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 217 records · Page 12Linked to original sources
PubMed1996

On-line EEG classification during externally-paced hand movements using a neural network-based classifier.

EEGs of 6 normal subjects were recorded during sequences of periodic left or right hand movement. Left or right was indicated by a visual cue. The question posed was: 'Is it possible to move a cursor on a monitor to the right or left side using the EEG signals for cursor control?' For this purpose the EEG during performance of hand movement was analyzed and classified on-line. A neural network in form of a learning vector quantizertion (LVQ) with an input dimension of 16 was trained to classify EEG patterns from two electrodes and two time windows. After two training sessions on 2 different days, 4 subjects showed a classification accuracy of 89-100%. For two subjects classification was not possible. These results show that in general movement specific EEG-patterns can be found, classified in real time and used to move a cursor on a monitor to the left or right. On-line EEG classification is necessary when the EEG is used as input signal to a brain computer interface (BCI). Such a BCI can be a help for handicapped people.

Adult↗
PubMed1998

Pattern of occlusal contacts in lateral positions: canine protection and group function validity in classifying guidance patterns.

STATEMENT OF PROBLEM: The concept of canine protection and group function lack consistency in the definitions and examining methods, and a valid system for evaluating and classifying occlusal contact patterns has not been established. PURPOSE: This study assessed the use of canine protection and group function in classifying occlusal guidance in the natural dentition. MATERIAL AND METHODS: Occlusal contacts of 86 young adults were examined with shim stock in regulated lateral positions, 0.5,1,2 and 3 mm from the maximum intercuspation. The patterns of occlusal contacts varying with the lateral position were described. RESULTS: Focusing on the working-side contact only, most contact patterns belonged to group function, and a few to canine protection. Focusing on both the working and nonworking side contacts, nearly half the contact patterns were those other than canine protection and group function and were classified into balanced occlusion. CONCLUSION: The validity of the classification system using canine protection and group function is questionable. A new classification system of occlusal guidance is desirable.

Adult↗
PubMed2003

Metabonomics classifies pathways affected by bioactive compounds. Artificial neural network classification of NMR spectra of plant extracts.

The biochemical mode-of-action (MOA) for herbicides and other bioactive compounds can be rapidly and simultaneously classified by automated pattern recognition of the metabonome that is embodied in the 1H NMR spectrum of a crude plant extract. The ca. 300 herbicides that are used in agriculture today affect less than 30 different biochemical pathways. In this report, 19 of the most interesting MOAs were automatically classified. Corn (Zea mays) plants were treated with various herbicides such as imazethapyr, glyphosate, sethoxydim, and diuron, which represent various biochemical modes-of-action such as inhibition of specific enzymes (acetohydroxy acid synthase [AHAS], protoporphyrin IX oxidase [PROTOX], 5-enolpyruvylshikimate-3-phosphate synthase [EPSPS], acetyl CoA carboxylase [ACC-ase], etc.), or protein complexes (photosystems I and II), or major biological process such as oxidative phosphorylation, auxin transport, microtubule growth, and mitosis. Crude isolates from the treated plants were subjected to 1H NMR spectroscopy, and the spectra were classified by artificial neural network analysis to discriminate the herbicide modes-of-action. We demonstrate the use and refinement of the method, and present cross-validated assignments for the metabolite NMR profiles of over 400 plant isolates. The MOA screen also recognizes when a new mode-of-action is present, which is considered extremely important for the herbicide discovery process, and can be used to study deviations in the metabolism of compounds from a chemical synthesis program. The combination of NMR metabolite profiling and neural network classification is expected to be similarly relevant to other metabonomic profiling applications, such as in drug discovery.

Herbicides↗
PubMed2000

Classifying environmental pollutants: Part 3. External validation of the classification system.

