SEARCH · PubMed Health

Results for “Clustering 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 703 records · Page 39Linked to original sources

PubMed1993

Criteria for anatomical compatibility of the total artificial heart: computerized three-dimensional modeling of the cardiovascular anatomy.

A quantitative study of cardiovascular anatomy was performed by obtaining three-dimensional reconstructions from regular computed tomography scan images in 15 patients, all candidates for heart transplantation. Volumetric estimates of the cardiovascular structures were obtained from these three-dimensional reconstructions using data directly related to total artificial heart (TAH) implantations. By using computerized three-dimensional modeling of these structures, reproducible measurements of the parameters defining the shape and the anatomical connections of the intrathoracic space available for TAH implantation could be derived. The results are intended to be used for both technical and clinical applications such as computer-assisted drawing of the pericardial cavity and the anatomical connections (useful for improving the design of TAH) and combined statistical calculations (multiple regressions, cluster algorithm) of the measurement results, which will then enable the best selection to be made among two or three TAH models for each patient.

Aortography↗

PubMed1998

Computerized characterization of masses on mammograms: the rubber band straightening transform and texture analysis.

A new rubber band straightening transform (RBST) is introduced for characterization of mammographic masses as malignant or benign. The RBST transforms a band of pixels surrounding a segmented mass onto the Cartesian plane (the RBST image). The border of a mammographic mass appears approximately as a horizontal line, and possible speculations resemble vertical lines in the RBST image. In this study, the effectiveness of a set of directional textures extracted from the images before the RBST. A database of 168 mammograms containing biopsy-proven malignant and benign breast masses was digitized at a pixel size of 100 microns x 100 microns. Regions of interest (ROIs) containing the biopsied mass were extracted from each mammogram by an experienced radiologist. A clustering algorithm was employed for automated segmentation of each ROI into a mass object and background tissue. Texture features extracted from spatial gray-level dependence matrices and run-length statistics matrices were evaluated for three different regions and representations: (i) the entire ROI; (ii) a band of pixels surrounding the segmented mass object in the ROI; and (iii) the RBST image. Linear discriminant analysis was used for classification, and receiver operating characteristic (ROC) analysis was used to evaluate the classification accuracy. Using the ROC curves as the performance measure, features extracted from the RBST images were found to be significantly more effective than those extracted from the original images. Features extracted from the RBST images yielded an area (Az) of 0.94 under the ROC curve for classification of mammographic masses as malignant and benign.

Biopsy↗

PubMed2002

Genomic diversity within the genus Pediococcus as revealed by randomly amplified polymorphic DNA PCR and pulsed-field gel electrophoresis.

The genomic diversity of 33 previously assigned strains from six species within the genus Pediococcus was assessed by randomly amplified polymorphic DNA (RAPD) PCR and pulsed-field-gel electrophoresis (PFGE). The RAPD PCR patterns produced by two separate random primers, termed P1 (ACGCGCCCT) and P2 (ATGTAACGCC), were compared by the Pearson correlation coefficient and the unweighted pair group method with arithmetic averages clustering algorithm. Pattern variations between repeat samples set a strain discrimination threshold of less than 70% similarity. P1 and P2 primers alone and in combination produced 14, 21, and 28 distinct patterns, respectively. When each strain was assigned with a type strain with which it shared the highest level of similarity, both primers grouped 17 of the 27 strains to their proposed species. PFGE following genomic digestion with the restriction enzymes ApaI, NotI, and AscI produced 30, 32, and 28 distinct macrorestriction patterns, respectively. Specific DNA fragments within the NotI and AscI macrorestriction patterns for each strain were observed that allowed 27 of the 33 strains to be assigned to their proposed species. For example, following digestion with AscI, all Pediococcus parvulus strains were characterized by two DNA fragments, one of approximately 220 kb and another between 700 and 800 kb. The exceptions correlated with those observed with both RAPD PCR primers and included three P. damnosus and two P. pentosaceus strains that grew at temperatures regarded as nonpermissive for their proposed species but not for those with which they grouped.

