Search PubMed⌕ Search

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.

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

At least 1,783 records · Page 99Linked to original sources
PubMed2006

Lidar detection of underwater objects using a neuro-SVM-based architecture.

This paper presents a neural network architecture using a support vector machine (SVM) as an inference engine (IE) for classification of light detection and ranging (Lidar) data. Lidar data gives a sequence of laser backscatter intensities obtained from laser shots generated from an airborne object at various altitudes above the earth surface. Lidar data is pre-filtered to remove high frequency noise. As the Lidar shots are taken from above the earth surface, it has some air backscatter information, which is of no importance for detecting underwater objects. Because of these, the air backscatter information is eliminated from the data and a segment of this data is subsequently selected to extract features for classification. This is then encoded using linear predictive coding (LPC) and polynomial approximation. The coefficients thus generated are used as inputs to the two branches of a parallel neural architecture. The decisions obtained from the two branches are vector multiplied and the result is fed to an SVM-based IE that presents the final inference. Two parallel neural architectures using multilayer perception (MLP) and hybrid radial basis function (HRBF) are considered in this paper. The proposed structure fits the Lidar data classification task well due to the inherent classification efficiency of neural networks and accurate decision-making capability of SVM. A Bayesian classifier and a quadratic classifier were considered for the Lidar data classification task but they failed to offer high prediction accuracy. Furthermore, a single-layered artificial neural network (ANN) classifier was also considered and it failed to offer good accuracy. The parallel ANN architecture proposed in this paper offers high prediction accuracy (98.9%) and is found to be the most suitable architecture for the proposed task of Lidar data classification.

Algorithms↗
PubMed2003

The Wadsworth Center brain-computer interface (BCI) research and development program.

Brain-computer interface (BCI) research at the Wadsworth Center has focused primarily on using electroencephalogram (EEG) rhythms recorded from the scalp over sensorimotor cortex to control cursor movement in one or two dimensions. Recent and current studies seek to improve the speed and accuracy of this control by improving the selection of signal features and their translation into device commands, by incorporating additional signal features, and by optimizing the adaptive interaction between the user and system. In addition, to facilitate the evaluation, comparison, and combination of alternative BCI methods, we have developed a general-purpose BCI system called BCI-2000 and have made it available to other research groups. Finally, in collaboration with several other groups, we are developing simple BCI applications and are testing their practicality and long-term value for people with severe motor disabilities.

Academic Medical Centers↗
PubMed2003

Improving care for minorities: can quality improvement interventions improve care and outcomes for depressed minorities? Results of a randomized, controlled trial.

OBJECTIVE: Ethnic minority patients often receive poorer quality care and have worse outcomes than white patients, yet practice-based approaches to reduce such disparities have not been identified. We determined whether practice-initiated quality improvement (QI) interventions for depressed primary care patients improve care across ethnic groups and reduce outcome disparities. STUDY SETTING: The sample consists of 46 primary care practices in 6 U.S. managed care organizations; 181 clinicians; 398 Latinos, 93 African Americans, and 778 white patients with probable depressive disorder. STUDY DEIGN: Matched practices were randomized to usual care or one of two QI programs that trained local experts to educate clinicians; nurses to educate, assess, and follow-up with patients; and psychotherapists to conduct Cognitive Behavioral Therapy. Patients and physicians selected treatments. Interventions featured modest accommodations for minority patients (e.g., translations, cultural training for clinicians). DATA EXTRACTION METHODS: Multilevel logistic regression analyses assessed intervention effects within and among ethnic groups. PRINCIPAL FINDINGS: At baseline, all ethnic groups Latino, African American, white) had low to moderate rates of appropriate care and the interventions significantly improved appropriate care at six months (by 8-20 percentage points) within each ethnic group, with no significant difference in response by ethnic group. The interventions significantly decreased the likelihood that Latinos and African Americans would report probable depression at months 6 and 12; the white intervention sample did not differ from controls in reported probable depression at either follow-up. While the intervention significantly improved the rate of employment for whites and not for minorities, precision was low for comparing intervention response on this outcome. It is important to note that minorities remained less likely to have appropriate care and more likely to be depressed than white patients. CONCLUSIONS: Implementation of quality improvement interventions that have modest accommodations for minority patients can improve quality of care for whites and underserved minorities alike, while minorities may be especially likely to benefit clinically. Further research needs to clarify whether employment benefits are limited to whites and if so, whether this represents a difference in opportunities. Quality improvement programs appear to improve quality of care without increasing disparities, and may offer an approach to reduce health disparities.

