Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Dataset”

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 1,603 records · Page 89Linked to original sources
PubMed1998

The Theoretical and Experimental Status of the n-tuple Classifier.

A number of theoretical approaches related to the n-tuple classification system are reviewed including Kanerva's SDM, the n-tuple regression network, the Hamming distance framework and likelihood estimation. The limitations of these methods are pointed out and resemblances that exist between them are underlined. Large-scale experiments carried out on StatLog project datasets confirm the n-tuple method as a viable competitor to more popular methods due to its speed, simplicity and accuracy on the majority of a wide variety of classification problems. A further investigation into the failure of the method on certain datasets shows its inner workings and reveals two main problems: difficulties with highly skewed class priors and more importantly, a mismatch between the scales involved in generalization, the amount of training data available, and the volume of the region in which data is likely to exist. This highlights areas where improvements in the method are needed and further theoretical progress would be helpful.

Journal Article↗
PubMed1998

Three-dimensional imaging in aortic disease by lighthouse transesophageal echocardiography using intravascular ultrasound catheters. Comparison to three-dimensional transesophageal echocardiography and three-dimensional intra-aortic ultrasound imaging.

Two-dimensional (2D) transesophageal echocardiography (TEE) and 2D intravascular ultrasound (IVUS) imaging face their greatest limitation in visualizing aortic disease in patients. With the aid of three-dimensional (3D) image reconstruction, TEE and IVUS can potentially overcome this limitation but still provide only limited spatial appreciation in aortic disease because 3D imaging of the thoracic aorta requires a broader spatial visualization of the mediastinum than provided by both techniques. Moreover, for timely decision making about aortic disease TEE is limited by a large probe, which requires sedation. Therefore, we developed an approach called 3D lighthouse transesophageal echocardiography (LTEE) using a thin intravascular ultrasound catheter, which provides a full circumferential (360 degree) image and requires no sedation. The purpose of this study was to compare the feasibility and accuracy of 3D TEE, 3D IVUS, and 3D LTEE for obtaining spatial visualization of the thoracic aorta to detect aortic diseases in patients. 3D image datasets were obtained for 3D LTEE by a manual pullback of a 3.3 mm thick, 10 MHz intravascular ultrasound catheter positioned in the esophagus; for 3D TEE using a conventional 15 mm thick probe; and for 3D IVUS using a 2.6 mm thick, 20 MHz intravascular ultrasound catheter. In 12 consecutive patients, three with aortic dissection (two with type III, one with type I) and 11 with suspected artherosclerosis, we analyzed and compared spatial visualization of the thoracic aorta, 3D image quality, patient discomfort, and study time. Providing a 3D dataset of 360-degree tomographic images of the mediastinum, 3D LTEE was the only approach that allowed broad spatial visualization of the aortic arch (9 of 12 patients) with the detection of aortic dissection or atherosclerotic plaques. Spatial visualization of the aortic arch by 3D TEE was incomplete because of the relatively narrow 90-degree image sector. However, in other segments 3D image quality by 3D TEE was superior to 3D LTEE and 3D IVUS. Because of the thin catheter, patient discomfort (p < 0.0001) and examination time (p = 0.015) were significantly less for 3D LTEE compared with 3D TEE. 3D LTEE is a promising new technique for 3D imaging of the thoracic aorta and detection of aortic disease with improved spatial visualization and reduced patient discomfort compared with 3D TEE and 3D IVUS.

Aortic Dissection↗
PubMed1999

Impact of on-line endocardial border detection on determination of left ventricular volume and ejection fraction by transthoracic 3-dimensional echocardiography.