Animals↗

PubMed1998

Molecular epidemiology of extended-spectrum beta-lactamase-producing Klebsiella pneumoniae isolates in a district hospital in Taiwan.

Thirty-one of 104 clinical isolates of Klebsiella pneumoniae collected over a period of 8 months were found to be putative extended-spectrum beta-lactamase (ESBL) producers. Isoelectric focusing and an iodine overlay agar method were used for preliminary identification of the ESBLs. They were further identified by DNA sequencing. Seventy-one percent of the isolates were found to produce SHV-5. The variation in the ESBL patterns of these isolates was slight, with only five patterns being identified. The strains were typed by pulsed-field gel electrophoresis (PFGE), and 16 different genotypes were identified. When the PFGE patterns were analyzed by the algorithmic clustering method called the unweighted-pair group method using arithmetic averages, five clusters were found. However, significant genetic variations were found among 11 isolates and between each cluster. A plasmid of 36 kb was found in all clinical isolates and in the transconjugants. Our results indicate that the increase in the number of ESBL-producing K. pneumoniae isolates in this hospital is due mainly to the dissemination of a resistance plasmid rather than to the clonal spread of a few epidemic strains.

Ceftazidime↗

PubMed1995

Comparison of multilayer neural network and Nearest Neighbor Classifiers for handwritten digit recognition.

The basic Nearest Neighbor Classifier (NNC) is often inefficient for classification in terms of memory space and computing time needed if all training samples are used as prototypes. These problems can be solved by reducing the number of prototypes using clustering algorithms and optimizing the prototypes using a special neural network model. In this paper, we compare the performance of the multilayer neural network and an Optimized Nearest Neighbor Classifier (ONNC) for handwritten digit recognition applications. We show that an ONNC can have the same recognition performance as an equivalent neural network classifier. The ONNC can be efficiently implemented using prototype and variable ranking, partial summation and distance triangular inequality based strategies. It requires the same memory space as, but less, training time and classification time than the neural network.

Computers↗

PubMed2004

Obstructive sleep apnea: MR imaging volume segmentation analysis.

PURPOSE: To retrospectively determine airway wall motion with volume segmentation of transverse cine magnetic resonance (MR) images in children with obstructive sleep apnea (OSA). MATERIALS AND METHODS: Transverse fast gradient-echo cine MR images of the hypopharynx were obtained at 1.5 T in 31 children with OSA (eight girls, 23 boys; mean age, 11.3 years) and 21 children free of airway symptoms who underwent MR imaging for other clinical indications (11 girls, 10 boys; mean age, 3.5 years). Volume segmentation with a k-means clustering algorithm was applied to transverse cine MR images to quantify airway volumes at each time. Airway wall motion for each child was described with standard deviation and range. Coefficient of variance and normalized range, which are independent of airway size, were used to compare groups (Kruskal-Wallis test). RESULTS: Plots of airway volume over time demonstrated large fluctuations during respiration in children with OSA and minimal fluctuations in controls; findings were consistent with airway distention and airway collapse in OSA. Average airway transverse volume was larger in the group with OSA than in the control group (OSA group, 2.52 mL; control group, 0.936 mL; P <.001). Mean standard deviation (OSA group, 0.840 mL; control group, 0.17 mL; P <.001) and mean range of airway cross section (OSA group, 3.552 mL; control group, 0.864 mL; P <.001) were larger in the group with OSA. Coefficient of variance (OSA group, 0.32; control group, 0.17; P <.001) and normalized range (OSA group, 1.42; control group, 0.96; P <.001) indicate statistically significant difference in airway dynamics in children with OSA. CONCLUSION: Volume segmentation of transverse cine MR images of the hypopharynx aids in quantification of increased airway wall motion in children with OSA. Transverse MR imaging demonstrates both airway distention and collapse in children with OSA.

Child↗

PubMed1987

Impact of intensive venous sampling on characterization of pulsatile GH release.

