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 1,171 records · Page 65Linked to original sources
PubMed1997

Diagnostic and prognostic relevance of malignancy-associated changes in cervical smears.

OBJECTIVE: For approximately 15 years, malignancy-associated changes (MACs) have been consistently found by means of high-resolution cytometry in different tissues, especially in visually normal appearing cervical cells. Their biologic nature is not yet fully understood. The aim of this investigation was to assess the expression of MACs in cervical smears and to evaluate their prognostic relevance. STUDY DESIGN: This study was performed on normal intermediate cells obtained from 53 cytologically positive and 78 cytologically negative cervical smears. From a second sample, 31 cases showing negative cytology were selected for a prospective longitudinal study. Densitometric and texture features were generated, and MACs were described on the basis of multivariate discriminant analysis. RESULTS: Discrimination between positive and negative cases was possible, with a correct classification rate of approximately 80%. After a mean period of 29.5 months, we noted no statistically significant increase in the incidence of cervical intraepithelial neoplasia in the group of healthy but MAC-positive women as compared to those who were MAC negative. CONCLUSION: MACs were constantly expressed in the epithelium of the cervix. Although their prognostic relevance remains unclear, MACs play an important role in the effort to automate cervical cytology.

Adult↗
PubMed2002

Biological markers and possibilities for predicting therapeutic results in schizophrenia: a methodological contribution.

The aim of the research was to select some easily available and easily replicable biological markers that could be used as predictors of both acute and long-term therapy. A selection of state markers (structural computer tomography [CT] parameters, psychological parameters, tests concentrating on attention and memory, "soft signs"), trait markers (quantified electroencephalograph [QEEG], cortisolemia, prolactinemia [PRL], and their changes) and clinical symptoms (Positive and Negative Syndrome Scale [PANSS], Clinical Global Impression [CGI]) were determined in 52 hospitalised schizophrenic patients showing either an acute episode or an exacerbation. The evaluation was repeated after one year of outpatient therapy. The parameters studied so far are not sufficiently specific and sensitive; therefore, the prediction cannot be based on a single parameter. In order to overcome this shortcoming, we carried out the analysis by means of multidimensional statistics, applying discriminant analysis. The results of both a broader and a more specific approach are stated in the paper. In all the used examples of discriminant analysis, the optimum discriminator is cortisolemia or its changes after administering dexamethasone and structural CT parameters. The results indicate that combinations of vulnerability markers and state markers (cortisolemia) may be of predictive value and are compatible with the vulnerability-stress ethiopathogenetic model of schizophrenia.

Adult↗
PubMedDate not supplied

Cross-validation and discriminant validity of Adolescent Health Promotion Scale among overweight and nonoverweight adolescents in Taiwan.

This study used cross-validation and discriminant analysis to evaluate the construct and discriminant validity of Adolescent Health Promotion (AHP) scale between the overweight and nonoverweight adolescents in Taiwan. A cross-sectional survey method was used and 660 adolescents participated in this study. Cluster and discriminant analyses were used to analyze the data. Our findings indicate that the AHP is a valid and reliable scale to discriminate between the health-promoting behaviors of overweight and nonoverweight adolescents. For the total scale, cluster analyses revealed two distinct patterns, which we designated the healthy and unhealthy groups. Discriminate analysis supported this clustering as having good discriminant validity, as nonoverweight adolescents tended to be classified as healthy, while the overweight tended to be in the unhealthy group. In general, overweight adolescents practiced health-related behaviors at a significantly lower frequency than the nonoverweight. These included exercise behavior, stress management, life appreciation, health responsibility, and social support. These findings can be used to further develop and refine knowledge of adolescent overweight and related strategies for intervention.

Adolescent↗
PubMedDate not supplied

Characterization of balsamic vinegar by multivariate statistical analysis of trace element content.

