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 235 records · Page 13Linked to original sources
PubMed2006

Discriminative analysis of lip motion features for speaker identification and speech-reading.

There have been several studies that jointly use audio, lip intensity, and lip geometry information for speaker identification and speech-reading applications. This paper proposes using explicit lip motion information, instead of or in addition to lip intensity and/or geometry information, for speaker identification and speech-reading within a unified feature selection and discrimination analysis framework, and addresses two important issues: 1) Is using explicit lip motion information useful, and, 2) if so, what are the best lip motion features for these two applications? The best lip motion features for speaker identification are considered to be those that result in the highest discrimination of individual speakers in a population, whereas for speech-reading, the best features are those providing the highest phoneme/word/phrase recognition rate. Several lip motion feature candidates have been considered including dense motion features within a bounding box about the lip, lip contour motion features, and combination of these with lip shape features. Furthermore, a novel two-stage, spatial, and temporal discrimination analysis is introduced to select the best lip motion features for speaker identification and speech-reading applications. Experimental results using an hidden-Markov-model-based recognition system indicate that using explicit lip motion information provides additional performance gains in both applications, and lip motion features prove more valuable in the case of speech-reading application.

Algorithms↗
PubMedDate not supplied

Quantitative electroencephalography and regional cerebral blood flow: discriminant analysis between Alzheimer's patients and healthy controls.

Forty-two patients with probable Alzheimer's disease (AD) and 18 elderly healthy controls underwent quantitative EEG (qEEG) and regional cerebral blood flow (rCBF; 133Xe clearance) examinations. Correlations were sought between relative qEEG band powers and percent rCBF values in a posterior temporoparietal region of interest in either hemisphere. Moreover, stepwise discriminant analysis was applied to study the accuracy of the two techniques in differentiating AD from healthy ageing. rCBF and qEEG were correlated with one another, especially in the right hemisphere (p values ranging from <0.001 to <0.01). Significant correlations were found between Mini Mental State Examination (MMSE) and relative power of both the 2- to 6-Hz and the 6.5- to 12-Hz bands on either side (p < 0.001), and between MMSE and left rCBF (p < 0.005), while the correlation with right rCBF was poorer (p < 0.02). The statistical procedure identified the right values of both examinations for the discriminant analysis. Sensitivity of qEEG and rCBF employed together was 88% and specificity 89%, with a total accuracy of 88.3%. The unrecognized patients (n = 5) were affected by mild AD (4 scoring 3 at the Global Deterioration Scale and 1 scoring 4). qEEG alone showed an accuracy of 77% in the whole group and of 69% in mild AD, and rCBF alone an accuracy of 75% in the whole group and of 71% in mild AD. It is concluded that qEEG and rCBF examinations employed together are accurate tools to differentiate AD from normal ageing, although a lower degree of accuracy is achieved in mildly demented patients.

Aged↗
PubMed2000

Discriminant analysis models for early detection of glaucomatous optic disc changes.

AIM: To evaluate and compare four different mathematical formulas for the early detection of morphometric optic nerve head changes in chronic open angle glaucoma. METHODS: The optic nerve heads of 161 patients with perimetrically defined glaucomatous optic nerve damage and of 194 normal subjects were examined by confocal laser scanning tomography. Using four formulas of linear discriminant analysis and the optic cup shape measure as the single optic disc variable, the predictive power of each of these methods was examined to differentiate between the normal eyes and the glaucoma eyes. RESULTS: The highest predictive power had an optic disc sector based formula, in particular in eyes with medium and large optic discs. This optic disc sector based formula was the one with the best agreement with the other formulas examined. It achieved a better predictability than any single optic disc variable evaluated. CONCLUSIONS: Combining quantitative optic disc variables by discriminant analysis functions, the predictive power of semiautomatic quantitative optic nerve head evaluation can be improved by providing the ophthalmologist with a diagnostic score for the detection of glaucomatous optic nerve damage. Because of the pattern of glaucomatous neuroretinal rim loss, an optic disc sector based discriminant formula may have a higher diagnostic precision than other formulas in detecting early glaucomatous damage.

