Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Discriminant Analysis”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 523 records · Page 29Linked to original sources
PubMed1997

Discriminant analysis of brain imaging data identifies subjects with early Alzheimer's disease.

In vivo functional brain imaging provides an opportunity to quantify and localize functional deficits associated with Alzheimer's disease (AD), in relation to dementia severity and heterogeneous cognitive profiles. Such imaging also provides a basis for distinguishing AD from other causes of dementia and for making an early diagnosis of disease. One imaging modality that can elucidate AD is positron emission tomography (PET), which is used to measure regional cerebral metabolic rates for glucose (rCMRglc) and regional cerebral blood flow (rCBF). Resting-state measurements with PET, when related to cognitive profiles in longitudinal studies, indicate that specific cognitive defects are preceded and predicted by reductions in rCMRglc in regions subserving the cognitive functions tested. Metabolic reductions and right/left metabolic asymmetries can be used to convert a "possible" to a "probable" diagnosis of AD by the National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer's Disease and Related Disorders Association (NINCDS-ADRDA) criteria. Furthermore, discriminant analyses of PET metabolic patterns can identify patients at risk for AD with mild memory deficits as having probable AD. In the future, stimulation PET studies should augment the power of this discriminant analysis.

Aged↗
PubMed2002

Bayesian framework for least-squares support vector machine classifiers, gaussian processes, and kernel Fisher discriminant analysis.

The Bayesian evidence framework has been successfully applied to the design of multilayer perceptrons (MLPs) in the work of MacKay. Nevertheless, the training of MLPs suffers from drawbacks like the nonconvex optimization problem and the choice of the number of hidden units. In support vector machines (SVMs) for classification, as introduced by Vapnik, a nonlinear decision boundary is obtained by mapping the input vector first in a nonlinear way to a high-dimensional kernel-induced feature space in which a linear large margin classifier is constructed. Practical expressions are formulated in the dual space in terms of the related kernel function, and the solution follows from a (convex) quadratic programming (QP) problem. In least-squares SVMs (LS-SVMs), the SVM problem formulation is modified by introducing a least-squares cost function and equality instead of inequality constraints, and the solution follows from a linear system in the dual space. Implicitly, the least-squares formulation corresponds to a regression formulation and is also related to kernel Fisher discriminant analysis. The least-squares regression formulation has advantages for deriving analytic expressions in a Bayesian evidence framework, in contrast to the classification formulations used, for example, in gaussian processes (GPs). The LS-SVM formulation has clear primal-dual interpretations, and without the bias term, one explicitly constructs a model that yields the same expressions as have been obtained with GPs for regression. In this article, the Bayesian evidence framework is combined with the LS-SVM classifier formulation. Starting from the feature space formulation, analytic expressions are obtained in the dual space on the different levels of Bayesian inference, while posterior class probabilities are obtained by marginalizing over the model parameters. Empirical results obtained on 10 public domain data sets show that the LS-SVM classifier designed within the Bayesian evidence framework consistently yields good generalization performances.

Artificial Intelligence↗
PubMed1989

Characterization and grouping of Trypanosoma brucei brucei, T.b. gambiense and T.b. rhodesiense by quantitative DNA-cytofluorometry and discriminant analysis.

For characterization and differentiation 13 stocks of Trypanosoma brucei spp. were used for a quantitative cytofluorometric determination of their DNA fluorescence intensities by two base-pair-specific fluorochromes. T. b. gambiense could be distinguished from T. b. brucei by an about 20% smaller G-C, and A-T content in the nuclear DNA and a 4% greater G-C and 15% greater A-T content of the kinetoplast DNA. T. b. gambiense could be differentiated from T. b. rhodesiense by a 20% smaller G-C and a 3.8% smaller A-T nuclear DNA content and a 2% greater G-C and a 18% greater A-T DNA content in the kinetoplast. After chromomycin (G-C specific) staining the DNA ratio kinetoplast/nucleus of T. b. gambiense was 6.8, of T. b. brucei 5.2 and of T. b. rhodesiense 5.4. The corresponding values after DAPI (A-T specific) application were 20.8 for T. b. gambiense, 14.4 for T. b. brucei and 16.6 for T. b. rhodesiense. With the fluorescence intensities a discriminant analysis has been computed. After chromomycin application the gambiense stocks could be separated from T. b. brucei and T. b. rhodesiense by this method with a hit rate of 100%. Such perfect separation could not be observed between T. b. brucei and T. b. rhodesiense. Even if most of them were classified into the corresponding subgroups some trypanosomes would nevertheless pass over into the brucei or rhodesiense subgroup and vice versa.(ABSTRACT TRUNCATED AT 250 WORDS)

Animals↗
PubMed2002

Glass classification by linear discriminant analysis of LA-ICP-MS data.

