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 721 records · Page 40Linked to original sources

PubMed2006

Local dimensionality reduction and supervised learning within natural clusters for biomedical data analysis.

Inductive learning systems were successfully applied in a number of medical domains. Nevertheless, the effective use of these systems often requires data preprocessing before applying a learning algorithm. This is especially important for multidimensional heterogeneous data presented by a large number of features of different types. Dimensionality reduction (DR) is one commonly applied approach. The goal of this paper is to study the impact of natural clustering--clustering according to expert domain knowledge--on DR for supervised learning (SL) in the area of antibiotic resistance. We compare several data-mining strategies that apply DR by means of feature extraction or feature selection with subsequent SL on microbiological data. The results of our study show that local DR within natural clusters may result in better representation for SL in comparison with the global DR on the whole data.

Algorithms↗

PubMed2004

Ridge-based vessel segmentation in color images of the retina.

A method is presented for automated segmentation of vessels in two-dimensional color images of the retina. This method can be used in computer analyses of retinal images, e.g., in automated screening for diabetic retinopathy. The system is based on extraction of image ridges, which coincide approximately with vessel centerlines. The ridges are used to compose primitives in the form of line elements. With the line elements an image is partitioned into patches by assigning each image pixel to the closest line element. Every line element constitutes a local coordinate frame for its corresponding patch. For every pixel, feature vectors are computed that make use of properties of the patches and the line elements. The feature vectors are classified using a kappaNN-classifier and sequential forward feature selection. The algorithm was tested on a database consisting of 40 manually labeled images. The method achieves an area under the receiver operating characteristic curve of 0.952. The method is compared with two recently published rule-based methods of Hoover et al. and Jiang et al. The results show that our method is significantly better than the two rule-based methods (p < 0.01). The accuracy of our method is 0.944 versus 0.947 for a second observer.

Algorithms↗

PubMed2004

Machine printed text and handwriting identification in noisy document images.

In this paper, we address the problem of the identification of text in noisy document images. We are especially focused on segmenting and identifying between handwriting and machine printed text because: 1) Handwriting in a document often indicates corrections, additions, or other supplemental information that should be treated differently from the main content and 2) the segmentation and recognition techniques requested for machine printed and handwritten text are significantly different. A novel aspect of our approach is that we treat noise as a separate class and model noise based on selected features. Trained Fisher classifiers are used to identify machine printed text and handwriting from noise and we further exploit context to refine the classification. A Markov Random Field-based (MRF) approach is used to model the geometrical structure of the printed text, handwriting, and noise to rectify misclassifications. Experimental results show that our approach is robust and can significantly improve page segmentation in noisy document collections.

Algorithms↗

PubMed2006

Shapeme histogram projection and matching for partial object recognition.

Histograms of shape signature or prototypical shapes, called shapemes, have been used effectively in previous work for 2D/3D shape matching and recognition. We extend the idea of shapeme histogram to recognize partially observed query objects from a database of complete model objects. We propose representing each model object as a collection of shapeme histograms and match the query histogram to this representation in two steps: 1) compute a constrained projection of the query histogram onto the subspace spanned by all the shapeme histograms of the model and 2) compute a match measure between the query histogram and the projection. The first step is formulated as a constrained optimization problem that is solved by a sampling algorithm. The second step is formulated under a Bayesian framework, where an implicit feature selection process is conducted to improve the discrimination capability of shapeme histograms. Results of matching partially viewed range objects with a 243 model database demonstrate better performance than the original shapeme histogram matching algorithm and other approaches.

Algorithms↗

PubMed2005

An incremental approach to genetic-algorithms-based classification.

Incremental learning has been widely addressed in the machine learning literature to cope with learning tasks where the learning environment is ever changing or training samples become available over time. However, most research work explores incremental learning with statistical algorithms or neural networks, rather than evolutionary algorithms. The work in this paper employs genetic algorithms (GAs) as basic learning algorithms for incremental learning within one or more classifier agents in a multiagent environment. Four new approaches with different initialization schemes are proposed. They keep the old solutions and use an "integration" operation to integrate them with new elements to accommodate new attributes, while biased mutation and crossover operations are adopted to further evolve a reinforced solution. The simulation results on benchmark classification data sets show that the proposed approaches can deal with the arrival of new input attributes and integrate them with the original input space. It is also shown that the proposed approaches can be successfully used for incremental learning and improve classification rates as compared to the retraining GA. Possible applications for continuous incremental training and feature selection are also discussed.

Algorithms↗

PubMed2005

An empirical comparison of combinations of evolutionary algorithms and neural networks for classification problems.

