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 973 records · Page 54Linked to original sources
PubMed2001

Using conjoint analysis to examine the importance of hearing aid attributes.

OBJECTIVE: To introduce conjoint analysis (CA) as a method to explain choice decisions of hearing aid users and to measure the contributions of different hearing aid attributes involved. DESIGN: The study was conducted using CA, a method that is well established in marketing, but has not been used for hearing aid research. CA examines the contributions (importance) various attributes make to overall preference judgements. It is based on the premises that objects (i.e., hearing aids) can be described as different combinations of several attributes. The relative importance of the attributes is estimated from integral judgements of hypothetical hearing aids. The advantages of CA are a more realistic presentation of complete products instead of isolated attributes and the consideration of trade-offs between the features. The following attributes selected from 12 features in a prestudy were included in the application: speech perception in quiet, speech perception in noise, handling, sound quality, localization, and feedback. 93 randomly chosen experienced hearing aid wearers participated in the survey. RESULTS: It was found that the attributes concerning speech perception were by far most important. Speech discrimination in quiet and in noise were almost equally significant and contributed with 56% to the preference decisions of the subjects. The remaining attributes showed values in the range of 10% to 12%. When comparing various subgroups dependent on age, sex, hearing loss, satisfaction, etc., significant differences concerning the preferences were found. Elder and female subjects attached greater importance of the attribute "handling" whereas younger patients judged "speech in noise" more important. In general, subjects with greater hearing loss considered speech intelligibility in quiet to be more important than in noisy surroundings. Moreover, sound quality seems to be a significant indicator for satisfaction with the hearing aid. CONCLUSIONS: This study proposes CA as a survey technique to examine individual preferences for hearing aids. Several assumptions about the importance of hearing aid features for different groups of users could be confirmed quantitatively. Segmentation showed valid results depending on socio-demographic and user-specific parameters. The method used and the results are of interest for audiologists as well as the hearing aid industry. They provide valuable tools for future hearing aid design and should be considered in the development of self-assessment inventories.

Correction of Hearing Impairment↗
PubMed2001

Automated melanoma recognition.

A system for the computerized analysis of images obtained from ELM has been developed to enhance the early recognition of malignant melanoma. As an initial step, the binary mask of the skin lesion is determined by several basic segmentation algorithms together with a fusion strategy. A set of features containing shape and radiometric features as well as local and global parameters is calculated to describe the malignancy of a lesion. Significant features are then selected from this set by application of statistical feature subset selection methods. The final kNN classification delivers a sensitivity of 87% with a specificity of 92%.

Algorithms↗
PubMed1994

Clonotypic dominance and variable gene elements of pathogenic anti-DNA autoantibodies from a single patient with lupus.

This study explores the usage and diversity of the variable gene elements expressed by human lupus antibodies to DNA bearing the 0-81 idiotype, a marker of pathogenic anti-DNA autoantibodies. Rather than studying DNA-specific clonotypes from different patients, a panel of idiotype positive anti-DNA autoantibody-secreting clones from a single individual were analysed. By cloning and nucleotide-sequencing the heavy-chain variable gene segments, evidence was found for dominance of clonotypic patterns. Also noted was a high rate of diversification among the variable (VH), diversity (DH) and junctional (JH) gene segments utilized, with a pattern of mutations indicative of antigenic selection. These features suggest that the clones secreting the lupus pathogenic autoantibodies have been selected over multiple generations through an affinity-maturation process that is reminiscent of antigen-driven immune responses.

Adult↗
PubMed2006

[Fuzzing pattern recognition study on Raman spectrum of tumor peripheral tissue].

On the basis of some theories about fuzzing pattern recognition, the present article studied the data preprocessing of the Raman spectrum of tumor peripheral tissue, and feature extraction and selection. According to these features the authors improved the leaning towards the bigger membership function of trapezoidal distribution. The authors built the membership function of Raman spectrum of tumor peripheral tissue which belongs to malignant tumor on the basis of 40 specimens, and designed the classifier. The test of other 40 specimens showed that the discrimination of malignant tumor is 82.4%, while that of beginning tumor is 73.9%.

Humans↗
PubMed1996

Search images: selective attention to specific visual features of prey.

