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 1,045 records · Page 58Linked to original sources
PubMed1998

Medical audit after 26,711 breast imaging studies: improved rate of detection of small breast carcinomas (classified as Tis or T1a,b)

BACKGROUND: Patients with breast carcinomas 1 cm or smaller are known to have the highest long term survival rates. This pathologic measurement was used as a quality indicator of the performance of different reporting radiologists. METHODS: A total of 26,711 diagnostic breast imaging studies were performed over a period of 5 years. Within the first 18 months, several general radiologists reported mammograms, whereas during the rest of the evaluated period only 1 radiologist was responsible for all mammography reporting. Radiologic findings leading to open biopsy were correlated with the histologic results. The cancer sizes from the histologic reports were recorded, and the number of Tis, T1a, and T1b cancers were correlated with all cancers found at open biopsy. The change in the relative percentage of small cancers was evaluated, along with the positive predictive value (PPV). RESULTS: Between 1991 and 1995, 26,711 breast imaging studies were performed. This resulted in 2698 open biopsies with 620 breast carcinomas detected. In the first 2 years, the percentages of small cancers detected (Tis or T1a,b) were 27.2% and 25.7%, respectively. In the third, fourth, and fifth years, the percentages increased to 38.8%, 34.5%, and 38.8%, respectively. At the same time, the PPV increased slightly, from 0.23 in 1991 to 0.27 in 1995. CONCLUSIONS: The detection of curable early stage breast carcinomas requires the dedication and commitment of a small group of radiologists who are willing to spend most of their time on this single subject. This will increase considerably the number of early stage cancers found and reduce the number of false-positive diagnoses. [See editorial counterpoint on pages 2430-1 and reply to counterpoint on pages 2432-3.]

Breast Neoplasms↗
PubMed1998

Antisocial personality disorder and psychopathy: diagnostic dilemmas in classifying patterns of antisocial behavior in sentencing evaluations.

Antisocial Personality Disorder (APD) and PCL-R psychopathy are critically examined regarding their application to sentencing determinations. PCL-R psychopathy is emerging in the literature as a more useful forensic diagnostic construct than APD, which appears flawed by multiple weaknesses. These include shifting diagnostic criteria, innumeracy problems, absence of symptom weighting, temporal instability, and the equivalence of some symptoms with substance abuse disorders. Additionally, APD overdiagnosis may result from inattention to issues of social context, trauma history, and symptom pervasiveness. Neither objective nor projective personality testing reliably differentiates APD. Finally, an APD diagnosis does not always indicate criminal, much less incorrigible criminal behavior. By contrast, PCL-R psychopathy results are strongly predictive of criminal behavior and violent recidivism for Caucasian males through mid-life residing in the community. Emerging research with the PCL-R regarding other important populations and contexts is promising but generalization is currently limited.

Antisocial Personality Disorder↗
PubMedDate not supplied

Near-optimal region selection for feature space reduction: novel preprocessing methods for classifying MR spectra.

We introduce a global feature extraction method specifically designed to preprocess magnetic resonance spectra of biomedical origin. Such preprocessing is essential for the accurate and reliable classification of diseases or disease stages manifest in the spectra. The new method is genetic algorithm-guided. It is compared with our enhanced version of the standard forward selection algorithm. Both seek and select optimal spectral subregions. These subregions necessarily retain spectral information, thus aiding the eventual identification of the biochemistry of disease presence and progression. The power of the methods is demonstrated on two biomedical examples: the discrimination between meningioma and astrocytoma in brain tissue biopsies, and the classification of colorectal biopsies into normal and tumour classes. Both preprocessing methods lead to classification accuracies over 97% for the two examples.

Algorithms↗
PubMed1999

Classifying symbiotic proteins from Bradyrhizobium japonicum into functional groups by proteome analysis of altered gene expression levels.

