reduced network size and training time.

Algorithms↗

PubMed2007

Implementation of the near-field signal redundancy phase-aberration correction algorithm on two-dimensional arrays.

Near-field signal-redundancy (NFSR) algorithms for phase-aberration correction have been proposed and experimentally tested for linear and phased one-dimensional arrays. In this paper the performance of an all-row-plus-two-column, two-dimensional algorithm has been analyzed and tested with simulated data sets. This algorithm applies the NFSR algorithm for one-dimensional arrays to all the rows as well as the first and last columns of the array. The results from the two column measurements are used to derive a linear term for each row measurement result. These linear terms then are incorporated into the row results to obtain a two-dimensional phase aberration profile. The ambiguity phase aberration profile, which is the difference between the true and the derived phase aberration profiles, of this algorithm is not linear. Two methods, a trial-and-error method and a diagonal-measurement method, are proposed to linearize the ambiguity profile. The performance of these algorithms is analyzed and tested with simulated data sets.

Algorithms↗

PubMed2004

Partially supervised learning using an EM-boosting algorithm.

Training data in a supervised learning problem consist of the class label and its potential predictors for a set of observations. Constructing effective classifiers from training data is the goal of supervised learning. In biomedical sciences and other scientific applications, class labels may be subject to errors. We consider a setting where there are two classes but observations with labels corresponding to one of the classes may in fact be mislabeled. The application concerns the use of protein mass-spectrometry data to discriminate between serum samples from cancer and noncancer patients. The patients in the training set are classified on the basis of tissue biopsy. Although biopsy is 100% specific in the sense that a tissue that shows itself to have malignant cells is certainly cancer, it is less than 100% sensitive. Reference gold standards that are subject to this special type of misclassification due to imperfect diagnosis certainty arise in many fields. We consider the development of a supervised learning algorithm under these conditions and refer to it as partially supervised learning. Boosting is a supervised learning algorithm geared toward high-dimensional predictor data, such as those generated in protein mass-spectrometry. We propose a modification of the boosting algorithm for partially supervised learning. The proposal is to view the true class membership of the samples that are labeled with the error-prone class label as missing data, and apply an algorithm related to the EM algorithm for minimization of a loss function. To assess the usefulness of the proposed method, we artificially mislabeled a subset of samples and applied the original and EM-modified boosting (EM-Boost) algorithms for comparison. Notable improvements in misclassification rates are observed with EM-Boost.

Algorithms↗

PubMed1991

Counting algorithms for linkage: correction to Morton and Collins.

In a recent paper, Morton & Collins (1990) claimed: (1) that the Lander-Green algorithm for genetic linkage analysis is not the EM algorithm for finding the maximum likelihood map; and (2) that a proposed alternative algorithm does have these properties. Here, we show that these assertions are both incorrect: the Lander-Green algorithm is an EM algorithm, while the Morton-Collins algorithm is not. We note that Morton and Collins concur with these conclusions.

Algorithms↗

PubMed2000

The development and validation of an algorithm for real-time computerised fetal heart rate monitoring in labour.

OBJECTIVE: To develop and validate a computerised algorithm for the interpretation of the characteristics of fetal heart rate monitoring in labour. DESIGN: Prospective observational study. SETTING: Labour ward in a tertiary hospital. SAMPLE: Intrapartum cardiotocograms from 24 pregnancies. METHODS: A computerised algorithm was developed to assess the fetal heart baseline rate, variability, the number of accelerations and the number of decelerations. Twenty five minute segments of cardiotocograms were interpreted by the algorithm and also by seven expert reviewers independently. The reviewers were unaware of the outcome of labour. The reliability of the characteristics of cardiotocography and the validity of the computerised algorithm were assessed using the intraclass correlation coefficient and weighted kappa statistic for continuous and ordinal variables respectively. RESULTS: The inter rater reliability of the baseline fetal heart rate and the number and type of decelerations was good (intraclass correlation coefficient 0.93, 0.93 and 0.79, respectively). The reliability of baseline variability (kappa = 0.27) and accelerations (intraclass correlation coefficient = 0.27) was poor. The computerised algorithm had good agreement with the reviewers for the baseline fetal heart rate (intraclass correlation coefficient 0.91 to 0.98) and the number of decelerations (intraclass correlation coefficient 0.82 to 0.91), but was less valid as regards the number of late decelerations (intraclass correlation coefficient 0.68 to 0.85) and the number of accelerations (intraclass correlation coefficient 0.06 to 0.80), and was invalid as regards baseline variability (kappa 0.00 to 0.34). CONCLUSIONS: The high level of validity of the computerised algorithm for the estimation of the baseline fetal heart rate and the number of decelerations justifies its further technical development.

Adolescent↗

PubMed1988

Heart rate correlation, response time and effect of previous exercise using an advanced pacing rate algorithm for temperature-based rate modulation.

