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 379 records · Page 21Linked to original sources

PubMed1991

Analysis of methotrexate treatment effect in a longitudinal observational study: utility of cluster analysis.

We studied 235 patients with rheumatoid arthritis (RA) beginning therapy with methotrexate utilizing a k-means clustering algorithm. Four groups were identified: mild RA (Group 3), very severe RA (Group 4), and 2 groups intermediate in severity (Groups 1 and 2). Group 2, the largest of the clusters (n = 89), appeared to have greater tolerability of RA as measured by severity and psychological variables, and took the drug almost twice as long as other groups, although improvement was not greater nor side effects fewer. All groups improved over a mean of 1.9 years, and the degree of improvement was not related to the initial severity classification. Improvement occurred almost equally in all clusters, and the relative ranking of the groups was maintained at study closure.

Arthritis, Rheumatoid↗

PubMed1989

A cluster analysis of institutionalized mentally retarded individuals.

The scores of two samples of 200 subjects each on five factors of a measure of adaptive behavior were subjected to cluster analysis using two hierarchical agglomerative clustering methods. Three clusters were found that proved stable across clustering methods, subject samples, and time points 3 years apart. In addition, cluster membership was shown to be meaningfully associated with a number of demographic characteristics and neurological and sensory handicaps. The three-cluster solution thus possesses some desirable features, but its practical utility remains to be demonstrated.

Activities of Daily Living↗

PubMed1991

[Application of cluster analysis to the classification of Radix scutellariae].

Ten samples of Radix Scutellariae from different habitats were analyzed by thin layer chromatography. From the analysis numerical characteristic features were obtained that represent the quality difference of samples. All of the multivariate data were treated with mathematic method of cluster analysis. Scutellaria bai alensis included in Chinese Pharmacopoeia was clearly differentiated from the other four varieties of herbs (S. rehderiana, S. viscidula, S. amoena, S. likiangensis).

Chromatography, Thin Layer↗

PubMedDate not supplied

Identification of groupings of graph theoretical molecular descriptors using a hybrid cluster analysis approach.

There is an abundance of structural molecular descriptors of various forms that have been proposed and tested over the years. Very often different descriptors represent, more or less, the same aspects of molecular structures and, thus, they have diminished discriminating power for the identification of different structural features that might contribute to the molecular property, or activity of interest. Therefore, it is essential that noncorrelated descriptors be employed to ensure the wider and the less inflated possible coverage of the chemical space. The most usual approach for reducing the number of descriptors and employing noncorrelated (or orthogonal) descriptors involves principal component analysis (PCA) or other factor analytical techniques. In this work we present an approach for determining relationships (groupings) among 240 graph-theoretical descriptors, as a means for selecting nonredundant ones, based on the application of cluster analysis (CA). To remove inherent biases and particularities of different CA algorithms, several clustering solutions, using these algorithms, were "hybridized" to obtain a reliable and confident overall solution concerning how the interrelationships within the data are structured. The calculated correlation coefficients between descriptors were used as a reference for a discussion on the different CA methods employed, and the resulted clusters of descriptors were statistically analyzed for deriving the intercorrelations between the different operators, weighting schemes and matrices used for the computation of these descriptors.

Cluster Analysis↗

PubMed2004

[Cluster analysis in geographical epidemiology: the use of several statistical methods and comparison of their results].

BACKGROUND: The increasing interest in environmental epidemiology has been followed by the development of many statistical tests for detecting disease clustering near a point source. The objectives of this study were to compare several tests to detect disease clustering, among which modelisation using Markov Chain Monte Carlo methods. METHODS: We compared six statistical methods for detecting disease clustering of bladder cancer around an industrial centre of Isère (France) for the period 1983-1997: Stone's test, score test, and two log-linear modelisations (with and without corrections for extra-Poisson variations) using two ways of parameters estimation (maximum likelihood and Markov Chain Monte Carlo methods). RESULTS: The results of the Stone test and the score test are not in favour of a higher risk of bladder cancer around the considered point source. The conclusions brought by the log linear modelisations are the same, but the results obtained using the Markov Chain Monte Carlo Method are very dependant of prior distributions determined for the different parameters. CONCLUSION: Markov Chain Monte Carlo methods, which allow taking into account complex geographical effects, seem well adapted to cluster analysis in geographical epidemiology. However, they remain difficult to implement.

