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 325 records · Page 18Linked to original sources
PubMed2002

DNA profiles and numeric histogram classifiers in nephrogenic adenoma.

BACKGROUND: The malignant potential of nephrogenic adenoma is still a matter of controversy and therapeutic regimens of this morphologic entity range from partial, even total cystectomy to watchful waiting. The objective of the current study was to evaluate several robust image cytometry-DNA histogram classifiers and to search among those for factors that separate a biologically nonaggressive metaplastic lesion from lesions with increased malignant potential. METHODS: The study included bladder irrigation specimens, 23 preceding transurethral resection of nephrogenic adenoma and 24 preceding resection of papillary bladder carcinoma. Feulgen-stained nuclei were imported to a static image analysis system, and densitometric data were interpreted by two different software programs. Histograms were described numerically by DNA index, 2c deviation index, and by 5c/9c-exceeding and euploid polyploidy rates. In addition, an interpretation algorithm based on a dual parameter analysis with an integrated automatic threshold was used. RESULTS: The numeric classification of DNA histograms of patients suffering from nephrogenic adenoma resulted in DNA indices between 0.91 and 1.15. The 2c deviation indices ranged from 0.03 to 0.43, and the 5c exceeding rates ranged from 0.0 to 1.58. None of the measurements showed nuclei exceeding 9c. The p25-75 ranges of 2c deviation indices in nephrogenic adenoma and papillary urothelial carcinoma did not overlap. These findings might be explained by minor proliferative activity in nephrogenic adenoma. Euploid polyploidy rates less than 5% confirm this explanation. Risk analysis documented high risk only for those patients with nephrogenic adenomas who had proven transitional cell carcinoma in their history. CONCLUSIONS: DNA estimation by image cytometry of urinary bladder irrigation specimens appears able to separate papillary bladder lesions. The method detects those lesions with higher malignant potential but is limited in separating entities with low malignant potential. Comparison of the discriminative power of robust numeric DNA classifiers reveals the 2c deviation index superior to the widely used DNA index and the 5c exceeding rate in this material.

Adenoma↗
PubMed2002

Classifying local disease recurrences after breast conservation therapy based on location and histology: new primary tumors have more favorable outcomes than true local disease recurrences.

BACKGROUND: To distinguish true local recurrences (TR) from new primary tumors (NP) and to assess whether this distinction has prognostic value in patients who develop ipsilateral breast tumor recurrences (IBTR) after breast-conserving surgery and radiotherapy. METHODS: Between 1970 and 1994, 1339 patients underwent breast-conserving surgery at The University of Texas M. D. Anderson Cancer Center for ductal carcinoma in situ or invasive carcinoma. Of these patients, 139 (10.4%) had an IBTR as the first site of failure. For the 126 patients with clinical data available for retrospective review, we classified the IBTR as a TR if it was located within 3 cm of the primary tumor bed and was of the same histologic subtype. All other IBTRs were designated NP. RESULTS: Of the 126 patients, 48 (38%) patients were classified as NP and 78 (62%) as TR. Mean time to disease recurrence was 7.3 years for NP versus 5.6 years for TR (P = 0.0669). The patients with NP had improved 10-year rates of overall survival (NP 77% vs. TR 46%, P = 0.0002), cause-specific survival (NP 83% vs. TR 49%, P = 0.0001), and distant disease-free survival (NP 77% vs. TR 26%, P < 0.0001). Patients with NP more often developed contralateral breast carcinoma (10-year rate: NP 29% vs. TR 8%, P = 0.0043), but were less likely to develop a second local recurrence after salvage treatment of the first IBTR (NP 2% vs. TR 18%, P = 0.008). CONCLUSIONS: Patients with NP had significantly better survival rates than those with TR, but were more likely to develop contralateral breast carcinoma. Distinguishing new breast carcinomas from local disease recurrences may have importance in therapeutic decisions and chemoprevention strategies. This is because patients with new carcinomas had significantly lower rates of metastasis than those with local disease recurrence, but were more likely to develop contralateral breast carcinomas.

Actuarial Analysis↗
PubMed1981

The look-up table: a classifier for cell sorters.