To characterize vinegars according to the types prescribed by Italian regulations, 8 trace elements (Cr, Mn, Co, Ni, Cu, Zn, Cd, and Pb) were determined. The data collected were successively elaborated by 3 statistical techniques: linear principal component analysis (LPCA), linear discriminant analysis (LDA), and cluster analysis (CA). LDA and LPCA best classified and discriminated the 3 types of vinegar under study, separating traditional balsamic vinegars from the other 2 types, nontraditionally aged balsamic vinegars and common vinegars. The latter 2 types were appreciably distinguished only by LDA through bidimensional analysis of discriminant scores.

Acetic Acid↗
PubMed1996

Sex determination by discriminant function analysis of lateral radiographic cephalometry.

The present work is an attempt to develop a new method to determine sex from the skull with lateral radiographic cephalometry and discriminant function analysis. The superciliary ridges, frontal sinuses, external occipital protuberance, and mastoid processes were adopted as objects of lateral radiographic cephalometric measurements. With discriminant functions created from 18 established cephalometric variables, a total of 100 cases were classified into two sexual groups with 100% accuracy in a random sample of Taiwanese adults. Therefore, we may obtain a much greater reliability of sex determination from skulls according to this newly developed technique.

Adult↗
PubMed2006

Metacarpophalangeal pattern profile analysis: useful diagnostic tool for differentiating between dyschondrosteosis, Turner syndrome, and hypochondroplasia.

PURPOSE: To assess the value of the metacarpophalangeal pattern profile (MCPP) analysis as a diagnostic tool for differentiating between patients with dyschondrosteosis, Turner syndrome, and hypochondroplasia. MATERIAL AND METHODS: Radiographic and clinical data from 135 patients between 1 and 51 years of age were collected and analyzed. The study included 25 patients with hypochondroplasia (HCP), 39 with dyschondrosteosis (LWD), and 71 with Turner syndrome (TS). Hand pattern profiles were calculated and compared with those of 110 normal individuals. Pearson correlation coefficient (r) and multivariate discriminant analysis were used for pattern profile analysis. Pattern variability index, a measure of dysmorphogenesis, was calculated for LWD, TS, HCP, and normal controls. RESULTS: Our results demonstrate that patients with LWD, TS, or HCP have distinct pattern profiles that are significantly different from each other and from those of normal controls. Discriminant analysis yielded correct classification of normal versus abnormal individuals in 84% of cases. Classification of the patients into LWD, TS, and HCP groups was successful in 75%. The correct classification rate was higher (85%) when differentiating two pathological groups at a time. Pattern variability index was not helpful for differential diagnosis of LWD, TS, and HCP. CONCLUSION: Patients with LWD, TS, or HCP have distinct MCPPs and can be successfully differentiated from each other using advanced MCPP analysis. Discriminant analysis is to be preferred over Pearson correlation coefficient because it is a more sensitive and specific technique. MCPP analysis is a helpful tool for differentiating between syndromes with similar clinical and radiological abnormalities.

Adolescent↗
PubMed2003

Near-infrared Raman spectroscopy of human coronary arteries: histopathological classification based on Mahalanobis distance.

OBJECTIVE: In this study, near-infrared Raman spectroscopy (NIRS) was used for evaluation of human atherosclerotic lesions using a simple algorithm based on discriminant analysis. The Mahalanobis distance was used to classify the clustered spectral features extracted from NIRS of a total of 111 arterial fragments of human coronary arteries. BACKGROUND DATA: Raman spectroscopy has been used for diagnosis of a variety of diseases. For real-time applications, it is important to have a simple algorithm that could perform fast data acquisition and analysis. The ultimate goal is to obtain a feasible diagnosis, which discriminates various atherosclerotic lesions with high sensitivities and specificities. MATERIALS AND METHODS: Non-atherosclerotic (NA) arteries, atherosclerotic plaques without calcification (NC), and atherosclerotic plaques with classification (C) were obtained and scanned with an NIR Raman spectrometer with 830-nm laser excitation. An algorithm based on the discriminant analysis using the Mahalanobis distance of the clustered spectral features was used for tissue classification into three categories: Na, NC, and C. RESULTS: Human coronary arteries exhibit different spectral signatures depending on different bio-chemicals present in each tissue type such as collagen, cholesterol, and calcium hydroxyapatite, respectively. It is shown that our algorithm has a maximum sensitivity and specificity of 85% and 89%, respectively, for the diagnosis of the NA tissue type, 85% and 89% for the NC tissue type, and 100% and 100% for the C tissue type. CONCLUSION: An algorithm (with a minimum of mathematical and computational requirements) based on the discriminant analysis of spectral features has been developed to classify atherosclerotic lesions with high sensitivities and specificities.

