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 595 records · Page 33Linked to original sources
PubMed1995

Multivariate statistical comparison of autoantibody-repertoires (western blots) by discriminant analysis.

A procedure for the quantification and comparison of complex (auto-) antibody repertoires of many individuals is described. It is based on multivariate statistical analyses of densitometric data of individual staining patterns, in this case Western blots. Implementation of algorithms can be based on standard software and hardware components. This procedure allows (i) the quantitative assessment of the reproducibility and reliability of any electrophoretic procedure, such as for the production of the immunological substrate, (ii) the statistical comparison of any group of staining patterns, e.g., derived from patients with autoimmune diseases or normal controls, (iii) the identification of the bands that contribute most to the differences between such groups, and (iv) the determination whether an unknown individual sample belongs to a known group. A statistical analysis of autoantibody repertoires has important possible applications: it detects spontaneous or therapeutically induced changes of repertoires, it identifies regions of interest, it supports the isolation of relevant antigens, and it will also be helpful in the diagnosis of autoimmune diseases.

Algorithms↗
PubMed1998

Information processing in schizophrenia and bipolar disorder: a discriminant analysis.

Schizophrenia and bipolar disorder are two distinct categories of mental disorders in the DSM-IV. However, it is often difficult to make a differential diagnosis because of the overlapping symptoms. A potential adjunct in the classification of schizophrenia and bipolar disorder is the application of information processing models, as patients with schizophrenia and possibly those with bipolar disorder have information processing deficits. A study was conducted in which a computerized battery of information processing tasks (called COGLAB) was administered to three participant groups: patients with schizophrenia, patients with bipolar disorder, and normal controls. The tasks included the Mueller-Lyer illusion, reaction time, size estimation, a variant of the Wisconsin Card Sorting Test, backward masking, and Asarnow continuous performance. Discriminant analyses were used to investigate the differences among the three groups. Results indicated that COGLAB correctly classified 75.5% of the cases of schizophrenia and bipolar disorder. The Mueller-Lyer illusion and the number of perseverative errors on the card sort most powerfully discriminated the two groups.

Bipolar Disorder↗
PubMed1993

[Noninvasive diagnosis of pulmonary hypertension in patients with COPD using regression and discriminant analysis].

The evidence of elevated non-invasive parameters in prediction of pulmonary hypertension was examined in 168 patients with chronic obstructive pulmonary disease (COPD). Forced vital capacity (FVC), Tiffeneau-test (FEV1), oxygen partial pressure, x-ray ascertained diameter of right descending branch of pulmonary artery (RDB), myocardial scintigraphy, and right ventricular ejection fraction showed significant differences between patients with and without pulmonary hypertension. Neither of them alone allows prediction of pulmonary pressure exactly, only in combination non-invasive parameters are usable. In non-invasive diagnosis of catheterisation we suggest to use a regression function including FEV1, FVC and RDB by an ascertained specificity of 80% and sensitivity of 75%. To classify the patients according to normal and increased pulmonary artery pressure, FEV1, RDB and myocardial scintigraphy by a discriminant function with specificity/sensitivity of 80%, are relevant parameters. Accurate diagnosis of latent pulmonary hypertension requires right heart catheterisation. The domain of non-invasive diagnosis is screening and therapy monitoring of pulmonary hypertension, or if right heart catheterisation is contraindicated.

Adult↗
PubMed1991

Prognostic signs in the evolution of premature thelarche by discriminant analysis.

Since premature thelarche (PT) can be a first sign of precocious puberty (PP), the aim of our study was to identify simple items in the course of the first 6 months of follow-up that could help predict if PT would evolve to PP. Thirty-two girls with PT were studied. First evaluation included bone age (BA), basal estradiol, FSH, LH and prolactin. GnRH was performed in 15 subjects and BA was checked at 6 month intervals in 30. Based on clinical outcome after a mean follow-up of 33.4 +/- 16.5 (SD) months, patients were divided into 2 groups: Group I (G-I) included subjects whose breast development either remained unchanged, increased or regressed; Group II (G-II) included subjects who progressed to PP. The multivariate combination of the items which was able to best discriminate between the two groups was chosen in predicting the evolution of PT. The items considered included four variables available at the time of diagnosis [chronological (CA) at onset less than 3 years, basal FSH, basal LH and BA/CA ratio] and two additional variables after a 6-month follow-up (delta BA/delta CA and growth velocity); 88% of G-I and 14% of G-II had CA less than 3 yr. Basal FSH levels were elevated in both G-I (7.6 +/- 3.0 mIU/ml) and G-II (12.1 +/- 4.1) with respect to controls (2.6 +/- 1.2); however, approximately 20% of G-I had low FSH levels. Basal LH levels were consistently higher in G-II (8.0 +/- 1.3 mIU/ml) than in G-I (2.9 +/- 1.5) or controls (2.8 +/- 1.2). Although initial BA was advanced (greater than 2SD) in 21% of G-I and in all of G-II, an acceleration of BA was seen only in G-II. The mean growth velocity of G-I (44.1 +/- 31.5%) was significantly less than G-II (92 +/- 32%; p less than 0.0025). With the help of the discriminant equations derived from data obtained at diagnosis and during the first 6 months of follow-up, all subjects with isolated premature thelarche could be sharply distinguished from those who subsequently progressed to precocious puberty.(ABSTRACT TRUNCATED AT 250 WORDS)

