SEARCH · PubMed Health

Results for “Classification 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 1,603 records · Page 89Linked to original sources

PubMedDate not supplied

Accuracy of administrative data for identifying patients with pneumonia.

The goal of this study was to determine the accuracy and the impact of 5 different claims-based pneumonia definitions. Three International Classification of Diseases, Version 9, (ICD-9), and 2 diagnosis-related group (DRG)-based case identification algorithms were compared against an independent, clinical pneumonia reference standard. Among 10748 patients, 272 (2.5%) had pneumonia verified by the reference standard. The sensitivity of claims-based algorithms ranged from 47.8% to 66.2%. The positive predictive values ranged from 72.6% to 80.8%. Patient-related variables were not significantly different from the reference standard among the 3 ICD-9-based algorithms. DRG-based algorithms had significantly lower hospital admission rates (57% and 65% vs 73.2%), lower 30-day mortality (5.0% and 5.8% vs 10.7%), shorter length of stay (3.9 and 4.1 days vs 5.6 days), and lower costs (USD $4543 and USD $5159 vs USD $8585). Claims-based identification algorithms for defining pneumonia in administrative databases are imprecise. ICD-9-based algorithms did not influence patient variables in our population. Identifying pneumonia patients with DRG codes is significantly less precise.

Aged↗

PubMed2005

Reproducible clusters from microarray research: whither?

MOTIVATION: In cluster analysis, the validity of specific solutions, algorithms, and procedures present significant challenges because there is no null hypothesis to test and no 'right answer'. It has been noted that a replicable classification is not necessarily a useful one, but a useful one that characterizes some aspect of the population must be replicable. By replicable we mean reproducible across multiple samplings from the same population. Methodologists have suggested that the validity of clustering methods should be based on classifications that yield reproducible findings beyond chance levels. We used this approach to determine the performance of commonly used clustering algorithms and the degree of replicability achieved using several microarray datasets. METHODS: We considered four commonly used iterative partitioning algorithms (Self Organizing Maps (SOM), K-means, Clutsering LARge Applications (CLARA), and Fuzzy C-means) and evaluated their performances on 37 microarray datasets, with sample sizes ranging from 12 to 172. We assessed reproducibility of the clustering algorithm by measuring the strength of relationship between clustering outputs of subsamples of 37 datasets. Cluster stability was quantified using Cramer's v2 from a kXk table. Cramer's v2 is equivalent to the squared canonical correlation coefficient between two sets of nominal variables. Potential scores range from 0 to 1, with 1 denoting perfect reproducibility. RESULTS: All four clustering routines show increased stability with larger sample sizes. K-means and SOM showed a gradual increase in stability with increasing sample size. CLARA and Fuzzy C-means, however, yielded low stability scores until sample sizes approached 30 and then gradually increased thereafter. Average stability never exceeded 0.55 for the four clustering routines, even at a sample size of 50. These findings suggest several plausible scenarios: (1) microarray datasets lack natural clustering structure thereby producing low stability scores on all four methods; (2) the algorithms studied do not produce reliable results and/or (3) sample sizes typically used in microarray research may be too small to support derivation of reliable clustering results. Further research should be directed towards evaluating stability performances of more clustering algorithms on more datasets specially having larger sample sizes with larger numbers of clusters considered.

Cluster Analysis↗

PubMedDate not supplied

Decomposition and quantitative analysis of clinical electromyographic signals.

Procedures for the quantitative analysis of clinical electromyographic (EMG) signals detected simultaneously using selective or micro and non-selective or macro electrodes are presented. The procedures first involve the decomposition of the micro signals and then the quantitative analysis of the resulting motor unit action potential trains (MUAPTs) in conjunction with the associated macro signal. The decomposition procedures consist of a series of algorithms that are successively and iteratively applied to resolve a composite micro EMG signal into its constituent MUAPTs. The algorithms involve the detection of motor unit action potentials (MUAPs), MUAP clustering and supervised classification and they use shape and firing pattern information along with data dependent assignment criteria to obtain robust performance across a variety of EMG signals. The accuracy, extent and speed with which a set of 10 representative 20-30 s, concentric needle detected, micro signals could be decomposed are reported and discussed. The decomposition algorithms had a maximum and average error rate of 2.5% and 0.7%, respectively, on average assigned 88.7% of the detected MUAPs and took between 4 to 8 s. Quantitative analysis techniques involving average micro and macro MUAP shapes, the variability of micro MUAPs shapes and motor unit firing patterns are described and results obtained from analysis of the data set used to evaluate the decomposition algorithms are summarized and discussed.

