Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “classifier”

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 541 records · Page 30Linked to original sources
PubMedDate not supplied

[Response to treatment and survival of adults with acute leukemia classified with immunophenotyping and cytochemistry].

We studied the response to treatment and survival of 30 adults with acute lymphocytic leukemia (ALL) and 19 with acute non lymphoid leukemia (ANLL) classified on basis of immunophenotype (monoclonal antibodies) and cytochemistry. For the ALL cases 70% corresponded to common ALL (CALLA positive), 23% to B lymphocytes and 7% to T cells. We had 68% of the ANLL patients classified as myeloid, 21% as hybrid (positive both myeloid and lymphoid markers) and 11% as undifferentiated. We analyzed demographic data (gender and age), basic laboratory values (hemoglobin, leucocytes, platelets and cytomorphology in peripheral blood and bone marrow) using the French-American-British classification, and found no statistically significant differences between ALL and ANLL. Three of four patients (75%) with hybrid ANLL achieved complete remission (CR), while 46% of cases with myeloid ANLL and none of the subjects with undifferentiated ANLL reached CR; these differences were not statistically significant. Patients with common ALL had a median survival (SV) of 499 days, for B cell ALL it was of 212 days, and for T cell ALL of 285 days. Our data suggest that: a) expression of lymphoid markers in patients with ANLL is probably associated with a higher CR ratio, and b) SV in adults with common ALL seems to be longer than in those with B and T cell ALL.

Adolescent↗
PubMed1994

The restricted surgical relevance of morphologic criteria to classify systemic-pulmonary collateral arteries in pulmonary atresia with ventricular septal defect.

Now that systemic-pulmonary collateral arteries are used for unifocalization in patients with pulmonary atresia and ventricular septal defect, the question arises whether morphologic criteria of these collateral arteries could help to provide better results. In an attempt to classify the morphologic features of systemic-pulmonary collateral arteries, we studied 31 heart-lung autopsy specimens with pulmonary atresia and ventricular septal defect. The course of the systemic-pulmonary collateral arteries (origin, branching pattern, and connections with systemic and central pulmonary arteries) was related to their histologic characteristics. The results show that systemic-pulmonary collateral arteries cannot be classified according to their course related to the trachea and the main branches of the bronchi. The histologic features of these collateral arteries vary along their course to the lungs. Nearly all systemic-pulmonary collateral arteries contain a muscular or a musculoelastic segment. One type of collateral artery (complex loop anastomoses) is completely muscular and resembles a bronchial artery. Nutritive branches (bronchial arteries) arise from all histologic types of systemic-pulmonary collateral artery segments. The size and number of intimal proliferations in muscular, elastic, and musculoelastic segments did not differ significantly. In 29 of 31 cases a ductus arteriosus did not coexist with large collateral arteries (two cases unknown). It is concluded that a classification of large systemic-pulmonary collateral arteries based on morphologic features results in a highly variable system, which does not facilitate decisions for the suitability of these arteries for unifocalization procedures. The variability of the systemic-pulmonary collateral arteries corresponds with the recent embryologic finding that during development, collateral artery formation is possible during extended periods.

Adolescent↗
PubMed1994

Classifying and identifying servers for biomedical information retrieval.

Useful retrieval of biomedical information from network information sources requires methods for organized access to those information sources. This access must be organized in terms of the information content of information sources and in terms of the discovery of the network location of those information sources. We have developed an approach to providing organized access to information sources based on a scheme of hierarchical classifiers and identifiers of the servers providing access to those information sources. This approach uses MeSH tree numbers as both classifiers and identifiers of servers. MeSH tree numbers are used to indicate the information content of servers, and also as OSF/DCE server identifiers. This allows the identity and location of a server providing access to a given information source to be determined from the information classification of that information source.

Computer Communication Networks↗
PubMed1994

Practices of county medical examiners in classifying deaths as on the job.

