SEARCH · Search PubMed

Results for “Feature selection”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 919 records · Page 51Linked to original sources

PubMed2004

An algorithm for the detection of faces on the basis of Gabor features and information maximization.

We propose an algorithm for the detection of facial regions within input images. The characteristics of this algorithm are (1) a vast number of Gabor-type features (196,800) in various orientations, and with various frequencies and central positions, which are used as feature candidates in representing the patterns of an image, and (2) an information maximization principle, which is used to select several hundred features that are suitable for the detection of faces from among these candidates. Using only the selected features in face detection leads to reduced computational cost and is also expected to reduce generalization error. We applied the system, after training, to 42 input images with complex backgrounds (Test Set A from the Carnegie Mellon University face data set). The result was a high detection rate of 87.0%, with only six false detections. We compared the result with other published face detection algorithms.

Algorithms↗

PubMed1991

Brain potential signs of feature processing during auditory selective attention.

We recorded event-related brain potentials (ERPs) to random dichotic tone sequences as subjects attended to tone bursts of a designated pitch (250, 1000 or 4000 Hz) and ear of delivery. The effects of attention were isolated as negative difference (Nd) waves by subtracting ERPs to ignored tones from ERPs to the same tones when either one or both features were attended. Early sensory components of the ERP changed tonotopically in scalp distribution, while the distributions of Nd waves were feature-specific (pitch processing differed from location processing) but not tonotopic. At longer latencies, Nd waves specific to feature-conjunction operations were isolated. These began 40-50 ms after Nds to isolated cues and continued for hundreds of ms.

Acoustic Stimulation↗

PubMed2000

Novel, potent and selective chimeric FXa inhibitors featuring hydrophobic P1-ketoamide moieties.

Judicious combination of P-region sequences of highly potent anticoagulant proteins including NAP5, NAP6, Ecotin, and Antistasin with SAR from small molecule FXa inhibitors led to a series of chimeric inhibitors of formula 1a-j. We report herein the design, synthesis, and biological activity of this novel family of FXa inhibitors that express both high in vitro potency and superb selectivity against related serine proteases.

Antifibrinolytic Agents↗

PubMed2004

Feature-based attention increases the selectivity of population responses in primate visual cortex.

BACKGROUND: Attending to the spatial location or to nonspatial features of visual stimuli can modulate neuronal responses in primate visual cortex. The modulation by spatial attention changes the gain of sensory neurons and strengthens the representation of attended locations without changing neuronal selectivities such as directionality, i.e., the ratio of responses to preferred and anti-preferred directions of motion. Whether feature-based attention acts in a similar manner is unknown. RESULTS: To clarify this issue, we recorded the responses of 135 direction-selective neurons in the middle temporal area (MT) of two macaques to an unattended moving random dot pattern (the distractor) positioned inside a neuron's receptive field while the animals attended to a second moving pattern positioned in the opposite hemifield. Responses to different directions of the distractor were modulated by the same factor (approximately 12%) as long as the attended direction remained unchanged. On the other hand, systematically changing the attended direction from a neuron's preferred to its anti-preferred direction caused a systematic change of the attentional modulation from an enhancement to a suppression, increasing directionality by about 20%. CONCLUSIONS: The results show that (1) feature-based attention exerts a multiplicative modulation upon neuronal responses and that the strength of this modulation depends on the similarity between the attended feature and the cell's preferred feature, in line with the feature-similarity gain model, and (2) at the level of the neuronal population, feature-based attention increases the selectivity for attended features by increasing the responses of neurons preferring this feature value while decreasing responses of neurons tuned to the opposite feature value.

Animals↗

PubMed2003

Dynamics of feature binding during object-selective attention.

Objects in the environment may be attended selectively and perceived as unified ensembles of their constituent features. To investigate the timing and cortical localization of feature-integration mechanisms in object-based attention, recordings of event-related potentials and magnetic fields were combined with functional MRI while subjects attended to one of two superimposed transparent surfaces formed by arrays of dots moving in opposite directions. A spatiotemporal analysis revealed evidence for a rapid increase in neural activity localized to a color-selective region of the fusiform gyrus when the surface moving in the attended direction displayed an irrelevant color feature. These data provide support for the "integrated-competition" model of object-selective attention and point to a dynamic neural substrate for the rapid binding process that links relevant and irrelevant features to form a unified perceptual object.