Adult↗
PubMed1976

A discriminant analysis approach to the identification of human chromosomes.

The identification of the various types of the 46 chromosomes in a normal human cell is formulated as a discriminant analysis problem. Derived are expressions for posterior probabilities, and rules are given for identification where the number of chromosomes of each type is exactly known. The resulting model is applied to experimental data from chromosomes of the Denver B-group. The results using the developed model compare favorably with those of the standard discriminant analysis approach.

Chromosomes↗
PubMed1993

Cholescintigraphy and biochemical tests in cholecystitis--an evaluation with discriminant analysis.

The biochemical values of 76 patients with suspected cholecystitis were subjected to discriminant analysis. The final diagnoses, i.e. acute cholecystitis, chronic cholecystitis and non-biliary disease, were used as the grouping variable. Cholescintigraphy identified patients with acute cholecystitis. Routine preoperative biochemical tests were found to be of limited value. Only alkaline phosphatase was of help in predicting common-duct stones, especially in patients with acute cholecystitis. The conclusion is that many biochemical tests presently in common use could as well be dispensed with.

Acute Disease↗
PubMed1989

[Prediction of operative mortality by a discriminant analysis for organ functions in patients with esophageal cancer--organ function index].

An "organ function index" (OFI) predicting the risk of operative mortality was presented. OFI was estimated on the basis of dysfunction of the systemic organs in patients with esophageal cancer. The pulmonary, cardiac, hepatic and renal functions were assessed by 23 parameters in 108 patients when they were admitted. Operative death was defined as death due to operative complications occurring within 120 days after esophagectomy or by-pass operation. For a discriminant analysis, patients were limited to those in the early period (from October 1981 until December 1985) when the incidence of operative mortality was relatively higher and the parameters were also limited to statistically evaluable ones. Then, a discriminant analysis was performed using data on 18 parameters of four organs in 35 patients each of whom had no deficit in these data. Operative death occurred in 8 out of these 35 patients. Based on the data, an equation to calculate OFI was generated. It consisted of 7 parameters regarding pulmonary, hepatic and renal functions. The values of OFI less than zero predicted no operative death while those more than zero did predict operative death. The prediction rate on presence or absence of operative mortality by this equation was 91.4% in 35 patients. For clinical application, the predictable risk of operative mortality based on OFI was classified as high (OFI less than 1.4), intermediate (0 less than OFI less than 1.4), or low (OFI less than 0).(ABSTRACT TRUNCATED AT 250 WORDS)

Discriminant Analysis↗
PubMedDate not supplied

A multiple discriminant analysis of smoking status and health-related attitudes and behaviors.

Using multiple discriminant analysis, we examined several health-related attitudes and behaviors (HABs) simultaneously across groups of university students differing by smoking status (n = 1,077). Nine HABs were considered: health responsibility, exercise, nutrition, interpersonal support, stress management, alcohol consumption, drug use, caffeine consumption, and safety practices. Overall, the findings indicated that HABs, particularly those involving substance use, differed among the smoking and nonsmoking groups. Furthermore, current smokers, former smokers, and nonsmokers represented a continuum of less healthful to more healthful attitudes and behaviors. In general, compared to men, women exhibited more positive HABs with respect to interpersonal support, health responsibility, alcohol consumption, and drug use, but less positive HABs with respect to stress management. More complex relationships emerged in comparisons of occasional and regular smokers, light and heavy smokers, and consonant and dissonant smokers. We discuss implications of the findings for smoking intervention programs.

Adolescent↗
PubMed1993

[Preoperative grading of renal cell carcinoma by plasma immunosuppressive acidic protein level and tumor diameter using discriminant analysis].

