PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Discriminant 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 1,153 records · Page 64Linked to original sources
PubMedDate not supplied

Specialized PC software package for creation of computer systems supporting partial diagnostics based on numerical results of medical examinations.

Computer scientists creating computer systems supporting partial diagnostics based on numerical results of medical examinations always follow a similar method. Why not algorithmize this procedure? This was the main reason for preparation of the specialized PC software presented in the paper. The software is addressed to clinicians-scientists and enables such users to create by themselves (with no need of programming) practically useful computer systems for differential diagnosis of chosen diseases. These generated systems are assigned to an end-user, i.e. any physician interested in a partial diagnosis of the considered type. They perform classification of patients on the basis of numerical results of the examinations selected for the training databases. The software is based on a statistical approach; multi-dimensional variance analysis and discriminant analysis methods have been used. The pilot version of the software (experimentally implemented in a Warsaw clinic) and an outline of planned work are presented.

Algorithms↗
PubMed2001

Interest of a multibiomarker approach in the assessment of freshwater ecosystem quality: laboratory and field studies.

The aims of this study were to validate several biochemical parameters as biomarkers of pollution in the freshwater bivalve Corbicula fluminea and to underline the interest of a multibiomarker approach in environmental biomonitoring. The study was divided into a laboratory exposure to 4 doses of trichloroethylene, toluene, cadmium chloride or a coal tar fraction for 5 days and a field exposure for one week in 5 sites surrounding an industrial effluent outlet. Whatever the product was, parameters that exhibited significant responses were mainly parameters related to oxidative stress and components of phase I metabolism. As a result of discriminant analysis, doses were clearly discriminated from the control and from each other. Likewise, products were discriminated from each other, based on results of the whole parameter responses obtained for the first dose. Concerning the field experiment, all biochemical parameters assayed exhibited significant responses for sites located downstream of the effluent outlet, compared to the upstream reference site. Through a discriminant analysis, sites could be distinguished from each other in terms of pollution intensity. In order to characterise pollution at a qualitative level, further laboratory and field studies are needed to obtain typical profiles for the main pollutants present in freshwater ecosystems.

Animals↗
PubMedDate not supplied

Recognition of the bouquet of Chinese spirits by artificial neural network analysis.

Forty-six samples of Chinese spirits, whose bouquets were determined by sensory evaluations, and 17 compounds characteristic of the flavors determined by gas chromatography/gas chromatography-mass spectrometry (GC/GC-MS), were subjected to neural network analysis and their corresponding factor scores developed. To make the bouquet recognition more efficient, an improved artificial back-propagation neural network (BPNN) was applied. In each kind of data, the BPNN was trained repeatedly until the error rate was less than the predetermined threshold error; then the trained network was applied to the test set that was not involved in the training process to establish the validity of the network, and a correct prediction rate of 100% was obtained. The BPNN provided a correlation between the data offered from sensory evaluations and the data of chemical compositions determined by instrumental analysis. The BPNN approach is feasible regardless of whether the crude data or the factor scores are used; however, recognition results were better with the latter than with the former. In a comparison of all the results obtained by BPNN, cluster analysis, and discriminant analysis, the method of artificial neural network analysis appeared to be the optimal technique for recognizing the bouquet of Chinese spirits.

Alcoholic Beverages↗
PubMed2001

[Analysis of factors influencing on Anisakis specific IgE antibodies].

Anisakis-specific IgE (AsIgE) was examined in 84 cases of acute urticaria (AU) and 100 cases of atopic dermatitis (AD). Gender, age, disease category (AU or AD) and serum IgE value of the patients were compared with AsIgE. Univariate analysis showed that there was significant correlation between age and AsIgE, and between serum IgE and AsIgE. Serum IgE was significantly higher in AD than in AU. Age was significantly higher in AU than in AD. AsIgE was positive value in 26 cases (31%) of AU and 25 cases (25%) of AD. There were no significant differences between them. In these cases, however, AsIgE was significantly higher in AU than in AD. Percentage of patients with high AsIgE value (> class 2) was significantly higher in AU (20 cases, 24%) than in AD (3 cases, 3%). Multivariate analysis using discriminant analysis and logistic regression analysis showed that the most influential factor on AsIgE was disease category (relative risk = 16.87), and the second was age (relative risk = 1.05). Serum IgE and gender were considered not to influence on AsIgE. It is possible that Anisakis or related antigens have something to do with urticaria in AU patients with high AsIgE value.