Although annual United States occupational injury fatalities range between 7,000 and 10,400, consistent rules to determine which deaths are "occupational" do not exist. Fifty-seven North Carolina county medical examiners (MEs), responsible for more than 50% of all medical examiner cases in 1990, received our questionnaire. Fifty-three (93%) responded, classifying 22 scenarios as on-the-job deaths and indicating usual classification practices and information sources. Agreement varied among the scenarios, but those involving transportation and nonpaid workers elicited particularly inconsistent responses. Fifty-six percent of medical examiners have a general rule for determining on-the-job status, but deaths associated with motor vehicles, farming, and occupations other than the decedent's usual job were classified most inconsistently. The lack of standard definitions of "job," "work," and "on-the-job" is apparent in classification decisions. Certain work situations need special consideration.

Accidents, Occupational↗
PubMed1994

A neural network classifier for cerebral perfusion imaging.

UNLABELLED: Artificial neural networks have been applied to a variety of pattern recognition tasks in medical imaging and have been shown to be a powerful classification tool. The potential usefulness to discriminate normal from abnormal cerebral perfusion patterns was investigated. METHODS: Cerebral perfusion imaging with 99mTc-labeled hexamethylpropyleneimine oxime was performed on 52 normal control subjects, 29 patients with clinically diagnosed Alzheimer's disease (AD) and 25 patients with chronic cocaine polydrug abuse. Each study was registered and scaled to a common anatomic coordinate system, yielding 120 standardized cortical regions. A back-propagation neural network classifier based on regional perfusion was used to classify normal and abnormal perfusion patterns. The neural network was trained to discriminate patients with AD from age-matched normal controls and cocaine polydrug abuse patients from normal controls. The performance of the neural network in these two tasks was evaluated quantitatively by receiver operating characteristic (ROC) analysis using cross-validation. RESULTS: For patients with AD, the area under the ROC curve was 0.93 +/- 0.04. When testing with the cocaine polydrug abuser data, the area under the ROC curve was 0.89 +/- 0.04. CONCLUSION: Neural networks provide a potentially useful tool in the decision-making task to discriminate patients with AD and cocaine abuse from normal controls.

Adult↗
PubMed1997

Grossly punched-out lesions in the aorto-iliac region can be histologically classified as false, pseudo-false, or disguised aneurysm.

Aneurysms are morphologically classified as true or false based on the nature of their walls. True aneurysms are composed of all or parts of layers of the vessel. False aneurysms are the result of rupture and their walls have only fibrous tissues. The orifice of false aneurysms is narrow relative to the aneurysmal diameter and thus they are grossly or angiographically referred to as punched-out lesions. Hence false aneurysms present with punched-out lesions, but in reverse, are all of punched-out lesions false aneurysms? We experienced some cases of punched-out lesions which histologically contained traces of elastin, and the purpose of this report was to histologically investigate grossly punched-out lesions. We examined 671 elderly autopsy cases, and a total of 21 grossly punched-out lesions in the aorto-iliac region were selected. They were histologically classified as false, "pseudo-false", or "disguised" aneurysm. False aneurysms were found in 3 patients (0.45%), and were histologically mycotic. A total of 5 "pseudo-false" aneurysms were found in 3 patients (0.45%). They histologically contained traces of elastin, and thus they were categorised in true aneurysms. A total of 13 "disguised" aneurysms were found in 6 patients (0.89%). They were true fusiform aneurysms with an eccentric thrombus, on which a fibrin-cap formed a narrow orifice. Partial sections are insufficient for diagnosis; cross-sections are necessary. To the best of our knowledge, there have been no reports of "pseudo-false" or "disguised" aneurysms in the aorto-iliac region.

Aged↗
PubMedDate not supplied

Discrimination between migraine patients and normal subjects based on steady state visual evoked potentials: discriminant analysis and artificial neural network classifiers.

Fifty-one migraine patients and 19 control subjects were examined by steady state visual evoked potentials (SSVEPs) procedure. The aim of this study was to develop a discriminant analysis and an artificial neural network (NN) classifier in order to discriminate between migraneurs during attack-free periods and normal subjects. Discriminant analysis correctly classified 72.5% of migraine patients with a false positive rate of 36.8%. The NN method had a sensitivity of 100% with a false positive rate of 15%. The results of this study confirm SSVEP pattern as a marker of migraine and demonstrate that NNs could be a useful method in the statistical analysis of topographic EEG data.