Black or African American↗
PubMed1983

Structural conservation in globular proteins.

Some well sequenced classes of homologous proteins have been analysed in the light of their representative tertiary structures to reveal the nature of structural conservation during protein evolution. Within 'the sphere of influence' around conserved residue sites preferential association of other conservative sites has been observed to be a feature of natural selection valid for all residue types. An information theory approach evolved to examine residue variability reveals that even some non-conservative sites scaffold clusters of high biochemical specificity and intermediate folding units.

Amino Acid Sequence↗
PubMed1980

Comparative studies of squirrel monkeys (Saimiri) and titi monkeys (Callicebus) in travel tasks.

Squirrel and titi monkeys were observed in a series of experiments in which the subjects' task was to move to a distant goal along above-ground pathways. The pathways were entirely visible to the subjects in all experiments. However, visual cues along the pathways (in Experiment I) and physical and spatial properties of the pathways (in Experiments II and III) were varied systematically in order to determine what effect features had upon selection of travel paths for monkeys of each species. Marked performance differences between the species were evident in these experiments, including differences in latency to move past the choice point, proportion of trials in which the shortest route was chosen first, and changes over test sessions in frequency of initial choice of the shortest route. In particular, titis tended to move past the choice point more slowly than squirrel monkeys; to pay more attention to distant properties of the pathways prior to making a decision, especially after experience in the test setting; and to prefer habitual pathways when these were available, whereas squirrel monkeys preferred novel routes when these were available. The relative "optimality" of decision making in these tasks in relation to species-typical modes of traveling and foraging in natural habitats is discussed. An alternative view of decision making, in which optimality is not assumed to be the only decision-making strategy, is suggested as an appropriate vehicle for further investigation into the sources of short-term variability in choice behavior.

Animals↗
PubMed1998

Quantitative microscopic anatomy, illustrated by its potential role in furthering understanding of the processes of structuring the developing human cerebral cortex.

In this study, we searched for patterns in selected, quantitative microscopic features of the developing postnatal human cortex for 35 cytoarchitectural areas at eight age points from birth to 72 months. These data come from the largest extant survey of the microscopic features of the developing postnatal human cerebral cortex (JL Conel, 1939-67). In contrast to Jacobson's proposal that cortical surface area increases in proportion to brain weight, Conel's data show that surface area increases as brain weight2/3, with the maximal rate of increase in both brain weight and cortical surface area occurring from 1 to 3 months. We computed the numbers of cortical neurons per cortical layer under 1 mm2 of cortical surface (neurons/layer per mm2 column) and divided these values by the total neuron number/mm2 column. For all areas, these data show plateaux in most of the layers for periods of months to years, often preceded and followed by changes in their neuron number in a sinusoidal fashion. The age points of the maxima and minima of such laminar values differ across the 35 cortical areas, indicating that their sinusoids are phase-shifted with respect to each other. Ranking the six layers in each area at each age point by their neuron number/layer per mm2 column shows that, by 72 months, the first areas to receive thalamic auditory or visual input (primary sensory and unimodal association cortices) have the most neurons in layer 4 and in either layer 3 or 6. In contrast, by 72 months, other areas have the most neurons in layers 3 and 6, with the primary motor cortex reaching this ranking earlier than any other area. For temporal and parietal association areas, layers 2 (short cortico-cortical connections) and 4 (thalamo-cortical connections) have the most neurons from birth to 6 months, whereas layers 3 (long cortico-cortical connections) and 6 (cortico-thalamic connections) have the most neurons by 72 months. The quantitative, statistically non-random patterns demonstrated by our analyses suggest that hierarchical correlations between such structural changes and age-specific behavioral acquisitions exist during the first 72 months of postnatal development.

Cerebral Cortex↗
PubMed1994

The definition of pre-eclampsia.