The effect of sampling intensity on quantitative properties of pulsatile growth hormone (GH) release was tested using an objective, statistically based pulse detection algorithm (Cluster). Seven normal young men had blood withdrawn at 5-min intervals for 24 h. The number of GH peaks/24 h (mean +/- SE) detected in the 5-min series (5.93 +/- 0.66) was significantly greater than the number of peaks documented in the constituent 10-min (3.45 +/- 0.28), 15-min (2.79 +/- 0.31), 20-min (2.86 +/- 0.64), 30-min (2.5 +/- 0.36), 45-min (2.21 +/- 0.21), and 60-min (1.93 +/- 0.23) series. The increased number of peaks detected with 5-min sampling reflected high-frequency pulsatile GH release occurring within the major GH secretory episodes. Both the mean widths and areas associated with peaks identified in the 5-min series were smaller than those documented with less intensive sampling. Peak amplitude did not change with sampling intensity. These data suggest that the major secretory episodes of GH release in normal young men encompass high-frequency GH secretory activity. That these high-frequency GH secretory events have not previously been described probably reflects the relatively infrequent sampling paradigms (e.g., every 20-30 min) commonly used in the past.

Adult↗

PubMed2000

Anti-inflammatory effects of triptolide in human bronchial epithelial cells.

Triptolide (PG490, 97% pure) is a diterpenoid triepoxide with potent anti-inflammatory and immunosuppressive effects in transformed human bronchial epithelial cells and T cells (Qiu D, Zhao G, Aoki Y, Shi L, Uyei A, Nazarian S, Ng JC-H, and Kao PN. J Biol Chem 274: 13443-13450, 1999). Triptolide, with an IC(50) of approximately 20-50 ng/ml, inhibits normal and transformed human bronchial epithelial cell expression of interleukin (IL)-6 and IL-8 stimulated by phorbol 12-myristate 13-acetate (PMA), tumor necrosis factor-alpha, or IL-1 beta. Nuclear runoff and luciferase reporter gene assays demonstrate that triptolide inhibits IL-8 transcription. Triptolide also inhibits the transcriptional activation, but not the DNA binding, of nuclear factor-kappa B. A cDNA array and clustering algorithm analysis reveals that triptolide inhibits expression of the PMA-induced genes tumor necrosis factor-alpha, IL-8, macrophage inflammatory protein-2 alpha, intercellular adhesion molecule-1, integrin beta(6), vascular endothelial growth factor, granulocyte-macrophage colony-stimulating factor, GATA-3, fra-1, and NF45. Triptolide also inhibits constitutively expressed cell cycle regulators and survival genes cyclins D1, B1, and A1, cdc-25, bcl-x, and c-jun. Thus anti-inflammatory, antiproliferative, and proapoptotic properties of triptolide are associated with inhibition of nuclear factor-kappa B signaling and inhibition of genes known to regulate cell cycle progression and survival.

Algorithms↗

PubMed2005

Functional organization of ganglion cells in the salamander retina.

Recently, we reported a novel technique for recording all of the ganglion cells in a retinal patch and showed that their receptive fields cover visual space roughly 60 times over in the tiger salamander. Here, we carry this analysis further and divide the population of ganglion cells into functional classes using quantitative clustering algorithms that combine several response characteristics. Using only the receptive field to classify ganglion cells revealed six cell types, in agreement with anatomical studies. Adding other response measures served to blur the distinctions between these cell types rather than resolve further classes. Only the biphasic off type had receptive fields that tiled the retina. Even when we attempted to split these classes more finely, ganglion cells with almost identical functional properties were found to have strongly overlapping spatial receptive fields. A territorial spatial organization, where ganglion cell receptive fields tend to avoid those of other cells of the same type, was only found for the biphasic off cell. We further studied the functional segregation of the ganglion cell population by computing the amount of visual information shared between pairs of cells under natural movie stimulation. This analysis revealed an extensive mixing of visual information among cells of different functional type. Together, our results indicate that the salamander retina uses a population code in which every point in visual space is represented by multiple neurons with subtly different visual sensitivities.

Action Potentials↗

PubMed2006

Proteomic analysis of the cerebrospinal fluid of patients with Creutzfeldt-Jakob disease.