This study was performed to determine whether use of on-line automated border detection (ABD) could reduce data analysis time for 3-dimensional echocardiography (3DE) while maintaining accuracy of 3DE in measures of left ventricular (LV) volumes and ejection fraction (EF). The study proceeded in 2 phases. In the validation phase, 20 subjects were examined with the use of 3DE and of monoplane 2-dimensional (2D) ABD. Results were compared with the reference standard of magnetic resonance imaging (MRI). In the test phase, 20 subjects underwent two 3DE studies (once with images optimized for visual border definition and once with images optimized for ABD border tracking) and a conventionally used 2D ABD study. For 3DE, volumes and EF were determined with the use of manually traced borders and ABD. Analysis times were recorded with a digital stopwatch. In the validation phase, 3DE and MRI results correlated very well (r = 0.99) without systematic differences. Comparison of 2D ABD with MRI showed good correlation for LV volumes (r >/= 0.90) and EF (r = 0.85) despite significant underestimation. For the test phase, Acoustic Quantification-optimized 3-dimensional datasets underestimated end-diastolic volume and EF relative to visually optimized 3-dimensional datasets regardless of whether borders were hand-traced or ABD was used. However, correlations ranged from r = 0.96 to r = 0.98 for LV volumes and 0.88 to 0.91 for LV EF and were superior to those for 2D ABD. Data analysis times decreased moderately with the use of ABD, but scan times increased; total study times were unchanged. Use of on-line ABD with 3DE reduces data analysis time and is more accurate than conventional monoplane 2D ABD but results in underestimation of LV volumes and EF. Additional automated postprocessing techniques may be required to obtain accurate measures, consistently using 3DE in conjunction with on-line ABD.

Adult↗
PubMed2002

A comparison of published head and neck stage groupings in laryngeal cancer using data from two countries.

The combination of T, N, and M classifications into stage groupings is meant to facilitate a number of activities including: the estimation of prognosis and the comparison of therapeutic interventions among similar groups of cases. We tested the UICC/AJCC fifth edition stage grouping and six other TNM-based groupings proposed for head and neck cancer for their ability to meet these expectations in laryngeal cancer using data from Ontario, Canada, and the area of Southern Norway surrounding Oslo. We defined four criteria to assess each grouping scheme: (1) the subgroups defined by T, N, and M comprising a given group within a grouping scheme have similar survival rates (hazard consistency); (2) the survival rates differ among the groups (hazard discrimination); (3) the prediction of cure is high (outcome prediction); and (4) the distribution of patients among the groups is balanced. We previously identified or derived a measure for each criterion, and the findings were summarized using a scoring system. The range of scores was from 0 (best) to 7 (worst). The data sets were population-based, with 861 cases from Ontario and 642 cases from Southern Norway. Clinical stage assignment was used and the outcome of interest was cause-specific survival. Summary scores across the seven schemes had similar ranges: 0.9 to 5.1 in Ontario and 1.8 to 5.7 in Southern Norway, but the ranking varied. Summing the scores across the two datasets, the TANIS-7 scheme (Head & Neck 1993;15:497-503) ranked first, and was ranked high in both datasets (first and second, respectively). The UICC/AJCC scheme ranked sixth out of seven schemes, and its ranking was fifth and seventh, respectively. UICC/AJCC stage groupings were defined without empirical investigation. When tested, this scheme did not perform best. Our results suggest that the usefulness of the TNM system could be enhanced by optimizing the design of stage groupings through empirical investigation.

Carcinoma, Squamous Cell↗
PubMed2003

Techniques for knowledge discovery in existing biomedical databases: estimation of individual aging effects in cognition in relation to dementia.

New interest is being expressed in the systematic application of modeling techniques to existing datasets. Under the rubric of Knowledge Discovery in Databases (KDD) large databases are being exploited for commercial and scientific purposes. This article reviews the development and applications of KDD techniques to dementia, using the longitudinal Canadian Study of Health and Aging dataset. KDD has demonstrated usefulness at the group level. For example, as in the course of functional impairment between Alzheimer's disease and no cognitive impairment suggest damage control-protection mechanisms for the former compared with noncompensated random accumulation of deficits for the latter. At the individual level, KDD suggests that more precise diagnosis seems possible as well as individual life expectancy prediction. Biomedical databases appear to hold the potential for novel insights when explored by systematic modeling.

Aged↗
PubMedDate not supplied

VoxeLine: a software program for 3D real-time visualization of biomedical images.