Adult↗
PubMed1998

Application of fuzzy-classifier system to coronary artery disease and breast cancer.

This paper presents an application of a genetic-algorithm-based representation of fuzzy rules for the classification of coronary artery disease data and breast cancer data. The performance of this fuzzy classifier for classification of coronary artery disease and breast cancer data is evaluated. In this study the concept of fuzzy if-then has been applied of rules proposed by Ishibuchi et al. for a multi dimensional data classification problem which leads to higher classification power. The fitness value of each fuzzy if-then rule was determined by the numbers of correctly and wrongly classified training patterns for that rule. The classification power on real world data for coronary artery disease and breast cancer was thus demonstrated by computer simulations.

Algorithms↗
PubMed2003

Chronic acquired demyelinating symmetric polyneuropathy classified by pattern of weakness.

OBJECTIVES: To study a representative group of patients with chronic acquired symmetric demyelinating polyneuropathies, and to evaluate classification by pattern of weakness and by presence of immunoglobulin monoclonal protein (M protein). METHODS: In Vest-Agder County, Norway, an unselected population of patients with chronic symmetric polyneuropathies who fulfill electrodiagnostic criteria for demyelination are registered in a database and followed up prospectively. Data were taken from the database on April 2, 2001. Patients with proximal as well as distal weakness were classified as having chronic inflammatory demyelinating polyradiculoneuropathy (CIDP), and patients with only distal symptoms as having distal acquired demyelinating symmetric polyneuropathy (DADS). RESULTS: A total of 29 patients had chronic acquired symmetric demyelinating polyneuropathy; 15 had CIDP and 14 had DADS. The 2 categories differed regarding spinal protein level (mean +/- SD, 0.102 +/- 0.060 g/dL in CIDP vs 0.065 +/- 0.029 g/dL in DADS; P =.05); clinical course (remitting in 6 of 13 patients with CIDP vs 0 of 14 with DADS; P =.02); disability score at diagnosis (mean +/- SD, 3.3 +/- 1.0 in CIDP vs 1.9 +/- 0.6 in DADS; P<.001) and at peak of symptoms (mean +/- SD, 3.6 +/- 1.1 in CIDP vs 2.3 +/- 0.6 in DADS; P<.001); and response to immunosuppressive treatment (11 of 12 patients with CIDP vs 2 of 7 with DADS; P =.01). An M protein was detected in 8 patients (3 with CIDP and 5 with DADS). Patients with polyneuropathy with and without M protein were similar in clinical features, course, disability, and treatment response. CONCLUSION: Classification by presence or absence of proximal weakness separates patients with chronic acquired symmetric demyelinating polyneuropathy into groups that are different in clinical course, disability, and treatment response.

Adult↗
PubMed1997

Adequacy of interviews vs checklists for classifying childhood psychiatric disorder based on parent reports.

BACKGROUND: The advantages and disadvantages of lay-administered structured interviews and self-administered problem checklists for estimating prevalence and associated features of childhood psychiatric disorder have attracted little comment. This article compares the scientific adequacy of these 2 instruments for classifying DSM-III-R categories of childhood psychiatric disorder in general population samples. METHODS: Study data are from parental assessments of 251 children aged 6 to 16 years participating in a 2-stage measurement evaluation study. Reliability and validity were compared between the Diagnostic Interview for Children and Adolescents (the structured interview in the study) and the revised Ontario Child Health Study scales (the self-administered problem checklist used in the study). RESULTS: Reliability estimates based on the kappa statistic were comparable for the 2 instruments and ranged from 0.21 (conduct disorder) to 0.70 (depression) on the lay interview and from 0.17 (depression) to 0.61 (oppositional defiant disorder) on the self-administered checklist. Validity coefficients tended to favor the checklist categories, but only marginally. CONCLUSIONS: On balance, differences in reliability and validity were small between the 2 instruments. These differences would appear to have no discernible impact on the knowledge about prevalence and associated features of disorder generated by use of such instruments in general population surveys.

Adolescent↗
PubMed1997

Classifying sex biased congenital anomalies.