Acute Disease↗
PubMed2004

Karyometric features of low and high grade glial tumors as compared with their MRI appearance.

OBJECTIVE: To compare the results of a magnetic resonance imaging (MRI) grading designed to identify low and high grade gliomas with karyometry used as a tool to grade primary brain tumors. STUDY DESIGN: A consecutive series of 23 primary brain tumors was selected for this study. The neuroradiologist, not knowing the histologic diagnoses, divided the cases into low and high grade categories on the basis of the following 7 features: border sharpness, heterogeneity without contrast, cavitation, contrast enhancement, hypervascularity, mass effect and perifocal T2 hyperintensity. To each feature was given a numerical value, ranging from 1 to 3. All the cases were reviewed and classified by the same pathologist, blinded to the MRI diagnosis. Two hundred nuclei per case were recorded, and 93 karyometric features related to nuclear area, total optical density and chromatin distribution were analyzed for each nucleus. Statistical analysis included discriminant analysis, Kruskal-Wallis test, nonsupervised learning algorithm P-index and Beale statistic. RESULTS: Ten cases were classified as low grade on the basis of their MRI features. The corresponding histopathologic diagnoses were: grade 2 astrocytoma in 2 cases and grade 2 oligodendroglioma in 8 cases. An MRI diagnosis of high grade tumor was made in 13 cases. In 10 cases it was confirmed by the histopathologic diagnosis (3 grade 3 astrocytomas, 1 grade 3 oligodendroglioma and 6 glioblastomas). In the remaining 3 cases the histologic examination revealed a low grade tumor, 1 grade 2 astrocytoma and 2 grade 2 oligodendrogliomas. For the purposes of the karyometric analysis the cases were allocated to the low or high grade category according to their histologic diagnosis (13 cases low grade and 10 cases high grade). Nuclei from low and high grade tumors showed clearly different karyometric characteristics. The oligodendroglioma nuclei had abnormality values close to the low grade standard, while the astrocytoma nuclei were a highly dispersed group with characteristics indicative of a higher degree of nuclear abnormality than the oligodendroglioma nuclei. The results of karyometric analysis showed that grade 2 tumors, corresponding to the low grade group, form a rather distinct category from grade 3 and 4 tumors belonging to the high grade group. CONCLUSION: The results of MRI grading based on a series of features that are routinely assessed by the neuroradiologist to reach a final diagnosis correlate highly with the histopathologic diagnosis. Karyometry can be a useful adjunct to histologic grading.

Astrocytoma↗
PubMed2004

[The abnormity and clinical value of tyrosine kinase signaling pathway in lung adenocarcinoma].

OBJECTIVE: To study the abnormity and clinical value of tyrosine kinase signaling pathway in lung adenocarcinoma. METHODS: Total RNA was extracted form 10 lung adenocarcinoma samples and matched normal tissues. cDNA was labeled fluorescence during reverse transcription and hybridized on 20 slides of microarray with 13 824 genes. After washed and scanned, the cy3:cy5 ratios were obtained by computer analysis. Each experiment was repeated two times and the mean value was chosen. Cluster analysis and discriminant analysis were used to mine the pool data. The results were combined with clinical prognostic factors for further analysis. RESULTS: According to the abnormity of tyrosine kinase signaling pathway in lung adenocarcinoma, sample cluster divided the 10 samples into 3 groups and gene cluster divided all genes into 5 groups. Different prognosis samples had different characteristic genes. The patients with the best prognosis overexpressed some genes related to immunity and inhibition proliferation, and those with the worst prognosis overexpressed some genes related to proliferation while the genes related to immunity and inhibition proliferation were expressed at a low level. CONCLUSION: There are 3 type abnormity of tyrosine kinase signaling pathway in lung adenocarcinoma that are related to different biological behavior and different prognosis.