The architecture and implementation of VoxeLine, a new interactive environment for display and analysis of 2D and 3D images in real-time, is discussed. This modular software project comprises two main parts: a user part (without programming expertise) and a programming part which permits its adaptation to specific problems. VoxeLine has the ability to deal with almost all sorts of data types encountered in the biomedical field (e.g. images, vectors). Another important feature is its ability to show datasets in all directions without duplicating data into the main memory. This feature allows VoxeLine to be used on machines with limited memory capacities and power. Real-time 3D manipulations (10 Hz for a 256 x 256 x 124 MRI dataset) are possible on a classic monoprocessor architecture such as a personal computer.

Color↗
PubMed2002

A novel, direct spatio-temporal approach for analyzing fMRI experiments.

We introduce a novel approach to couple temporal similarity with spatial neighborhood information. This is achieved by concatenating the K nearest, spatially contiguous neighbors of a pixel time-course (TC) of T time-instances. This produces a new TC of (K+1)T time instances. Depending on how "nearest" is defined, we have various options. Strictly spatial nearness means augmenting a given TC by its K nearest neighbors in some canonical spatial order. A more powerful and flexible option is to order the TCs to be concatenated according to their temporal similarity to the central voxel TC. For this study, we have chosen Pearson's cross-correlation coefficient as the measure of similarity. For more than a single neighbor, two concatenation options are possible. The direct ordering option requires that the TCs to be concatenated be spatially contiguous to the central pixel. The more flexible indirect option merely demands that one of a chain of temporally similar TCs be spatially connected to the central pixel. We also apply the temporal similarity criterion to the more conventional spatial (median) filtering, and show that it gives superior result to a strict spatial filtering. The method is tested and verified on a null fMRI dataset onto which we superposed two types of "activations" with known temporal behavior and spatial location. It is also applied to a real dataset containing visual activation. We also propose a strategy, based on the flexibility of the method, to determine a consensus, "core" set of activations.

Brain↗
PubMedDate not supplied

Mining knowledge for HEp-2 cell image classification.

HEp-2 cells are used for the identification of antinuclear autoantibodies (ANAs). They allow for recognition of over 30 different nuclear and cytoplasmic patterns, which are given by upwards of 100 different autoantibodies. The identification of the patterns has recently been done manually by a human inspecting the slides with a microscope. In this paper, we present results on the analysis and classification of cells using image analysis and data mining techniques. Starting from a knowledge acquisition process with a human operator, we developed an image analysis and feature extraction algorithm. The collection of the dataset was done based on an expert's image reading and based on the automatic extracted features. A dataset containing 132 features for each entry was set up and given to a data mining algorithm to find out the relevant features among this large feature set and to construct the classification knowledge. The classifier was evaluated by cross validation. The results gave the expert new insights into the necessary features and the classification knowledge and show the feasibility of an automated inspection system.

Algorithms↗
PubMed2003

A combined neural network and decision trees model for prognosis of breast cancer relapse.

The prediction of clinical outcome of patients after breast cancer surgery plays an important role in medical tasks such as diagnosis and treatment planning. Different prognostic factors for breast cancer outcome appear to be significant predictors for overall survival, but probably form part of a bigger picture comprising many factors. Survival estimations are currently performed by clinicians using the statistical techniques of survival analysis. In this sense, artificial neural networks are shown to be a powerful tool for analysing datasets where there are complicated non-linear interactions between the input data and the information to be predicted. This paper presents a decision support tool for the prognosis of breast cancer relapse that combines a novel algorithm TDIDT (control of induction by sample division method, CIDIM), to select the most relevant prognostic factors for the accurate prognosis of breast cancer, with a system composed of different neural networks topologies that takes as input the selected variables in order for it to reach good correct classification probability. In addition, a new method for the estimate of Bayes' optimal error using the neural network paradigm is proposed. Clinical-pathological data were obtained from the Medical Oncology Service of the Hospital Clinico Universitario of Málaga, Spain. The results show that the proposed system is an useful tool to be used by clinicians to search through large datasets seeking subtle patterns in prognostic factors, and that may further assist the selection of appropriate adjuvant treatments for the individual patient.

Adult↗
PubMed2003

Brain volumes characterisation using hierarchical neural networks.