The reasons for sex biases in congenital anomalies that arise before structural or hormonal dimorphisms are established has long been unclear. A review of such disorders shows that patterning and tissue anomalies are female biased, and structural findings are more common in males. This suggests different gender dependent susceptibilities to developmental disturbances, with female vulnerabilities focused on early blastogenesis/determination, while males are more likely to involve later organogenesis/morphogenesis. A dual origin for some anomalies explains paradoxical reductions of sex biases with greater severity (i.e., multiple rather than single malformations), presumably as more severe events increase the involvement of an otherwise minor process with opposite biases to those of the primary mechanism. The cause for these sex differences is unknown, but early dimorphisms, such as differences in growth or presence of H-Y antigen, may be responsible. This model provides a useful rationale for understanding and classifying sex-biased congenital anomalies.

Congenital Abnormalities↗
PubMed1999

Improvement of the Mair scoring system using structural equations modeling for classifying the diagnostic adequacy of cytology material from thyroid lesions.

The scoring system developed by Mair et al. (Acta Cytol 1989;33:809-813) is frequently used to grade the quality of cytology smears. Using a one-factor analytic structural equations model, we demonstrate that the errors in measurement of the parameters used in the Mair scoring system are highly and significantly correlated. We recommend the use of either a multiplicative scoring system, using linear scores, or an additive scoring system, using exponential scores, to correct for the correlated errors. We suggest that the 0, 1, and 2 points used in the Mair scoring system be replaced by 1, 2, and 4, respectively. Using data on fine-needle biopsies of 200 thyroid lesions by both fine-needle aspiration (FNA) and fine-needle capillary sampling (FNC), we demonstrate that our modification of the Mair scoring system is more sensitive and more consistent with the structural equations model. Therefore, we recommend that the modified Mair scoring system be used for classifying the diagnostic adequacy of cytology smears. Diagn. Cytopathol. 1999;21:387-393.

Biopsy↗
PubMed1998

The validity of the DSM-IV scheme for classifying bulimic eating disorders.

OBJECTIVE: This study was designed to assess the validity of the DSM-IV scheme for classifying recurrent binge eating. METHOD: A general population sample of 250 young women with recurrent binge eating was recruited using a two-stage design. Information on their eating habits and associated psychopathology was obtained by personal interviews. Subjects were reassessed 1 year later. RESULTS: The diagnosis of bulimia nervosa had good descriptive and predictive validity. On present state features it was not possible to distinguish binge-eating disorder from the nonpurging subtype of bulimia nervosa. However, these groups differed in their outcome at 1 year. Within eating disorder not otherwise specified (EDNOS), there was a subgroup of subjects with milder symptoms which were relatively unstable over time. DISCUSSION: The findings suggest that bulimic eating disorders exist on a continuum of clinical severity, from bulimia nervosa purging type (most severe), through bulimia nervosa nonpurging type (intermediate severity), to binge-eating disorder (least severe). The data on outcome support retaining a distinction between nonpurging bulimia nervosa and binge-eating disorder.

Adolescent↗
PubMed1980

Adjective Check List self and ideal self correlates of MMPI profiles classified according to the Meehl-Dahlstrom rules.

Classified MMPI profiles of 61 male and 119 female Italian psychiatric outpatients as neurotic (N = 100), psychotic (N = 45), or indeterminate (N = 35) by means of the Meehl-Dahlstrom rules. These classifications were uninfluenced by age or sex. Real and ideal self descriptions on the Adjective Check List (ACL) also were obtained. Thirteen of the 24 ACL scales scored on the real protocols differentiated significantly (P < .05) among the three subgroups. Adequacy of personal adjustment, as inferred from these differences, was poorest for patients with "psychotic" MMPI profiles, next poorest for the indeterminates, and best for those in the neurotic category. Personal Adjustment was the lowest ACL scale for all three subgroups. Only one ideal self scale differentiated significantly among the three subgroups. Also, the ideal self profile for the total sample of 180 patients was almost perfectly correlated with that for a sample of 229 nonpatients. Descriptions of the real self appear to be related systematically and meaningfully to psychiatric status as indicated by the Meehl-Dahlstrom, rules, whereas descriptions of the ideal self are not associated with diagnosis.