BACKGROUND: Redman and Jefferies have proposed a revised definition of pre-eclampsia which is based on absolute blood pressure levels and an increment from the baseline in the first half of pregnancy. There is no requirement for proteinuria. This definition should facilitate the distinction between nonproteinuric pre-eclampsia and other causes of nonproteinuric gestational hypertension, such as chronic essential hypertension. OBJECTIVES: 1. To determine whether the blood pressure criteria of Redman and Jefferies can select women with characteristics of pre-eclampsia from the group of women with gestational hypertension by the current criteria of the International Society for the Study of Hypertension in Pregnancy (ISSHP). 2. To determine the level of agreement between the classification system proposed by Redman and Jefferies and that of the ISSHP. DESIGN: A prospective study. SETTING: Obstetric unit, Dudley Road Hospital, Birmingham, UK. SUBJECTS: Six hundred and ninety-two healthy nulliparous women and 11 women with chronic hypertension antedating pregnancy. MAIN OUTCOME MEASURES: 1. Differences in maternal characteristics and obstetric outcome among women with gestational (nonproteinuric) hypertension by the ISSHP criteria, meeting (and failing to meet) the Redman and Jefferies' blood pressure criteria. 2. The proportion of women classified as normal, proteinuric pre-eclampsia, and chronic hypertension on the basis of both Redman and Jefferies' criteria and the current ISSHP criteria. RESULTS: There were 55 women with gestational hypertension alone by the ISSHP criteria, of whom 33 met Redman and Jefferies' blood pressure criteria for pre-eclampsia. This group of 33 women had characteristics of nonproteinuric pre-eclampsia, compared with the remaining 22 women in the ISSHP gestational hypertension category who had characteristics of chronic hypertension. The group of 33 were significantly younger and less obese, had significantly lower blood pressure at their first antenatal visit and their obstetric outcome was poorer. The Redman and Jefferies' blood pressure criteria identified as normal 99.5% (95% CI, 98.6% to 99.9%) of women who were also characterised as normal on the basis of the ISSHP criteria (622/625). There were 12 women with proteinuric pre-eclampsia by the ISSHP criteria of whom 11 (92%; 95% CI, 62% to 99.8%) met Redman and Jefferies' blood pressure criteria for pre-eclampsia. None of the 11 women with chronic hypertension antedating pregnancy met these criteria. CONCLUSIONS: In this population the blood pressure criteria for pre-eclampsia proposed by Redman and Jefferies select women with features of pre-eclampsia (i.e., proteinuria and relatively poor outcome) and, in particular, they enable a distinction to be made between nonproteinuric pre-eclampsia and other causes of gestational hypertension.

Adult↗
PubMed1991

Opioid activity of dermenkephalin analogues in the guinea-pig myenteric plexus and the hamster vas deferens.

1. To elucidate the structural features required for selective and potent action of dermenkephalin at the delta-opioid receptor, a series of analogues of dermenkephalin and dermorphin were tested for their effectiveness in depressing electrically-evoked contractions of the vas deferens of the hamster (delta-opioid receptors) and the guinea-pig myenteric plexus-longitudinal muscle preparation (mu- and kappa-opioid receptors). 2. Dermenkephalin was more selective and more potent at delta-receptors than the delta-ligand [D-Pen2, D-Pen5]-enkephalin. The responses to dermenkephalin in the hamster vas deferens were increased by addition of peptidase inhibitors; the maximum effect was obtained with 3 microM thiorphan. 3. [L-Met2]-dermenkephalin had 0.2% and [L-Ala2]-dermorphin 0.01% of the agonist activity of the corresponding endogenous peptides which have D-amino acids in position 2. The pharmacological activity of these analogues was unaffected by inhibition of peptidases. This emphasizes the role that the D-configuration plays in determining the bioactive folding of these highly active peptides. 4. Dermenkephalin-(1-6)-NH2 was more potent at delta-receptors than at mu-receptors whereas, dermenkephalin-(1-4)-NH2 is a selective mu-agonist, having no activity at delta-receptors. 5. Substitution of the C-terminal tripeptide of dermorphin with the C-terminal tripeptide of dermenkephalin abolished the mu-receptor preference of dermorphin. The resulting hybrid peptide, Tyr-D-Ala-Phe-Gly-Leu-Met-Asp-NH2 was as potent as dermenkephalin at delta-receptors. A shift towards a preference for delta-receptors was obtained when the C-terminal tetrapeptide of dermorphin was replaced by the C-terminal tetrapeptide of dermenkephalin. 6. Substitution of Asp by Asn in position 7 of dermenkephalin caused an increase in mu-receptor potency and a decrease in delta-receptor potency, resulting in a 20 fold decrease in mu-receptor selectivity. Dermenkephalin-(1-6)-NH2 and [Asn7]-dermenkephalin have almost identical delta-receptor agonist potencies and ratios of IC50 in the myenteric plexus to IC50 in the hamster vas deferens. 7. The results obtained emphasise the importance of a negative charge at the C-terminus of dermenkephalin for selectivity at the delta-opioid receptor. Furthermore, the hydrophobic residues Leu5 and Met6 may be critical in ensuring tight binding to the receptor which results in high agonist potency.