To predict histological grade of renal cell carcinoma preoperatively, an equation using plasma immunosuppressive acidic protein (IAP) level and tumor diameter was proposed. Between April 1977 and December 1991, 281 patients, 195 males and 86 females, aged between 20 and 84-year-old, were operated at our institutions. Multivariate analysis was performed to detect relationship between tumor grade and various factors that have continuous quantity. Among plasma components tested, plasma IAP levels showed the best correlation with grade (r value = 0.526, n = 157). Tumor diameter also showed correlation with grade (r value = 0.476, n = 273). Therefore, these two factors were used as variates. Discriminant analysis was applied to differentiate the patients with grade 1 or grade 2 tumor from those with grade 3 ones. From this analysis the linear function was obtained; z = 6.9429 - 0.0048x1 - 0.4118x2 (x1: IAP (micrograms/ml), x2: diameter (cm)). Tumor was assumed to be grade 1 or grade 2 when z value was positive, and it was assumed to be grade 3 when z value was negative. Overall accuracy was 87.7%. Positive predictive value of grade 1 or grade 2 tumor, and grade 3 tumor was 91.7% and 77.8%, respectively. Sensitivity was 90.9% for grade 1 or grade 2 tumor, and 79.5% for grade 3 tumor. As a result, it's possible to estimate tumor grade before operation using discriminant analysis in patients with renal cell carcinoma.

Adult↗
PubMed1990

Discriminant analysis of urethral pressure profilometry data for the diagnosis of genuine stress incontinence.

Urethral pressure profilometry (UPP) has been advocated for the diagnosis of genuine stress incontinence (GSI) but no published data exist to define clearly the criteria of measurement that should be used. The aim of this study was to examine the value of UPP in the diagnosis of GSI. Data from 102 urodynamically normal women and 70 women with GSI were compared. Thirty UPP variables from the resting and stress profiles were examined. The urodynamic diagnosis was made on the basis of a pad test, uroflowmetry and videocystourethrography. Each UPP variable was examined singly and thereafter all the measurements were examined by discriminate analysis. The single most discriminatory UPP variable was 'area under the stress profile' but the overlap between normal and GSI patients was too great to allow the test to be used diagnostically. Discriminate analysis resulted in correct classification in 78% of cases. On the basis of these results, UPP is not an accurate test for the diagnosis of GSI.

Discriminant Analysis↗
PubMed2000

Classification of antibiotic resistance patterns of indicator bacteria by discriminant analysis: use in predicting the source of fecal contamination in subtropical waters.

The antibiotic resistance patterns of fecal streptococci and fecal coliforms isolated from domestic wastewater and animal feces were determined using a battery of antibiotics (amoxicillin, ampicillin, cephalothin, chlortetracycline, oxytetracycline, tetracycline, erythromycin, streptomycin, and vancomycin) at four concentrations each. The sources of animal feces included wild birds, cattle, chickens, dogs, pigs, and raccoons. Antibiotic resistance patterns of fecal streptococci and fecal coliforms from known sources were grouped into two separate databases, and discriminant analysis of these patterns was used to establish the relationship between the antibiotic resistance patterns and the bacterial source. The fecal streptococcus and fecal coliform databases classified isolates from known sources with similar accuracies. The average rate of correct classification for the fecal streptococcus database was 62.3%, and that for the fecal coliform database was 63.9%. The sources of fecal streptococci and fecal coliforms isolated from surface waters were identified by discriminant analysis of their antibiotic resistance patterns. Both databases identified the source of indicator bacteria isolated from surface waters directly impacted by septic tank discharges as human. At sample sites selected for relatively low anthropogenic impact, the dominant sources of indicator bacteria were identified as various animals. The antibiotic resistance analysis technique promises to be a useful tool in assessing sources of fecal contamination in subtropical waters, such as those in Florida.

Animals↗
PubMed2003

Prediction of dihydrofolate reductase inhibition and selectivity using computational neural networks and linear discriminant analysis.