Action Potentials↗

PubMed2006

Performance analysis of LVQ algorithms: a statistical physics approach.

Learning vector quantization (LVQ) constitutes a powerful and intuitive method for adaptive nearest prototype classification. However, original LVQ has been introduced based on heuristics and numerous modifications exist to achieve better convergence and stability. Recently, a mathematical foundation by means of a cost function has been proposed which, as a limiting case, yields a learning rule similar to classical LVQ2.1. It also motivates a modification which shows better stability. However, the exact dynamics as well as the generalization ability of many LVQ algorithms have not been thoroughly investigated so far. Using concepts from statistical physics and the theory of on-line learning, we present a mathematical framework to analyse the performance of different LVQ algorithms in a typical scenario in terms of their dynamics, sensitivity to initial conditions, and generalization ability. Significant differences in the algorithmic stability and generalization ability can be found already for slightly different variants of LVQ. We study five LVQ algorithms in detail: Kohonen's original LVQ1, unsupervised vector quantization (VQ), a mixture of VQ and LVQ, LVQ2.1, and a variant of LVQ which is based on a cost function. Surprisingly, basic LVQ1 shows very good performance in terms of stability, asymptotic generalization ability, and robustness to initializations and model parameters which, in many cases, is superior to recent alternative proposals.

Algorithms↗

PubMed1992

Multiple dipole modeling and localization from spatio-temporal MEG data.

An array of biomagnetometers may be used to measure the spatio-temporal neuromagnetic field or magnetoencephalogram (MEG) produced by neural activity in the brain. A popular model for the neural activity produced in response to a given sensory stimulus is a set of current dipoles, where each dipole represents the primary current associated with the combined activation of a large number of neurons located in a small volume of the brain. An important problem in the interpretation of MEG data from evoked response experiments is the localization of these neural current dipoles. We present here a linear algebraic framework for three common spatio-temporal dipole models: i) unconstrained dipoles, ii) dipoles with a fixed location, and iii) dipoles with a fixed orientation and location. In all cases, we assume that the location, orientation, and magnitude of the dipoles are unknown. With a common model, we show how the parameter estimation problem may be decomposed into the estimation of the time invariant parameters using nonlinear least-squares minimization, followed by linear estimation of the associated time varying parameters. A subspace formulation is presented and used to derive a suboptimal least-squares subspace scanning method. The resulting algorithm is a special case of the well-known MUltiple SIgnal Classification (MUSIC) method, in which the solution (multiple dipole locations) is found by scanning potential locations using a simple one dipole model. Principal components analysis (PCA) dipole fitting has also been used to individually fit single dipoles in a multiple dipole problem. Analysis is presented here to show why PCA dipole fitting will fail in general, whereas the subspace method presented here will generally succeed. Numerically efficient means of calculating the cost functions are presented, and problems of model order selection and missing moments are discussed. Results from a simulation and a somatosensory experiment are presented.

Algorithms↗

PubMed2002

Positron emission tomography partial volume correction: estimation and algorithms.

Partial volume effects in positron emission tomography (PET) lead to quantitative under- and over-estimations of the regional concentrations of radioactivity in reconstructed images and corresponding errors in derived functional or parametric images. The limited resolution of PET leads to "tissue-fraction" effects, reflecting underlying tissue heterogeneity, and "spillover" effects between regions. Addressing the former problem in general requires supplementary data, for example, coregistered high-resolution magnetic resonance images, whereas the latter effect can be corrected for with PET data alone if the point-spread function of the tomograph has been characterized. Analysis of otherwise homogeneous region-of-interest data ideally requires a combination of tissue classification and correction for the point-spread function. The formulation of appropriate algorithms for partial volume correction (PVC) is dependent on both the distribution of the signal and the distribution of the underlying noise. A mathematical framework has therefore been developed to accommodate both of these factors and to facilitate the development of new PVC algorithms based on the description of the problem. Several methodologies and algorithms have been proposed and implemented in the literature in order to address these problems. These methods do not, however, explicitly consider the noise model while differing in their underlying assumptions. The general theory for estimation of regional concentrations, associated error estimation, and inhomogeneity tests are presented in a weighted least squares framework. The analysis has been validated using both simulated and real PET data sets. The relations between the current algorithms and those published previously are formulated and compared. The incorporation of tensors into the formulation of the problem has led to the construction of computationally rapid algorithms taking into account both tissue-fraction and spillover effects. The suitability of their application to dynamic and static images is discussed.

Algorithms↗

PubMed2005

High-throughput pedigree drawing.