The classification of pharmaceutical and common glasses (ampoules, infusion bottles, cylinders, lead crystal and bottle glass) has been performed means by of chemometric methods. For this purpose intensity data of glass examinations received by laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) were applied. A Nd:YAG laser with 10 Hz repetition rate in the Q-switch mode at its 4th harmonic (266 nm) was used to get mass spectrometric data of the glasses. 13 isotopes (= variables) were used for measurements (7Li, 11B, 23Na, 24Mg, 27Al, 28Si, 39K, 42Ca, 47Ti, 57Fe, 123Sb, 137Ba, 207Pb). The glass samples represented four types of glasses: 1. borosilicate glass (type "Duran") 2. borosilicate glass (type "Fiolax") 3. soda-lime glass 4. lead glass. Calculations were made by using the SAS statistical software. All results were obtained on the assumption that the underlying data are normally distributed. By applying linear discriminant data analysis a classification of all types of the glasses was possible (13 variables included). The use of the method of "stepwise discriminant analysis" reduced the minimal required number of variables to eight. All types of glasses could be distinguished faultlessly. Besides a graphic system consisting of the isotopes 24Mg-47Ti-137Ba was developed. By means of this system optical distinction of glass classes was possible. The developed method allows to distinguish several types of pharmaceutical and common glasses on the basis of intensity data of mass spectrometric measurement without calibration experiments and linear regression.

Algorithms↗
PubMed2003

Minimum distance probability discriminant analysis for mixed variables.

Minimum distance probability (MDP) is a robust discriminant algorithm based on a distance function. In this article, we generalize the use of MDP to the case of mixed (continuous and categorical) variables by means of the individual-score (IS) distance. This distance assumes an underlying parametric model and is based on the score transformation of the data. We have adapted it to the usual case of ignoring the distribution of the whole set of observed variables, but assuming that some knowledge about the marginal distributions is available. Finally, MDP with IS distance (IS-MDP) is compared with other discriminant methods (including those designed for mixed data) in several examples and simulations. IS-MDP is shown to be the most efficient method according the leave-one-out criterion.

Breast Neoplasms↗
PubMed2005

Sex determination by discriminant analysis: an evaluation of the reliability of patella measurements.

Recently, a method which estimates the sex of unknown skeletal remains by discriminant function analysis of the patella has received much recognition. With seven simple measurements and a sexing accuracy of almost 84% this morphometric approach would be very suitable both in the forensic and anthropological context. In order to put this method to the test, 52 unfractured and non-pathological left and right patellae (26 males, 26 females) from a prehistoric skeletal sample were analyzed. In addition, 30 patellae were randomly selected which were not part of the original reference series. The first 15 (4 males, 11 females) had the same sample origin as the 30 target specimen but were of mediocre preservational state. The second 15 belonged to an archaeological sample of different provenience (8 males, 7 females). The results of the current study indicate that the morphometric assessment of this sesamoid is indeed well-suited as the results of the previous study could be duplicated, when sample size was not taken into consideration (85%). However, when the issue of small sample size was systematically addressed via a jack-knife procedure or the calculation of Delta-value, the resultant accuracies dropped to 78 and 74%, respectively. It is suggested that two confounding factors--age and sex of the individual--may be held accountable. There are slight, whilst not statistically significant tendencies to misclassify female and older specimen.

Anthropometry↗
PubMed1997

[Predictive value of the study of cerebro-spinal fluid in neurological diseases. Discriminant analysis of 733 cases].

INTRODUCTION: The study of cerebrospinal fluid (CSF) gives very valuable data for the diagnosis of certain neurological disorders. Objective. To discover the discriminative value of different tests and their usefulness in the study of neurological disorders. MATERIAL AND METHODS: Discriminative multivariate analysis was used for the group of parameters obtained from study of CSF and serum in a sample of 733 patients. BO was included, CSF cytobiochemical data and five of the equations used for calculation of the synthesis of intrathecal IgG. RESULTS: The best discriminative function was expressed as the intrathecal IgG with normal BHE. The formula of Tourtellotte was noteworthy for overestimation of the synthesis of intrathecal IgG when the albumin index indicated BHE rupture. Prospective reclassification of the 733 cases showed that these tests are well worthwhile to rule out immunologically mediated disorders (93% of the cases were correctly reclassified); they suggest the diagnosis in multiple sclerosis (66% correctly reclassified) and give very poor reliability when trying to distinguish between inflammatory infectious processes. CONCLUSIONS: In spite of their undoubted use in supporting the clinical diagnosis, the tests used for study of the synthesis of IgG, even complemented by cytobiochemical examination of the CSF, are not in themselves sufficient for use as a discriminative test in the differential diagnosis of neurological processes.

Albumins↗
PubMed2001

Differentiation between pigmented Spitz naevus and melanoma by digital dermoscopy and stepwise logistic discriminant analysis.