The advent of whole genome sequences has brought with it a vast number of new potential proteins whose function is unknown. We describe an approach to sorting proteins into functional groups by comparative two-dimensional (2-D) gel mapping of cells grown under different physiological conditions. Computerized image analysis selects the proteins whose expression levels change significantly for subsequent identification by mass spectrometry. The protein groupings are further subdivided by directed alteration of their expression levels (e.g., by gene inactivation) and following the changes in the expression pattern of the mutants. We have applied this approach to study the regulation of micro- and anaerobically induced genes including the genes involved in nitrogen fixation in the symbiotic bacterium Bradyrhizobium japonicum. The results obtained show that in addition to the two known regulons controlled by the transcription factors NifA and FixK2, a third set of proteins may exist in B. japonicum which are induced by anaerobic conditions and are regulated independently. The approach can be applied generally and can be used to build up functional relationship maps of genomes. Protein identification by mass spectrometry was shown to be vital to the interpretation of the expression analysis since 15% of the 2-D gel spots contained more than one protein.

Amino Acid Sequence↗
PubMed1979

Atypical lipoma, atypical intramuscular lipoma, and well differentiated retroperitoneal liposarcoma: a reappraisal of 30 cases formerly classified as well differentiated liposarcoma.

Thirty cases formerly diagnosed as Grade I lipogenic liposarcoma (well differentiated liposarcoma) were reviewed. The basic histologic pattern in all was that of adult fat modified by the presence of cells with enlarged, hyperchromatic nuclei; in most cases there was also a component of myxoid and/or fibrous tissue. The length of follow-up ranged from two to 30 years. Nine of the tumors were located in the subcutaneous layer. None of these recurred after excision, not even those which were simply "shelled out," and none metastasized. The term "atypical lipoma" is proposed for this group. Thirteen were located within or between muscles of the limbs, limb girdles, and head and neck. Nine of these recurred at least one, but there were no metastases and no deaths due to tumor. These were designated "atypical intramuscular lipoma." The remaining eight originated in the retroperitoneum. Although none of these patients developed metastases, five suffered inoperable recurrence and three died as a result of the neoplasm. It is suggested that the term "well differentiated retroperitoneal liposarcoma" be retained for cases of this type.

Adult↗
PubMed1980

Estrogen receptor concentrations in 269 cases of histologically classified human breast cancer.

This report provides a histological review of 264 female primary breast cancers analyzed for estrogen receptor protein (ER). We also describe 5 cases of male breast cancer all of which bound estradiol specifically. Generally the ER concentrations were lower in the tumors of premenopausal women than in those of postmenopausal women. Three types of cancer with specific morphological features were shown to have ER concentrations that differed significantly from the other types of cancer. Medullary and comedo carcinoma had very low and papillary carcinoma very high levels of ER. No obvious trend in ER concentration was found in unspecified ductal carcinoma, colloid carcinoma, nor in lobular carcinoma. A positive correlation was demonstrated between ER content and degree of differentiation in ductal carcinoma. Cancer with lymphoid infiltration generally showed low ER levels.

Aged↗
PubMed1981

An approach to classifying prognostic factors related to survival experience for non-Hodgkin's lymphoma patients: based on a series of 982 patients: 1967-1975.

The survival experience of 982 non-Hodgkin's lymphoma patients registered at Princess Margaret Hospital, Toronto, between 1967 and 1975, was studied. Prognostic groups were obtained by means of a classification procedure based on standard statistical techniques; the variables utilized in the classification were ones of known reproducibility which could be measured with little inconvenience to the patient. The results show that these prognostic groups give better information than the Ann Arbor staging classification in the sense that the survival curves are clearly separated and "good prognosis" and "poor prognosis" groups are clearly identified. Implications for therapy planning are briefly discussed.

Aged↗
PubMed1987

Incidence of lymphoma in the US classified by the working formulation.

The incidence of lymphoma in the US and Puerto Rico among 13,600 patients is reported by histologic subtypes according to the Working Formulation. The most frequent histologic types were intermediate grades (five per 100,000 among whites, three per 100,000 among blacks, and two per 100,000 among Puerto Rican Hispanics). Low-grade types were next in frequency (2.7 per 100,000 in whites, 1.5 per 100,000 in blacks, and one per 100,000 in Puerto Rican Hispanics). More than 95% of patients had low-or intermediate-grade lymphomas, and of these, intermediate-grade lymphomas occurred in 65% of patients. High-grade types were infrequent in all ethnic groups. The incidence among men was significantly greater than among women in all ethnic groups. There was a distinct peak for small noncleaved cell type among white boys but not girls. For all other histologic types, there was a conspicuous absence of a young adult component, with incidence increasing steadily with age. No evidence for seasonal fluctuation in month of diagnosis was found. This delineation of incidence by histologic groupings, sex, and ethnic group will facilitate future studies using the Working Formulation.