In 3 experiments, pigeons (Columba livia) searched a digitized image of a gravel patch on a computer monitor for cryptic grains. Experiment 1 demonstrated that the bird's ability to detect a type of grain improved over successive encounters, and detection of alternative targets was attenuated when a search image was active. Experiment 2 demonstrated search-image effects independently for the 2 grains. Perception was biased to detect wheat grains after wheat encounters and was biased to detect beam grains after bean encounters. Experiment 3 demonstrated that when a search image was activated, selective attention was heightened to the visual features of the grain used to discriminate if from the multicolor gravel background. These results provide strong support for the view that search images are selective attention to specific visual features of cryptic prey.

Animals↗
PubMed1991

Selective and divided attention during visual discriminations of shape, color, and speed: functional anatomy by positron emission tomography.

Positron emission tomography (PET) was used to identify the neural systems involved in discriminating the shape, color, and speed of a visual stimulus under conditions of selective and divided attention. Psychophysical evidence indicated that the sensitivity for discriminating subtle stimulus changes in a same-different matching task was higher when subjects selectively attended to one attribute than when they divided attention among the attributes. PET measurements of brain activity indicated that modulations of extrastriate visual activity were primarily produced by task conditions of selective attention. Attention to speed activated a region in the left inferior parietal lobule. Attention to color activated a region in the collateral sulcus and dorsolateral occipital cortex, while attention to shape activated collateral sulcus (similarly to color), fusiform and parahippocampal gyri, and temporal cortex along the superior temporal sulcus. Outside the visual system, selective and divided attention activated nonoverlapping sets of brain regions. Selective conditions activated globus pallidus, caudate nucleus, lateral orbitofrontal cortex, posterior thalamus/colliculus, and insular-premotor regions, while the divided condition activated the anterior cingulate and dorsolateral prefrontal cortex. The results in the visual system demonstrate that selective attention to different features modulates activity in distinct regions of extrastriate cortex that appear to be specialized for processing the selected feature. The disjoint pattern of activations in extravisual brain regions during selective- and divided-attention conditions also suggests that preceptual judgements involve different neural systems, depending on attentional strategies.

Adult↗
PubMed1995

Assessment and selection considerations: ESRD patient and family education materials and media.

Education has long been an integral component of the care of end-stage renal disease patients and their families. The health care environment of the 1990s led to increase in the numbers and acuity of patients and a decrease in professional staff and reimbursement. A basic understanding of the patient education process, assessment of educational materials, adult learning principles, and empowerment is essential for the health care professional responsible for assessing and selecting patient education materials and media. This article provides an overview of the education process, assessment of materials and media (content, scope and depth, readability, clarity and design features), and selection of materials and media based on learner needs and characteristics. The features of four types of materials and media that affect selection and purchasing decisions for the individual patient or for an agency or group of patients are also presented. Special considerations about developing in-house materials are addressed throughout the article.

Family↗
PubMedDate not supplied

Automatic identification of biological microorganisms using three-dimensional complex morphology.

We propose automated identification of microorganisms using three-dimensional (3-D) complex morphology. This 3-D complex morphology pattern includes the complex amplitude (magnitude and phase) of computationally reconstructed holographic images at arbitrary depths. Microscope-based single-exposure on-line (SEOL) digital holography records and reconstructs holographic images of the biological microorganisms. The 3-D automatic recognition is processed by segmentation, feature extraction by Gabor-based wavelets, automatic feature vector selection by graph matching, training rules, and a decision process. Graph matching combined with Gabor feature vectors measures the similarity of complex geometrical shapes between a reference microorganism and unknown biological samples. Automatic selection of the training data is proposed to achieve a fully automatic recognition system. Preliminary experimental results are presented for 3-D image recognition of Sphacelaria alga and Tribonema aequale alga.

Algorithms↗
PubMed2001

Epidemiology of selected mycobacteria that infect humans and other animals.

This paper provides a summary of salient clinical and epidemiological features of selected mycobacterial diseases that are common to humans and other animals. Clinical and diagnostic issues are discussed and related to estimates of the incidence and prevalence of these diseases among humans. Source of infection, route of transmission and control measures are also presented. The mycobacteria discussed in this paper are Mycobacterium bovis, M. ulcerans, M. leprae and M. avium complex, although this is by no means a complete list of the mycobacteria common to humans and other animals. Certain generalities can be made regarding these species of mycobacteria and their occurrence in humans and other animals; firstly, current understanding of the epidemiology and control of many of the resultant diseases is incomplete; secondly, environmental sources other than animal reservoirs may play a role in transmission (with M. leprae perhaps being the exception); and thirdly, the incidence and prevalence of these diseases in many countries of the world are unclear, principally because of the complexity of diagnosis and lack of reporting systems.