Adult↗

PubMedDate not supplied

Risk of cardiovascular events associated with selective COX-2 inhibitors.

Atherosclerosis is a process with inflammatory features and selective cyclooxygenase 2 (COX-2) inhibitors may potentially have antiatherogenic effects by virtue of inhibiting inflammation. However, by decreasing vasodilatory and antiaggregatory prostacyclin production, COX-2 antagonists may lead to increased prothrombotic activity. To define the cardiovascular effects of COX-2 inhibitors when used for arthritis and musculoskeletal pain in patients without coronary artery disease, we performed a MEDLINE search to identify all English-language articles on use of COX-2 inhibitors published between 1998 and February 2001. We also reviewed relevant submissions to the US Food and Drug Administration by pharmaceutical companies. Our search yielded 2 major randomized trials, the Vioxx Gastrointestinal Outcomes Research Study (VIGOR; 8076 patients) and the Celecoxib Long-term Arthritis Safety Study (CLASS; 8059 patients), as well as 2 smaller trials with approximately 1000 patients each. The results from VIGOR showed that the relative risk of developing a confirmed adjudicated thrombotic cardiovascular event (myocardial infarction, unstable angina, cardiac thrombus, resuscitated cardiac arrest, sudden or unexplained death, ischemic stroke, and transient ischemic attacks) with rofecoxib treatment compared with naproxen was 2.38 (95% confidence interval, 1.39-4.00; P =.002). There was no significant difference in cardiovascular event (myocardial infarction, stroke, and death) rates between celecoxib and nonsteroidal anti-inflammatory agents in CLASS. The annualized myocardial infarction rates for COX-2 inhibitors in both VIGOR and CLASS were significantly higher than that in the placebo group of a recent meta-analysis of 23 407 patients in primary prevention trials (0.52%): 0.74% with rofecoxib (P =.04 compared with the placebo group of the meta-analysis) and 0.80% with celecoxib (P =.02 compared with the placebo group of the meta-analysis). The available data raise a cautionary flag about the risk of cardiovascular events with COX-2 inhibitors. Further prospective trial evaluation may characterize and determine the magnitude of the risk.

Anti-Inflammatory Agents, Non-Steroidal↗

PubMed2004

From magic bullets to designed multiple ligands.

Increasingly, it is being recognised that a balanced modulation of several targets can provide a superior therapeutic effect and side effect profile compared to the action of a selective ligand. Rational approaches in which structural features from selective ligands are combined have produced designed multiple ligands that span a wide variety of targets and target classes. A key challenge in the design of multiple ligands is attaining a balanced activity at each target of interest while simultaneously achieving a wider selectivity and a suitable pharmacokinetic profile. An analysis of literature examples reveals trends and insights that might help medicinal chemists discover the next generation of these types of compounds.

Chemistry, Pharmaceutical↗

PubMed1990

DNH deoxyribonucleohelicates: self assembly of oligonucleosidic double-helical metal complexes.

Nucleic acids, because of their key biological role, are prime targets for the design of either analogues that may mimic some of their features or of complementary ligands that may selectively bind to and react with them for regulation or reaction. Whereas there has been much work on the latter topic since the elucidation of the double-helical structure of DNA, comparatively little has been done on structural and/or functional models, probably owing to the lack of self-organizing molecular systems. Here we present a class of artificial systems, the nucleohelicates, which are of interest from both points of view because they combine the double-helical structure of the double-stranded metal complexes, the helicates, with the selective interaction features of nucleic-acid bases. These functionalized species allow the study of structural effects on the formation of the double helix and on the binding to other entities, in particular to nucleic acids.

DNA↗

PubMed2001

Computerized classification of suspicious regions in chest radiographs using subregion Hotelling observers.