Family trees have long been a valuable visual tool for geneticists in identifying clusters of inherited traits and genotypes. As more data are collected, drawing the graphs by hand becomes impractical and, for this reason, we have developed the pedigree software CraneFoot. It can process any family graph with minimal computational cost by making a pedigree transformation that enables the use of a linear node positioning algorithm. The program is designed for automated drawing to printed media and efficient visual classification of genetically interesting families from large data sets. It also incorporates a robust pedigree topology check with detailed error messages.

Algorithms↗

PubMedDate not supplied

Application of four-layer neural network on information extraction.

This paper applies neural network to extract marsh information. An adaptive back-propagation algorithm based on a robust error function is introduced to build a four-layer neural network, and it is used to classify Thematic Mapper (TM) image of Zhalong Wetland in China and then extract marsh information. Comparing marsh information extraction results of the four-layer neural network with three-layer neural network and the maximum likelihood classifier, conclusion can be drawn as follows: the structure of the four-layer neural network and the adaptive back-propagation algorithm based on the robust error function is effective to extract marsh information. The four-layer neural network adopted in this paper succeeded in building the complex model of TM image, and it avoided the problem of great storage of remotely sensed data, and the adaptive back-propagation algorithm speeded up the descending of error. Above all, the four-layer neural network is superior to the three-layer neural network and the maximum likelihood classifier in the accuracy of the total classification and marsh information extraction.

Algorithms↗

PubMed2005

Automated determination of bubble grades from Doppler ultrasound recordings.

INTRODUCTION: One of the difficulties in the development of automated algorithms for the detection of bubbles in Doppler ultrasound recordings is that expert labels are only available on an aggregate basis, i.e., the expert provides a single label for a recording which may contain many bubbles. It is thus very difficult to determine whether an algorithm is correctly identifying the actual bubbles or simply identifying the correct number of events, but mislabeling some events that are not due to bubbles. METHODS: The analysis presented here shows that the classification probabilities for the detection of bubble events and other artifacts can be determined if a large number of recordings are available. DISCUSSION: Using a half-integer scoring system from 0-4 gives a bias of approximately 1-2% and a standard deviation that varies with the number of event sequences available, dropping from approximately 7.5% for 100 60-s recordings to 3% for 1000 60-s recordings. These values are larger if an integer scoring system is used, but using a scale finer than half-integers confers no extra benefit due to the fact that the expert labels the whole recording rather than individual bubbles. CONCLUSIONS: It is thus possible to estimate the classification probabilities with a reasonably high degree of accuracy, but difficult to show that one bubble detection algorithm is superior to another to any degree of statistical significance. Expert labels can be used to validate, but not to compare the performance of bubble detection algorithms.

Algorithms↗

PubMed1977

[Interpretation of pulse curves by means of the Walsh analysis].

A method for classification of medical data using the Walsh transformation is demonstrated. The Walsh spectrum was obtained by the algorithm of Andrews-Kane-Pratt. The spectral points were used to declare the signals normal or abnormal. The example described in this paper shows that in the case of pulse wave the classification is successful in 93 percent.

Carotid Arteries↗

PubMed2001

High order statistics for image classification.

This paper addresses the problem of image classification using local information which is aggregated to provide global representation of different image classes. Local information is adaptively extracted from an image database using Independent Component Analysis (ICA) which provides a set of localized, oriented, and band-pass filters selective to independent features of the images. Local representation using ICA techniques has been previously investigated by several researchers. However, very little work has been done on further use of these representations to provide more complex and global description of images. In this paper, we present an algorithm which uses the energy of a minimal set of ICA filters to provide class-specific signatures which are shown to be strongly discriminant. Computer simulations are carried on two image databases, one consisting of five classes--referred to as categories--(buildings, rooms, mountains, forests and beaches) and one consisting of a set of 30 objects from multiple views for viewpoint invariant object recognition. The classification performance of the algorithm using both Independent and Principal Component Analyses are reported and discussed.

Image Processing, Computer-Assisted↗

PubMed1992

Generation of a substructure library for the description and classification of protein secondary structure. I. Overview of the methods and results.

Protein secondary structure has been typically classified into four major classes--alpha-helices, extended strands, reverse turns, and loops. Available methods for secondary structure analysis utilize predefined structure templates to search for structural matches among proteins. By this approach a significant portion of a proteins backbone conformation is assigned to one of a limited number of conformations or, if unassigned, to random coil. To expand our ability to describe protein secondary structure, we have developed an algorithm that operates independently of a predefined structure template. The procedure uses two geometric descriptors, the linear distance and the backbone dihedral angle, to represent the conformation form the alpha-carbon coordinates. The algorithm functions by searching for conformationally equivalent, contiguous fragments without regard to secondary structural classification and is thus independent of the complexity of the backbone fold. The result is a library of conformationally equivalent structure fragments that exhibit some novel characteristics. The library contains features that reproduce the major secondary structure classes as well as defining conformations previously described only as random or undefined conformations. Additionally, the library defines several subclassifications of beta-strands. We present here a validation of this method and a presentation and discussion of the most significant results. In a second study, we report the results of application of this method to spectra-structure correlations in Fourier transform infrared spectroscopy.

