SEARCH · PubMed Health

Results for “cluster analysis”

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 217 records · Page 12Linked to original sources

PubMed1993

Describing the homeless mentally ill: cluster analysis results.

Presented descriptive data on a group of homeless, mentally ill individuals (N = 108) served by a two-site demonstration project, funded by NIMH. Comparing results with those from other studies of this population produced some differences and some similarities. Cluster analysis techniques were applied to the data, producing a 4-group solution. Data validating the cluster solution are presented. It is suggested that the cluster results provide a more meaningful and useful method of understanding the descriptive data. Results suggest that while the population of individuals served as homeless and mentally ill is quite heterogeneous, many have well-developed functioning skills--only one cluster, making up 35.2% of the sample, fits the stereotype of the aggressive, psychotic individual with skill deficits in many areas. Further discussion is presented concerning the implications of the cluster analysis results for demonstrating contextual effects and thus better interpreting research results from other studies and assisting in future services planning.

Adult↗

PubMed1996

Subpopulation of dogs with severe brain parenchymal beta amyloidosis distinguished with cluster analysis.

A study of the brains of 30 dogs, mongrels from 6.5 to 26.5 years of age, revealed that all dogs older than 13 years of age develop amyloid-beta-positive plaques. Cluster analysis based on the age of the dogs and the numerical density of amyloid-positive plaques stained with monoclonal antibody 4G8 (17-24aa) revealed that the population of old dogs consists of two subpopulations: one with a very low (0.8/mm2 on average) and other with a high (19.2/mm2 on average) numerical density of plaques. These two groups (19.5 and 19.1 years of age, respectively) appear to emerge from the younger group (12.2 years of age on average), with moderate (2.2/mm2 on average) numerical density of 4G8-positive plaques. These data may indicate that only a portion of the mongrel population (43%) is susceptible to amyloidosis beta or that only this severely affected subpopulation was exposed to a factor or factors inducing this pathology and developed severe cortical amyloidosis that correlates with age. Dog plaques are only of the diffuse type, with nonfibrillar, thioflavin S-, and Congo red-negative amyloid in all groups distinguished by cluster analysis. Only from 10% of 4G8-positive plaques in the mildly affected group to 29% in the severely and 37% in the moderately affected group are Bielschowsky positive. In the younger, moderately affected group, 6E10 (1-17aa)-positive plaques prevail. In the two old groups with severe and weak changes, almost all 4G8-positive plaques are also 6E10-positive. Carboxy-terminal region immunocytochemistry reveals that BC42-positive plaques are numerous, whereas BC40-positive plaques are few or absent. The differences in the silver-positivity of plaques and their immunoreactivity in both the amino- and carboxy-terminal regions may reflect differences in amyloid-beta deposition and resolution. Dog parenchymal amyloidosis beta appears to be a model for the study of diffuse plaques.

Amyloid beta-Peptides↗

PubMed2000

Analysis of input functions from different arterial branches with gamma variate functions and cluster analysis for quantitative blood volume measurements.

Regional cerebral blood volume (rCBV) provides valuable information about the nature and progress of diseases of the central nervous system. While relative rCBV maps can be derived directly from dynamic susceptibility contrast data, the arterial input function (AIF) has to be measured for absolute rCBV quantification. For determination of the AIF pixels located completely within a feeding artery must be selected. However, by using a region-of-interest (ROI) based selection some confounding effects can occur, especially if single shot echo planar imaging (EPI) with low spatial resolution is used. In this study we analyzed the influence of partial volume effects and spatial misregistration due to frequency shifts induced by paramagnetic contrast agents. We analyzed AIFs from the internal carotid artery (ICA), the vertebral artery (VA) and the middle cerebral artery (MCA) using gamma variate function based parameterization. The concentration time curves (CTC) of several pixels which were selected on the basis of strong signal drop appeared distorted during the bolus passage. Moreover, the amplitudes of input functions derived from the MCA were smaller by a factor of three as compared to those of the ICA and VA. Simulations revealed that these effects can be attributed to a spatial shift of the vessel along phase-encoding direction during the passage of the bolus. We therefore developed a procedure for a pixel selection based on cluster analysis which classifies pixels according to the parameters of the fitted gamma variate functions. This approach accounted for misregistration of the vessel and yielded very consistent results for a group of normal subjects.