We propose to investigate the use of subregion Hotelling observers (SRHOs) in conjunction with perceptrons for the computerized classification of suspicious regions in chest radiographs for being nodules requiring follow up. Previously, 239 regions of interest (ROIs), each containing a suspicious lesion with proven classification, were collected. We chose to investigate the use of SRHOs as part of a multilayer classifier to determine the presence of a nodule. Each SRHO incorporates information about signal, background, and noise correlation for classification. For this study, 225 separate Hotelling observers were set up in a grid across each ROI. Each separate observer discriminates an 8 by 8 pixel area. A round robin sampling scheme was used to generate the 225 features, where each feature is the output of the individual observers. These features were then rank ordered by the magnitude of the weights of a perceptron. Once rank ordered, subsets of increasing number of features were selected to be used in another perceptron. This perceptron was trained to minimize mean squared error and the output was a continuous variable representing the likelihood of the region being a nodule. Performance was evaluated by receiver operating characteristic (ROC) analysis and reported as the area under the curve (Az). The classifier was optimized by adding additional features until the Az declined. The optimized subset of observers then were combined using a third perceptron. A subset of 80 features was selected which gave an Az of 0.972. Additionally, at 98.6% sensitivity, the classifier had a specificity of 71.3% and increased the positive predictive value from 60.7% to 84.1 %. Preliminary results suggest that using SRHOs in combination with perceptrons can provide a successful classification scheme for pulmonary nodules. This approach could be incorporated into a larger computer aided detection system for decreasing false positives.

Algorithms↗

PubMed1986

Dental and social effects of malocclusion and effectiveness of orthodontic treatment: a strategy for investigation.

Study of the effectiveness of orthodontic treatment is complicated by a variety of factors, including the need for longitudinal, multidisciplinary appraisal and difficulties of randomisation, control and sample composition. This report outlines a recent initiative adopting a longitudinal, observational approach with a cohort of approximately 1000 12-yr-olds, selected by disproportionate stratified sampling in order that occlusal features of low prevalence, but high orthodontic interest, would be well represented. In the event, 4810 subjects were listed and after certain exclusions, 3420 children were screened. Using preselected screening criteria, 663 with specific occlusal features were selected and an additional 355 children with nonspecific features were randomly allocated on a pro-rata basis. This gave a final study population of 1018. Baseline dental data included the recording of caries, periodontal status and mandibular function. Study models were obtained and analysed using a new standardised technique for recording the alignment of individual teeth. Photographs of the face and dentition were collected and a rating system developed to score dental and facial attractiveness. Social-psychological data were recorded via self-rating questionnaires, interviews, parental questionnaires, teacher questionnaires and peer ratings.

Child↗

PubMed1998

Resistance of the human colon carcinoma cell line HCT-8 to methotrexate results in selection of cells with features of enterocytic differentiation.

Results obtained previously with the human colon carcinoma cell line HT-29 have shown that the ability of the cells to develop resistance against methotrexate (MTX) or 5-fluorouracil is restricted to cells committed to differentiate. With the aim of investigating whether this observation is cell type-specific or more general, we have extended our studies to another colon cell line, HCT-8. We have compared HCT-8 parental cells and the MTX-resistant subline HCT8-MTX using transmission electron microscopy and immuno-fluorescence detection of markers of cell polarity and differentiation. Post-confluent parental HCT-8 cells appeared highly heterogeneous and occurred in clusters of piled-up cells in which the majority were unpolarized and undifferentiated, with a minority exhibiting features of enterocyte-like cells. In contrast, HCT8-MTX cells formed domes and appeared as a monolayer of polarized cells with tight junctions and a discrete apical brush border which expressed villin, dipeptidylpeptidase-IV, CEA and the epithelial mucin MUC1. Together, our results suggest that, as in HT-29 cells, induction of resistance to MTX of HCT-8 cells results in the selection of differentiated cell types.

Animals↗

PubMed1995

Feature-specific effects of selective visual attention.

Four experiments were conducted to quantify the effect of performing a foveal discrimination task on sensitivity for a peripheral grating. The observer's primary task was to discriminate either the spatial frequency or orientation of successive foveal Gabor patches. On a third of the trials they also performed a secondary task to detect the presence of a near-threshold grating in the periphery. We find that sensitivity for detection of the peripheral grating depends on the similarity of the spatial frequencies and orientations between the foveal and peripheral stimuli. Importantly, sensitivity is also affected by which feature is being discriminated in the central task. Because the detectability of the peripheral grating is different when different features of the central stimuli are discriminated, we suggest that the effects on sensitivity are due to feature-specific attention and not simply to passive interactions between filters with similar tuning properties.