In order to validate a classification system for the prediction of the toxic effect concentrations of organic environmental pollutants to fish, all available fish acute toxicity data were retrieved from the ECETOC database, a database of quality-evaluated aquatic toxicity measurements created and maintained by the European Centre for the Ecotoxicology and Toxicology of Chemicals. The individual chemicals for which these data were available were classified according to the rulebase under consideration and predictions of effect concentrations or ranges of possible effect concentrations were generated. These predictions were compared to the actual toxicity data retrieved from the database. The results of this comparison show that generally, the classification system provides adequate predictions of either the aquatic toxicity (class 1) or the possible range of toxicity (other classes) of organic compounds. A slight underestimation of effect concentrations occurs for some highly water soluble, reactive chemicals with low log K(ow) values. On the other end of the scale, some compounds that are classified as belonging to a relatively toxic class appear to belong to the so-called baseline toxicity compounds. For some of these, additional classification rules are proposed. Furthermore, some groups of compounds cannot be classified, although they should be amenable to predictions. For these compounds additional research as to class membership and associated prediction rules is proposed.

Animals↗
PubMed1999

Quantitation of multiple gene expression by in situ hybridization autoradiography: accurate normalization using Bayes classifier.

In the method of in situ hybridization autoradiography, quantitative comparisons among multiple mRNA signals have proven difficult for many reasons, attributable both to technical factors (e.g. different probe specific activities) as well as to large differences in the patterns and levels of expression of different genes in pathologic states. Here we report a standardized normalization procedure for in situ hybridization autoradiography, employing a Bayes classifier, which permits the comparison of multiple mRNA probes. Autoradiograms of different probes in individual animals are first digitized and converted to units of radioactivity. Next, pixel-distribution histograms are generated for each mRNA signal. The Bayes classifier is then used to establish an optimal threshold to distinguish activated and non-activated pixels. This threshold also defines the minimal level of mRNA expression. The maximal mRNA signal is defined as the mean + 3 SD of the activated pixel distribution. We then use a linear transformation to convert each pixel from absolute activity to percentage of maximal mRNA signal for that particular probe. The normalized autoradiographic images can then be averaged to represent group trends and can be compared by standard statistical methods. We illustrate this normalization procedure using in situ hybridization autoradiography for three genes (GADD45, HSP70 and MAP2) expressed in the brains of rats studied at various recirculation times following transient (2 h) middle cerebral artery occlusion. The Bayes classifier is reviewed and its analytical application is presented. Step-by-step examples of intermediate steps are presented, construction of averaged data sets, and pixel-based statistical comparisons among expressed genes.

Animals↗
PubMed2002

Classifying the probability of urinary incontinence in psychogeriatric nursing home patients.

Urinary incontinence (UI) frequently occurs in psychogeriatric nursing home patients. In general the personnel involved in the care for these patients act on incontinence noted. Patients are not monitored or classified according to likelihood or severity of incontinence. This study was conducted to develop and validate a model for the classification of the likelihood of UI in demented nursing home patients. A multi-center cross-sectional study was conducted using data on clinical and functional status of 692 subjects. Subjects were subdivided in a Derivation set of 532 patients and a Validation set of 160 patients. The data were ascertained with questionnaires completed by physicians and nursing staff. All psychogeriatric wards (25) of four Dutch nursing homes were included. Using univariate logistic regression analysis on the derivation set we identified correlates of UI among 22 clinical and functional patient characteristics. Subsequently, we developed a classification model for prevalent UI, including independent patient characteristics by means of multivariable logistic regression. Next, we stratified patients into groups with varying likelihood's of UI based on the model developed. Subsequently, we transformed the model to an easy applicable classification rule for the identification of patient subgroups with high or low likelihood on UI. Finally, the rule was validated on the validation set. The independent multivariate factors associated with urinary incontinence were impaired ADL and mobility, diminished alertness and fecal impaction. After transforming the regression model to an easy classification rule, the scores ranged from 0 to 7. The area under the curve was 0.88 (95% Confidence Interval (CI): 0.85-0.91) in the derivation set. In the validation set a similar area under the curve was obtained (0.90 (95% CI: 0.85-0.95)). Among subjects with none of the associated factors the rule classified 0.5% as incontinent patients. In case all associated factors were present the proportion classified as incontinent increased to 91%. In conclusion, the developed classification rule provides means to stratifying nursing home patients according to their likelihood of being incontinent of urine.