Adolescent↗
PubMed1985

Determination of sex with a discriminant analysis of new pelvic bone measurements: Part II.

The pelves of 100 white skeletons were measured on both sides for the following: (1) length from the superiormost aspect of the pubic symphysis to the nearest rim of the acetabulum (PS-A), (2) length from the highest point of the pubic tubercle to the nearest rim of the acetabulum (PT-A), (3) acetabular diameter (AD), (4) the vertical distance from the anterior aspect of the ischial tuberosity to the farthest rim of the acetabulum (IT-A), and (5) greatest femur head diameter. From these, three indices were derived: AD/PS-A (acetabulum/pubis index), AD/PT-A (acetabular diameter/pubic tubercle-acetabular rim index), and IT-A/PS-A (ischium-acetabulum height/pubic symphysis-acetabular rim index). The left AD/PS-A ratio and left IT-A height proved statistically to be of greatest discriminating value. Using these two variables, a discriminant function was derived which correctly separated 98% of our sample. The acetabulum/pubis ratio alone correctly assigned 95%. With either the discriminant function analysis of two variables or the acetabulum/pubis index as a single predictor, 97% of our sample of known sex was correctly identified if all specimens that fell within a doubtful or overlapping range of values were sorted by femur head diameter.

Anthropology, Physical↗
PubMed1997

Frequency domain discriminant analysis of the electroencephalogram.

A frequency domain generalization of the classical quadratic discriminant function was applied to the problem of classifying alpha-band multichannel electroencephalogram recordings in three task conditions. The data consisted of 41-channel recordings obtained in eyes closed, verbal, and spatial task conditions. Classifier performance was measured by deriving a decision rule from a training sample of 42 recordings and then applying the obtained rule to a test sample of 46 recordings. The proportion of correct classification was .93 in the training sample and .85 in the test sample. The classifier performed better when based on the complete cross-spectral matrix than when restricted to power spectrum variables. Classification based on a subset of 16 leads reduced the overall proportion of correct classification to .79 in the training sample and to .70 in the test sample.

Adolescent↗
PubMed1992

Magnesium, total calcium, phosphorus, copper, and zinc in plasma and erythrocytes of venous cord blood from infants of diabetic mothers: comparison with a reference group by logistic discriminant analysis.

Concentrations of magnesium (Mg), total calcium (Ca), phosphorus (P), copper (Cu), and zinc (Zn) were investigated in plasma (Pl) and erythrocytes (Erc) of venous cord blood of 44 infants of diabetic mothers (IDMs). These same concentrations plus total glycohemoglobin and fructosamine were determined at delivery in a subset of 15 mothers of these infants. Mineral results for IDMs were compared with those for 66 apparently healthy newborns. The duration of gestation in the two groups was significantly different (P < 0.001). After adjustment for gestational age, the mean (+/- SD) differences between groups were significant for birth weight, head circumference, Erc-Mg (1.71 +/- 0.17 for IDMs vs 1.76 +/- 0.15 mmol/L for control subjects), Pl-Ca (1.96 +/- 0.32 vs 2.48 +/- 0.22 mmol/L), Pl-P (1.99 +/- 0.40 vs 1.57 +/- 0.25 mmol/L), and Erc-Cu (10.9 +/- 2.41 vs 12.9 +/- 3.00 mumol/L), but not for Erc-Zn (33.0 +/- 18.3 vs 40.4 +/- 13.6 mumol/L). The variable that best discriminated between the two infant groups after adjustment for gestational age was Pl-Ca. In the 15 mothers, Pl-Mg (0.67 +/- 0.07 mmol/L) and Pl-Ca (1.66 +/- 0.21 mmol/L) concentrations were low, Pl-Zn (9.81 +/- 3.40 mumol/L) was normal, and Pl-Cu (33.5 +/- 10.7 mumol/L) was above normal. Correlations between total glycohemoglobin and mineral values of the mothers or paired IDM mineral values were not significant. The concentration of Pl-Ca was positively correlated with Erc-Cu (P < 0.001) and Pl-Cu (P < 0.05) in the comparison group newborns but not in the IDMs.