Attention↗

PubMed2004

Data mining and genetic algorithm based gene/SNP selection.

OBJECTIVE: Genomic studies provide large volumes of data with the number of single nucleotide polymorphisms (SNPs) ranging into thousands. The analysis of SNPs permits determining relationships between genotypic and phenotypic information as well as the identification of SNPs related to a disease. The growing wealth of information and advances in biology call for the development of approaches for discovery of new knowledge. One such area is the identification of gene/SNP patterns impacting cure/drug development for various diseases. METHODS: A new approach for predicting drug effectiveness is presented. The approach is based on data mining and genetic algorithms. A global search mechanism, weighted decision tree, decision-tree-based wrapper, a correlation-based heuristic, and the identification of intersecting feature sets are employed for selecting significant genes. RESULTS: The feature selection approach has resulted in 85% reduction of number of features. The relative increase in cross-validation accuracy and specificity for the significant gene/SNP set was 10% and 3.2%, respectively. CONCLUSION: The feature selection approach was successfully applied to data sets for drug and placebo subjects. The number of features has been significantly reduced while the quality of knowledge was enhanced. The feature set intersection approach provided the most significant genes/SNPs. The results reported in the paper discuss associations among SNPs resulting in patient-specific treatment protocols.

Algorithms↗

PubMed2006

Specificity of molecular interactions in transient protein-protein interaction interfaces.

In this study, we investigate what types of interactions are specific to their biological function, and what types of interactions are persistent regardless of their functional category in transient protein-protein heterocomplexes. This is the first approach to analyze protein-protein interfaces systematically at the molecular interaction level in the context of protein functions. We perform systematic analysis at the molecular interaction level using classification and feature subset selection technique prevalent in the field of pattern recognition. To represent the physicochemical properties of protein-protein interfaces, we design 18 molecular interaction types using canonical and noncanonical interactions. Then, we construct input vector using the frequency of each interaction type in protein-protein interface. We analyze the 131 interfaces of transient protein-protein heterocomplexes in PDB: 33 protease-inhibitors, 52 antibody-antigens, 46 signaling proteins including 4 cyclin dependent kinase and 26 G-protein. Using kNN classification and feature subset selection technique, we show that there are specific interaction types based on their functional category, and such interaction types are conserved through the common binding mechanism, rather than through the sequence or structure conservation. The extracted interaction types are C(alpha)-- H...O==C interaction, cation...anion interaction, amine...amine interaction, and amine...cation interaction. With these four interaction types, we achieve the classification success rate up to 83.2% with leave-one-out cross-validation at k = 15. Of these four interaction types, C(alpha)--H...O==C shows binding specificity for protease-inhibitor complexes, while cation-anion interaction is predominant in signaling complexes. The amine ... amine and amine...cation interaction give a minor contribution to the classification accuracy. When combined with these two interactions, they increase the accuracy by 3.8%. In the case of antibody-antigen complexes, the sign is somewhat ambiguous. From the evolutionary perspective, while protease-inhibitors and sig-naling proteins have optimized their interfaces to suit their biological functions, antibody-antigen interactions are the happenstance, implying that antibody-antigen complexes do not show distinctive interaction types. Persistent interaction types such as pi...pi, amide-carbonyl, and hydroxyl-carbonyl interaction, are also investigated. Analyzing the structural orientations of the pi...pi stacking interactions, we find that herringbone shape is a major configuration in transient protein-protein interfaces. This result is different from that of protein core, where parallel-displaced configurations are the major configuration. We also analyze overall trend of amide-carbonyl and hydroxyl-carbonyl interactions. It is noticeable that nearly 82% of the interfaces have at least one hydroxyl-carbonyl interactions.

Animals↗

PubMedDate not supplied

Visual recognition based on temporal cortex cells: viewer-centred processing of pattern configuration.