Brain↗

PubMed1987

Variation in international cancer mortality: factor and cluster analysis.

Mortality rates for cancers of 13 sites in 34 countries were analysed using two data reduction techniques, factor and cluster analysis. Factor analysis identified two independent underlying factors which appear to influence cancer mortality patterns. The first factor, which appears to be related to affluence, may represent the combined effects of high fat diets and cigarette smoking common in developed countries. The second factor may reflect the common consumption of beverages of a high tannin content such as tea, red wine and 'mate' as well as the smoking or chewing of black tobacco. Two factor scores were computed for each country, and the countries were then ranked according to their scores on each factor. Cluster analysis aggregated countries into seven distinct groups using these factor scores as the clustering criteria. Each of the groups thus defined displays a distinctive profile of site-specific cancer mortality rates. This methodology shows promise as a means of summarizing large sets of data on morbidity and mortality from a variety of cancers (and possibly other chronic diseases as well) in diverse populations.

Humans↗

PubMedDate not supplied

Cluster analysis of nutritional and immunological indicators for identification of high risk surgical patients.

In spite of the many anthropometric, biohumoral, and immunologic parameters employed in the nutritional assessment of hospitalized patients, it is difficult in clinical practice to evaluate accurately the degree and type of malnutrition and to assess the prognostic significance of this determination. The purpose of this study is to evaluate nutritional status of surgical patients by means of cluster analysis in orderr to identify different nutritional patterns and to evaluate their clinical and prognostic significance. Nutritional assessment of 71 surgical patients was carried out at admission, and the sets of data were evaluated by means of cluster analysis. Four clusters with different nutritional patterns were identified. The incidence of clinical variables (type of disease, postoperative sepsis, palliative procedures, mortality at 6 months, etc.) in each cluster was determined in order to evaluate their clinical and prognostic significance. Cluster 1 showed minor variations of the indicators, including most of the controls presented the lowest incidence of sepsis, palliative procedures, and mortality at 6 months. It was then considered as a reference group representative of the normal nutritional condition at our institution. The other three clusters showed major variations of nutritional indicators and represent poorer risk clinical conditions. Sepsis, palliative procedures and mortality rate were significantly more frequent in these clusters (p less than 0.05, p less than 0.001, p less than 0.05). A different distribution in the clusters was recorded in gastrointestinal tract cancers and other neoplasms. Only the incidence of gastrointestinal tract cancers increases progressively in the clusters with poorer prognosis, suggesting that this type of neoplasia is more frequently associated with major changes of nutritional status.

Adult↗

PubMedDate not supplied

Using cluster analysis for medical resource decision making.

Escalating costs of health care delivery have in the recent past often made the health care industry investigate, adapt, and apply those management techniques relating to budgeting, resource control, and forecasting that have long been used in the manufacturing sector. A strategy that has contributed much in this direction is the definition and classification of a hospital's output into "products" or groups of patients that impose similar resource or cost demands on the hospital. Existing classification schemes have frequently employed cluster analysis in generating these groupings. Unfortunately, the myriad articles and books on clustering and classification contain few formalized selection methodologies for choosing a technique for solving a particular problem, hence they often leave the novice investigator at a loss. This paper reviews the literature on clustering, particularly as it has been applied in the medical resource-utilization domain, addresses the critical choices facing an investigator in the medical field using cluster analysis, and offers suggestions (using the example of clustering low-vision patients) for how such choices can be made.

Algorithms↗

PubMedDate not supplied

An attempt to validate Akiskal's classification of chronic depression using cluster analysis.