Adolescent↗

PubMed2001

[Grey correlation and cluster analysis on relationship between Cercidiphyllum japonicum community and its environment].

The relationship between 10 years old C. japonicum plantation forest community and its environmental conditions was studied by grey correlation method and cluster analysis. The results showed that there existed closer relations among soil organic matter, available Ca, Fe, Mg and N. Except for N and Mn, all soil available elements tested had no direct impact on the absorption and accumulation of corresponding elements by the community. The tree species in different habitats was found to absorb soil elements selectively to some extent. The factors affecting the basal diameter, DBH, height and biomass of C. japonicum could be lined as soil chemical characteristics > soil physical properties > climate factors, in which, available P had little impact on the growth of C. japonicum. It is suggested that C. japonicum could be restored in the area with higher air moisture and with fertile porous soil between elevation of 2200-2700 m.

Cluster Analysis↗

PubMed2002

Post-combat syndromes from the Boer war to the Gulf war: a cluster analysis of their nature and attribution.

OBJECTIVES: To discover whether post-combat syndromes have existed after modern wars and what relation they bear to each other. DESIGN: Review of medical and military records of servicemen and cluster analysis of symptoms. DATA SOURCES: Records for 1856 veterans randomly selected from war pension files awarded from 1872 and from the Medical Assessment Programme for Gulf war veterans. MAIN OUTCOME MEASURES: Characteristic patterns of symptom clusters and their relation to dependent variables including war, diagnosis, predisposing physical illness, and exposure to combat; and servicemen's changing attributions for post-combat disorders. RESULTS: Three varieties of post-combat disorder were identified-a debility syndrome (associated with the 19th and early 20th centuries), somatic syndrome (related primarily to the first world war), and a neuropsychiatric syndrome (associated with the second world war and the Gulf conflict). The era in which the war occurred was overwhelmingly the best predictor of cluster membership. CONCLUSIONS: All modern wars have been associated with a syndrome characterised by unexplained medical symptoms. The form that these assume, the terms used to describe them, and the explanations offered by servicemen and doctors seem to be influenced by advances in medical science, changes in the nature of warfare, and underlying cultural forces.

Cluster Analysis↗

PubMed2002

A cluster analysis of symptom patterns and adjustment in Vietnam combat veterans with chronic posttraumatic stress disorder.

This study investigated whether a subgroup of veterans with malignant posttraumatic stress syndrome, as described by Rosenheck (1985) and Lambert et al. (1996), could be identified via cluster analysis within two samples of Vietnam veterans with combat-related posttraumatic stress disorder (PTSD). In the initial subsample (n = 157), four clusters were identified, including a subgroup that scored consistently higher on measures of interpersonal violence and current physical problems. Similar results were found in the cross-validation subsample (n = 156). These results provide support for the theoretical concept of malignant PTSD and suggest that veterans with chronic PTSD are not homogenous. Whereas some manifest extreme levels of both functional impairment and PTSD symptomatology, others exhibit markedly less functional impairment despite manifesting clinically significant levels of PTSD. Clinicians can consider this heterogeneity in their treatment decisions.

Adjustment Disorders↗

PubMed2003

[Identification of Aaron's beard with clustering analysis by attenuation reflection-Fourier transform infrared spectrometry].

OBJECTIVE: To establish a new method to discriminate Aaron's beard by Fourier transformation infrared spectrometry. METHODS: Attenuation Reflection-Fourier transformation infrared spectrometry with clustering analysis was used to the identification of Aaron's beard. RESULTS: There were obvious differences in Aaron's beard. The results are consistent with that of morphologic study. CONCLUSION: This method is rapid, simple and economical, and can be used to the quality control.

Cluster Analysis↗

PubMed1994

Stressful jobs and non-stressful jobs: a cluster analysis of office jobs.

The purpose of the study was to determine if office jobs could be characterized by a small number of combinations of stressors that could be related to job-title information and self-report of psychological strain. Two-hundred-and-sixty-two office workers from three public service organizations provided data on nine job stressors and seven indicators of psychological strain. Using cluster analysis on the nine stressors, office jobs were classified into three clusters. The first cluster included jobs with high skill utilization, task clarity, job control and social support and low future ambiguity, but also high on job demands such as quantitative work-load, attention and work pressure. The second cluster included jobs with high demands and future ambiguity and low skill utilization, task clarity, job control and social support. The third cluster was intermediary between the first two clusters. The three clusters were related to job-title information. The second cluster was the highest on a range of psychological strain indicators, while the other two clusters were high on certain strain indicators but low on others. The study showed that office jobs could be characterized by a small number of combinations of stressors that were related to job-title information and psychological strain.