Algorithms↗

PubMed1995

A 3D sequence-independent representation of the protein data bank.

Here we address the following questions. How many structurally different entries are there in the Protein Data Bank (PDB)? How do the proteins populate the structural universe? To investigate these questions a structurally non-redundant set of representative entries was selected from the PDB. Construction of such a dataset is not trivial: (i) the considerable size of the PDB requires a large number of comparisons (there were more than 3250 structures of protein chains available in May 1994); (ii) the PDB is highly redundant, containing many structurally similar entries, not necessarily with significant sequence homology, and (iii) there is no clear-cut definition of structural similarity. The latter depend on the criteria and methods used. Here, we analyze structural similarity ignoring protein topology. To date, representative sets have been selected either by hand, by sequence comparison techniques which ignore the three-dimensional (3D) structures of the proteins or by using sequence comparisons followed by linear structural comparison (i.e. the topology, or the sequential order of the chains, is enforced in the structural comparison). Here we describe a 3D sequence-independent automated and efficient method to obtain a representative set of protein molecules from the PDB which contains all unique structures and which is structurally non-redundant. The method has two novel features. The first is the use of strictly structural criteria in the selection process without taking into account the sequence information. To this end we employ a fast structural comparison algorithm which requires on average approximately 2 s per pairwise comparison on a workstation. The second novel feature is the iterative application of a heuristic clustering algorithm that greatly reduces the number of comparisons required. We obtain a representative set of 220 chains with resolution better than 3.0 A, or 268 chains including lower resolution entries, NMR entries and models. The resulting set can serve as a basis for extensive structural classification and studies of 3D recurring motifs and of sequence-structure relationships. The clustering algorithm succeeds in classifying into the same structural family chains with no significant sequence homology, e.g. all the globins in one single group, all the trypsin-like serine proteases in another or all the immunoglobulin-like folds into a third. In addition, unexpected structural similarities of interest have been automatically detected between pairs of chains. A cluster analysis of the representative structures demonstrates the way the "structural universe' is populated.

Algorithms↗

PubMed1996

Texture analysis of differently reconstructed PET images.

The insufficient number of projections and the statistics of radioactive decay are supposed to lead to different textures in the images reconstructed by the diverse algorithms used in positron emission tomography (PET). In our study, four reconstruction methods were examined, including the filtered backprojection and three algebraic approaches. Grey-level morphological operators were used to extract the texture. To quantify the observations, we used the texture parameters directionality, fractal dimension, and lacunarity. By increasing the number of iteration steps, each of the algebraic reconstruction methods evolve the texture from coarse to fine granularities, but surprisingly, during the iteration process all methods follow a quite similar 'trail' of possible textures within the active zone. The texture of the termination points of the movement have been compared with the result that the 'hils' and 'valleys' of the activity distribution are located in the same positions. The texture derived from the filtered backprojection is included in the set of textures produced by the algebraic methods. Artefacts are restricted to zones without activity. We found that the texture is independent of the number of counts when more than ten counts are registered in each pixel within the examined region of interest. A high reliability in the reconstructed images is given. All methods allow texture based tissue classification.

Algorithms↗

PubMed2002

Accuracy of three classifiers of acute gastrointestinal syndrome for syndromic surveillance.

ICD-9-coded emergency department (ED) diagnoses and free-text triage diagnoses are routinely collected data elements that have potential value for public health surveillance and early detection of epidemics. We constructed and measured performance of three classifiers for the detection of cases of acute gastrointestinal syndrome of public health significance: one used ICD-9-coded ED diagnosis as input data; the other two used free-text triage diagnosis. We measured the performance of these classifiers against the expert classification of cases based on review of ED reports. The sensitivity of the ICD-9-code classifier was 0.32, and the specificity was 0.99. The sensitivity of a naïve Bayes classifier using triage diagnoses was 0.63, the specificity was 0.94, and the area under the ROC curve was 0.82. A bigram Bayes classifier had sensitivity 0.38, specificity 0.94, and area under the ROC of 0.69. We conclude that a naive Bayes classifier of free-text triage diagnosis data provides more sensitive and earlier detection of cases of acute gastrointestinal syndrome than either a bigram Bayes classifier or an ICD-9 code classifier. The sensitivity achieved should be sufficient for syndromic surveillance system designed to detect moderate to large epidemics.