A numerical taxonomy program that overcomes several problems often associated with cluster analysis was used in an attempt to validate Akiskal's classification of chronic depression. Whereas two of the four groups produced by this technique demonstrated features of either Akiskal's character spectrum disorder or chronic secondary depression, one group appeared to subsume both Akiskal's chronic primary depression and subaffective dysthymia. Within this group, early-onset patients exhibited characteristics of subaffective dysthymia and late-onset patients demonstrated features of chronic primary depression. Family history of depression and family history of alcohol dependence were potent discriminators between groups. There were points of similarity between the typology and several classifications of episodic depression, for example, Winokur's classification of unipolar depression and other typologies derived from multivariate analysis.

Adolescent↗

PubMed1986

Codon usage in yeast: cluster analysis clearly differentiates highly and lowly expressed genes.

Codon usage data has been compiled for 110 yeast genes. Cluster analysis on relative synonymous codon usage revealed two distinct groups of genes. One group corresponds to highly expressed genes, and has much more extreme synonymous codon preference. The pattern of codon usage observed is consistent with that expected if a need to match abundant tRNAs, and intermediacy of tRNA-mRNA interaction energies are important selective constraints. Thus codon usage in the highly expressed group shows a higher correlation with tRNA abundance, a greater degree of third base pyrimidine bias, and a lesser tendency to the A+T richness which is characteristic of the yeast genome. The cluster analysis can be used to predict the likely level of gene expression of any gene, and identifies the pattern of codon usage likely to yield optimal gene expression in yeast.

Base Composition↗

PubMed2004

Quantification of human atherosclerotic plaques using spatially enhanced cluster analysis of multicontrast-weighted magnetic resonance images.

One of the current limitations of magnetic resonance imaging (MRI) is the lack of an objective method to classify plaque components. Here we present a cluster analysis technique that can objectively quantify and classify MR images of atherosclerotic plaques. We obtained three-dimensional (3D) images from 12 human coronary artery specimens on a 9.4T imaging system using multicontrast-weighted fast spin-echo (T1-, proton density-, and T2-weighted) imaging with an isotropic voxel size of 39 micro. Spatially enhanced cluster analysis (SECA) was performed on multicontrast MR images, and the resulting segmentation was evaluated against histological tracings. To visualize the overall structure of plaques, the MR images were rendered in 3D. The specimens exhibited lesions of American Heart Association (AHA) plaque classification types I-VI. Both MR images and histological sections were independently reviewed, categorized, and compared. Overall, the classification obtained from the cluster-analyzed MR and histopathology images showed very good agreement for all AHA types (92%, Cohen's kappa = 0.89, P < 0.0001). All plaque types were identified and quantified by SECA with a high degree of correlation between cluster-analyzed MR and manually traced histopathology data. MRI combined with SECA provides an objective method for atherosclerotic plaque component characterization and quantification.

Analysis of Variance↗

PubMed2003

[Typing of hemorrhagic fever with renal syndrome by cluster analysis].

OBJECTIVE: To make an inquiry into method of typing of hemorrhagic fever with renal syndrome (HFRS). METHOD: Average monthly rates were calculated on the basis of data from 1995 to 1999, then cluster analysis was carried out to type out endemic areas. RESULTS: Compared with the results of 36 surveillance spots from 1980 to 1992, twenty-four surveillance spots had the same results (66.7%). Twenty-three surveillance spots had the same results with the original data in 1999 (82.1%). CONCLUSION: HFRS incidences increased in spring or in summer, but decreased in autumn or in winter. Cluster analysis seemed to be a supplementary method in distinguishing the epidemic types for HFRS.

Animals↗

PubMed2002

Cluster analysis of activity-time series in motor learning.