Journal Article↗
PubMedDate not supplied

Supracricoid partial laryngectomy with cricohyoidoepiglottopexy for "early" glottic carcinoma classified as T1-T2N0 invading the anterior commissure.

PURPOSE: "Early" glottic squamous cell carcinoma classified as T1-T2N0 with anterior commissure invasion is conventionnaly managed with vertical partial laryngectomy (VPL) or radiation therapy (RT). At our insitution, in the early 1980s, vertical partial laryngectomy was progressively replaced by supracricoid partial laryngectomy with cricohyoidoepiglottopexy (SCPL-CHEP). The medical files and operative charts of 62 patients with "early" glottic carcinoma classified as T1-T2N0 invading the anterior commissure, consecutively managed with cricohyoidoepiglottopexy, were retrospectively reviewed to ascertain whether any conclusions could be drawn regarding this treatment modality. MATERIALS AND METHODS: Survival, local control, nodal recurrence, distant metastasis, and metachronous second primary tumor estimate was analyzed using the Kaplan-Meier life table method. RESULTS: The 3- and 5-year actuarial survival estimate was 93.3% and 86.5%, respectively. The 3- and 5-year actuarial local control estimate was 98.2%. The only patient with local recurrence was successfully salvaged with RT resulting in an overall 100% local control rate and laryngeal preservation rate. The 3- and 5-year actuarial nodal recurrence estimate was 1.8%. The 3- and 5-year actuarial distant metastasis estimate was 0% and 2%, respectively. Aspiration related completion total laryngectomy and permanent tracheostomy never occurred. CONCLUSION: The present retrospective study suggests that cricohyoidoepiglottopexy for glottic carcinoma classified as T1-T2 invading the anterior commissure resulted in higher local control rates and overall laryngeal preservation rate when compared with historical series using either VPL or RT. Further series are warranted to confirm our results.

Adult↗
PubMed1997

An extension of stochastic curtailment for incompletely reported and classified recurrent events: the Multicenter Study of Hydroxyurea in Sickle Cell Anemia (MSH).

The Multicenter Study of Hydroxyurea in Sickle Cell Anemia (MSH), a double-blind randomized clinical trial, compared the frequency of acute vaso-occlusive (painful) crises during 2 yr of follow-up in 299 patients randomly assigned to hydroxyurea or placebo. Most patients had more than one crisis; all crises reported were included in the primary outcome analysis. A total of 7,229 follow-up medical contact reports were classified as crises/not crises by a Crisis Review Committee. Because of the time required to report, document, and classify contacts, interim analyses were prepared with incomplete data. If a stopping boundary were crossed, early termination could be advised only after assessing the potential impact of the incomplete data. In an extension of stochastic curtailment methods, simulation procedures were used to estimate the probability of detecting differences when group crisis rates projected to the end of the study were compared using a rank test. To account for medical contacts not yet reported and the future occurrence of crises, Poisson process models assuming no treatment effect on crisis rates were used for these simulations. The number of unclassified contacts that would be classified as crises was simulated as a binomial random variable. These methods may be useful for interim monitoring in other studies of recurrent events with ongoing event reporting and classification.

Anemia, Sickle Cell↗
PubMed1998

Design and test of neural networks and statistical classifiers in computer-aided movement analysis: a case study on gait analysis.