Currently used classifiers for flow cell sorters are described and the look-up table is introduced as such. An experimental setup of a look-up table classifier added to a commercially available cell sorter, under control of a microcomputer is described. The contents of the look-up table are obtained from contours of regions of interest. These regions are defined by the user who designates the areas on a video color display of the bivariate flow data. Since most cell clusters form elliptic or ovoid shapes in a multi-dimensional feature space this method enables a better matching of window shapes to cluster outlines, compared to conventional sort decision circuitry.

Cell Separation↗
PubMed1981

The effect of abnormal cell proportion on specimen classifier performance.

In two previous papers we developed formulas relating the performance (error rates) of a two-class specimen classifier to the performance of a preceding two-class classifier and the number of cells examined (K. R. Castleman and B. S. White, Analytical and Quantitative Cytology 2:117, 1980 and K. R. Castleman and B. S. White, Cytometry 1:156, 1980.). This analysis assumed a certain proportion (p) of abnormal cells on an abnormal specimen. In this paper we examine what happens when a system designed assuming one value of p is presented with a positive specimen having a different abnormal cell proportion. We show that the specimen false negative error rate increases dramatically as p decreases below the design value, and conversely. This suggests that the specimen classification performance of a particular system should be quoted only with reference to the abnormal cell proportion of the specimens used for testing.

Cell Separation↗
PubMed1987

Improved model for specimen classification based on single-cell classifiers.

We consider probabilistic models for specimen classification procedures based on systems which classify individual cells as normal or abnormal. The models which we consider generalize those discussed previously by Castleman and White (Anal. Quant. Cytol. 2:117-122, 1980; Cytometry 2:155-158, 1981) and by Timmers and Gelsema (Cytometry 6:22-25, 1985). In particular, they include the biologically plausible possibility that the specimen contains cells which are intermediate between the extremes of normal and abnormal. We find that if these additional cells occur differentially in normal and abnormal specimens, then specimen classification can become substantially more efficient when the cell classifier has different error rates for these cells.

Cervix Uteri↗
PubMed2002

Normal mode analysis of macromolecular motions in a database framework: developing mode concentration as a useful classifying statistic.

We investigated protein motions using normal modes within a database framework, determining on a large sample the degree to which normal modes anticipate the direction of the observed motion and were useful for motions classification. As a starting point for our analysis, we identified a large number of examples of protein flexibility from a comprehensive set of structural alignments of the proteins in the PDB. Each example consisted of a pair of proteins that were considerably different in structure given their sequence similarity. On each pair, we performed geometric comparisons and adiabatic-mapping interpolations in a high-throughput pipeline, arriving at a final list of 3,814 putative motions and standardized statistics for each. We then computed the normal modes of each motion in this list, determining the linear combination of modes that best approximated the direction of the observed motion. We integrated our new motions and normal mode calculations in the Macromolecular Motions Database, through a new ranking interface at http://molmovdb.org. Based on the normal mode calculations and the interpolations, we identified a new statistic, mode concentration, related to the mathematical concept of information content, which describes the degree to which the direction of the observed motion can be summarized by a few modes. Using this statistic, we were able to determine the fraction of the 3,814 motions where one could anticipate the direction of the actual motion from only a few modes. We also investigated mode concentration in comparison to related statistics on combinations of normal modes and correlated it with quantities characterizing protein flexibility (e.g., maximum backbone displacement or number of mobile atoms). Finally, we evaluated the ability of mode concentration to automatically classify motions into a variety of simple categories (e.g., whether or not they are "fragment-like"), in comparison to motion statistics. This involved the application of decision trees and feature selection (particular machine-learning techniques) to training and testing sets derived from merging the "list" of motions with manually classified ones.

Databases, Protein↗
PubMed2001

Biopsy correlates of abnormal cervical cytology classified using the Bethesda system.