Adenocarcinoma↗
PubMed1990

[Respiratory function abnormalities and pneumonia in HIV-positive patients. A prospective study of 112 patients].

Pulmonary function studies are often limited to the alone measurement of transfer lung factor for CO (TLCO) in screening for pneumonia in patients with Human Immunodeficiency Virus (HIV) infection. We prospectively measured pulmonary function tests (PFT) in 112 HIV seropositive patients. The population consisted of two groups: on one hand, a group free of clinical and radiological abnormalities, on the other hand, one with respiratory symptoms and/or abnormal chest X-Ray, with or without overt pneumonitis. For this latter group, a fiberoptic bronchoscopy with bronchoalveolar lavage was routinely performed in addition to PFT. In case of pneumonitis, PFT showed a restrictive disease and a reduced TLCO. The specificity of this functional pattern was however weaker in the subgroup of drug abusers than in the non-drug addicts. This difference was above all linked to a low TLCO value in the subgroup of drug addicts without pulmonary complications. Multivariate statistical analysis, including discriminant analysis, maintained the same sensibility and improved specificity of PFT in diagnosis of pneumonia, especially if the analysis takes the existence of drug abuse into account. Moreover, initial PFT, performed before any lung disease, improved the sensibility of the screening. The results are discussed in relation to new tests proposed for the screening of pneumonitis in HIV positive patients. At the present time, PFT seems to be useful and enables one to understand natural functional evolution.

Adult↗
PubMed1990

[Study on prognosis of threatened abortion assessed by multivariate analysis].

The serum concentrations of various hormones which included hPL, progesterone(P), estradiol(E2) and 17 alpha-hydroxyprogesterone (17P) in serum, and hCG in urine were measured simultaneously in patients with threatened abortion to evaluate the prognosis. Ultrasound scanning was also undertaken. One hundred and thirty-six patients at 5 to 8 weeks of gestation were divided into two groups retrospectively. Group A consisted of 99 patients with a good prognosis and group B consisted of 37 patients who aborted spontaneously. The mean concentrations of hormones, and ultrasound findings (GS, CRL and FHM) were not significantly different in both groups at 5 weeks. An attempt was made to apply the data to the computation in multivariate analysis. By discriminant analysis, 100% reliability was achieved in separating the two groups according to hormone levels only at this stage of pregnancy. At 6 weeks, the hCG, P and E2 values were statistically higher in group A than B. Because of the overlapping of the data, principal component analysis was attempted. Two groups were roughly separated. The values for factor loading indicated that GS, CRL and 17P effective in making the separation possible. At 7 and 8 weeks of gestation, the mean concentrations of hCG, hPL, P and E2 were significantly higher in A than B. The factor loading values indicated that ultrasound findings were essential in separating into two groups. These results suggested that hormonal analysis may be useful in assessing the prognosis in threatened abortion in the early stage of gestation (5 and 6 weeks), while ultrasound findings make an accurate prediction for the prognosis possible after 7 weeks of gestation.

Abortion, Threatened↗
PubMed1989

Prognostic significance (relapse, non-relapse) of nuclear shape parameters in lymph node negative breast cancer.

A series of 15 postmenopausal patients with stage I invasive breast cancer (pT1-2, pNO, cMO) of uniform histologic type (ductal, NOS) was selected from the Basel Cancer Registry for retrospective analysis. A group of seven patients with a minimal disease-free survival of 9 years was compared to a group of eight patients with clinical and/or histopathological relapse within 8 years of mastectomy. Computer-assisted morphometry of cancer nuclei in routine histological material was used to investigate possible correlations of clinical outcome with nuclear size and shape parameters. Significant differences were found for certain shape descriptors by nonparametric statistical analysis. Stepwise discriminant analysis was applied to find the best combination of parameters separating the two prognostic groups. Application of the 'leaving one out method' showed correct classification in 6 of 8 recurrent cases and in all non-recurrent cases. The number of patients is small, therefore the results should be carefully interpreted.