Algorithms↗
PubMed1990

[Discriminant function analysis for evaluating the status of coronary heart disease risk].

We investigated to discriminate those individuals categorized by 1. obesity, 2. hypercholesterolemia, 3. hypertension, 4. low maximal oxygen uptake, 5. an abnormal electrocardiogram reflecting ischemic patterns, and/or 6. real sedentary life, from relatively healthier individuals without coronary heart disease (CHD) risk factors. One hundred and six Japanese women, aged 30 to 72 years, all of whom were in the postabsorptive state, were recruited in a series of tests for anthropometric and physiologic profiles both during the resting state and during the submaximal-maximal cycling exercise. Subjects were categorized into two groups--those who possessed four or more of the above 1, 2, 3, 4, 5, and 6 (high-CHD-risk group, n = 15) and apparently healthy individuals with a minimum number of risk factors (low-CHD-risk group, n = 83). Analyses of the data revealed that a combination of 8 variables extracted from among original 25 variables accurately classified 13/15 (87%) of high-CHD-risk group and 77/83 (93%) of low-CHD-risk group (mean = 90/98 or 92%) into their respective groups. The 8 variables were double product, Katsura index, waist girth, chest girth, TG, TC, and skinfold thicknesses at the subscapular and abdominal sites. Subsequent t-test identified significant differences between groups not only for VO2max, SBP and TC but also for DBP, LDLC, TG, Hb, HR, and HRmax. Most of these differences were of a much greater magnitude compared to the existing difference in chronological age. These findings suggest the usefulness and importance of anthropometric and blood lipid variables in the explanation of differences in the health status between high-CHD-risk women and their counterparts.

Adult↗
PubMed2002

Diagnosis of female bladder outlet obstruction and relevance of the parameter area under the curve of detrusor pressure during voiding: preliminary results.

PURPOSE: The diagnosis of bladder outlet obstruction in women by pressure flow study may be difficult because there are several definitions of bladder outlet obstruction, several parameters and no standard cutoffs. We evaluated the ability of pressure flow studies to separate women into unobstructed, equivocal and obstructed groups. MATERIALS AND METHODS: In a prospective study 85 women with lower urinary tract symptoms underwent clinical evaluation, including physical examination, voiding cystourethrography, endoscopy, flow rate and post-void residual volume measurement. A pressure flow study was performed 15 days later. The pressure flow study parameters were maximum flow, post-void residual volume, detrusor pressure at maximum flow, vesical pressure at maximum flow, area under the curve of detrusor pressure during voiding and area under the curve of detrusor pressure during voiding adjusted for voided volume. After considering the clinical evaluation 2 urologists classified the patients into 3 groups, namely unobstructed, equivocal and obstructed, as the traditional classification. Linear discriminant analysis was then performed using the traditional classification and pressure flow study data. RESULTS: Mean patient age was 55 years (range 18 to 83). According to the traditional classification there were 36 unobstructed, 28 equivocal and 21 obstructed cases. Significant differences were noted in all pressure flow study parameters (analysis of variance p <0.05). Linear discriminant analysis showed that area under the curve of detrusor pressure during voiding adjusted for voided volume was the most statistically discriminating parameter. Of the cases 86%, 36% and 57% were identically categorized by the traditional and area under the curve of detrusor pressure during voiding adjusted for voided volume parameter classifications in the unobstructed, equivocal and obstructed groups, respectively. The other pressure flow study parameters showed less satisfactory results. CONCLUSIONS: Area under the curve of detrusor pressure during voiding adjusted for voided volume appears to be the most discriminating urodynamic parameter of female bladder outlet obstruction. Other studies are needed to test the reliability and validity of this new parameter.