Calcium↗
PubMedDate not supplied

[The use of discriminant analysis for the differential diagnosis of periodontal diseases].

Basing on rheo- and photoplethysmography combined with mathematical methods and computers the authors have developed a method for the differential diagnosis of periodontal diseases (gingivitis, periodontitis). Diagnostic studies carried out in 200 patients have shown a high probability (at least 92 percent) of the method for the differential diagnosis of periodontal diseases.

Adolescent↗
PubMed2003

Recognition of eukaryotic promoters using a genetic algorithm based on iterative discriminant analysis.

A new approach to recognizing promoter regions of eukaryotic genes is proposed and illustrated by an example of Drosophila melanogaster. The essence of its novelty is in realizing the genetic algorithm to search for optimal partition of a promoter region into local nonoverlapping fragments and selection of the most significant dinucleotide frequencies for the fragments obtained. The method developed was applied to recognizing TATA-containing (TATA+) and DPE-containing (DPE+) promoters of Drosophila melanogaster genes. The program for promoter recognition is included into the GeneExpress system, section RegScan (http://wwwmgs.bionet.nsc.ru/mgs/programs/proga/).

Algorithms↗
PubMed2001

TMJ disorders and myogenic facial pain: a discriminative analysis using the McGill Pain Questionnaire.

Our aim was to assess the discriminative capacity of the McGill Pain Questionnaire (MPQ) in patients with temporomandibular joint (TMJ) disorders or with myogenous facial pain (MP). The MPQ was administered to 57 TMJ and 28 MP patients who were also asked to assess the level of pain using the Visual Analog Scale (VAS). Weighted MPQ item scores, subscale Pain Rating Indexes (PRI), total PRI and the number of words chosen were calculated. Mean scores were tested for significant differences (Student's t-test), and the frequency with which each descriptor was chosen by the patients in both groups was analyzed. Data were also processed through two systems based on a counter-propagation neural network: the Self-Organizing Map (SOM) system, and a cluster-like analysis. In the MP group, 16 out of 20 mean MPQ item scores and all mean PRI and VAS scores were significantly higher than those in the TMJ group. There was a marked difference in descriptor choice. In the TMJ group the following descriptors were chosen by 25% or more of the patients: tiring, troublesome, nagging, sore, tender, and aching. In the MP group the descriptors most frequently chosen were: 'exhausting' (57%), 'punishing' (50%), and pulling (47%). SOM analysis distributed the two groups in the two halves of the map: only two out of 28 MP cases (7%) and 12 out of 57 TMJ cases (21%) were misplaced. The cluster-like analysis based on the 20 MPQ item scores correctly recognized 94.73% of TMJ patients and 89.28% of MP patients. In conclusion, the MPQ consistently discriminated between TMJ and MP patients. Although the higher affective scores in the MP patients may be partly induced by higher levels of anxiety in these patients, the data convincingly show that the system's discriminative capacity relates to all MPQ subscores and to the majority of the MPQ items. Moreover, within the same item, the choice of verbal descriptors varies consistently between the two groups of patients.

Adult↗
PubMedDate not supplied

Diagnosis of dementias using partial least squares discriminant analysis.

Twenty patients with Alzheimer's disease (AD), 34 with senile dementia of the Alzheimer type (SDAT) and 12 with vascular dementia (VAD) were carefully investigated. The diagnoses were made according to the DSM-III-R criteria and for AD/SDAT also according to the NINCDS-ADRDA criteria. The data set consisted of 29 descriptor variables obtained at the examination, and the data were fitted to a partial least squares model using dummy variables for the response variables (i.e. clinical diagnoses). The distribution of the scores over the descriptor variable space revealed adequate classification regarding all three diagnoses, with some overlap between SDAT and AD and between VAD and SDAT. The study demonstrates a clear difference in symptomatology between AD and SDAT with no influence of age per se or the severity of dementia.

Aged↗
PubMed1993

[Evaluation of the quality of a decision-making rule in discriminant analysis].