Adolescent↗
PubMed1992

A flow cytometric study of non-small cell lung cancer classified as T1N0.

Surgical resection currently offers the best chance for cure of non-small cell lung cancer but its efficacy is limited by subsequent tumor recurrence. Even the most favorable cancers (T1N0 tumors) recur 20% to 30% of the time within 5 years and there is currently no way to anticipate precisely which tumors will recur. To test whether DNA flow cytometric study might be useful in this regard, the authors performed a retrospective case-control study of 102 tumors (51 recurrent cases and 51 controls) from a prospective registry of patients with completely resected, meticulously staged T1N0 non-small cell carcinomas. Unbiased relative hazard ratios of recurrence were estimated for ploidy and proliferative rate, as well as for tumor histologic type and clinical variables. Ploidy abnormalities were slightly more common among cases (67%) than controls (57%) but this difference was not statistically significant. Estimation of proliferative rates was possible for 85 tumors but there was no significant difference between cases and controls and proliferative rates were not prognostic of recurrence. In multivariate analyses, the observed predictive value for each of the flow cytometric parameters was modest at best and smaller than that seen for tumor histologic type. These results suggest that flow cytometric analysis has limited value in guiding management of patients with early stage non-small cell carcinoma.

Carcinoma, Non-Small-Cell Lung↗
PubMed2001

Frequency distributions of breast cancer characteristics classified by estrogen receptor and progesterone receptor status for eight racial/ethnic groups.

BACKGROUND: The National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) cancer registries have been collecting data regarding estrogen receptor (ER) and progesterone receptor (PR) status in breast cancer since 1990. The current study reports on some of these data for eight racial/ethnic groups. METHODS: Stratified by ER and PR status, the frequency distributions of 112,588 breast cancer cases diagnosed between 1992--1997 in 11 SEER cancer registries were examined by age at diagnosis, stage at diagnosis, histologic grade, and tumor type for white, black, Hispanic, Japanese, Chinese, Filipino, Native Hawaiian, and American Indian and Alaska Native (AI/AN) females. RESULTS: For each racial/ethnic group, the percentage of ER positive (+)/PR+ was > ER-PR- > ER+PR- > ER-PR+ tumors. For the two major ER/PR groups, the ER+PR+ tumors were different from the ER-PR- tumors in several ways. For white females, there were differences in the age distributions, stage at diagnosis, and histologic grade. For black females, the differences involved the age distributions and tumor grades. For Hispanic and Japanese females, there were differences with regard to the age distributions and tumor grades. For Filipino, Chinese, and AI/AN females, the tumor stages and grades differed. For Native Hawaiians, the histologic tumor grades were different. CONCLUSIONS: For each racial/ethnic group, the ER/PR status appeared to divide breast cancer patients into two or more subgroups with unique tumor characteristics. In general, ER status appeared to have the greatest impact on delineating these subgroups, whereas in some cases, PR status was able to modify the subgroups further. It is hoped that reporting these tumor characteristics by ER/PR status for each racial/ethnic group will spur more investigation into the significance of ER/PR status in each racial/ethnic group.

Adolescent↗
PubMed2001

Comparison of five clustering algorithms to classify phytoplankton from flow cytometry data.

BACKGROUND: Artificial neural networks (ANNs) have been shown to be valuable in the analysis of analytical flow cytometric (AFC) data in aquatic ecology. Automated extraction of clusters is an important first stage in deriving ANN training data from field samples, but AFC data pose a number of challenges for many types of clustering algorithm. The fuzzy k-means algorithm recently has been extended to address nonspherical clusters with the use of scatter matrices. Four variants were proposed, each optimizing a different measure of clustering "goodness." METHODS: With AFC data obtained from marine phytoplankton species in culture, the four fuzzy k-means algorithm variants were compared with each other and with another multivariate clustering algorithm based on critical distances currently used in flow cytometry. RESULTS: One of the algorithm variants (adaptive distances, also known as the Gustafson--Kessel algorithm) was found to be robust and reliable, whereas the others showed various problems. CONCLUSIONS: The adaptive distances algorithm was superior in use to the clustering algorithms against which it was tested, but the problem of automatic determination of the number of clusters remains to be addressed.

Algorithms↗