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 757 records · Page 42Linked to original sources
PubMed1988

[Multivariate discriminant analysis of EEG power spectra in acute schizophrenic patients and healthy probands].

The electroencephalogram (EEG) during information processing is influenced by specific changes in brain electrical activity. Based on the theory of a disturbed information processing in schizophrenics we analysed auditory stimulus induced EEG changes by Fast Fourier Transformation. The most important of the significant stimulation dependent EEG power changes were measured in the 0.5-3.5 Hz and 10-13.5 Hz frequency bands in the left parietal lead. In a multivariate analysis the separation of the subjects examined into acute schizophrenics and normals was incorrect in only 21% of the cases (resubstitution rate): using the pi-method an error of 31% was estimated.

Acute Disease↗
PubMed1988

[Results of coronary artery bypass surgery. Surgical mortality, survival and follow-up functional state. Multivariate discriminant analysis of relative risk factors].

Clinical results of coronary artery bypass surgery, have been evaluated by analyzing operative mortality, late survival, late functional results and their related risk factors. Four hundred and thirty-seven consecutive patients who underwent coronary artery bypass surgery between January, 1979, and December, 1983, were the clinical material of this study. The gender of patients was male in 89% of the cases; age ranged from 34 to 78 years with a mean of 54.8 +/- 8.2 (SD). Patients with combined surgical procedures were excluded. Follow-up averaged 55.7 months; 404 survivors have been contacted (97% of the total study population, if operative and late deaths are added); 218 patients underwent a control exercise test; postoperative employment status was determined for 242 individuals. The operative mortality was 5.49% (24 patients). Death was due to cardiac causes in 75% of the cases. The overall actuarial survival rate was 85 +/- 1.9% after 5 years and 83.4 +/- 2.2% after 7 years. When non-cardiac related deaths were excluded the actuarial rates were 88 +/- 1.7% and 87.2 +/- 1.9% after 5 and 7 years respectively. Actuarial freedom from all ischemic events (cardiac related death, late myocardial infarction or recurrence of angina) was 66.1 after five years, and was 70.2% if operative deaths were excluded. Actuarial freedom from recurrence of angina for patients alive at follow-up was 78.7% after five years, actuarial freedom from myocardial infarction was 93.5%. The mortality hazard showed a diphasic response, being higher perioperatively and constant in the following 5 years of follow-up. All ischemic events, both singularly and together, showed an accelerated rate of occurrence at the first and after the fifth postoperative year, the slope of the curve being quite flat during the 1 to 5 year interval. The control exercise test was negative for 62.8% of the patients, positive for 33.5% and undeterminable in 3.7%. Employment status was postoperatively unaffected in 49.6% of the cases, while 27.3% of the patients retired: the remaining individuals had already retired before surgery. Statistical analysis (stepwise logistic regression) identified age (p = 0.002) and cross-clamp time (p = 0.016) as significant risk factors of operative mortality. The ejection fraction showed a value close to statistical significance (p = 0.06).(ABSTRACT TRUNCATED AT 400 WORDS)

Actuarial Analysis↗
PubMed1986

[The role of CEA, CA 19-9 and TPA in the diagnosis of pancreatic cancer. Evaluation by discriminant analysis].

In order to assess the relative value of CA 19-9, Tissue Polypeptide Antigen (TPA) and Carcinoembryonic Antigen (CEA), evaluated alone and in combination, in diagnosing pancreatic malignancy, serum CA 19-9, TPA and CEA were determined in 25 control subjects (CS), 26 pancreatic cancer (PC), 23 chronic pancreatitis (CP) and 21 benign extra-pancreatic diseases (EPD). The three markers were able to allocate the subjects correctly in 56.8% of the cases (CS 100%, PC 73.1%, CP 17.4%, EPD 28.6%). Sensitivity, specificity and diagnostic accuracy in detecting pancreatic cancer were respectively: 77%, 91% and 68% for CA 19-9; 92%, 75% and 67% for TPA; 50%, 84% and 34% for CEA; 73%, 91% and 64% for the three parameters evaluated simultaneously. CA 19-9 and TPA appear to be useful indices of pancreatic cancer with a satisfactory specificity when related to chronic pancreatitis; their diagnostic value seems to be comparable and better than that of CEA; the combination of these markers does not improve the results obtained by CA 19-9 or TPA alone.

Adult↗