There are numerous combinations of neural networks (NNs) and evolutionary algorithms (EAs) used in classification problems. EAs have been used to train the networks, design their architecture, and select feature subsets. However, most of these combinations have been tested on only a few data sets and many comparisons are done inappropriately measuring the performance on training data or without using proper statistical tests to support the conclusions. This paper presents an empirical evaluation of eight combinations of EAs and NNs on 15 public-domain and artificial data sets. Our objective is to identify the methods that consistently produce accurate classifiers that generalize well. In most cases, the combinations of EAs and NNs perform equally well on the data sets we tried and were not more accurate than hand-designed neural networks trained with simple backpropagation.

Algorithms↗

PubMedDate not supplied

Streamline predicates.

Predicates are functions that return Boolean values. They are an essential tool in computer science. A close look at flow feature definitions reveals that they can be seen as point predicates that tell if a specific feature exists at a certain point. Besides the information about features, scientists and engineers like to know the overall behavior of all streamlines in the flow, typically in the connection with the important features in their application domain. We call this a structure definition for the flow. A successful example for a structure definition is flow topology. In this paper, we present streamline predicates as functions that tell the user about the connection between streamlines and features selected by the user. This means answers to questions like: Which streamlines flow through a given vortex, separation bubble, or shock wave? It can be shown that streamline predicates may refine flow topology so that it also reveals questions about vortices in 3D.

Algorithms↗

PubMed2026

Large-Scale Plasma Proteomics Identifies Early Molecular Deviations and Improves Risk Prediction for Heart Failure Among Individuals With Obesity.

AIMS: Heart failure (HF) is a major global public health challenge, with obesity being one of its key risk factors. Although several HF risk prediction models have been developed in the general population, few are specifically tailored to individuals with obesity. This underscores the urgent need for precise biomarkers to improve individual risk stratification and enable personalized prevention strategies. We aimed to develop and validate a plasma proteomics-based protein risk score (PRS) to predict incident HF among individuals with obesity. MATERIALS AND METHODS: We analysed 9831 participants with obesity (BMI ≥ 30 kg/m2) from the UK Biobank with baseline measurements of 2911 circulating proteins and up to 16 years of follow-up. Multivariable Cox regression identified proteins associated with incident HF after comprehensive covariate adjustment. A PRS was constructed using LASSO regression and evaluated in a held-out test set. Protein trajectories before HF onset were reconstructed using LOESS modelling. To enhance clinical feasibility, a minimal protein panel was identified using LightGBM with forward feature selection. RESULTS: A total of 727 participants developed HF during follow-up. Multivariable cox analyses identified 578 proteins significantly associated with HF. LASSO regression further selected 81 proteins to build the PRS, which showed a strong association with HF risk in both training (HR 3.57; 95% CI 3.19-4.00) and test cohorts (HR 2.45; 95% CI 2.20-2.74). Adding the PRS improved prediction beyond age and sex (ΔC = 0.091) and beyond the Pooled Cohort Equations to Prevent Heart Failure (PCP-HF) model (ΔC = 0.052), with consistent gains in NRI and IDI. Proteomic deviations were detectable up to 16 years before diagnosis. A four-protein panel (GDF15, NT-proBNP, TNFRSF10B, CTHRC1) achieved robust discrimination (AUC 0.789), outperforming NT-proBNP alone (AUC 0.695) and complementing the PCP-HF model (combined AUC 0.803). DISCUSSION: Large-scale plasma proteomics substantially improves HF risk prediction in individuals with obesity and reveals long-standing molecular alterations preceding clinical onset. A simplified four-protein panel maintains robust predictive accuracy and provides a practical approach for the early detection and targeted prevention of obesity-related HF.

Humans↗

PubMed2026

AI In Leukemia Diagnostics: Complementing the Pathologist's Role.

Artificial intelligence (AI) is reshaping every stage of leukemia diagnostics, from digital morphology and multiparameter flow cytometry to next-generation sequencing, multi-omics analysis, and emerging computational frontiers such as quantum-inspired feature selection. This review outlines how contemporary AI tools can automate labor-intensive quantitation, flag diagnostically salient patterns, and standardize interpretation, while the pathologist or hematologist retains authority over validation, context-specific integration, and clinical decision-making. We present an illustrative "human-in-the-loop" workflow that embeds AI modules within current laboratory information systems, emphasizing points where expert oversight mitigates algorithmic bias and resolves discordant findings. We further map the validator-integrator role across morphology, flow cytometry, and genomic/multi-omic interpretation and provide practical training competencies and use cases for AI-assisted hematopathology. Beyond technical deployment, the article addresses the educational transformation required for sustainable adoption. Drawing on international competency frameworks, including the Digital Health Competencies in Medical Education Framework and recently proposed AI-specific Entrustable Professional Activities, we map core skills that future hematopathologists must master: data-science literacy, critical appraisal of AI outputs, and ethical governance. We highlight evaluated training models such as the Pathology Informatics Essentials for Residents curriculum, Stanford Artificial Intelligence in Machine and Imaging workshops, and College of American Pathologists bootcamps and propose integration strategies adaptable across resource settings. By pairing rigorous validation with targeted education, AI can elevate rather than eclipse the diagnostic role of the leukemia specialist, enabling more timely, reproducible, and personalized patient care.