A temperature-based algorithm to produce pacing rate that resembles chronotropic response to activity was developed. Measurement criteria for the algorithm included workload dependent rate increases with activity and response time within 60 seconds of exercise onset. To evaluate the algorithm, right ventricular blood temperature was recorded during rest and treadmill exercise in 25 patients with implanted Kelvin 500 pacemakers (Cook Pacemaker). Patients included 16 males and nine females, ages 44-81 (mean 72). Indications for pacing were sinus node disease, atrioventricular block and atrial fibrillation with slow ventricular response. Temperature changes reflected physical activity as well as emotional stress. The algorithm was based on the rate of change (dT/dt), the relative change (delta T) and the baseline history (T) of temperature. At exercise onset, a rapid, brief drop in temperature (dT/dt) typically occurred due to peripheral vasodilation, causing prompt increase in pacing rate. As exercise continued, the increase in metabolic rate caused dT/dt as well as delta T to increase, further increasing pacing rate. After exercise, temperature returned to resting level which correspondingly decreased the pacing rate. Sensitivity of the algorithm to temperature variations, and the upper and lower pacing rate limits were programmable to adapt to individual patient needs. The rates produced by the algorithm mimicked intrinsic rate response for various activity levels and produced a mean response time of 16 seconds from exercise onset. Previous exercise had no significant effect on response time. Correlation between normal chronotropic response and simulated pacing rate from five exercise tests was 0.92. These results show good specificity and refute the statement that blood temperature yields a slow response.

Adult↗

PubMed1990

Endless-loop tachycardias: description and first clinical results of a new fully automatic protection algorithm.

Endless-loop tachycardia (ELT) is one of the most common pacemaker mediated tachycardia. An innovative ELT protection algorithm has proven to be clinically effective. A new improved version that will eliminate the need to program any parameter is now under clinical evaluation. Nine patients entered the study: six men and three women, aged 52 +/- 22 years. This automatic algorithm needs only 10 cycles to detect and confirm an ELT. Three hundred thirty-three ELTs lasting more than 9 cycles have been induced and analyzed. The total results are the following: mean duration: 6.7 sec +/- 3.1; mean ELT rate: 137 +/- 21.9 bpm, mean programmed upper rate limit (URL): 142.5 +/- 26.5 bpm (Only 70% of ELTs presented rates equal to programmed URL). (1) ELTs reduced by postventricular atrial refractory period (PVARP) extension on one cycle: 291 ELTs (87%). ELT rate: 128.5 +/- 18.2 bpm. (2) Retrograde block: algorithm operation may induce a retrograde block due to a short atrioventricular delay (AVD) applied during the confirmation phase to discriminate an ELT from a stable sinus rhythm. Thirty-two ELTs (10%) have been reduced and detected on a retrograde block occurrence. (3) Algorithm failure due to an unstable ventriculoatrial conduction time (VACT) even at fixed rate or to a retrograde Wenckebach behavior on AVD reduction during the confirmation phase. A total of 10 algorithms failed to detect or confirm an ELT have been recorded (3%). Mean duration: 8.2 +/- 4.2 sec, mean ELT rate: 148.9 +/- 14.3 bpm. This new fully automatic algorithm has reduced 97% of ELTs, including high rate episodes (100-175 bpm).(ABSTRACT TRUNCATED AT 250 WORDS)

Algorithms↗

PubMed1993

Simultaneously collected monopolar and discrete bipolar electrograms: comparison of activation time detection algorithms.

Designation of the time of local activation is fundamental to electrophysiological mapping. In normal myocardium, the minimum slope in extracellular monopolar (MP) electrograms has been linked through simultaneous intracellular and extracellular recordings to phase 0 of the action potential. However, no similar correlation has been demonstrated for a parameter from bipolar (BP) electrograms, now commonly used during electrophysiological studies and intraoperative mapping. The purpose of this work is to compare the activation time, determined according to several common algorithms applied to BP electrograms, with the time of the minimum slope in MP electrograms. Simultaneous normal epicardial MP and BP electrograms were acquired from subjects undergoing surgery for Wolf-Parkinson-White Syndrome and from dogs. The activation time in BP electrograms was defined by four algorithms: (1) peak (P); (2) greatest absolute slope (S); (3) zero crossing of the segment containing the greatest slope (FZC); and (4) morphological (M). Each was compared to the time of the minimum slope in the simultaneously recorded MP response. The incidence of outliers was tabulated. The distribution of activation times computed using each BP algorithm was statistically different from the distribution of activation times derived from MP electrograms. M performed best (absolute difference: 2.6 +/- 2.9 msec; cor coef: 0.9925 in man). The M, P, FZC, and S algorithms produced 3.2%, 3.5%, 4.7% and 4.7% outliers, respectively. The overall performance of a morphologically based algorithm is superior to simplistic BP algorithms based only on slope or peak.(ABSTRACT TRUNCATED AT 250 WORDS)

Algorithms↗

PubMed1997

A simple electrocardiographic algorithm for detecting ventricular tachycardia.