So far, only the detection of 14-3-3 proteins in cerebrospinal fluid (CSF) has been accepted as diagnostic criterion for Creutzfeldt-Jakob disease (CJD). However, this assay cannot be used for screening because of the high rate of false-positive results, whereas patients with variant CJD are often negative for 14-3-3 proteins. The aim of this study was to compare the spot patterns of CSF by 2-dimensional polyacrylamide gel electrophoresis (2D-PAGE) to search for a CJD-specific spot pattern. We analyzed the CSF of 28 patients [11 CJD, 9 Alzheimer's disease (AD), 8 nondemented controls (NDC)] employing 2D-PAGE which was optimized for minimal volumes of CSF (0.1 ml; 7-cm strips). All samples were run at least three times, gels were silver stained and analyzed by an analysis software and manually revised. We could consistently match 268 spots which were then compared between all groups. By the use of 5 spots, we were able to differentiate CJD from AD or NDC with a sensitivity of 100%. CJD could also be distinguished from both groups by using a heuristic clustering algorithm of 2 spots. We conclude that this proteomic approach can differentiate CJD from other diseases and may serve as a model for other neurodegenerative diseases.

14-3-3 Proteins↗

PubMed1998

Time course of ADCw changes in ischemic stroke: beyond the human eye!

BACKGROUND AND PURPOSE: Using newly developed computerized image analysis, we studied the heterogeneity of apparent diffusion coefficient of water (ADCw) values in human ischemic stroke within 10 hours of onset. METHODS: Echo-planar trace diffusion-weighted images from 9 patients with focal cortical ischemic stroke were obtained within 10 hours of symptom onset. An Iterative Self-Organizing Data Analysis (ISODATA) clustering algorithm was implemented to segment different tissue types with a series of DW images. ADCw maps were calculated from 4 DW images on a pixel-by-pixel basis. The segmented zones within the lesion were characterized as low, pseudonormal, or high, expressed as a ratio of the mean+/-SD of ADCw of contralateral noninvolved tissue. RESULTS: The average ADCW in the ischemic stroke region within 10 hours of onset was significantly depressed compared with homologous contralateral tissue (626.6+/-76.8 versus 842.9+/-60.4x10(-6) mm2/s; P<0.0001). Nevertheless, ISODATA segmentation yielded multiple zones within the stroke region that were characterized as low, pseudonormal, and high. The mean proportion of low:pseudonormal:high was 72%:20%:8%. CONCLUSIONS: Despite low average ADCW, computer-assisted segmentation of DW MRI detected heterogeneous zones within ischemic lesions corresponding to low, pseudonormal, and high ADCw not visible to the human eye. This supports acute elevation of ADCw in human ischemic stroke and, accordingly, different temporal rates of tissue evolution toward infarction.

Acute Disease↗

PubMed1988

Fasting enhances growth hormone secretion and amplifies the complex rhythms of growth hormone secretion in man.

Studies in man have shown that the episodic release of growth hormone (GH) is infrequent and erratic, and unlike that in the rat does not appear to have discernible ultradian periodicities. However, these observations in nonfasted subjects may be invalid since mixed nutrients have unpredictable effects on GH release. Moreover, in the fed state basal GH levels are frequently undetectable, thus rendering the identification of low amplitude pulses unreliable. Accordingly, the 24-h pulsatile pattern of GH secretion obtained from repetitive venous sampling in six normal adult male subjects was examined during a control fed day and during the first and fifth days of a 5-d fast. The GH data were analyzed using two distinct methods: a discrete pulse detection algorithm (Cluster analysis) and Fourier expansion time-series, which allows fixed periodicities of secretory activity to be resolved. The 5-d fast resulted in a significant increase in discrete GH pulse frequency (5.8 +/- 0.7 vs. 9.9 +/- 0.7 pulses/24 h, P = 0.028), 24 h integrated GH concentration (2.82 +/- 0.50 vs. 8.75 +/- 0.82 micrograms.min/ml; P = 0.0002), and maximal pulse amplitude (5.9 +/- 1.1 vs. 12.3 +/- 1.6 ng/ml, P less than 0.005). While multiple low-amplitude sinusoidal periodicities were present on the control fed day, time-series analysis revealed enhancement of circadian and ultradian cycles on the first and fifth days of fasting. Concomitantly, fasting resulted in a decline (day 1 vs. day 5) in serum concentrations of somatomedin C (1.31 +/- 0.22 vs. 0.77 +/- 0.18 U/ml) and glucose (4.9 +/- 0.2 vs. 3.2 +/- 0.2 mmol/liter), and a marked rise in free fatty acid (0.43 +/- 0.12 vs. 1.55 +/- 0.35 mmol/liter) and acetoacetate (35 +/- 6 vs. 507 +/- 80 nmol/liter). We conclude that the acute nutritional status is an important determinant of spontaneous pulsatile GH secretion in man. Fast-induced enhancement of GH release is achieved through combined frequency (discrete pulses) and amplitude (sinusoidal periodicities) modulation. Such alterations in somatotropic hormone release may play an important role in substrate homeostasis during starvation.