Amino Acid Sequence↗
PubMed2006

A hybrid model for nonignorable dropout in longitudinal binary responses.

This article presents a likelihood-based method for handling nonignorable dropout in longitudinal studies with binary responses. The methodology developed is appropriate when the target of inference is the marginal distribution of the response at each occasion and its dependence on covariates. A "hybrid" model is formulated, which is designed to retain advantageous features of the selection and pattern-mixture model approaches. This formulation accommodates a variety of assumed forms of nonignorable dropout, while maintaining transparency of the constraints required for identifying the overall model. Once appropriate identifying constraints have been imposed, likelihood-based estimation is conducted via the EM algorithm. The article concludes by applying the approach to data from a randomized clinical trial comparing two doses of a contraceptive.

Algorithms↗
PubMed1994

Differentiation between malignant melanomas and benign melanocytic nevi by computerized DNA cytometry of imprint specimens.

Recently image analysis (IA) and DNA-cytophotometry (CP) have proved to be useful for the differentiation between benign and malignant melanocytic lesions on paraffin sections. Since, on sections, these procedures are very time-consuming, we tested in the present study whether IA of imprint specimens, which can be evaluated in less than 30 minutes, might also be sufficient. In 39 malignant melanomas (MM), 18 melanocytic nevi (MN), and 6 dysplastic nevi (DN), 12 different morphometric and DNA cytometric features were determined in 100 randomly selected nuclei. In univariate analysis, 5 features were found to be significantly different between the benign and malignant groups (p < 0.0001): mean value (MAREA) and standard deviation (SAREA) of nuclear area and the 80th, 90th, and 95th percentiles of DNA distribution. Using SAREA, the best univariate feature, 82.5% of the cases could be correctly separated. In multivariate analysis with a combination of three features--standard deviation of nuclear area (SAREA), mean DNA value (MDNA), and 95th percentile of DNA distribution (PERC95)--a correct diagnosis was achieved in 89.5% of the cases. Results obtained in the cases of DN indicated an increased proliferation, but did not allow the separation of DN from MM and MN. Since our technique allows a rapid analysis without loss of tissue, which might be important for histological analysis, and the classification rates are equal or still higher than reported in studies on sections, imprints of melanocytic lesions seem to be most appropriate for the calculation of DNA cytometric features as helpful diagnostic criteria in equivocal melanocytic lesions.

Adult↗
PubMed2006

[Ehlers-Danlos syndrome--20 years experience with diagnosis and classification at the university skin clinic of Heidelberg].

BACKGROUND: The Ehlers-Danlos syndrome encompasses a group of hereditary disorders of the connective tissue, characterized by hyperextensible skin, joint hypermobility; and varying degrees of vessel and tissue fragility. The main forms are classical, hypermobile, vascular, kyphoscoliotic A/B, arthrochalasis A/B and dermatosparaxis types. PATIENTS AND METHODS: We report our experience in diagnosis and classification of Ehlers-Danlos-syndrome, especially with the combination of clinical and morphological criteria, at the Department of Dermatology of the University of Heidelberg with more than 600 patients between 1984 and 2004. RESULTS: We classified those types of EDS which are characterized by regular and characteristic ultrastructural changes in the dermal components, primarily collagen, including the classical, hypermobile and vascular types as well as the less-common arthrochalasis and dermatosparaxis types. The combination of clinical and morphologic features facilitates the selection of candidate genes for molecular genetic investigation. CONCLUSIONS: Besides the skin, skeleton and vessels, many other organ systems such as eyes and intestine, can be affected in Ehlers-Danlos syndrome. Accordingly, interdisciplinary cooperation (pediatrics, surgery, orthopedics, rheumatology, neurology, genetics) is necessary. As the connective tissue of the skin is accessible for biopsy and diagnostic investigation, dermatologists should be trained in the diagnostic approach and classification of this syndrome.