The purpose of this study was to determine whether a simple ECG algorithm could be developed for predicting susceptibility to ventricular tachyarrhythmias (VT) as defined by sustained spontaneous or inducible VT. Two different QT dispersion algorithms were determined by the difference between the longest and shortest QT interval measured in three orthogonal leads (I, aVF, V1; QTD3), and at least 11 of 12 leads (QTD12) from the 12-lead ECG. These QT dispersion algorithms were investigated (with and without the QRS duration from the 12-lead ECG) and compared to the signal-averaged ECG (SAECG) in order to determine their sensitivity and specificity for detecting VT. Only patients who underwent SAECG and were referred for programmed electrical stimulation were included in this study. A positive SAECG was defined by filtered QRS duration > 114 ms, and/or low amplitude signal duration > 38 ms, and/or root mean square voltage in the last 40 ms of < 20 microV. Sixty patients were enrolled in this study with a mean age of 63 +/- 2 years. Fifty-five percent of the patients had coronary artery disease. A simple ECG algorithm consisting of the sum of QTD3 plus the QRS duration had a sensitivity and specificity of 90% and 63%, respectively, wheras the SAECG had a sensitivity and specificity of 60% and 63%, respectively (P = 0.022). We conclude that a simple ECG algorithm is more sensitive than the SAECG for predicting VT. This algorithm combines two easily measured variables obtained from the 12-lead ECG, and can easily be performed without expensive computer equipment.

Adolescent↗

PubMed1998

An algorithm for automatic measurement of stimulation thresholds: clinical performance and preliminary results.

We have developed an algorithmic method for automatic determination of stimulation thresholds in both cardiac chambers in patients with intact atrioventricular (AV) conduction. The algorithm utilizes ventricular sensing, may be used with any type of pacing leads, and may be downloaded via telemetry links into already implanted dual-chamber Thera pacemakers. Thresholds are determined with 0.5 V amplitude and 0.06 ms pulse-width resolution in unipolar, bipolar, or both lead configurations, with a programmable sampling interval from 2 minutes to 48 hours. Measured values are stored in the pacemaker memory for later retrieval and do not influence permanent output settings. The algorithm was intended to gather information on continuous behavior of stimulation thresholds, which is important in the formation of strategies for programming pacemaker outputs. Clinical performance of the algorithm was evaluated in eight patients who received bipolar tined steroid-eluting leads and were observed for a mean of 5.1 months. Patient safety was not compromised by the algorithm, except for the possibility of pacing during the physiologic refractory period. Methods for discrimination of incorrect data points were developed and incorrect values were discarded. Fine resolution threshold measurements collected during this study indicated that: (1) there were great differences in magnitude of threshold peaking in different patients; (2) the initial intensive threshold peaking was usually followed by another less intensive but longer-lasting wave of threshold peaking; (3) the pattern of tissue reaction in the atrium appeared different from that in the ventricle; and (4) threshold peaking in the bipolar lead configuration was greater than in the unipolar configuration. The algorithm proved to be useful in studying ambulatory thresholds.

Adult↗

PubMed1999

A new defibrillator discrimination algorithm utilizing electrogram morphology analysis.

Inappropriate therapies delivered by implantable cardioverter defibrillators (ICDs) for supraventricular arrhythmias remain a common problem, particularly in the event of rapidly conducted atrial fibrillation or marked sinus tachycardia. The ability to differentiate between ventricular tachycardia and supraventricular arrhythmias is the major goal of discrimination algorithms. Therefore, we developed a new algorithm, SimDis, utilizing morphological features of the shocking electrograms. This algorithm was developed from electrogram data obtained from 36 patients undergoing ICD implantation. An independent test set was evaluated in 25 patients. Recordings were made in sinus rhythm, sinus tachycardia, and following the induction of ventricular tachycardia and atrial fibrillation. The arrhythmia complex is defined as wide if the duration is at least 30% greater than the template in sinus rhythm. For narrow complexes, four maximum and minimum values were measured to form a 4-element feature vector, which was compared with a representative feature vector during normal sinus rhythm. For each rhythm, any wide complex was classified as ventricular tachycardia. For narrow complexes, the second step of the algorithm compared the electrogram with the template, computing similarity and dissimilarity values. These values were then mapped to determine if they fell within a previously established discrimination boundary. On the independent test set, the SimDis algorithm correctly classified 100% of ventricular tachycardias (27/27), 98% of sinus tachycardias (54/55), and 100% of episodes of atrial fibrillation (37/37). We conclude that the SimDis algorithm yields high sensitivity (100%) and specificity (99%) for arrhythmia discrimination, using the computational capabilities of an ICD system.

Algorithms↗