Acetoacetates↗

PubMed1997

Luteinizing hormone pulse characteristics in depressed women.

OBJECTIVE: Luteinizing hormone (LH) pulse characteristics in depressed and normal women were compared to determine whether hypothalamic dysregulation in depression extends to the hypothalamic-pituitary-gonadal axis. METHOD: The subjects were 10 depressed and 13 normal comparison women admitted to a clinical research center. For each woman, an intravenous line was started and blood was withdrawn every 10 minutes for 8 hours. Blood samples were assayed for LH and LH pulse characteristics determined by using the computerized cluster algorithm of Veldhuis and Johnson. RESULTS: The depressed women differed significantly from the comparison women in LH pulse amplitude, rhythmicity, and area under the curve. CONCLUSIONS: Major depressive disorder is associated with abnormal regulation of luteinizing hormone. Gonadotropin regulation may provide a hormonal link between major depressive disorder and impaired fertility.

Adolescent↗

PubMed1977

Automated classification of cells in electron microscopic images of lymphoreticular tissue.

In classifying cells in tissue sections, one must consider the fact that only random projections of cells and of subcellular structures are available in the two-dimensional image. Therefore, measurement values that solely reflect the size of such projections cannot be taken on their own as a basis for cell classification. More complex morphologic features such as shape, texture and distribution pattern of cells and their components should be analyzed. Using cell nuclei as an example, the relationship between such features and geometric measurement values is evaluated. It can be shown that a well balanced combination of geometric parameters provides a suitable basis for reproducing the visual preclassification of lymphocytes in tissue sections. Moreover, using a cluster algorithm, which allows different levels of similarity to be defined, a hierarchical sequence of subclusters turns out, indicating the heterogeneity of the visually determined cell classes. Whether or not these subclusters can be correlated to functionally defined subpopulations of lymphocytes remains a matter for further investigation.

Autoanalysis↗

PubMed2004

RAG: RNA-As-Graphs web resource.

BACKGROUND: The proliferation of structural and functional studies of RNA has revealed an increasing range of RNA's structural repertoire. Toward the objective of systematic cataloguing of RNA's structural repertoire, we have recently described the basis of a graphical approach for organizing RNA secondary structures, including existing and hypothetical motifs. DESCRIPTION: We now present an RNA motif database based on graph theory, termed RAG for RNA-As-Graphs, to catalogue and rank all theoretically possible, including existing, candidate and hypothetical, RNA secondary motifs. The candidate motifs are predicted using a clustering algorithm that classifies RNA graphs into RNA-like and non-RNA groups. All RNA motifs are filed according to their graph vertex number (RNA length) and ranked by topological complexity. CONCLUSIONS: RAG's quantitative cataloguing allows facile retrieval of all classes of RNA secondary motifs, assists identification of structural and functional properties of user-supplied RNA sequences, and helps stimulate the search for novel RNAs based on predicted candidate motifs.

Computational Biology↗

PubMed2006

PhyloPat: phylogenetic pattern analysis of eukaryotic genes.