Academic Medical Centers↗
PubMed1993

Structure of the retinoid X receptor alpha DNA binding domain: a helix required for homodimeric DNA binding.

The three-dimensional solution structure of the DNA binding domain (DBD) of the retinoid X receptor alpha (RXR alpha) was determined by nuclear magnetic resonance spectroscopy. The two zinc fingers of the RXR DBD fold to form a single structural domain that consists of two perpendicularly oriented helices and that resembles the corresponding regions of the glucocorticoid and estrogen receptors (GR and ER, respectively). However, in contrast to the DBDs of the GR and ER, the RXR DBD contains an additional helix immediately after the second zinc finger. This third helix mediates both protein-protein and protein-DNA interactions required for cooperative, dimeric binding of the RXR DBD to DNA. Identification of the third helix in the RXR DBD thus defines a structural feature required for selective dimerization of the RXR on hormone response elements composed of half-sites (5'-AGGTCA-3') arranged as tandem repeats.

Amino Acid Sequence↗
PubMed2004

ECGSIM: an interactive tool for studying the genesis of QRST waveforms.

BACKGROUND: Discussion about the selection of diagnostic features of the ECG and their possible interpretation would benefit from a model of the genesis of these signals that has a sound basis in electrophysiology as well as in physics. Recent advances in computer technology have made it possible to build a simulation package whereby the genesis of ECG signals can be studied interactively. DESIGN: A numerical method was developed for computing ECG signals on the thorax, as well as electrograms on both endocardium and epicardium. The source representation of the myocardial electric activity is the equivalent double layer. The transfer factors between the electric sources and the resulting potentials on the heart surface as well as on the body surface were computed using a realistic thorax model. RESULTS AND CONCLUSION: The resulting transfer factors were implemented in a simulation program. The program allows the user to make interactive changes in the timing of depolarisation and repolarisation on the ventricular surface, as well as changing the local source strength, and to inspect or document the effect of such changes instantaneously on electrograms and body surface potentials, visualised by waveforms as well as by potential maps and movies. The entire simulation package can be installed free of charge from www.ecgsim.org.

Computer Simulation↗
PubMed1995

Clinicoradiological and pathological correlations in patients with solitary cysticercus granuloma and epilepsy: focus on presence of the parasite and oedema formation.

A study of the clinical, radiological, and pathological correlations in 43 patients with solitary cysticercus granuloma and epilepsy focused on factors that might help in predicting the presence of the parasite in the granuloma and those that might influence the formation of oedema around the granuloma. The duration of symptoms (< six months and > or = six months) and CT morphology of the granuloma (ring and disc, type A; nodular lesion, type B) were studied as factors that could possibly predict the presence of the parasite in the granuloma. The influence of sex of the patient and the presence of a neutrophilic response in the granuloma on the intensity of oedema around the lesion as seen on CT was also studied. The pathological features were studied in the excised granulomas. The intact or degenerated form of the cysticercus was evident in 22 of 43 specimens. Neither the duration of seizures (P = 0.17) nor the type of lesion on CT (P = 0.16) was predictive of the presence of the parasite in the granuloma. The sex of the patient (P = 0.51) and the neutrophilic response in the specimen (P = 0.73) did not correlate with the degree of oedema on CT indicating that neither of these host factors was a major determinant of oedema production. The findings point to the varied and unpredictable natural history of solitary cysticercus granulomas and the complex nature of host-parasite interactions in individual patients. The inability to predict the presence of the parasite in the granuloma on the basis of the clinical or radiological features precludes a selection of patients with such lesions for cysticidal drug treatment.

Adolescent↗
PubMed1984

Quantitative evaluation of a transport-enrichment medium for Campylobacter fetus.