Adolescent↗
PubMed2002

Evolution of the aroma profile of sherry wine vinegars during an experimental aging in wood.

Changes in the aroma profile of five Sherry wine vinegars submitted to an experimental static aging in wood were followed along 24 months. Eighteen volatile compounds were determined by GC-FID. The results were subjected to multivariate analyses: principal component analysis and linear discriminant analysis. The aroma profile of vinegar can be useful to discriminate vinegars produced from different substrates or with different aging times. During the experimental aging, volatile compounds such as methyl acetate, methanol, diacetyl, and gamma-butyrolactone underwent significant concentration increases. Moreover, the initial ethanol content of vinegars is a factor in the final aromatic richness. The formation of ethyl acetate stood out in samples with an initial ethanol content of approximately 2 alcoholic degrees.

4-Butyrolactone↗
PubMed1988

[Possibility of the current segregation analysis to discriminate between monogenic and multifactorial types of inheritance of traits. The effect of the structure of family data on model robustness and power of the analysis].

The influence of sampling designs for robustness of the autosomal major locus model and the multifactorial model as well as possibility of segregation analysis to discriminate these models was studied. Nuclear families and 3-generation pedigrees were considered. It was found that robustness of models increased, when the size of sibships in nuclear families grows and when configuration of pedigrees is complicated. The resolution power of the analysis is always increased with size elevation of sibships, the highest effect of the analysis being observed for sibships of the size 3 or 4. Consideration of new generations is only advisable, if attracting sibs of these generations, the resolution power being increased, provided that the parameters of models are of high value.

Genetics, Population↗
PubMed1994

Multivariate analysis of health status scores: chronic airway disorders and the MOS SF-36.

Multivariate analysis of variance (MANOVA) with follow-up canonical discriminant analysis may be used to interpret differences in health-related quality of life measured by the Medical Outcome Study Short Form 36 (MOS SF-36). Due to the moderate correlations between the 8 health dimensions of the SF-36, MANOVA is theoretically a more appropriate method than traditional univariate approaches for detecting group differences on the SF-36. Additionally, canonical discriminant analysis presents a novel approach to understanding the relationship between health dimensions of the SF-36 and model-independent variables. Results from the MANOVA and canonical discriminant analysis provide evidence of the sensitivity of the SF-36 in cross-sectional, self-reported data. Significant differences in health status (alpha less than or equal to 0.05) were found for the variables of age, and primary physician visits, and between levels of disease severity, type of breathing problem, whether patients had seen a specialist or not, use of emergency room, the comorbid states of depression and arthritis, and income. No significant differences in health status were reported between males and females or racial groups.

Chronic Disease↗
PubMed1994

Discriminant function analysis of body surface potential maps in acute myocardial infarction.

Using a newly developed 64-electrode portable mapping device, QRS and ST-T isointegral maps were compared in 194 control subjects and 101 patients. One hundred ninety-four control subjects (mean age, 48 years; 120 men) with no history of cardiac disease were selected randomly and mapped. One hundred one patients (mean age, 62 years; 77 men) were mapped at presentation of chest pain suggestive of first myocardial infarction (MI); all patients had classic 12-lead electrocardiographic findings--46 with anterior and 55 with inferior MI. The diagnosis was confirmed in all cases by a significant rise in serial cardiac enzymes. The mean delay between onset of chest pain to map recording was 163 minutes. Of the 101 patients, 78 were first mapped outside the hospital. Applying discriminant function analysis to the isointegral measurements made on the control subjects and on the first map of MI patients achieved a correct classification of 97% of the control subjects (189 of 194) and 72% of the anterior (33 of 46) and 76% of the inferior (42 of 55) MI groups. This preliminary study suggests that discriminant function analysis, based on isointegral maps, not only provides a method of separating control subjects from MI patients but that it can also differentiate between types of infarct. Further studies are required to improve the predictive values of discriminant function and to extend the methodology to assess both the site and size of MI.

Adult↗
PubMed1983

Discriminant function analysis of LD/BD children scores on the WISC-R.