Acute Disease↗

PubMed1998

Time-frequency analysis using the matching pursuit algorithm applied to seizures originating from the mesial temporal lobe.

OBJECTIVES: The ability to analyze patterns of recorded seizure activity is important in the localization and classification of seizures. Ictal evolution is typically a dynamic process with signals composed of multiple frequencies; this can limit or complicate methods of analysis. The recently-developed matching pursuit algorithm permits continuous time-frequency analyses, making it particularly appealing for application to these signals. The studies here represent the initial applications of this method to intracranial ictal recordings. METHODS: Mesial temporal onset partial seizures were recorded from 9 patients. The data were analyzed by the matching pursuit algorithm were continuous digitized single channel recordings from the depth electrode contact nearest the region of seizure onset. Tine frequency energy distributions were plotted for each seizure and correlated with the intracranial EEG recordings. RESULTS: Periods of seizure initiation, transitional rhythmic bursting activity, organized rhythmic bursting activity and intermittent bursting activity were identified. During periods of organized rhythmic bursting activity, all mesial temporal onset seizures analyzed had a maximum predominant frequency of 5.3-8.4 Hz with a monotonic decline in frequency over a period of less than 60 s. The matching pursuit method allowed for time-frequency decomposition of entire seizures. CONCLUSIONS: The matching pursuit method is a valuable tool for time-frequency analyses of dynamic seizure activity. It is well suited for application to the non-stationary activity that typically characterizes seizure evolution. Time-frequency patterns of seizures originating from different brain regions can be compared using the matching pursuit method.

Algorithms↗

PubMed2006

Interpretable gene expression classifier with an accurate and compact fuzzy rule base for microarray data analysis.

An accurate classifier with linguistic interpretability using a small number of relevant genes is beneficial to microarray data analysis and development of inexpensive diagnostic tests. Several frequently used techniques for designing classifiers of microarray data, such as support vector machine, neural networks, k-nearest neighbor, and logistic regression model, suffer from low interpretabilities. This paper proposes an interpretable gene expression classifier (named iGEC) with an accurate and compact fuzzy rule base for microarray data analysis. The design of iGEC has three objectives to be simultaneously optimized: maximal classification accuracy, minimal number of rules, and minimal number of used genes. An "intelligent" genetic algorithm IGA is used to efficiently solve the design problem with a large number of tuning parameters. The performance of iGEC is evaluated using eight commonly-used data sets. It is shown that iGEC has an accurate, concise, and interpretable rule base (1.1 rules per class) on average in terms of test classification accuracy (87.9%), rule number (3.9), and used gene number (5.0). Moreover, iGEC not only has better performance than the existing fuzzy rule-based classifier in terms of the above-mentioned objectives, but also is more accurate than some existing non-rule-based classifiers.

Algorithms↗

PubMed2003

Symposium part 1: Should the Bethesda System terminology be used in diagnostic surgical pathology?: Point.

During the past 50 years, the histologic classification of cervical intraepithelial neoplasia (CIN) has evolved to incorporate the entire spectrum of genital papillomavirus infections, segregating those lesions with the higher risk of containing prototypic high-risk human papillomavirus types, and recently has meshed with treatment algorithms that include loop electrical excision procedures and follow-up alone. This review describes a classification system that divides CIN into categories of low-grade (CIN 1) and high-grade (CIN 2 and CIN 3). To successfully apply this system, the practitioner must efficiently exclude nonneoplastic entities and base the distinction of CIN 1 from CIN 2/3 on criteria that recognize the effects of viral oncogenes on replicating cells. This is achieved by basing the diagnosis of CIN 1 on uniform polarized epithelial growth, low mitotic index, low mitotic counts, and minimal parabasal cell anisokaryosis and coarse chromatin and CIN 2 on the presence of these features or abnormal mitoses. Simply put, the definition of CIN 2 (or higher) is the presence of atypical immature cells in the biopsy that if seen in a cytologic smear would merit a diagnosis of high-grade squamous intraepithelial lesion. In essence, a successful two-grade system requires careful application of cytologic criteria in a histologic milieu. This model is illustrated in a set of 25 images that underscore the importance of excluding benign changes (with the appropriate use of biomarkers), segregating unusual variants of low-grade squamous intraepithelial lesions, and identifying the morphologic transition to high-grade squamous intraepithelial lesions (CIN 2 or CIN 3) with an acceptable level of reproducibility.

Female↗