BACKGROUND: Phylogenetic patterns show the presence or absence of certain genes or proteins in a set of species. They can also be used to determine sets of genes or proteins that occur only in certain evolutionary branches. Phylogenetic patterns analysis has routinely been applied to protein databases such as COG and OrthoMCL, but not upon gene databases. Here we present a tool named PhyloPat which allows the complete Ensembl gene database to be queried using phylogenetic patterns. DESCRIPTION: PhyloPat is an easy-to-use webserver, which can be used to query the orthologies of all complete genomes within the EnsMart database using phylogenetic patterns. This enables the determination of sets of genes that occur only in certain evolutionary branches or even single species. We found in total 446,825 genes and 3,164,088 orthologous relationships within the EnsMart v40 database. We used a single linkage clustering algorithm to create 147,922 phylogenetic lineages, using every one of the orthologies provided by Ensembl. PhyloPat provides the possibility of querying with either binary phylogenetic patterns (created by checkboxes) or regular expressions. Specific branches of a phylogenetic tree of the 21 included species can be selected to create a branch-specific phylogenetic pattern. Users can also input a list of Ensembl or EMBL IDs to check which phylogenetic lineage any gene belongs to. The output can be saved in HTML, Excel or plain text format for further analysis. A link to the FatiGO web interface has been incorporated in the HTML output, creating easy access to functional information. Finally, lists of omnipresent, polypresent and oligopresent genes have been included. CONCLUSION: PhyloPat is the first tool to combine complete genome information with phylogenetic pattern querying. Since we used the orthologies generated by the accurate pipeline of Ensembl, the obtained phylogenetic lineages are reliable. The completeness and reliability of these phylogenetic lineages will further increase with the addition of newly found orthologous relationships within each new Ensembl release.

Algorithms↗

PubMed2002

Bacteria classification using Cyranose 320 electronic nose.

BACKGROUND: An electronic nose (e-nose), the Cyrano Sciences' Cyranose 320, comprising an array of thirty-two polymer carbon black composite sensors has been used to identify six species of bacteria responsible for eye infections when present at a range of concentrations in saline solutions. Readings were taken from the headspace of the samples by manually introducing the portable e-nose system into a sterile glass containing a fixed volume of bacteria in suspension. Gathered data were a very complex mixture of different chemical compounds. METHOD: Linear Principal Component Analysis (PCA) method was able to classify four classes of bacteria out of six classes though in reality other two classes were not better evident from PCA analysis and we got 74% classification accuracy from PCA. An innovative data clustering approach was investigated for these bacteria data by combining the 3-dimensional scatter plot, Fuzzy C Means (FCM) and Self Organizing Map (SOM) network. Using these three data clustering algorithms simultaneously better 'classification' of six eye bacteria classes were represented. Then three supervised classifiers, namely Multi Layer Perceptron (MLP), Probabilistic Neural network (PNN) and Radial basis function network (RBF), were used to classify the six bacteria classes. RESULTS: A [6 x 1] SOM network gave 96% accuracy for bacteria classification which was best accuracy. A comparative evaluation of the classifiers was conducted for this application. The best results suggest that we are able to predict six classes of bacteria with up to 98% accuracy with the application of the RBF network. CONCLUSION: This type of bacteria data analysis and feature extraction is very difficult. But we can conclude that this combined use of three nonlinear methods can solve the feature extraction problem with very complex data and enhance the performance of Cyranose 320.

Algorithms↗

PubMed1990

Variations of pulsatile growth hormone release in healthy short prepubertal boys.

Overnight growth hormone (GH) concentrations obtained by frequent venous sampling of 20 healthy, short prepubertal boys were evaluated using the objective pulse detection algorithm, CLUSTER. The resulting pulsatile characteristics were compared with those of 11 healthy prepubertal boys of normal stature and with those of nine prepubertal children with documented GH deficiency. Although no significant differences of pulsatile GH release were found between the normal and short subjects, a subset of the short prepubertal boys with significantly delayed skeletal ages had subnormal sum of GH pulse areas and sum of GH pulse amplitudes. The finding of a significant correlation in all subjects between growth velocity and the sum of GH pulse amplitudes is important, as the results are compatible with the hypothesis that alterations of amplitude-modulated GH release underlie the pathophysiology of suboptimal growth in some short prepubertal children.

Adolescent↗