Adult↗

PubMed2001

Prognosis of renal amyloidosis: a clinicopathological study using cluster analysis.

Progression of renal amyloidosis is associated with severe proteinuria or nephrotic syndrome, and various mechanisms have been postulated to explain these complications. We studied the acceleration of proteinuria and reduced renal function by cluster analysis using clinical parameters, renal histological findings, type of renal amyloidosis and follow-up data. We divided 97 cases into three groups of renal amyloidosis. Accelerated progression correlated with serum creatinine (s-Cr) levels at renal biopsy and histological grade of renal damage by amyloid deposition (p < 0.0001). The most influential prognostic factors (s-Cr level > or =2.0 mg/dl) were tubulointerstitial and vascular damage induced by amyloid deposition at biopsy (odds ratio 96.9 and 69.2, respectively). In addition, we found amyloidosis type amyloid associated (AA) correlated with more amyloid-mediated vascular and tubulointerstitial damage than amyloidosis type amyloid light chain (AL) (p < 0.001, p < 0.01, respectively). Proteinuria and nephrotic syndrome were more severe in cases of amyloidosis AL than in amyloidosis AA (p = 0.076). In conclusion, less tubulointerstitial and vascular damage was caused by amyloid deposition; this was slowly progressive. Amyloid AA was detected in tubulointerstitial tissue and vessels more frequently than amyloid AL. Heavy proteinuria and/or nephrosis were not indicators of rapid progression.

Adult↗

PubMed2006

Phenon cluster analysis as a method to investigate epidemiological relatedness between sources of Campylobacter jejuni.

AIMS: To develop a method for assessing the relative epidemiological significance of possible infection sources for human campylobacteriosis. METHODS AND RESULTS: Using fluorescent amplified fragment length polymorphism (AFLP), 243 apparently epidemiologically unrelated Campylobacter jejuni isolates were genotyped (77 human, 46 cattle, 49 pet and 71 poultry isolates). In total 136 different phena were identified, of which 48 were clusters grouping at least two isolates. Isolates from different sources were frequently clustered together, underlining the high degree of source mixing and the lack of host specificity of C. jejuni. The phena were classified into different phenon types according to the sources of the isolates they contained. The occurrence of these phenon types was analysed using an area-proportional Euler diagram to describe epidemiological relatedness among C. jejuni isolates. Group separation statistics revealed that 43% of analysed human isolates expressed maximum similarity to other human isolates, 9% to cattle isolates, 21% to pet isolates and 27% to poultry isolates; these results were in accordance with the pattern observed in the phenon cluster analysis. CONCLUSIONS: Based on the grouping of strains into molecular similarity clusters, ecological patterns between sources can be investigated. SIGNIFICANCE AND IMPACT OF THE STUDY: This approach is a new methodological contribution to establish the relative epidemiological significance of concurrent infection sources.

Animals↗

PubMed1982

An approach to the definition of periodontal disease syndromes by cluster analysis.

Clinical syndromes of 22 untreated patients with advanced destructive periodontal disease were analyzed using cluster analysis. Clinical characteristics coded for each patient included age, sex, measures of gingival inflammation, plaque, suppuration, pocket depth, attachment level, extent and pattern of bone loss, rate of change in pocket depth, and correlation coefficients between certain clinical measurements. Microbiologic features included darkfield enumeration of 10 morphologically distinct forms of organisms which were removed from the three sites showing the most advanced destruction in each patient, as well as viable counts of specific microbial groups from the same teeth using elective and selective media. Serum antibody levels were determined by the ELISA technique to 13 species of subgingival microorganisms. The Gower coefficient was used to estimate similarity between patients and clusters were formed using an average unweighted linkage sort. Three distinct patient clusters were observed with greater than 70% average intra-cluster similarity. One subject did not fall into any of the patient clusters. The features which defined and differentiated the clusters were found to include age of subject, extent and patterns of bone loss, percent of sites showing change in pocket depth and attachment level, percent of small motile rods, intermediate spirochetes and fusiforms and serum IgG levels against Bacteroides gingivalis, Selenomonas sputigena and a Wolinella strain.