OBJECTIVE: The goal of this study was to determine the colposcopic findings underlying cytologic abnormalities classified according to the Bethesda system. METHODS: Women undergoing colposcopy for abnormal cytology at an urban teaching hospital between July 1, 1996 and December 31, 1999 had Papanicolaou smears repeated. Results were compared both with biopsy histology and with the worst histology reported after 8-26 months of follow-up. kappa statistics and Spearman's rho were calculated to determine the degree of agreement. RESULTS: Colposcopy was performed for 2263 (94%) women. Referral and repeat Pap smears were reported identically in 493 (25%) of the 1962 women with results for both. No AGUS (atypical glandular cells of uncertain significance) smears were confirmed on repeat smear, and after excluding AGUS, agreement within one grade was found in 1305 of 1854 (70%). Among the 1842 women with squamous cytologic abnormalities, biopsy revealed a lesion more severe than that suggested by referral cytology in 577 (31%) and a less severe lesion in 648 (35%); exact correspondence was found in only 646 (35%). Of 317 women with ASCUS (atypical squamous cells of uncertain significance) on referral Pap smear, a negative repeat smear, and a specific biopsy result, 95 (30%) had true negative histology, while 148 (47%) had condyloma, 56 (18%) had cervical intraepithelial neoplasia (CIN) 1, 8 (3%) had CIN 2, 10 (3%) had CIN 3, and none had cancer. Comparison of repeat smear and colposcopic biopsy yielded a kappa statistic of 0.16. CONCLUSIONS: Cytology classified according to the Bethesda system does not accurately predict histologic diagnosis.

Adolescent↗
PubMed1996

The use of amino acid patterns of classified helices and strands in secondary structure prediction.

Elements of secondary structure from known protein three-dimensional structures have been classified with respect to their environments in tertiary structures. The size of the solvent-inaccessible face of an alpha-helix and the accessibility patterns on the two sides of a beta-strand have been used to classify the secondary structures. For each class, we have derived a sequence template, giving the amino acid propensity at each position. A prediction is made by calculating the compatibility of segments of polypeptide sequence against templates for each type of secondary structure. This method predicts not only position of a secondary structure in a protein sequence but also the orientation of the secondary structure with respect to the core of the protein tertiary structure. A jack-knife test is applied to 78 proteins of known structure solved at better than 2 A resolution. It shows that this method predicts between 13% and 17% better than the methods of Lim, GOR and Chou and Fasman at the level of secondary structure. The orientations of inaccessible faces are predicted within 50 degrees of correct value for about two-thirds of alpha-helices.

Adenylate Kinase↗
PubMed2005

Efficient learning by combining confidence-rated classifiers to incorporate unlabeled medical data.

In this paper, we propose a new dynamic learning framework that requires a small amount of labeled data in the beginning, then incrementally discovers informative unlabeled data to be hand-labeled and incorporates them into the training set to improve learning performance. This approach has great potential to reduce the training expense in many medical image analysis applications. The main contributions lie in a new strategy to combine confidence-rated classifiers learned on different feature sets and a robust way to evaluate the "informativeness" of each unlabeled example. Our framework is applied to the problem of classifying microscopic cell images. The experimental results show that 1) our strategy is more effective than simply multiplying the predicted probabilities, 2) the error rate of high-confidence predictions is much lower than the average error rate, and 3) hand-labeling informative examples with low-confidence predictions improves performance efficiently and the performance difference from hand-labeling all unlabeled data is very small.

Algorithms↗
PubMed1996

Carcinoembryonic antigen, tissue polypeptide antigen and neuron-specific enolase pleural levels used to classify small-cell and non-small-cell lung cancer patients by discriminant analysis.

The classification of lung cancer into small-cell lung cancer (SCLC) and non-small-cell lung cancer (NSCLC) is essential for disease prognosis and treatment. For this purpose, we have tried to optimize the use of three tumour markers determined on pleural effusions, to differentiate SCLC from NSCLC by means of a canonic variable, generated by discriminant analysis, including subjects with histologically proven lung cancer. Discriminant analysis was performed by using carcinoembryonic antigen, neuron-specific enolase and tissue polypeptide antigen pleural levels, determined in 65 consecutive and unselected patients, histologically classified as 49 NSCLC and 16 SCLC. To validate the formula generated, a control group of 37 lung cancer patients (10 SCLC and 27 NSCLC), enrolled subsequently, was employed. Applying the discriminant analysis to SCLC and NSCLC patients a good classification was obtained (92% rate of correct classification). The aforementioned formula, applied to the validation group, showed a 92% rate of correct classification. This method, which is rapid, inexpensive and routinely applicable to malignant pleural effusions, may be reliably used to classify lung cancer patients.

Adult↗
PubMed1994

Comparison of the polymerase region of small round structured virus strains previously classified in three antigenic types by solid-phase immune electron microscopy.