OBJECTIVE: To describe the methods for designing and testing diagnostic systems in movement analysis and to verify the clinical usefulness of neural networks and statistical classifiers in a case study. DESIGN: Connectionist and statistical models trained and tested with measured data. BACKGROUND: A basic need in rehabilitation and related fields is to efficiently manage the vast information obtained from a movement analysis laboratory. Many studies have dealt with the interpretation of measured variables in order to correlate objective descriptors to the presence and/or severity of specific neuromusculoskeletal disorders or their consequences. This traditional analytical approach has been complemented in the last decade by new non-linear classification tools called neural networks. METHODS: A gait analysis study on 148 lower limb arthrosis patients and 88 age-matched control subjects. Pathological and healthy gait patterns obtained from force plates wer discriminated by means of multilayer perceptrons and statistical classifiers. RESULTS: Ten input features were enough to train a multilayer perceptron with six hidden neurons. The discrimination rate of the neural net was 80% after cross-validation, significantly higher (P<0.05) than the performance of a Bayes quadratic classifier (about 75%). A great variance due to a small cross-validation set could be demonstrated. CONCLUSIONS: Strict statistical requirements must be observed for designing a neural network. Although these models attain a better performance than conventional statistical approaches, the benefits they bring are sometimes not sufficient to justify their use. Furthermore, clinicians routinely involved in critical decisions may not consider such diagnostic systems reliable enough.

Journal Article↗
PubMed1998

Time domain analysis of embolic signals can be used in place of high-resolution Wigner analysis when classifying gaseous and particulate emboli.

The sample-volume length (SVL) of an embolic signal has previously been used to differentiate between gaseous and particulate emboli and has been calculated using high-resolution Wigner analysis. Although successful, this method of analysis is not widely available to other groups using transcranial Doppler ultrasound (TCD) to classify emboli. The SVL of embolic signals can also be calculated using time domain analysis, which is a far simpler method and potentially available to all TCD users. The aim of this study was to compare the SVL of embolic signals calculated using Wigner analysis and time domain analysis to assess whether or not time domain analysis can replace Wigner analysis to classify emboli. In total, 215 particulate and 100 gaseous emboli were recorded onto digital audiotape and analysed off-line. Two SVLs for each embolic signal were calculated by measuring embolic duration and velocity in the time domain and with Wigner analysis. Receiver operator characteristic (ROC) curves were plotted to assess the optimum SVL threshold for each method, and levels of sensitivity and specificity were defined. The optimum SVL threshold using Wigner analysis was 1.28 cm, yielding 93% sensitivity and 97% specificity. Using time domain analysis, the optimum threshold was 1.12 cm, yielding 90% sensitivity and 96% specificity. The methods were compared statistically (chi2) using their optimum thresholds, and were found not to be statistically different for classifying particles p=0.283) or gaseous emboli (p=0.700). This study has shown that the SVL of embolic signals, used to differentiate particulate from gaseous emboli, can be calculated more simply in the time domain, which yields as accurate results as calculating the SVL using Wigner analysis.

Blood Flow Velocity↗
PubMed2003

Air classifier technology (ACT) in dry powder inhalation. Part 2. The effect of lactose carrier surface properties on the drug-to-carrier interaction in adhesive mixtures for inhalation.

The effect of carrier surface properties on drug particle detachment from carrier crystals during inhalation with a special test inhaler with basic air classifier has been studied for mixtures containing 0.4% budesonide. Carrier crystals were retained in the classifier during inhalation and subsequently examined for the amount of residual drug (carrier residue: CR). Carrier surface roughness and impurity were varied within the range of their appearance in standard grades of lactose (Pharmatose 80, 100, 110, 150 and 200M) by making special sieve fractions. It was found that roughness and impurity, both per unit calculated surface area (CSA), tend to increase with increasing mean fraction diameter for the carrier. Drug re-distribution experiments with two different carrier sieve fractions with distinct mean diameters showed that the amount of drug per CSA (drug load) in the state of equilibrium is highest for the coarsest fraction. This seems to confirm that surface carrier irregularities are places where drug particles preferentially accumulate. However, a substantial increase in surface roughness and impurity appears to be necessary to cause only a minor increase in CR at an inspiratory flow rate of 30 l/min through a classifier. At 60 l/min, CR is practically independent of the carrier surface properties. From the difference in CR between 30 and 60 l/min, it has been concluded that particularly the highest adhesive forces (for the largest drug particles) in the mixture are increased when coarser carrier fractions (with higher rugosity) are used. Not only increased surface roughness and impurities may be responsible for an increase in the adhesive forces between drug and carrier particles when coarser carrier fractions are used, but also bulk properties may play a role. With increasing mean carrier diameter, inertial and frictional forces during mixing are increased too, resulting in higher press-on forces with which the drug particles are attached to carrier crystals and to each other.