Animals↗
PubMed1998

Image analysis and pattern recognition for computer supported skin tumor diagnosis.

A new approach to computer supported recognition of melanoma and naevocytic naevi based on high resolution skin surface profiles is presented. Profiles are generated by sampling an area of 4 x 4 mm2 at a resolution of 125 sample points per mm with a laser profilometer at a vertical resolution of 0.1 micron. With image analysis algorithms Haralick's texture parameters, Fourier features and features based on fractal analysis are extracted. Genetic algorithms are employed successfully to select good feature subsets for the following classification process. As quality measure for feature subsets, the error rate of the nearest neighbor classifier estimated with the leaving-one-out method is used. Classification is performed with feed forward back-propagation network and the nearest neighbor classifier. Classification performance of the neural classifier is optimized using different topologies, learning parameters and pruning algorithms. The best neural classifier achieved an error rate of 4.5% and was found after network pruning. The best result with an error rate of 2.3% was obtained with the nearest neighbor classifier.

Algorithms↗
PubMed2004

Effects of lossy image compression on quantitative image analysis of cell nuclei.

OBJECTIVE: To investigate whether statistically significant changes occur in quantitative image analysis of cell nuclei when lossy image compression techniques are used. STUDY DESIGN: Thirty-five stoichiometric, Feulgen-stained samples of rat hepatocytes, human thyroid and ovarian cancer cell nuclei were used. Image analysis was performed by a computerized system that used AutoCyte LINK V1.1.1.56 software (Burlington, North Carolina, U.S.A.) for image acquisition and Zeiss Vision KS 400 V3.0 software (Oberkochen, Germany) for quantitative image analysis. After lossy JPEG compression of acquired images at different quality levels, some densitometric features were selected and measurements performed. RESULTS: We observed that nearly all the standard densitometric features showed statistically significant changes when images were compressed with the lossy JPEG algorithm. However, most invariant densitometric moment features remained free of statistically significant changes. CONCLUSION: The standard densitometric measurements that we used do not tolerate lossy compression. However, analyses using invariant densitometric features may be performed on images compressed with lossy JPEG, resulting in simpler, less expensive systems demanding less network bandwidth.

Animals↗
PubMedDate not supplied

Microsatellite status and immunohistochemical features of ovarian clear-cell carcinoma.

BACKGROUND: Ovarian clear-cell carcinoma (OCC) is known to have a poor prognosis and selected genetic features of OCC remain unknown. We investigated microsatellite instability (MSI) and the expression of the DNA mismatch repair-related protein, p53. MATERIALS AND METHODS: MSI was examined by polymerase chain reaction using mono-, di-, tri- and tetranucleotide repeat markers, and hMSH2, hMLH1, hMSH6, MSH3 and p53 were determined immunohistochemically in 24 cases of OCC. RESULTS: A total of 9 (37.5%) cases exhibited MSI. Two cases (8.3%) exhibited MSI-H in mononucleotide repeat loci with the negative expression of hMLH1, while another 7 cases (29.2%) exhibited selected trinucloetide repeat MSI (MSI-TR). Of these MSI-TR cases, 4 cases (57.1%) were determined to be negative for MSH3, while hMSH2, hMSH6, MSH3 and p53 expressions were normal. CONCLUSION: Our findings suggest that MSI-TR would be a feature indicating the microsatellite status in OCC, and that the loss of MSH3 expression may promote MSI-TR.

Adaptor Proteins, Signal Transducing↗
PubMed2005

Learning best features for deformable registration of MR brains.

This paper presents a learning method to select best geometric features for deformable brain registration. Best geometric features are selected for each brain location, and used to reduce the ambiguity in image matching during the deformable registration. Best geometric features are obtained by solving an energy minimization problem that requires the features of corresponding points in the training samples to be similar, and the features of a point to be different from those of nearby points. By incorporating those learned best features into the framework of HAMMER registration algorithm, we achieved about 10% improvement of accuracy in estimating the simulated deformation fields, compared to that obtained by HAMMER. Also, on real MR brain images, we found visible improvement of registration in cortical regions.

Algorithms↗
PubMed1995

Endoscopic features of Helicobacter pylori-related gastritis.