Neuroimaging studies of learning focus on brain areas where the activity changes as a function of time. To circumvent the difficult problem of model selection, we used a data-driven analytic tool, cluster analysis, which extracts representative temporal and spatial patterns from the voxel-time series. The optimal number of clusters was chosen using a cross-validated likelihood method, which highlights the clustering pattern that generalizes best over the subjects. Data were acquired with PET at different time points during practice of a visuomotor task. The results from cluster analysis show practice-related activity in a fronto-parieto-cerebellar network, in agreement with previous studies of motor learning. These voxels were separated from a group of voxels showing an unspecific time-effect and another group of voxels, whose activation was an artifact from smoothing.

Adult↗

PubMedDate not supplied

Documenting patterns of nursing interventions using cluster analysis.

Use of inferential statistics in research applications of the Nursing Intervention Classification has been rare, yet use of these statistical techniques is needed to answer questions related to intervention patterns. Using data from a descriptive study of 3,733 visits documented by 19 adult nurse practitioner students, hierarchical agglomerative cluster analysis was used to determine whether meaningful nursing intervention patterns could be depicted. Eight intervention clusters were derived, replicated, and validated. Clusters of intervention classes differed in the type of nursing and medical diagnosis, amount of time the nurse spent during the visit, cost of visit, and the age of the patient. Clustering supported the utility of standardized nursing diagnosis and intervention typologies to identify actual practice patterns of adult nurse practitioner students. Cluster analysis is a valuable data analysis strategy when analyzing multiple related variables.

Adolescent↗

PubMed1998

A cluster analysis of not-in-treatment drug users at risk for HIV infection.

The purpose of the analysis described here was to classify not-in-treatment drug users participating in the National Institute on Drug Abuse (NIDA)-sponsored Cooperative Agreement study into several "homogeneous" HIV risk groups using cluster analysis. Data for this analysis (N=17,778) were collected at 19 study sites in the United States and Puerto Rico. Measures selected for the cluster analysis were limited to (a) current drug use and HIV risk behaviors, (b) mutually exclusive behaviors, (c) behaviors directly related to HIV risk, and (d) behaviors that were not statistically rare. Eight homogeneous HIV risk clusters were produced. Crack cocaine use was the most distinguishing feature of three clusters. Another three clusters were distinguishable by drug injection and needle use practices. Two additional clusters could not be grouped with either the crack- or the injection-dominant clusters. Prostitution was the most distinguishing risk behavior of one of these clusters, and extremely high drug injection frequencies and relative rates of risky needle use characterized the other. Composition of the clusters varied significantly by gender, race/ethnicity, educational attainment, and drug use characteristics. In addition, perceptions and behaviors initiated to reduce the chances of becoming infected with HIV varied by cluster. Subjects in the crack-predominant clusters reported low perceptions of the chances of getting AIDS. Perceptions of the chances of becoming infected with HIV among subjects in the injection-predominant clusters were strongly related to injection frequency. Seroprevalence was also related to cluster. Higher rates of HIV infection were evident among the injection-predominant clusters, and higher rates were related to frequency of injection and the rate of risky needle use. Among the crack-predominant clusters, the relationship between drug use and sexual behaviors and HIV infection was less clear.

Acquired Immunodeficiency Syndrome↗

PubMed2000

[Cluster analysis methods appropriate for classification of drought-resistant wheat ecotypes].

The appropriate cluster analysis methods for classification of drought-resistant wheat ecotypes were investigated, based on the analysis of 21 morphologic and agronomic characters of 15 winter wheat cultivars. According to the proximity degree to the results from experts experiences, the cluster results using original data were orderly better than those using Varimax orthogonal rotation, Promax skew intersection rotation, and principal component analysis with phenotypic mean's correlation matrix. The results using Euclidean distance were greater than those using Mahalanobis distance. The clustering methods of correspondent analysis and fuzzy cluster were better than those of nearest distance, furthest distance or group average method (UPGMA). Among all cluster results from various methods combined, the two cluster results from the fuzzy cluster using original data and from the correspondence analysis method were closest to the results of experts' experience. Based on both comparisons of results and examination of the performance of different strategies with several statistical properties, the two methods mentioned above were also acceptable.