Adhesives↗
PubMed1999

The eigenspace separation transform for neural-network classifiers.

This paper presents a linear transform that compresses data in a manner designed to improve the performance of a neural network used as a binary classifier. The classifier is intended to accommodate data distributions that may be non-normal, may have equal class means, may be multimodal, and have unknown a priori probabilities for the two classes. The transform, which is called the eigenspace separation transform, allows the reduction of the size of a neural network while enhancing its generalization accuracy as a binary classifier.

Journal Article↗
PubMedDate not supplied

Evaluation of a neural network classifier for pancreatic masses based on CT findings.

We have investigated a neural network classifier based on CT findings extracted by a radiologist for the differential diagnosis between the pancreatic ductal adenocarcinoma and mass-forming pancreatitis, and compared its classification performance with that of Bayesian analysis, Hayashi's quantification method II, and radiologists. The three computerized classification methods were designed to classify categorized CT findings extracted by a radiologist, and were trained and tested on 71 cases. There was comparable performance between the neural the network, the Bayesian analysis, Hayashi's quantification method II, and the radiologists, in classifying pancreatic carcinoma and inflammatory mass.

Adult↗
PubMed1998

Phenotypic and genotypic diversity of organisms previously classified as maltose positive Pasteurella multocida.

Fifteen isolates tentatively classified as maltose positive Pasteurella multocida have been characterized in 79 biochemical tests and by ribotyping using HindIII and HpaII for digestion of DNA. Phenotypic and genotypic results were analysed using the computer programmes NTSYS and GelCompar, respectively. Two strains were classified as maltose positive P. multocida ssp. multocida while six strains were classified as maltose positive P. multocida ssp. septica. The remaining strains clustered with P. volantium and P. gallinarum, but remained unclassified. With the exception of a single isolate correlation was observed between phenotypic and genotypic results. The unclassified isolates which represented three different sources were heterogenous according to both phenotypic and genotypic results. The findings obtained support the 16S rRNA sequencing results indicating that the genus Pasteurella sensu stricto might represent two or more genera.

Animals↗
PubMed2002

Validation using sensitivity and target transform factor analyses of neural network models for classifying bacteria from mass spectra.

Temperature constrained cascade correlation networks (TCCCNs) are computational neural networks that configure their own architecture, train rapidly, and give reproducible prediction results. TCCCN classification models were built using the Latin-partition method for five classes of pathogenic bacteria. Neural networks are problematic in that the relationships among the inputs (i.e., mass spectra) and the outputs (i.e., the bacterial identities) are not apparent. In this study, neural network models were constructed that successfully classified the targeted bacteria and the classification model was validated using sensitivity and target transformation factor analysis (TTFA). Without validation of the classification model, it is impossible to ascertain whether the bacteria are classified by peaks in the mass spectrum that have no causal relationships with the bacteria, but instead randomly correlate with the bacterial classes. Multiple single output network models did not offer any benefits when compared to single network models that had multiple outputs. A multiple output TCCCN model achieved classification accuracies of 96 +/- 2% and exhibited improved performance over multiple single output TCCCN models. Chemical ionization mass spectra were obtained from in situ thermal hydrolysis methylation of freeze-dried bacteria. Mass spectral peaks that pertain to the neural network classification model of the pathogenic bacterial classes were obtained by sensitivity analysis. A significant number of mass spectral peaks that had high sensitivity corresponded to known biomarkers, which is the first time that the significant peaks used by a neural network model to classify mass spectra have been divulged. Furthermore, TTFA furnishes a useful visual target as to which peaks in the mass spectrum correlate with the bacterial identities.