A data set of 345 dihydrofolate reductase inhibitors was used to build QSAR models that correlate chemical structure and inhibition potency for three types of dihydrofolate reductase (DHFR): rat liver (rl), Pneumocystis carinii (pc), and Toxoplasma gondii (tg). Quantitative models were built using subsets of molecular structure descriptors being analyzed by computational neural networks. Neural network models were able to accurately predict log IC(50) values for the three types of DHFR to within +/-0.65 log units (data sets ranged approximately 5.5 log units) of the experimentally determined values. Classification models were also constructed using linear discriminant analysis to identify compounds as selective or nonselective inhibitors of bacterial DHFR (pcDHFR and tgDHFR) relative to mammalian DHFR (rlDHFR). A leave-N-out training procedure was used to add robustness to the models and to prove that consistent results could be obtained using different training and prediction set splits. The best linear discriminant analysis (LDA) models were able to correctly predict DHFR selectivity for approximately 70% of the external prediction set compounds. A set of new nitrogen and oxygen-specific descriptors were developed especially for this data set to better encode structural features, which are believed to directly influence DHFR inhibition and selectivity.

Animals↗
PubMed1992

Lipids and other risk factors selected by discriminant analysis in symptomatic patients with supra-aortic and peripheral atherosclerosis.

BACKGROUND: Different patterns of risk factors might be related to the involvement of specific vascular districts by atherosclerosis. In this sense, many investigations have addressed coronary artery disease, whereas extracoronary atherosclerosis has received less extensive attention. METHODS AND RESULTS: Vascular risk factors, with particular attention to lipid parameters (total cholesterol [TC]; triglycerides; high density lipoprotein cholesterol [HDL-C], HDL2 and HDL3 cholesterol [HDL2-C, HDL3-C]), were evaluated by means of univariate and multivariate (discriminant) analysis in a group of 169 patients (128 men and 41 women; mean ages, 58 +/- 7 and 62 +/- 7 years, respectively) with clinically and angiographically demonstrated atherosclerosis of the supra-aortic trunk and/or lower limbs. Patients with coronary artery disease were excluded from this study. The control group consisted of 140 age- and sex-matched individuals. By univariate analysis, smoking was more closely associated with peripheral atherosclerosis, whereas blood pressure was higher in patients with supra-aortic disease. Unrecognized diabetes mellitus was a frequent finding in patients with peripheral disease. The percentage of hyperlipidemias was fourfold higher in patients than in control subjects, with differences consisting of higher triglycerides and lower HDL-C, HDL2-C, and HDL3-C concentrations. By discriminant analysis, high correct classification (CC) rates were achieved in the various patient subgroups on the basis of variables selected from the statistical function. In male patients with peripheral disease, the variables HDL-C, smoking, diastolic blood pressure, uric acid, and glucose, in that order, yielded a CC in 90.4% of the cases; in female patients, smoking, TC/HDL-C, and body mass index gave a CC rate of 95.9%. In men with cerebral disease, the selected variables TC/HDL-C, diastolic blood pressure, and TC yielded a CC of 90.7%; in women, uric acid, TC/HDL-C, and fibrinogen levels produced a CC rate of 89.2%. CONCLUSIONS: Risk profiles in atherosclerosis of the supra-aortic trunks and lower limbs seem to differ in relation to gender and circulatory district involved. The importance of lipid parameters, in particular HDL-C, HDL2-C, and TC/HDL-C, as extracoronary risk factors is further confirmed.

Arteriosclerosis↗
PubMed1991

Classification of beef calves as protein-deficient or thermally stressed by discriminant analysis of blood constituents.