Humans↗

PubMed2005

Matching food security analysis to context: the experience of the Somalia food security assessment unit.

This case study reviews the experience of the Somalia Food Security Assessment Unit (FSAU) of operating a food security information system in the context of a complex emergency. In particular, it explores the linkages between selected features of the protracted crisis environment in Somalia and conceptual and operational aspects of food security information work. The paper specifically examines the implications of context characteristics for the establishment and operations of the FSAU field monitoring component and for the interface with information users and their diverse information needs. It also analyses the scope for linking food security and nutrition analysis and looks at the role of conflict and gender analysis in food security assessment work. Background data on the food security situation in Somalia and an overview of some key features of the FSAU set the scene for the case study. The paper is targeted at those involved in designing, operating and funding food security information activities.

Food Supply↗

PubMed1994

Automated tracing and volume measurements of neurons from 3-D confocal fluorescence microscopy data.

Three-dimensional (3-D) image analysis algorithms and experimental results that demonstrate the feasibility of fully automated tracing of neurons from fluorescence confocal microscopy data are presented. The input to the automated analysis is a set of successive optical slices that have been acquired using a confocal scanning laser microscope. The output of the system is a labelled graph representation of the neuronal topology that is spatially aligned with the 3-D image data. A variety of topological and metric analyses can be carried out using this representation. For instance, precise measurements of volumes, lengths, diameters and tortuosities can be made over specific portions of the neuron that are specified in terms of the graph representation. The effectiveness of the method is demonstrated for a set of sample fields featuring selectively stained neurons. Additional work will be needed to refine the method for unsupervised use with complex data involving multiple intertwined neurons and extremely fine dendritic structures.

Algorithms↗

PubMed1992

Oscillatory Neuronal Responses in the Visual Cortex of the Awake Macaque Monkey.

An important step in early visual processing is the segmentation of scenes. Features constituting individual objects have to be grouped together and segregated from those of other figures or the background. It has been proposed that this grouping could be achieved by synchronizing the fine temporal structure of responses from neurons excited by an individual figure. In the cat visual cortex evidence has been obtained that responses of feature-selective neurons have a distinctive oscillatory structure and can synchronize both within and across cortical areas, the synchronization depending on stimulus configuration. Here we investigate the generality of oscillatory responses and their synchronization and specifically whether these phenomena occur in extrastriate areas of the visual cortex of the awake behaving primate. We find in the caudal superior temporal sulcus of the macaque monkey (Macaca fascicularis) that adjacent neurons can synchronize their responses, in which case their discharges exhibit an oscillatory temporal structure. During such periods of local synchrony spatially separated cell groups can also synchronize their responses if activated with a single stimulus. These findings resemble those described previously for the cat visual cortex, except that in the awake monkey the oscillatory episodes tend to be of shorter duration and exhibit more variability of oscillation frequency.

Journal Article↗

PubMed2005

A neural signature of colour and luminance correspondence in bistable apparent motion.

The 'correspondence problem' refers to the ambiguity of apparent motion (AM) paths if several similar objects are displaced across successive displays. We investigated the effect of intrinsic object properties such as colour and luminance on AM paths, and used functional magnetic resonance imaging to localize neural correlates of correspondence matching in visual cortical regions. Human subjects looked at an AM display where two dots in diagonally opposite corners of an implicit rectangle were flashed in alternation with two dots in the other two corners, yielding spontaneous alternations between horizontal and vertical AM. The dots differed in colour or luminance, or were identical. Neural activity was analysed as a function of whether the perceived AM path matched the dots' colour or luminance, and was also compared to activity during bistable AM displays without correspondence cues. When AM paths matched colour and luminance cues, activity in early visual cortex was the same as during perception of uncued displays, whereas it was suppressed when perceived AM paths violated colour or luminance cues. Colour-sensitive extrastriate cortex (V4 complex) transiently activated whenever AM perception switched from a pattern violating colour correspondence to one consistent with colour. We propose that the neural correlate of correspondence in early visual cortex reflects regulatory mechanisms that flexibly gate early visual feature processing in accord with an overriding perceptual decision. Conversely, activation of feature-selective extrastriate regions depends on the type of cue used for correspondence matching and may reflect the salience of percepts that match in colour and motion.

Adult↗

PubMed2005

Hepatitis C virus infection.