Logistic regression (LR) and concept of the beta-confidence allocation rule were applied to predict the survival of dogs within the radiobiological experiment. The biochemical and haematological investigations in 86 dogs on the 2nd, 5th and 8th day post-external gamma-neutron irradiation (4.8-7.2 Gy) were used as explanatory variables for the prediction of more than 64 days survival. The influence of experimental conditions was monitored. Correctly predicted cases with LR models were proportioned as 81%, 85% and 93% on 2nd, 5th and 8th day respectively. Using beta-allocation rule, those correctly allocated with 0.50 confidence on the individual days were 75%, 81% and 85% of cases, that adds valuable information on the stability of the estimated classification rule. Method of beta-allocation also allows testing of whether individual observations are correctly assigned at a given significance level.

Animals↗
PubMed2000

Luminescence studies of the phagocyte response to endotoxin infusion into normal human subjects: multiple discriminant analysis of luminescence response and correlation with phagocyte morphologic changes and release of elastase.

A blood luminescence system (BLS) was employed to analyze blood phagocyte function in response to infusion of endotoxin (4 ng Escherichia coli lipopolysaccharide (LPS)/kg body weight) into 7 healthy human subjects. The subjects were closely monitored clinically, and extensive chemical, hematological and coagulation measurements were taken during the pretreatment, early (symptomatic, 1-8 h post-LPS), and late (asymptomatic, 12-48 h post-LPS) phases of acute inflammation. BLS assessment included measurement of basal and PMA-stimulated phagocyte oxidase and myeloperoxidase (MPO) activities, and also included measurement of circulating (COR) and PAF-primed maximum (MOR) opsonin receptor-dependent phagocytic activities. Basal oxidase activity peaked at T + 1 h and showed an additional peak at T + 24 h post-LPS. The COR activity also peaked at 1-2 h, but remained elevated through T + 24 h post-LPS, while the basal MPO activity peaked only once at T + 1 h. We concluded that while MPO evidence of phagocyte respiratory activation returned to baseline by T + 4 h, COR evidence of receptor expression (receptor alert) remained elevated through T + 24 h. During this early (0-8 h) period, elastase/alpha 1AT complex concentration peaked at T + 3-4 h and again at T + 8 h. Peak numbers of circulating polarized and vacuolated phagocytes also appeared at T + 3 h and 7 h. We concluded that there was biochemical and morphological evidence of continuing phagocyte activity beyond T + 4 h to T + 8 h, and that this corresponded with the period during which the subjects were symptomatic. In addition, the appearance of a second peak of basal oxidase activity at T + 24 h, multiple discriminant analyses of all the luminescence data, and the sustained elevation of lactate suggested that there was a later second stage (T + 12 h to 48 h) of the human response to endotoxin, during which time the subjects were asymptomatic.

Adult↗
PubMed1990

[Dilated cardiomyopathy. Multivariate discriminant analysis of main hemodynamic-angiographic indices].

The study group consisted of 47 consecutive patients (38 men and 9 women) aged between 16 and 56 years with dilated cardiomyopathy studied invasively between January 1980 and December 1986. Follow-up observation (40 +/- 29 months) showed that 28 patients were mildly symptomatic (group 1) and 19 patients were severely symptomatic (group 2: eight of them died due to intractable congestive heart failure). At univariate analysis, group 2 was characterized by higher pulmonary vascular resistance, larger end-diastolic and end-systolic left ventricular volumes, increased left ventricular mass, lower ratio of mass to volume index, depressed angiographic ejection fraction, lower ratio of end-systolic stress and peak systolic pressure to volume index. Multivariate analysis was used to determine which combination of factors might be a better effective predictor of prognosis in these patients: the most important factors were mass to volume index ratio (M/V) and end-systolic stress to volume index ratio (sigma TS/VTS). An equation was developed that can be applied to the prognosis of patients with dilated cardiomyopathy (F = 7.41573 M/V + 0.87805 sigma TS/VTS - 10.34571). A score system was devised on the basis of the summed-up value of these two parameters. When the patients were assigned (according to the score) to one of the previously mentioned two groups, the classification proved to be correct in 98% of the cases. Thus, these factors can accurately predict the development of congestive heart failure or the risk of death in patients with dilated cardiomyopathy, who are reasonable candidates for cardiac transplantation.

Adolescent↗
PubMed1995

Older adults' expectations about mental health counseling: a multivariate and discriminant analysis.

The present study aimed at identifying significant predictors of expectations about mental health counseling in a sample of highly educated elderly. This task was achieved by administering the full version of the Expectations About Counseling (EAC) questionnaire to fifty-seven retired professors. We first addressed the issue of the current elderly's under-utilization of formal counseling services, then conducted a literature review on the relationship between elderly's characteristics and their views of counseling. Specific hypotheses were formulated for each of the seven possible predictors of EAC scale scores. Previous counseling experience and marital status were significant predictors of EAC scale scores. Young-old adults (i.e., younger than 75 years of age) had received counseling experience significantly more than their older counterpart; their EAC scores, however, were not significantly different than those of old-old participants. Gender, area of residence, income and religiosity did not predict expectations about counseling significantly. Cell size heterogeneity for some predictors might have been responsible for lack of significance on additional factors. The article ends with a discussion of several clinical implications of the findings.