Linear discriminant functions hold promise for identifying either protein-deficient or cold-stressed calves based on blood constituents. For each of 2 yr 60 artificially bred Angus heifers were assigned randomly to a 2 x 2 factorial nutritional plan consisting of .32 or .96 kg/d of maternal CP and 8.7 or 12.2 Mcal/d of ME. The calves from these heifers were assigned randomly to environmental chambers set at either 0 or 21 degrees C in a repeated measures design. Linear discriminant functions were computed for 1 yr (training data) and then used to predict the classification of calves for the other year (validation data). Using the original data, the correct classifications of calves to the protein groups were 96, 80, 60, 59, 54, and 51% for blood samples obtained at 0, 12, 24, 36, 48, and 72 h of age, respectively. Using normalized data, corresponding correct classifications to protein groups were 94, 91, 80, 56, 54, and 52%. Results indicate that protein classification should use blood samples obtained within 12 h of age for reasonable success. For cold-stressed calves, correct classifications using original data were 47 (pre-exposure), 72, 54, 70, 67, and 66% for calves at 0, 12, 24, 36, 48, and 72 h of age, respectively. Corresponding correct classifications using normalized data were 54 (pre-exposure), 74, 70, 72, 69, and 77%. Cold stress could be detected after only 12 h of exposure; the time window for testing was much wider than for protein classification, but the classification generally was less discriminative.

Alkaline Phosphatase↗
PubMed2006

Dual-tracer dopamine transporter and perfusion SPECT in differential diagnosis of parkinsonism using template-based discriminant analysis.

UNLABELLED: Clinical differential diagnosis in parkinsonism can be difficult especially at early stages. We investigated whether combined perfusion and dopamine transporter (DAT) imaging can aid in the differential diagnosis of parkinsonian disorders: idiopathic Parkinson's disease (IPD), progressive supranuclear palsy (PSP), multiple system atrophy (MSA), dementia with Lewy bodies (LBD), and essential tremor (ET). METHODS: One hundred twenty-nine patients were studied, retrospectively (69 males; 24 MSA, 12 PSP, 8 LBD, 27 ET, and 58 IPD; mean disease duration, 3.5 +/- 3.7 y). Diagnosis was based on established clinical criteria after follow-up of 5.5 +/- 3.8 y in a university specialist movement disorders clinic. Group characterization was done using a categoric voxel-based design and, second, a predefined volume-of-interest approach along Brodmann areas (BA) and subcortical structures, including striatal asymmetry and anteroposterior indices. Stepwise forward discriminant analysis was performed with cross-validation (CV) using the leave-one-out technique. RESULTS: Characteristic patterns for perfusion and DAT were found for all pathologies. In the parkinson-plus group, MSA, PSP, and LBD could be discriminated in 100% (+CV) of the cases. When including IPD, discrimination accuracy was 82.4% (99% without CV). 2beta-Carbomethoxy-3beta-(4-iodophenyl)nortropane imaging as a single technique was able to discriminate between ET and neurodegenerative forms with an accuracy of 93.0% (+CV); inclusion of perfusion information augmented this slightly to 97.4% (+CV). CONCLUSION: Dual-tracer DAT and perfusion SPECT in combination with discrimination analysis allows an automated, accurate differentiation between the most common forms of parkinsonism in a clinically relevant setting.

Algorithms↗
PubMedDate not supplied

Using uncorrelated discriminant analysis for tissue classification with gene expression data.

The classification of tissue samples based on gene expression data is an important problem in medical diagnosis of diseases such as cancer. In gene expression data, the number of genes is usually very high (in the thousands) compared to the number of data samples (in the tens or low hundreds); that is, the data dimension is large compared to the number of data points (such data is said to be undersampled). To cope with performance and accuracy problems associated with high dimensionality, it is commonplace to apply a preprocessing step that transforms the data to a space of significantly lower dimension with limited loss of the information present in the original data. Linear Discriminant Analysis (LDA) is a well-known technique for dimension reduction and feature extraction, but it is not applicable for undersampled data due to singularity problems associated with the matrices in the underlying representation. This paper presents a dimension reduction and feature extraction scheme, called Uncorrelated Linear Discriminant Analysis (ULDA), for undersampled problems and illustrates its utility on gene expression data. ULDA employs the Generalized Singular Value Decomposition method to handle undersampled data and the features that it produces in the transformed space are uncorrelated, which makes it attractive for gene expression data. The properties of ULDA are established rigorously and extensive experimental results on gene expression data are presented to illustrate its effectiveness in classifying tissue samples. These results provide a comparative study of various state-of-the-art classification methods on well-known gene expression data sets.