We have used a reverse transcription-polymerase chain reaction with nested sets of primers to determine the nucleotide sequences of a 166 base pair segment of the RNA polymerase region of seven strains of small round structured viruses (SRSVs) from the United Kingdom. These SRSV strains were previously classified by solid-phase immune electron microscopy into three antigenic types--UK2, UK3 and UK4, which are comparable to the prototype strains Norwalk virus, Hawaii agent, and Snow Mountain agents, respectively. Based on their sequences, the seven strains from the United Kingdom could be divided into two groups. The first group included two strains of the UK2 type along with Norwalk virus and Southampton virus and the second group included three strains of UK3 and two strains of UK4 types. Viruses in the first group showed 75.3%-77.1% nucleotide and 89.1%-94.6% amino acid identity with Norwalk virus while those of the second group showed 60.8%-63.3% nucleotide and 67.3%-69.1% amino acid identity. Nucleotide and amino acid identity within the second group ranged between 91.6%-99.4% and 96.4%-100%, respectively. These results suggest that the SRSVs antigenically related with Norwalk virus, Hawaii agent, and Snow Mountain agent, can be classified into two genotypes on the basis of their sequences in the RNA polymerase region.

Adult↗
PubMed1994

The development of a decision support system for the pathological diagnosis of human cerebral tumours based on a neural network classifier.

This study describes the use of a topological mapping system in the classification of cerebral tumours and the development of a decision support system based upon that classifier. Fourteen pathological parameters from two hundred primary cerebral tumours are presented as vectors to a topological map. The map, consisting of a grid of neurones, learns the features of each tumour by means of a shortest Euclidean distance algorithm, after which self adaptation of the neurons occurs. An LVQ algorithm performs the final classification. Study of the map reveals that it can correctly classify tumors following their malignancy potential and their cytogenesis. The decision support system uses the network at its core and helps not only in reaching a diagnosis but also in finding the optimal way to reach that diagnosis. The usefulness of such a mapping system lies in the field of education, clinical research and medically acceptable cost reduction.

Algorithms↗
PubMed1991

[Schmuth's group findings--a simple system for classifying dysgnathias].

The advantages and disadvantages of Schmuth's findings, which classify eugnathic as well as dysgnathic dentition are opposed to Angle's classification. Using Schmuth's findings is a way of classifying without ignoring the physiological variability in the area of the first molars, which can lead to false diagnosis. 386 models were examined at the beginning of the patients' treatment. The results are discussed.

Dental Occlusion↗
PubMed1996

Classifying general medicine readmissions. Are they preventable? Veterans Affairs Cooperative Studies in Health Services Group on Primary Care and Hospital Readmissions.

OBJECTIVES: To describe a new quality assessment method used to classify the preventability of hospitalization in terms of patient, clinician, or system factors. DESIGN: The instrument was developed in two phases. Phase 1 was a prospective comparison of admitting residents' and their attending physicians' classifications of the perceived preventability of consecutive admissions to one Veterans Affairs Medical Center (VAMC) excluding admissions to the intensive care unit (ICU). In phase 2, a panel of 10 physicians rated 811 abstracted records of readmissions from nine VAMCs. SETTING: Nine VAMCs across the United States with varying degrees of university hospital affiliation. PATIENTS: Phase 1, 156 patients admitted to the general medicine service at the Durham VAMC. Phase 2, 514 patients accounting for 811 readmissions within 6 months of a general medicine service discharge at nine VAMCs. MEASUREMENTS AND MAIN RESULTS: Physicians used a checklist to record the reason for hospitalization, the preventability of the hospitalization, and, if preventable, a reason defining preventability, which was classified in terms of system, clinician, and patient factors. In phase 2, two physician panelists assessed preventability for each chart. When two panelists disagreed on the preventability of hospitalization, a third panelist, blind to the original assessments, rated the chart. In phase 1, residents and attending physicians rated 33% and 34% of admissions as preventable (kappa = 0.41), respectively. In phase 2, 277 (34%) of 811 readmissions were deemed preventable. Intraobserver accuracy for the assessment of preventability was 96% (kappa = 0.89). interobserver accuracy was 73% (kappa = 0.43). Hospital system factors accounted for 37% of preventable readmissions, clinician factors for 38%, and patient factors for 21%. The nine hospitals differed markedly in their profile of reasons for preventable readmissions (p = .005). CONCLUSIONS: Using a new method of determining the preventability of hospitalizations, we identified several factors that might avert hospitalizations. Focusing efforts to identify preventable hospitalizations may yield better methods for managing patients' total health care needs; however, the content of those efforts will vary by institution.