Breast Neoplasms↗
PubMedDate not supplied

Cellular changes in the mammary gland epithelium during the menstrual cycle. A computer image analysis study.

While examining fine needle aspirates of lesions, it became evident that preovulatory aspirates differed from postovulatory ones. Proliferation activity was noted in cells and in fragments aspirated during the luteal phase. In order to verify this observation, pairs of smears obtained during different phases of the menstrual cycle from eight fertile women were studied by computer image analysis. Using discriminant analysis based on nuclear features extracted from the digitized cell images, 84.87% of all cells were correctly classified as preovulatory or postovulatory, according to the smear of origin. When the classification was restricted to only the cells from each woman, a 100% correct identification of cell types was obtained, which supports the visual observation that there are cellular changes in the mammary epithelium correlated to the menstrual cycle. Since some proliferative changes may mimic atypia or malignancy, these findings suggest that it is advisable to perform needle aspirations of breast lesions during the preovulatory phase of the cycle.

Adolescent↗
PubMed1978

Noncompliance in children with renal transplants.

Fourteen patients (13 of them adolescents) interrupted immunosuppressive treatment following renal transplantation. Twelve were girls and two were boys. Six subsequently lost their allografts and eight had impaired renal function. Noncompliance was suspected when diminution in cushingoid features, unexplained weight loss, or changes in renal function occurred. Noncompliance was comfirmed by interview with psychosocial staff. Available psychosocial data from family interview and personality test obtained earlier as part of systematic follow-up study were analyzed to explore the reasons for noncompliance. Non compliant patient families had lower incomes, more fatherless households, and comunication difficulties within the family and with the medical establishment. Using a stepwise discriminant analysis, a discriminant function was derived which selected 13 of 14 noncompliant patients. Noncompliance may be a preventable cause of allograft failure. These data can aid in identifying high-risk patients and planning intervention programs.

Adolescent↗
PubMed1987

Preoperative prediction of extent and prognosis of gastric carcinoma by four serum proteins and erythrocyte sedimentation rate.

In 195 patients with gastric carcinoma the preoperative ESR and serum concentrations of IgG, C4, C1-INH and CEA varied significantly with the extent of disease. Extent of disease and prognosis were predicted from these variables by discriminant analysis. The discriminant rules were tested on the same patients in an unbiased way. Metastases or no metastases were correctly predicted in 75% of the patients. By an appropriate prior distribution 93% of the patients without metastases were identified. The disease extent was also predicted in subgroups of patients with and without metastases. Survival was correctly predicted preoperatively in 66% of the patients and 83% of the patients with a fair prognosis were identified. Of the patients preoperatively allocated to the non-survival group 94% did actually die during follow-up. When used in addition to other available information, our discriminant rules will contribute to the quality of the preoperative evaluation of patients with gastric carcinoma.

Adult↗
PubMed1986

A taxonomic analysis of multihospital systems.

Research to date on multihospital systems has proved largely uninformative, in part because similarities and differences among these organizations have not been addressed systematically. Through numeric classification, this article identifies populations of multihospital systems that share similar organizational attributes. Drawing on McKelvey's classification theory, 16 organizational characteristics of 160 multihospital systems are analyzed using a series of taxonomic techniques, including cluster analysis, multiple discriminant analysis, and analysis of coefficients of variation. Fifteen distinct subgroups of systems are identified and described, and their implications for organization research discussed.

Catholicism↗
PubMed1978

Classification of sleep stage with period analysis features derived from the EEG.