Algorithms↗
PubMed1981

Primary hyperparathyroidism. Changing clinical spectrum, prevalence of hypertension, and discriminant analysis of laboratory tests.

The clinical spectrum of 100 consecutive cases of surgically proved primary hyperparathyroidism treated from 1974 through 1978 was analyzed. Their laboratory test results were compared with 64 cases of other form of hypercalcemia using multivariate discriminant analysis. The clinical spectrum has dramatically shifted during the past three decades from renal calculi and bone disease to the asymptomatic patient discovered by routine serum chemical analysis. Hypertension was twice as common among hyperparathyroid patients as in the general population but failed to improve in 92% after parathyroidectomy. The most useful discriminant laboratory test in descending order of value were the serum chloride, serum calcium, hematocrit, serum phosphorus, and parathormone. Multivariate discriminant analysis of the serum calcium, phosphorus, chloride, and Hct provided a 98% degree of accuracy in separating hyperparathyroidism from other forms of hypercalcemia.

Adult↗
PubMed1984

Discriminant analysis of volatile fatty acids produced in culture medium: a novel approach to the identification of Pseudomonas species.

The volatile fatty acids produced in culture medium by 357 Pseudomonas strains belonging to eight species were determined quantitatively by GLC. The resultant chromatograms were submitted to discriminant analysis. Stable discriminant functions were computed and included in a computerized identification system which also involved some distinctive volatile fatty acids regarded as two-state qualitative characters (presence or absence characters). Using a test group of 249 strains belonging to the studied species, more than 89% of the identifications made by this system agreed with those made by conventional biochemical methods despite the relatively poor differentiation between P. putida and P. fluorescens. When the individual species within the matrices were weighted with prior probabilities reflecting results given by two simple biochemical tests, 96% of the 249 strains were correctly identified.

Chromatography, Gas↗
PubMed1991

Fibrous dysplasia vs adamantinoma of the tibia: differentiation based on discriminant analysis of clinical and plain film findings.

Differentiation between benign fibrous dysplasia and malignant adamantinoma of the tibia is challenging because of the impact the diagnosis has on the choice of treatment (none or extensive surgery). The histologic and pathologic similarities of the lesions and the controversial relationship between fibrous dysplasia, osteofibrous dysplasia, and adamantinoma complicate the matter. We found a large overlap of histologic features in lesions considered either fibrous dysplasia or osteofibrous dysplasia on the basis of the radiologic findings. The purpose of this study was to determine the value of the plain radiograph of the lower leg in combination with clinical findings to differentiate the benign from the malignant condition. The clinical symptoms, radiographs, and histologic slides of 46 patients with fibrous dysplasia and 22 with adamantinoma in the tibia were reviewed retrospectively. In only one of 12 patients with radiologic or histologic characteristics of osteofibrous dysplasia were both radiologic and histologic criteria for the diagnosis present. A linear discriminant analysis was performed on six clinical (age, spontaneous pain, pain after trauma, swelling only, pain and swelling, and bowing deformity) and 25 radiologic signs. Fibrous dysplasia and its variant osteofibrous dysplasia could be identified correctly in 87% (40 of 46 patients) and adamantinoma in 95% (21 of 22 patients) by using the patient's age and four radiologic signs. When results from the discriminant analysis of a randomized subgroup of patients (32) were used on the other subgroup (36 patients), fibrous dysplasia was correctly identified in 84% (21 of 25) and adamantinoma in 82% (nine of 11). Fibrous dysplasia is more prevalent than adamantinoma in a young patient, when radiographs show a ground-glass appearance and anterior bowing and when there is no multilayered periosteal reaction and moth-eaten destruction. When radiologic signs and the patient's age are combined, fibrous dysplasia and adamantinoma can be discriminated in a high percentage of patients.

Adolescent↗