A model of recognition is described based on cell properties in the ventral cortical stream of visual processing in the primate brain. At a critical intermediate stage in this system, 'Elaborate' feature sensitive cells respond selectively to visual features in a way that depends on size (+/- 1 octave), orientation (+/- 45 degrees) but does not depend on position within central vision (+/- 5 degrees). These features are simple conjunctions of 2-D elements (e.g. a horizontal dark area above a dark smoothly convex area). They can arise either as elements of an object's surface pattern or as a 3-D component bounded by an object's external contour. By requiring a combination of several such features without regard to their position within the central region of the visual image, 'Pattern' sensitive cells at higher levels can exhibit selectivity for complex configurations that typify objects seen under particular viewing conditions. Given that input features to such Pattern sensitive cells are specified in approximate size and orientation, initial cellular 'representations' of the visual appearance of object type (or object example) are also selective for orientation and size. At this level, sensitivity to object view (+/- 60 degrees) arises because visual features disappear as objects are rotated in perspective. Processing is thus viewer-centred and the neurones only respond to objects seen from particular viewing conditions or 'object instances'. Combined sensitivity to multiple features (conjunctions of elements) independent of their position, establishes selectivity for the configurations of object parts (from one view) because rearranged configurations of the same parts yield images lacking some of the 2-D visual features present in the normal configuration. Different neural populations appear to be selectively tuned to particular components of the same biological object (e.g. face, eyes, hands, legs), perhaps because the independent articulation of these components gives rise to correlated activity in different sets of input visual features. Generalisation over viewing conditions for a given object can be established by hierarchically pooling outputs of view-condition specific cells with pooling operations dependent on the continuity in experience across viewing conditions. Different object parts are seen together and different views are seen in succession when the observer walks around the object. The view specific coding that characterises the selectivity of cells in the temporal lobe can be seen as a natural consequence of selective experience of objects from particular vantage points. View specific coding for the face and body also has great utility in understanding complex social signals, a property that may not be feasible with object-centred processing.

Animals↗

PubMed1998

Visual marking of moving objects: a role for top-down feature-based inhibition in selection.

Recently, the authors presented evidence that new items can be prioritized for selection by the top-down attentional inhibition of old stimuli already in the field (visual marking; D. G. Watson & G. W. Humphreys, 1997). In this article the authors assess whether this inhibition extends to moving old items and test an alternative account of visual marking. Six experiments showed that old moving items could be inhibited provided they did not undergo abrupt property changes. Further, and in contrast to effects with static stimuli, the marking of old moving stimuli was based on inhibition applied at the level of a whole feature map, rather than at their locations. The results also rule out an alternative account of visual marking based on the top-down weighting of dynamic or static processing pathways.

Adolescent↗

PubMed2000

Mechanisms of visual object recognition studied in monkeys.

Cells in area TE of the inferotemporal cortex of the monkey brain selectively respond to various moderately complex object-features, and those responding to similar features cluster in a columnar region elongated vertical to the cortical surface. Although cells within a column respond to similar features, their selectivity is not identical. The data of optical imaging in TE have suggested that the borders between neighboring columns are not discrete but columns representing related features overlap one another. We have also found, by training adult monkeys for discrimination of a specific set of shapes, that such a long-term training increases the proportion of TE cells responding to the shapes used in the training even in the adult. The data suggested that TE plays important roles in discrimination of complex shapes and in visual expert learning of discriminating a certain class of objects in the adult.

Animals↗

PubMedDate not supplied

Identifying cytologic characteristics and grading endocervical columnar cell abnormalities. A study aided by high-definition television.

OBJECTIVE: To test the ability of cytotechnologists to recognize and accurately interpret selected architectural, cellular and nuclear features presented on a high-definition television (HDTV) and to make a reliable diagnosis with HDTV. STUDY DESIGN: A total of 1,122 features considered diagnostic of different endocervical columnar cell abnormalities were selected from 50 smears from 48 women with the help of a motor-driven-stage microscope by five observers who had knowledge of the final diagnosis. The selected and stored features were presented on an HDTV and evaluated in five successive sessions without knowledge of the final diagnosis. RESULTS: Specific types of features were correctly identified in a high number of cases. Considerable interobserver variability was demonstrated in the scoring of grades of expression of features. Overrated and under-rated monitor diagnoses were related to overvalued and undervalued features. From a group of 437 images that were correctly diagnosed by four or five observers, five features proved to be highly related to the correct diagnosis. CONCLUSION: Observers were capable of making a reliable diagnosis on features, selected by other observers, when presented on an HDTV. An overall correct diagnosis was made in 93% of cases.

Adenocarcinoma↗