Objective knowledge of tissue density distribution in CT/MRI brain datasets can be related to anatomical or neuro-functional regions for assessing pathologic conditions characterised by slight differences. The process of monitoring illness and its treatment could be then improved by a suitable detection of these variations. In this paper, we present an approach for three-dimensional (3D) classification of brain tissue densities based on a hierarchical artificial neural network (ANN) able to classify the single voxels of the examined datasets. The method developed was tested on case studies selected by an expert neuro-radiologist and consisting of both normal and pathological conditions. The results obtained were submitted for validation to a group of physicians and they judged the system to be really effective in practical applications.

Aged↗
PubMed1997

An artificial intelligence approach to motif discovery in protein sequences: application to steriod dehydrogenases.

MEME (Multiple Expectation-maximization for Motif Elicitation) is a unique new software tool that uses artificial intelligence techniques to discover motifs shared by a set of protein sequences in a fully automated manner. This paper is the first detailed study of the use of MEME to analyse a large, biologically relevant set of sequences, and to evaluate the sensitivity and accuracy of MEME in identifying structurally important motifs. For this purpose, we chose the short-chain alcohol dehydrogenase superfamily because it is large and phylogenetically diverse, providing a test of how well MEME can work on sequences with low amino acid similarity. Moreover, this dataset contains enzymes of biological importance, and because several enzymes have known X-ray crystallographic structures, we can test the usefulness of MEME for structural analysis. The first six motifs from MEME map onto structurally important alpha-helices and beta-strands on Streptomyces hydrogenans 20beta-hydroxysteroid dehydrogenase. We also describe MAST (Motif Alignment Search Tool), which conveniently uses output from MEME for searching databases such as SWISS-PROT and Genpept. MAST provides statistical measures that permit a rigorous evaluation of the significance of database searches with individual motifs or groups of motifs. A database search of Genpept90 by MAST with the log-odds matrix of the first six motifs obtained from MEME yields a bimodal output, demonstrating the selectivity of MAST. We show for the first time, using primary sequence analysis, that bacterial sugar epimerases are homologs of short-chain dehydrogenases. MEME and MAST will be increasingly useful as genome sequencing provides large datasets of phylogenetically divergent sequences of biomedical interest.

Alcohol Dehydrogenase↗
PubMed2002

Lessons learnt about early prostate cancer from large scale databases: population-based pearls of wisdom.

Prostate cancer is one of most common solid tumors in men and poses some of the most difficult problems in clinical research. Although many clinical research hypotheses in this condition have been explored using single center cases series and multi-center clinical trials, the results of these studies have often been equivocal, leaving many questions unanswered. Recently, investigators have utilized large, administrative datasets for prostate cancer research. These databases tend to include large numbers of patients from different geographic regions increasing their generalizability and statistical power. The goal of this report is to review lessons learnt about early prostate cancer using these data sources. In particular, we focus on the application of large, population-based datasets to address issues concerning the natural history of prostate cancer, the impact of race on outcomes in prostate cancer and the effectiveness of various treatments for localized disease. Information gathered from large, administrative databases will be helpful when counseling patients regarding their treatments options for localized prostate cancer and in identifying future directions for prostate cancer research.

Centers for Medicare and Medicaid Services, U.S.↗
PubMedDate not supplied

Chromosomal location effects on gene sequence evolution in mammals.

BACKGROUND: Nucleotide substitution rates and G + C content vary considerably among mammalian genes. It has been proposed that the mammalian genome comprises a mosaic of regions - termed isochores - with differing G + C content. The regional variation in gene G + C content might therefore be a reflection of the isochore structure of chromosomes, but the factors influencing the variation of nucleotide substitution rate are still open to question. RESULTS: To examine whether nucleotide substitution rates and gene G + C content are influenced by the chromosomal location of genes, we compared human and murid (mouse or rat) orthologues known to belong to one of the chromosomal (autosomal) segments conserved between these species. Multiple members of gene families were excluded from the dataset. Sets of neighbouring genes were defined as those lying within 1 centiMorgan (cM) of each other on the mouse genetic map. For both synonymous substitution rates and G + C content at silent sites, neighbouring genes were found to be significantly more similar to each other than sets of genes randomly drawn from the dataset. Moreover, we demonstrated that the regional similarities in G + C content (isochores) and synonymous substitution rate were independent of each other. CONCLUSIONS: Our results provide the first substantial statistical evidence for the existence of a regional variation in the synonymous substitution rate within the mammalian genome, indicating that different chromosomal regions evolve at different rates. This regional phenomenon which shapes gene evolution could reflect the existence of 'evolutionary rate units' along the chromosome.