Epiluminescence light microscopy (ELM) has proven useful in the diagnosis of pigmented skin lesions (PSLs). However, in some cases this technique does not sufficiently increase the diagnostic accuracy in distinguishing pigmented Spitz naevi (PSNs) from melanoma. With the aim of obviating these problems of qualitative interpretation, methods based on the mathematical analysis of PSLs, such as digital dermoscopy analysis (DDA), have recently been developed. In the present study we used a digital dermoscope (DBDermo-MIPS, Dell'Eva-Burroni) to analyse PSNs and melanomas with similar clinical and dermoscopic features for any correlation between variables and to determine its discriminating power with respect to histological diagnosis. The 100 lesions underwent histological examination by three experienced dermatopathologists and were identified as PSNs (43) or melanomas (57). Thirty-six parameters were identified as possible discriminating variables and were grouped in four categories: geometry, colour, texture, and islands of colour. Statistical analysis was used to identify the variables with the highest discriminating power. Stepwise discriminant analysis selected only four variables: entropy, minimum diameter, red lesion value and peripheral dark (the means of these variables were higher in melanomas than in PSNs). Thus the combined use of digital dermoscopy and stepwise logistic discriminant analysis made it possible to single out the best objective variables for distinguishing PSN and melanoma.

Dermatology↗
PubMedDate not supplied

[The use of discriminant analysis in the classification of the alcohol intake in a general adult population].

Different studies have highlighted the efficacy of some laboratory parameters as markers of the abuse of alcohol in alcoholics populations. The intent of the present study has been to verify if the same variables are able to discriminate the intake of alcohol also in a general population. The survey has been carried out on 510 subjects (224 males and 286 females), between 20 and 70 years old, living in a rural area in the North of Italy. Statistically relevant differences have been found amongst mean values of mean globular volume (VGM), containing hemoglobin (MCH), gamma-glutamil transferase (GGT), HDL cholesterol (ColHDL), tryglycerides, apoproteins-A (APO-A), serum glutamic-ossalacetic transaminase (SGOT), leukocyte, lead blood level in subgroups of population stratified according to drinking habits (heavy, moderate and medium drinkers). The discriminant function analysis, applied using all the 18 parameters as continuous variables, allows the correct classification of 57.84% of the cases amongst moderate, medium and heavy drinkers. The most influent selected variables result lead blood level, GGT, hematocrit and MCH. If we leave out the lead blood level (not weighted as a routine) the percentage falls to 52.35%. The analysis carried out on "extreme" groups (moderate and heavy drinkers) identifies as most influent the lead blood level, GGT, APO-A, and VGM variables, allowing the correct classification of 90.14% of the cases (83.39% excluding the lead blood level, selecting GGT, hematocrit, VGM and serum glutammic-piruvic transaminase. The casual reselection of samples confirms the obtained data, giving evidence of the independence of the used variables from particular characteristics of the sampled population.(ABSTRACT TRUNCATED AT 250 WORDS)

Adult↗
PubMed2002

Discriminant analysis of image karyometric variables in benign and malignant melanocytic skin lesions.

OBJECTIVE: To perform a quantitative analysis to identify which of 7 nuclear morphometry-related variables are of diagnostic value in distinguishing benign from malignant melanocytic skin lesions. STUDY DESIGN: At the Institute of Pathology, University of Nis, formalin-fixed, paraffin-embedded skin biopsies from 23 cases of benign nevi (18 intradermal and 5 junctional) and 25 cases of primary nodular malignant melanomas were retrieved. Specimens were routinely stained with hematoxylin and eosin and analyzed using a computer-assisted interactive image analysis system. Nuclear area, equivalent diameter, volume of equivalent sphere, perimeter, mean chord, circularity and integrated optical density were estimated after manual editing of binary images. RESULTS: In univariate analysis, 6 features were found to be significantly different between the benign and malignant groups (P < .0001); all measured nuclear variables (except circularity) were higher in malignant melanomas. No significant differences were found among lesions with respect to nuclear shape. Using discriminant function analysis, a correct diagnosis was achieved in 95.8% of benign nevi cases and 84.0% of malignant melanoma cases. The best discriminant variable was nuclear area. CONCLUSION: Image analysis is diagnostically relevant to the evaluation of melanocytic lesions of the skin. The area of the nucleus appeared to have potential for differentiating benign from malignant tumors and can be estimated in the course of routine histology.

Cell Nucleus↗
PubMed1990

The microscopic features of inflammatory abdominal aortic aneurysms: discriminant analysis.

Up to 15% of abdominal aortic aneurysms are designated as inflammatory. They are characterized by marked fibrous thickening of the aneurysmal wall, with the fibrosis extending into the adjacent retroperitoneum. Thirty-five abdominal aortic aneurysms were studied, 15 inflammatory and 20 atherosclerotic. Of the inflammatory group, 10 were symptomatic and five asymptomatic. For each resection specimen, 59 microscopic features (variables) were scored semi-quantitatively. Discriminant function analysis showed that endarteritis obliterans, fibrosis around nerves or ganglia at the outer margin of mural fibrosis, and the thickness of the combined fibrotic media and adventitia gave a satisfactory high discrimination between atherosclerotic and inflammatory aneurysms. When these three variables are used together, a histological diagnosis of inflammatory aneurysm can be made with an expected accuracy in excess of 80%.

Aged↗