Adult↗
PubMed2001

Classification of wheat varieties: use of two-dimensional gel electrophoresis for varieties that can not be classified by matrix assisted laser desorpiton/ionization-time of flight-mass spectrometry and an artificial neural network.

Analyzing a gliadin extract by matrix assisted laser desorption/ionization-time of flight-mass spectrometry (MALDI-TOF-MS) combined with an artificial neural network (ANN) is a suitable method for identification of wheat varieties. However, the ANN can not distinguish between all different wheat varieties. Two-dimensional polyacrylamide gel electrophoresis (2-D PAGE) was applied to three pairs of wheat varieties, which can not be classified correctly by ANN. By 2-D PAGE the varieties in the three pairs can be discriminated and these six wheat varieties can be separated from each other, which could not be separated by MALDI-TOF-MS and NN.

Electrophoresis, Gel, Two-Dimensional↗
PubMed2001

Cell fingerprinting: an approach to classifying cells according to mass profiles of digests of protein extracts.

We present a statistical framework for classifying cells according to the set of peptide masses obtained by mass spectrometric analysis of digestions of whole cell protein extracts. The digest is separated by high performance liquid chromatography (HPLC) coupled directly to a mass spectrometer either by an electrospray interface or by collection to a matrix-assisted laser desorption/ionization target plate. Here, the mass to charge ratio, intensity, and HPLC retention time of the peptides are measured. We have used defined bacterial strains to test this approach. For each bacterium, this process is repeated for extracts obtained at different points in the growth curve in order to try and define an invariant set of signals that uniquely identify the bacterium. This paper presents algorithms for the creation of this cell fingerprint database and develops a Bayesian classification scheme for deciding whether or not an unknown bacterium has a match in the database. Our initial testing based on a limited data set of three bacteria indicates that our approach is feasible. Via a jack-knife test, our Bayesian classification scheme correctly identified the bacterium in 67.8% of the cases.

Bacteria↗
PubMed1990

Bone marrow in vitro growth and cytogenetic studies in patients with FAB-classified primary myelodysplastic syndromes.

Thirty-eight consecutive patients with a FAB-classified primary myelodysplastic syndrome (MDS) were investigated for in vitro growth of colony-forming units for granulocyte-macrophage precursors (CFU-GM) and cytogenetic analysis of bone marrow cells. Abnormal CFU-GM growth was found in 30 patients (79%), and clonal chromosome abnormalities were found in 13 patients (34%). The eight patients who showed normal CFU-GM growth were either cytogenetically normal (n = 5), or had a 5q-deletion (n = 3) as single or dominating karyotypic abnormality. Among the 30 patients with reduced or no colony growth, ten patients had a clonal chromosome abnormality. Leukemia developed in eight patients. None of them grew any CFU-GM colonies, and three of them were cytogenetically abnormal at the time of diagnosis of MDS. Analysis of the bone marrow in vitro growth for CFU-GM and the karyotype in patients with MDS emphasizes the close relationship between these disorders and manifest acute leukemia. Subgroups of MDS may be defined by a cytogenetic classification (e.g., the 5q-syndrome), and the CFU-GM growth pattern can be of value for predicting leukemic transformation.

Aged↗
PubMedDate not supplied

Heterogeneity of the genome ancestry of individuals classified as White in the state of Rio Grande do Sul, Brazil.

One hundred nineteen individuals classified as White, living in different localities of the Brazilian state of Rio Grande do Sul, were studied in relation to the HVS-I region of the mitochondrial DNA (mtDNA). The male fraction of the sample (N = 74) was also tested for seven Y-chromosome polymorphisms. In a specific population (Veranópolis), a city characterized by a large influence of the Italian immigration of the 19th century, the results from the maternal and paternal sides indicated almost complete European ancestry. However, another sample identified as White, from different localities of Rio Grande do Sul, presented significant fractions of Native American (36%) and African (16%) mtDNA haplogroups. These results indicate that Brazilian populations are remarkably heterogeneous; while some present an overwhelming majority of transplanted European genomes, with a complete correspondence between physical appearance and ancestry, others reflect a history of extensive admixture with dissociation between physical appearance and ancestry.

Brazil↗