The enrichment feature of a selective serum-based transport medium for Campylobacter fetus was quantitatively examined. Preputial samples from artificial insemination bulls were spiked with known numbers of C fetus strains and inoculated into transport-enrichment medium (TEM). The survival and multiplication of these strains in TEM under different incubation periods and temperatures were assessed by plate counts. Mean enrichment values of 3.72 log and 4.42 log were observed after incubation at 37 degrees C for two and four days, respectively. There was no significant difference in the enrichment values between the C fetus subspecies venerealis strains and a C fetus subspecies fetus strain. Incubation of inoculated TEM vials at room temperature for up to two days neither improved the growth of C fetus nor affected its subsequent enrichment when the vials were reincubated at 37 degrees C. Comparison of the survival of C fetus with and without the use of TEM under simulated transport conditions demonstrated the superiority of TEM.

Animals↗
PubMed2001

Constructing a fuzzy rule-based system using the ILFN network and Genetic Algorithm.

In this paper, a method for automatic construction of a fuzzy rule-based system from numerical data using the Incremental Learning Fuzzy Neural (ILFN) network and the Genetic Algorithm is presented. The ILFN network was developed for pattern classification applications. The ILFN network, which employed fuzzy sets and neural network theory, equips with a fast, one-pass, on-line, and incremental learning algorithm. After trained, the ILFN network stored numerical knowledge in hidden units, which can then be directly interpreted into if then rule bases. However, the rules extracted from the ILFN network are not in an optimized fuzzy linguistic form. In this paper, a knowledge base for fuzzy expert system is extracted from the hidden units of the ILFN classifier. A genetic algorithm is then invoked, in an iterative manner, to reduce number of rules and select only discriminate features from input patterns needed to provide a fuzzy rule-based system. Three computer simulations using a simulated 2-D 3-class data, the well-known Fisher's Iris data set, and the Wisconsin breast cancer data set were performed. The fuzzy rule-based system derived from the proposed method achieved 100% and 97.33% correct classification on the 75 patterns for training set and 75 patterns for test set, respectively. For the Wisconsin breast cancer data set, using 400 patterns for training and 299 patterns for testing, the derived fuzzy rule-based system achieved 99.5% and 98.33% correct classification on the training set and the test set, respectively.

Algorithms↗
PubMed2001

High order statistics for image classification.

This paper addresses the problem of image classification using local information which is aggregated to provide global representation of different image classes. Local information is adaptively extracted from an image database using Independent Component Analysis (ICA) which provides a set of localized, oriented, and band-pass filters selective to independent features of the images. Local representation using ICA techniques has been previously investigated by several researchers. However, very little work has been done on further use of these representations to provide more complex and global description of images. In this paper, we present an algorithm which uses the energy of a minimal set of ICA filters to provide class-specific signatures which are shown to be strongly discriminant. Computer simulations are carried on two image databases, one consisting of five classes--referred to as categories--(buildings, rooms, mountains, forests and beaches) and one consisting of a set of 30 objects from multiple views for viewpoint invariant object recognition. The classification performance of the algorithm using both Independent and Principal Component Analyses are reported and discussed.

Image Processing, Computer-Assisted↗
PubMed2006

The prediction of bacterial transcription start sites using SVMs.

Identifying promoters is the key to understanding gene expression in bacteria. Promoters lie in tightly constrained positions relative to the transcription start site (TSS). In this paper, we address the problem of predicting transcription start sites in Escherichia coli. Knowing the TSS position, one can then predict the promoter position to within a few base pairs, and vice versa. The accepted method for promoter prediction is to use a pair of position weight matrices (PWMs), which define conserved motifs at the sigma-factor binding site. However this method is known to result in a large number of false positive predictions, thereby limiting its usefulness to the experimental biologist. We adopt an alternative approach based on the Support Vector Machine (SVM) using a modified mismatch spectrum kernel. Our modifications involve tagging the motifs with their location, and selectively pruning the feature set. We quantify the performance of several SVM models and a PWM model using a performance metric of area under the detection-error tradeoff (DET) curve. SVM models are shown to outperform the PWM on a biologically realistic TSS prediction task. We also describe a more broadly applicable peak scoring technique which reduces the number of false positive predictions, greatly enhancing the utility of our results.

Artificial Intelligence↗