Bacteria↗
PubMed2001

Classifying complications of interventional procedures: a survey of practicing radiologists.

PURPOSE: To determine the variability of radiologists' classification of complications from interventional procedures. MATERIALS AND METHODS: Fifteen test cases were selected from a database of morbidity and mortality cases that occurred in our department during the past 2 years. Ten cases were selected randomly, and five were chosen because of classification difficulties within our department. A survey with the case descriptions was presented to 145 SCVIR members via the World Wide Web and 48 were distributed to participants at a statewide angiography club meeting. Participants were asked to complete a short assessment of the their clinical background and to classify each case as "no complication," "minor complication," or "major complication." RESULTS: Thirty-eight percent (74 of 193) of the surveys were completed. Seventy percent (52 of 74) of the respondents were affiliated with an academic program, 12% (nine of 74) were affiliated with private practice groups, and 18% (13 of 74) claimed both academic and private affiliation. The consensus rate in classifying the complications for the randomly selected cases varied from 50% to 95%, with a median of 69%, and the consensus rate in classifying the selected cases varied from 46% to 95%, with a median of 85%. The lowest consensus rates occurred when (i) a significant procedural event was followed by a normal outcome, (ii) when a procedure was aborted, and (iii) when a significant event occurred but did not prolong hospital stay. CONCLUSION: Current criteria for reporting complications are associated with moderate rates of disagreement among interventional radiologists.

Adult↗
PubMed1991

Later rather than sooner: extralinguistic categories in the acquisition of Thai classifiers.

An experimental elicitation task with children between the ages of 1;8 and 11;3 shows that children learning Thai numerical classifiers begin with purely distributional information: specifically, (1) that classifiers must appear in the post-numeral position, and (2) that classifiers comprise a conventional, closed set of words. Semantic organizing features, such as salient features of the head noun's referent, appear later than these syntagmatic organizing features. Use of such semantic information is not an immature 'first guess' at grammatical categories, but rather, a necessary component of adult linguistic competence, because the categories are productive both for older children and for adults.

Child↗
PubMed1997

Distribution of cortical granules in bovine oocytes classified by cumulus complex.

The present study was conducted to examine distributional changes of cortical granules (CGs) during meiotic maturation and fertilisation in vitro and the developmental ability in bovine oocytes classified by cumulus cells. The oocytes were classified by the morphology of their cumulus cell layers as follows: class A, compact and thick; class B, compact but thin; class C, naked; and class D, expanded. Some of the oocytes were stained with Lens curinalis agglutinin (LCA) before and after maturation in vitro and after insemination, and then stained with orcein to observe their nuclear stages. The others were left in culture. Distributional patterns of the CGs were classified into four types: type I, CGs distributed in clusters; type II, CGs dispersed and partly clustered; type III, all CGs dispersed; and type IV, no CGs. Most of the oocytes before culture showed a type I pattern, but this decreased after maturation culture, whereas type III increased in class A. The oocytes of class B showed similar changes while the oocytes of class C did not. In class C, many oocytes showed type I after culture, indicating that cytoplasmic maturation was not completed. In class D, 80.4% of the oocytes exhibited type III before maturation culture, indicating that their cytoplasmic maturation was different from classes A-C. And about 70% of the class D oocytes were at the nuclear stage of germinal vesicle breakdown (GVBD) before culture. The developmental rates to blastocysts in classes A-D were 28.7%, 23.1%, 0.5% and 3.4% respectively.

Agglutinins↗