Adult↗
PubMed1992

An overview of principles for classifying brain tumors.

The purpose of this review is to clarify for the nonneuropathologist some of the confusing issues concerning the classification of brain tumors. Following a short discussion of the commonly used methods to diagnose brain tumors clinically (frozen section, light and electron microscopy, immunohistochemistry), the general principles of classifying neural tumors are presented. Grading of tumors on the basis of histological anaplasia, and the concept that tumor cells can be related to specific cytological stages of normal cellular development (cytogenetic classification) are presented. The World Health Organization system of classifying neural tumors is an attempt to develop a standardized classification scheme with as few interpretative controversies as possible, but it has required revision as new information has been gained. The major clinical and biological features of the commonest tumor groups are then discussed. It is unlikely that improvements in the classification of brain tumors will be based solely on histological information. Definitions of tumor entities that will provide more accurate prognoses and bases for effective therapy will require considerably more information at the molecular level than is currently available.

Astrocytoma↗
PubMed2006

Age and sex differences in the locomotor effect of repeated methylphenidate in rats classified as high or low novelty responders.

RATIONALE: Rats displaying high levels of activity in an inescapable novel environment (high responders; HR) are more sensitive to the locomotor effect of stimulant drugs than rats displaying low levels of activity (low responders; LR). OBJECTIVE: The current study determined the age- and sex-dependent locomotor effects of repeated methylphenidate in HR and LR rats. MATERIALS AND METHODS: Periadolescent and adult male and female Sprague-Dawley rats were first classified as HR or LR; rats were also classified as high or low novelty seekers based on free-choice preference for a novel environment. Locomotor activity was subsequently assessed after ten daily injections of methylphenidate (3 or 10 mg/kg s.c.) or saline. Fifteen days later, rats were challenged with saline and methylphenidate (10 mg/kg) over 2 days. RESULTS: During the repeated methylphenidate treatment phase, adult females showed greater methylphenidate-induced hyperactivity than adult males; there was no reliable difference in methylphenidate-induced hyperactivity between HR and LR rats of either age or sex. However, periadolescent male HR rats given repeated methylphenidate showed greater conditioned hyperactivity after the saline challenge than periadolescent male LR rats. Further, adult female HR rats given repeated methylphenidate showed greater conditioned hyperactivity and sensitization than adult female LR rats. In contrast, although free-choice novelty preference was greater among periadolescents than adults, individual differences in this variable did not predict the effect of repeated methylphenidate during any phase of the experiment. CONCLUSION: Although individual differences in response to inescapable novelty predict methylphenidate-induced conditioned hyperactivity and sensitization, this relationship is moderated by age and sex.

Age Factors↗
PubMed2005

Classifying honeys from the Soria Province of Spain via multivariate analysis.

A total of 73 different honeys from seven botanical origins [ling (Calluna vulgaris L.), heather (Erica sp.), rosemary (Rosmarinus officinalis L.), thyme (Thymus vulgaris L.), honeydew (Quercus sp.), spike lavender (Lavandula latifolia M.) and french lavender (Lavandula stoechas L.)] have been classified by applying discriminant analysis to their metal content data and other common physicochemical parameters. Fifteen minerals were identified and quantified using atomic emission spectroscopy (AES) for K and Na, and inductively coupled plasma atomic emission spectrometry (ICP-AES) for Mg, Ca, Al, Fe, Mn, Zn, B, Cu, Co, Cr, Ni, Cd and Pb. Moreover, eight physicochemical parameters were analysed following the Harmonised Methods of the International Honey Commision: ash content, moisture, insoluble matter, reducing sugars, apparent sucrose, diastase activity, free acidity and hydroxymethylfurfural. The honeys analysed were characterised and distinguished using chemometrics. ANOVA highlighted significant differences between the honeys in terms of the mean contents of all variables except apparent sucrose, HMF, Fe and Zn. Principal component analysis was used as a descriptive tool to visualise the data structure in two dimensions, finding relationships between variables and types of honey. Likewise, discriminant analysis, together with various methods (stepwise, forward and backward), was used to select the variables with the highest discriminating power, which allowed us to classify all of the botanical origins considered in this work, achieving a global success rate close to 90% following cross-validation.

Honey↗