BACKGROUND AND STUDY AIMS: It is still controversial whether certain endoscopic features can be used to diagnose Helicobacter pylori (Hp)-related gastritis. To clarify this issue, we performed two consecutive prospective studies. PATIENTS AND METHODS: In the first study, we tried to identify endoscopic criteria associated with Hp-related gastritis from a total of 66 predefined gastric features on endoscopy. These features were selected by a stepwise logistic regression analysis in 101 patients. The validity of these features gained from this first study was then evaluated in a second study in 86 patients (40 with Hp gastritis, 8 with Hp-unrelated gastritis and 38 with normal gastric mucosa). RESULTS: Three features, namely an abnormal antral surface texture, a mammillated corpus surface, and white antral erosions, were identified in the first study as independent predictors of Hp-related gastritis. However, the sensitivity and specificity of these three criteria, as assessed in the second study, were only 75% and 63% respectively. CONCLUSIONS: We conclude that it is not possible to diagnose Hp-related gastritis on the basis of the endoscopic appearance alone. The diagnosis should be based on other criteria, such as a rapid urease test, or a histological examination of gastric biopsies, or both.

Adult↗
PubMed1990

Quantification of histologic features of dysplastic nevi.

We quantified the histologic features of 100 consecutive dysplastic nevi. Although there is heterogeneity in the microscopic attributes of dysplastic nevi, certain features that are seen repeatedly should enable the histologic identification of these melanocytic neoplasms. These features include a central dermal nevocytic component with a peripheral extension of a junctional component, elongated epidermal rete ridges, bridging of nests of melanocytes at the dermo-epidermal junction, nests of melanocytes at the sides of rete ridges as well as at their bases, and concentric eosinophilic fibrosis. Cytologic features of melanocytes were not as valuable as architectural features of individual nevi in classifying these lesions. We also provide an atlas of selected histologic features that depicts some aspects of this heterogeneity.

Adolescent↗
PubMed1998

Quantitative characterization of texture used for identification of eggs of bovine parasitic nematodes.

This study investigates the use of texture, i.e. the grey level variation in digital images, as a basis for identification of strongylid eggs. Texture features were defined by algorithms applied to digital images of eggs from the bovine parasitic nematodes, Ostertagia ostertagi, Cooperia oncophora, and Oesophagostomum radiatum. The resulting data served to establish classification criteria by linear discrimination analysis, and the criteria were subsequently evaluated by cross-validations. From 25 texture features, ten features were selected by their significant discriminatory powers. Using a classification criterion based on these ten texture features, an average of 91.2% of eggs from the three species were correctly classified. All O. radiatum eggs were correctly classified, 11.8% of O. ostertagi and C. oncophora were reciprocally misclassified, and 2.9% of O. ostertagi were identified as O. radiatum. When the ten texture features were used singly an average of 51.2 to 37.9% of the species could be classified correctly. When texture was used together with the shape and size features, a higher percentage of eggs were correctly classified compared with the classification based on either texture, or shape and size. Hence, all O. radiatum were correctly classified as well as 88.3% of O. ostertagi and 91.2% of C. oncophora, resulting in an average of 93.1% correctly classified eggs. The rapid and accurate measurements of texture features may serve as a basis for identification or enhance performance of classification criteria based on egg shape/size.

Animals↗
PubMedDate not supplied

Selective IgA deficiency with recurrent vasculitis of the central nervous system.

We describe an unusual case of selective IgA deficiency complicated by recurrent vasculitis of the central nervous system (CNS). The patient suffered from two episodes of CNS vasculitis, one of which was located in the cerebrum and the other in the cerebellum. The vasculitic process resulted in brain tumor-like lesions shown by computed tomography. There was no evidence of associated connective tissue diseases. Vasculitis in other organs or tissues was not noted. This is the first detailed description in the English literature of pathologically proven CNS vasculitis in a patient with selective IgA deficiency. Our report demonstrates that isolated CNS angiitis can be a rare clinical feature of selective IgA deficiency.

Adult↗
PubMed1995

Selective IgA deficiency with unusual features: development of common variable immunodeficiency, Sjögren's syndrome, autoimmune hemolytic anemia and immune thrombocytopenic purpura.

We report on a girl with selective IgA deficiency and persistently low complement component 4 (C4) levels compatible with heterozygous C4 deficiency. Deterioration of her serum immunoglobulin levels and transition to common variable immunodeficiency were observed within a 5 year follow-up. She also developed Sjögren's syndrome, autoimmune hemolytic anemia and immune thrombocytopenic purpura. While these abnormalities have been described before in various combinations, to our knowledge, they have not been reported in a single individual.

Anemia, Hemolytic, Autoimmune↗