Adolescent↗

PubMedDate not supplied

[A sentinel physicians network in Castile and León: the use of cluster analysis to obtain a representative population].

In order to establish a sentinel physicians network in Castilla y León to collect systematically population-based morbidity data, a random sample of general practitioners (GP's) whose covered population was representative of the regional population was obtained. A cluster analysis with the ZBS (Zonas Básicas de Salud) was performed according to a list of variables considered important in the diseases incidence. Five clusters were obtained in the urban areas and twenty in the rural areas, where, after stratification, the GP's random sample was selected. The outcome of the distribution, within each cluster, between all Castilla y León GP's and the sentinel GP's did not show statistically significant differences. The statistically significant difference found between the age of all GP's and the 127 sentinel GP's, due to the voluntary participation, was not found in the comparison of the two populations covered by them, confirming the efficacy of this method in the selection of a representative population from a random sampling of population groups.

Cluster Analysis↗

PubMed2004

Cluster analysis of key diagnostic variables from two independent samples of eating-disorder patients: evidence for a consistent pattern.

INTRODUCTION: The optimal classification of eating disorders has been a matter of considerable debate. The present paper tackles this issue using cluster analysis with large independent samples of eating-disorder patients. METHOD: Two samples of adult female patients from Sweden (n = 631) and England (n = 472) were classified on the basis of 10 key clinical variables of primary significance for diagnosing eating disorders. A separate series of cluster analyses were conducted on each sample. RESULTS: Results suggested that a three-cluster solution was optimal in both samples. The first cluster ('generalized eating disorder') was characterized by high levels of eating-disorder psychopathology on all variables except weight and menstrual functioning. The second cluster ('anorexics') was typified by low weight, amenorrhoea and the absence of binge eating, and seemed to correspond to the clinical picture of anorexia nervosa. The third cluster ('overeaters') was characterized by high weight and moderate levels of binge eating and compensatory behaviour. CONCLUSIONS: Results suggest that patients presenting to eating-disorder services in different countries have clinical features that fall into very similar patterns. These patterns resemble, but are not identical to, existing diagnostic categories.

Adolescent↗

PubMed1989

[Dermatoglyphics parameters and cluster analysis of seven minority nationalities].

This paper reports the normal values of dermatoglyphics parameters of seven minority nationalities in Yunnan Province which are Bai, Blang, Yi, Hui, Lisu, Nu and Jinuo. The test of difference signification and cluster analysis show different parameters in several nationalities and the greatest most remarkable difference between Jinou and other nationalities. Han is very different from several nationalities. In each nationality, the symmetry pattern of same name finger or area is highly unanimous, the symmetry between left and right does not show random combination.

China↗

PubMed2006

A cluster analysis of people on Community Treatment Orders in Victoria, Australia.

This paper explores the clinical, social and demographic characteristics of 164 people on Community Treatment Orders (CTOs) in one area mental health service in Victoria, Australia. The results of an exploratory cluster analysis are presented to address the question of whether people on Community Treatment Orders can be categorised into statistically reliable, qualitatively distinct groupings. The data are presented in the context of key stakeholder perspectives on the current use and purpose of CTOs. Three stable clusters emerged and each potentially reflects how social dimensions, as well as clinical issues, influence decision making regarding the implementation of CTOs. These findings are important in the context of policy and practice in Victoria, where the use of CTOs is common practice, and orders are generally made for a 12 month period. The potential for improved targeting of CTOs and more specific treatment planning is identified.

Adolescent↗

PubMed2003

Quantitative image reconstruction in PET from emission data only using cluster analysis.

Quantitative image reconstruction in positron emission tomography requires attenuation correction. In case the attenuation correction is not measured separately, under certain conditions this can be determined from the emission data alone. We present a method based on cluster analysis that assumes only 3 empirical attenuation coefficients, i.e., 0.095 cm-1 for soft tissue, 0.02 cm-1 for lung, and 0 cm-1 for air. The subsequent image reconstruction takes place in an iterative fashion, through maximization of image likelihood. For the mathematical thorax phantom used in the present study, the results are comparable to those obtained after separate measurement of the attenuation correction.

Cluster Analysis↗