Investigated the performance of learning-disabled (LD) and behavior-disorder (BD) children and youth on the WISC-R. A stepwise discriminant function analysis and chi-square analysis were used to analyze the performance of 135 children and youth (83 males, 52 females) on the WISC-R. The sample consisted of 68 LD and 67 BD Ss, whose ages ranged from 81-1 month to 168-6 months with a mean age of 123-1 months. Results suggest that discriminant function analysis can be useful as one criterion in differentiating LD children from BD children. Implications are discussed.

Adolescent↗
PubMed1979

Determination of sex by discriminant function analysis of postcranial skeletal measurements.

Thieme and Schull [16] took measurements of seven postcranial traits from black skeletons of known sex to devise a method for accurately sexing human remains. With a linear discriminant function analysis of the data they obtained a 98.5% accuracy. The research reported here was designed to assess the accuracy of their method in the hands of other investigators using a sample of the same black population and its applicability to other populations of the same and different races. None of the samples used in this study were sexed with the same rate of accuracy either achieved or predicted by Thieme and Schull. However, two samples did reach 91% or better. Such results suggest that the methods employed have considerable value if used with appropriate precautions. Our multiple discriminant function analysis strongly indicates that neither side nor observer techniques affected the results. Sex accounts for a predominant fraction of the total variation among samples and race accounts for most of that remaining.

Anthropometry↗
PubMed1990

An analysis of the level of dental care in cardiac patients at risk for infective endocarditis.

Three hundred and thirty-eight cardiac patients, at risk of infective endocarditis, filled in a questionnaire consisting of a series of questions on their social, educational, familial, professional, medical and dental history and also on their level of dental care. Univariate (chi-square test) and stepwise discriminant analysis were applied to evaluate the factors related to or responsible for the presence or absence of a sufficient level of dental care, defined as at least one dentist visit per year and once daily teeth brushing. An acceptable level of dental care was present in one third of the dentate but almost nonexistent in edentulous patients. Univariate analysis revealed that in dentate patients level of education, knowledge of endocarditis prophylaxis, previous cardiac surgery, time since first heart operation and type of previous operation (artificial valve surgery) were significantly related to the level of dental care. Level of education, knowledge of endocarditis prophylaxis and time since first heart operation were withheld as independent factors by discriminant analysis. A discriminant score was calculated for each patient which permitted classification of the patients in two groups: one with a very high chance for absence of appropriate dental care and one with an equal chance for sufficient or insufficient care. Patients were also classified according to the number of risk factors. An answer to one of the 3 variables, selected by discriminant analysis, unfavorable to the presence of sufficient dental care, was considered a risk factor. Patients with two risk factors had a very high chance for the absence of sufficient dental care; in those with 0 or 1 factor, chances were divided between sufficient and insufficient dental care.

Adolescent↗
PubMed2003

Monitoring cell cycle distributions in living cells by videomicrofluorometry and discriminant factorial analysis.

BACKGROUND: The study of the cell cycle of living cells is often based on quantification of nuclear DNA. These studies may be improved by multifactorial analysis evaluating several parameters for each cell. METHODS: Single lymphoblastoid living cells were labeled with three fluorescent markers: Hoechst 33342 for nuclear DNA, Rhodamine 123 for mitochondria, and Nile Red for plasma membrane. Numerical image analysis allowed us to obtain, for each cell, morphological parameters (e.g., cell size, nuclear size, and shape) and functional information (e.g., nuclear DNA content, level of mitochondria energetic state, and the amount and properties of the plasma membrane) by fluorescence intensity. These parameters were used in a typological analysis that separated control cells into four groups. RESULTS: A discriminant factorial analysis (DFA) confirmed the four groups: G0-G1, S, G2+M, and polyploid cells called Gn. These groups were significantly different, with a classification probability of 0.9999; these control cells defined a learning population. Different populations of untreated and adriamycin-treated cells were analyzed as additional individuals within a DFA and were classified into the G0-G1, S, G2+M, and Gn groups by their probability of belonging to each of the groups. CONCLUSIONS: This approach is particularly efficient when it is used to determine variations in cellular properties and to objectively study cellular populations.

Cell Cycle↗