This paper details our experience with the use of regression analysis and discriminant analyses, combined with a Bayes classifier, to evaluate the usefulness of a series of features derived in real time from a single channel of EEG. These features were 1) the number of zero crossings, 2) the number of zero crossings of the first derivative of the EEG, and 3) the number of zero crossings of the second derivative and five additional "histogram" measures. Each histogram measure represented the number of occurrences of a zero crossing interval falling within an arbitrarily selected range of values. The distributions of several measures conditioned on sleep stage are presented along with night-to-night and subject-to-subject variations in the measures. These details provide a basis for understanding effectiveness and limitations of zero crossing measures and the regression and discriminant analysis techniques. Our results permitted a range of 57-90% accurate classification with just zero crossing measures with the cross-correlation between computer- and hand-scored nightly sleep patterns ranging from 0.63-0.94. We consider these to be good and acceptable accuracies for this type of analysis.

Electroencephalography↗
PubMed1984

[Pre- and postoperative accommodative convergence/accommodation (AC/A) ratio: preliminary study of computer-assisted analyses].

A group of 15 patients who underwent bilateral medial rectus muscle recession for esodeviation were analyzed for factors most closely correlated with the outcome. The data evaluated included age of onset of strabismus, refractive error, year of surgery, age of surgery, sex, preoperative esodeviation, and clinical assessment of preoperative heterophoria. Linear regression analysis was used to evaluate the relationships of these parameters. Computer-assisted logistic regression analysis and discriminant analysis were used to asses predictive variables for a postoperative increase in the AC/A ratio.

Accommodation, Ocular↗
PubMed1981

Annoyance by aircraft noise and fear of overflying aircraft in relation to attitudes toward the environment and community.

If, as has been suggested, expressions of annoyance attributable to aircraft noise may reflect, in part, fear of aircraft overflights and possible crashes, then residents of areas where crashes have occurred should express more annoyance. To test this hypothesis, 50 residents of an Albany, New York area where an aircraft crash producing fatalities recently occurred, and 50 residents of a comparable nearby area without such a history, were asked to respond to a "Quality of Life" questionnaire. Among the items were some designed to test annoyance by noise and fear of aircraft overflights. It was predicted that those in the crash area would express more fear and would more often identify aircraft as a noise source. Factor analysis and discriminant analysis results sustained these hypotheses. A near-replication was carried out in Louisville, Kentucky; results were much the same. For the crash-area groups there was association of aircraft fear and noise annoyance responses; this was true to an apparently lesser extent for non-crash-area groups. The greater annoyance of crash groups by aircraft community noise apparently does not carry over to assessment of aircraft noise in the laboratory.

Accidents, Aviation↗
PubMed1996

Nuclear morphometry in node-negative breast carcinoma.

OBJECTIVE: To determine whether nuclear morphometry can confirm or add useful information to classic clinicopathological prognosticators to identify the subpopulation of breast carcinoma patients with node-negative (N-) disease, at high risk of disease relapse. STUDY DESIGN: On the basis of results obtained by clinicopathologic evaluation of a group of patients with N- breast cancer, on a test group of 56 cases (32 patients disease free and 24 with relapse), we performed a morphometric analytical study of nuclei using the Shape Analytical Morphometry (SAM) software system; 20 nuclei for each case and 17 morphometric parameters for each nucleus were analyzed. RESULTS: The SAM system allowed us to quantify shape differences in nuclei in terms of contour irregularities and asymmetries along with evaluation of nuclear dimensions. Dimensional and analytic parameters were subjected to univariate (Student's t test) and multivariate (Hotelling's test) analysis. Multivariate discriminant analysis showed that an exact forecast of disease relapse could be made in 77% of patients with N- breast cancer by using a set of six both analytic and dimensional parameters. CONCLUSION: These results confirm that nuclear pleomorphism is the result of both contour irregularities and shape asymmetries and that even though they should be considered preliminary results, they stress the importance of quantitative shape evaluation.

Breast Neoplasms↗