Cluster Analysis↗

PubMed1992

Cluster analysis reveals at least three, and possibly five distinct handedness groups.

Multivariate techniques have used data from hand preference questionnaires to group hand preference items, but no attempt has been made to date to use multivariate analyses to group individuals in terms of handedness groups. This study analyzed the responses of 645 subjects on the Waterloo 60-item handedness questionnaire with a cluster analysis (BMDP) in order to determine the grouping of individuals in terms of hand preference patterns. Five distinct handedness groups were recognized by this procedure and a Discriminant Function Analysis revealed a very high accuracy of assigning individuals to the five groups. A cluster analysis of a shorter 14-item questionnaire suggested three distinct handedness groups, and the degree of accuracy of assigning individuals to these groups was also very high. As is the case with all multivariate techniques in neuropsychology, the question of whether the clusters form meaningful groupings awaits an answer in terms of their different neuropsychological properties.

Adult↗

PubMed2004

The importance of degree versus type of maltreatment: a cluster analysis of child abuse types.

The author conducted secondary data analysis of 3 previously reported studies (D. J. Higgins & M.P. McCabe, 1998, 2000b, 2003) to examine whether respondents are best classified according to their experience of separate maltreatment types (sexual abuse, physical abuse, psychological maltreatment, neglect, and witnessing family violence) or whether their experience reflects a single unifying concept: child maltreatment. The author conducted a cluster analysis of the combined dataset followed by a confirmatory discriminant function analysis. Finally, the differences in psychological adjustment between those classified into the 3 different clusters were examined as a test of the 3-cluster solution. The best cluster analysis solution grouped individuals according to the degree to which maltreatment behaviors were reported. Individuals classified into the high maltreatment cluster had significantly more adjustment problems than those in either the moderate or the low maltreatment clusters. The results showed that it may be more meaningful to talk about the degree of maltreatment (frequency and/or severity) experienced by the child rather than about the type.

Adult↗

PubMed2006

Cluster analysis and two-dimensional quantitative structure-activity relationship (2D-QSAR) of Pseudomonas aeruginosa deacetylase LpxC inhibitors.

Compounds from a wide variety of structural classes inhibit Pseudomonas aeruginosa deacetylase LpxC. However, a single unified understanding of the relationship between the structures and activities of these compounds still eludes the researchers. We report herein, the development of cluster analysis-based 2D-QSAR models for LpxC inhibition. Principal component analysis (PCA), hierarchical cluster analysis (HCA), and genetic function approximation (GFA) were employed for the development of the QSAR model. The conventional 2D-QSAR model derived for the complete set of three-structural classes had unsatisfactory predictability with a correlation coefficient (r(2)) of 0.703 and a cross-validated correlation coefficient (q(2)) of 0.584. Descriptor-based cluster analysis indicated that the three-structural classes of LpxC inhibitors studied belonged to two clusters. Separate QSAR models for these two clusters showed substantially improved predictability with r(2) values of 0.904 and 0.944 and q(2) values of 0.805 and 0.906, respectively. Thus, we expect that compared to the conventional model, our two QSAR models can be better used to preliminarily screen molecules from a diverse chemical space while searching for novel LpxC inhibitors.

Amidohydrolases↗

PubMed2001

Evidence for shared genetic programs from cluster analysis of hippocampal gene expression dynamics in development and response to injury.

Cluster analysis is a computational method that groups together similarly-shaped patterns. It may be applied to large-scale gene expression data to form new hypotheses regarding gene function. In the present study, we clustered the temporal expression patterns of genes expressed in the rat hippocampus during normal development and after a kainate-induced seizure injury at postnatal day 25. We found that two different methods, Euclidean hierarchical and K-means clustering, produced slightly different results, and concluded that different clustering methods may he used to complement one another. We also found that certain genes cluster together both during development and after seizure injury, consistent with the idea of sets of genes that act in concert under various conditions.

Animals↗