Aged↗
PubMed2003

Validity and reliability of a newly developed method for evaluating masticatory function using discriminant analysis.

The objective of the study was to assess validity and reliability of a newly developed method for evaluating mixing ability of masticatory function. A two-coloured paraffin wax cube was used as a test food in the mixing ability test. Eleven full dentate and 33 denture wearers participated in the validity assessment. They were divided into three groups according to their dental state. Validity was assessed by the Spearman's coefficient of correlations between the mixing ability test and the sieving method used as a 'gold standard' test of the masticatory performance. The subjects were asked to masticate the test foods for 5, 7, 10, 15, 20 and 30 strokes in each test. Significant correlations (P < 0.001) were obtained in natural dentition group (r(s) = 0.66), in denture-natural dentition group (r(s) = 0.72) and in denture-denture group (r(s) = 0.56). In the reliability assessment, three full dentate subjects were asked to repeat the mixing ability test a number of times. The intraclass correlation coefficients were 0.94 in interexaminer consistency, 0.98 in intra-examiner consistency, 0.89 in test-retest consistency and 0.99 in measuring consistency. These results indicated that the mixing ability test had the validity and the reliability for evaluation of the masticatory performance.

Adult↗
PubMedDate not supplied

Method of help for the diagnosis of the temporomandibular joint internal derangements. Discriminant analysis applied to the temporomandibular derangements.

OBJECTIVE: The purpose of the study is to find an objective method of help for the clinician in the diagnosis of the pathology of the temporomandibular joint, different of the image methods habitually utilized until this moment. MATERIAL AND METHOD: This study is based initially on the data obtained of a sample of 1164 patients with symptoms and/or signs of pathology of the temporomandibular joint. Nine different and excluding diagnostic groups settled down, according to the classification of the American Academy of Orofacial Pain (AAOP), in collaboration with the International Headache Society (IHS). We realized magnetic resonances to the patients and were selected those that adjust to the clinical criterion and of diagnosis for the image, and could only in a diagnostic group. Finally 449 patients were selected, 390 women and 59 men. RESULTS: The results obtained (expressed in percentage of well classified cases) by means of the proposed method were: Arthrosis 98.9%, Anterior Disk Displacement with Reduction (ADDR) 87.5%, Anterior Disk Displacement without Reduction (ADD) 100%, Capsulitis 100%, Disk Immobile (DIN) 97.9%, Hypermobility Condylar (HC)95.8%, Lateral Displacement Without Reduction (LD) 100%, Pathology Muscular (PM)100%, Disk Hipomobile (DHM) 86.4%. CONCLUSION: The proposed method reaches a fine percentage of successes in the diagnosis of these processes good enough, through its effectiveness as for its cost and should be considered an alternative in the diagnosis of temporomandibular derangements.

Diagnosis, Computer-Assisted↗
PubMed1994

[Discriminative analysis of dependent elderly between home care and institutionalized groups in Sendai].

Public health and welfare services were examined in 125 dependent elderly aged 70 to 89 years old from a randomly selected population of 1,249 from 3 different areas (central, residential and suburban) in Wakabayashi-ku, Sendai. Of the 125 dependent elderly, 91 (72.8%) were being cared for at home (the home group), and 34 (27.2%) were in hospitals or nursing homes (the resident group). Multivariate analyses using Hayashi's quantification method II were conducted for a comparison between the home group and the resident group. For the resident group relationships were found to "not having any family members to live with", and to "severe disability in ADL". Of the 91 subjects in the home group, 27 (29.7%) were visited by public health nurses (the visited group), and 64 (70.3%) were not (the not-visited group). Multivariate analyses using Hayashi's quantification method II were conducted for a comparison between the visited group and the not-visited group. For the visited group relation to husbands or wives being caregivers and to severe disability in ADL was found. Of the 91 in the home group, 15 (16.5%) received welfare services (the recipient group), and 76 (83.5%) did not (the non-recipient group). Again, multivariate analyses using Hayashi's quantification method II were conducted for a comparison between the recipient group and the non-recipient group. In the non-recipient group a relationship to not having caregivers or to caregivers being persons other than husbands, wives, children or grandchildren was seen.

Aged↗