Hepatitis C virus (HCV) infection is a major public health problem. Up to 3% of the world's population is infected with HCV, and at least 200 000 adults in the UK carry the virus. Of those exposed to HCV, 80% become chronically infected, and at least 30% of carriers develop chronic liver disease, including cirrhosis and hepatocellular carcinoma. This review provides an overview of selected features of the molecular biology and pathogenesis of HCV infection, and thereafter discusses in detail the epidemiology of HCV, the hepatic and extra-hepatic diseases caused by the virus, and the current treatment options for both acute and chronic virus infection. The special cases of healthcare workers, prison inmates and individuals coinfected with human immunodeficiency virus and HCV are considered in detail.

HIV Infections↗

PubMed2006

Fast identification of ten clinically important micro-organisms using an electronic nose.

AIMS: To evaluate the electronic nose (EN) as method for the identification of ten clinically important micro-organisms. METHODS AND RESULTS: A commercial EN system with a series of ten metal oxide sensors was used to characterize the headspace of the cultured organisms. The measurement procedure was optimized to obtain reproducible results. Artificial neural networks (ANNs) and a k-nearest neighbour (k-NN) algorithm in combination with a feature selection technique were used as pattern recognition tools. Hundred percent correct identification can be achieved by EN technology, provided that sufficient attention is paid to data handling. CONCLUSIONS: Even for a set containing a number of closely related species in addition to four unrelated organisms, an EN is capable of 100% correct identification. SIGNIFICANCE AND IMPACT OF THE STUDY: The time between isolation and identification of the sample can be dramatically reduced to 17 h.

Bacteria↗

PubMed1981

Spectral analysis of Doppler velocity patterns in normals and patients with carotid artery stenosis.

A computerized pattern recognition program was utilized to assess the predictive ability of various parameters obtained from the spectra of ultrasonic pulsed Doppler signals from the carotid arteries. The most accurate features selected by linear regression analysis were the natural log (ln) of the ratio of the mean velocity in the internal carotid artery compared to that in the common carotid artery, and the ln of the maximum velocity, the ln of the maximum frequency, and the square of the fractional broadening term, all of which were measured at peak systole in the internal carotid artery. Using the combination of the velocity ratio and the fractional broadening term, the average difference in the estimated percentage stenosis, as compared to that obtained by arteriography, was 12.8%.

Arteriosclerosis↗

PubMed1993

Effective photodynamic action by rhodamine 123 leading to photosensitized killing of Chinese hamster ovary cells in tissue culture and a proposed mechanism.

The effectiveness of rhodamine 123 (R123) as a photosensitizer of cell killing is relatively low and correlates with its inefficient production of singlet oxygen. The known selective retention of R123 in the mitochondria of epithelially derived carcinoma cells, however, is a selective feature that could lead to a more useful therapeutic ratio if photosensitizing effectiveness could be increased. Chinese hamster ovary (CHO) cells in tissue culture were therefore exposed to R123 shortly before and during illumination under conditions controlled for oxygen concentration and temperature. Effective photosensitization of cell killing, as judged by colony formation, was produced by 95% but not by 19% O2 during illumination of cells at 5 degrees C or 37 degrees C, and this was additionally enhanced at the sublethal temperature of 42 degrees C. Two CHO cell lines were examined; one line, CHO-AA8, was proficient in the repair of DNA damage and the parent to the second line, CHO-EM9, that was deficient in the repair of DNA strand breaks. Cells of both lines incorporated R123 to a similar degree and were similarly photosensitized by the presence of high oxygen concentration. Furthermore, plasma membrane damage as judged by the exclusion of trypan blue was not observed immediately after illumination in the presence of R123, but was seen in the presence of meso-tetra-(4-sulfonatophenyl)-porphine (TPPS4). The extent of damage to the plasma membrane by TPPS4 was greater in the presence of 95% compared to 19% O2 during illumination.(ABSTRACT TRUNCATED AT 250 WORDS)

Animals↗

PubMedDate not supplied

Improvement in specificity of ultrasonography for diagnosis of breast tumors by means of artificial intelligence.

A set of ultrasonograms of lesions from 200 patients between the ages of 14 and 93 years who underwent mammography followed by ultrasonographic examination and excisional biopsy has been studied with computer vision techniques to improve the ultrasonographic specificity of the diagnosis. Selected features representing the texture of the lesion were calculated and then classified by an artificial neural network. This network was biased toward correctly classifying all the malignant cases at the expense of some misclassification of the benign cases. The network diagnosed the malignant cases with 100% sensitivity and 40% specificity (compared with 0% specificity for the radiologists diagnosing the same set of cases in the breast imaging setting), and tests performed with a leave-one-out technique indicate that the network will generalize well to new cases. This suggests that methods based on neural network classification of texture features show promise for potentially decreasing the number of unnecessary biopsies by a significant amount in patients with sonographically identifiable lesions.

Adolescent↗