Animals↗
PubMed2001

In silico analysis of cancer through the Cancer Genome Anatomy Project.

The Cancer Genome Anatomy Project (CGAP) was designed and implemented to provide public datasets, material resources and informatics tools to serve as a platform to support the elucidation of the molecular signatures of cancer. This overview of CGAP describes the status of this effort to develop resources based on gene expression, polymorphism identification and chromosome aberrations, and we describe a variety of analytical tools designed to facilitate in silico analysis of these datasets.

Animals↗
PubMed1999

The development of an international surgical registry: the ECSUR project. The European Cardiac Surgical Registry.

OBJECTIVES: Cardiac surgery is the first of any therapeutic discipline to attempt to establish a major international registry on a patient-by-patient basis. Being the first necessarily imposes a challenge since no rules or guidelines exist. Possibilities and boundaries must be identified. METHODS: The EACTS database committee established ECSUR (The European Cardiac SUrgical Registry) with the aim of collecting patient-by-patient information from centres across Europe. It was decided to collect patient data in three areas; adult cardiac surgery, paediatric cardiac surgery and thoracic surgery. Currently, these three areas are in different stages of development, with multinational data collection now having already commenced in adult cardiac surgery. RESULTS: After just over a year of data collection ECSUR has now received data from 98 centres in 30 countries, encompassing approximately 100,000 patients. Three national cardiac surgical registries currently send their data to ECSUR, and agreement has been reached with three other countries for prospective data sharing during 1999. CONCLUSION: This paper describes how ECSUR was established, the on-going status of the project, and how the project has now developed a global perspective through a joint European and US initiative to undertake international dataset unification for surgeons worldwide, by the definition of core adult cardiac, paediatric cardiac and thoracic surgical datasets.

Data Collection↗
PubMed2003

Subcortical, cerebellar, and magnetic resonance based consistent brain image registration.

A new landmark-initialized segmentation and intensity-based (LI-SI) inverse-consistent linear elastic image registration algorithm is presented. This method uses manually identified landmarks, segmented volumetric (anatomical) structures, and normalized image signal intensity information to coregister datasets. The features used for image registration and evaluation include 35 cortical, cerebellar, and commissure landmarks manually identified by experts, subcortical and cerebellar regions defined semi-automatically by an artificial neural network and manually trimmed for validity by experts, and tissue classified images that were generated using a discriminant analysis of three magnetic resonance image sets representing T1, T2, and PD modalities. Four groups of results were computed for coregistering 16 datasets with the following registration techniques: rigid registration, extended Talairach registration, intensity-only inverse-consistent linear elastic registration, and the new LI-SI registration. Results are presented showing that relative overlap measurements increased as the dimensionality of the registration algorithm and amount of anatomical information increased. The average relative overlap improved from 0.53 for the rigid registration to 0.55 for the Talairach registration to 0.74 for the intensity-only and to 0.85 for the LI-SI algorithm. We showed a statistically significant improvement for all but one structure using the intensity-only algorithm compared to the Talairach registration. Furthermore, statistically significant improvements for all structures were achieved using the LI-SI algorithm compared to the intensity-only algorithm.

Adult↗
PubMed2003

The phylogenetics of Desmognathine salamander populations across the southern Appalachians.

Salamanders in the genus Desmognathus (Caudata: Plethodontidae) are distributed along an aquatic to terrestrial habitat gradient in the southern Appalachian Mountains. The spatial distribution of species is believed to have formed as aquatic ancestors displaced lineages by competition and predatory interactions into less optimal terrestrial habitats. Aquatic and terrestrial species may also display different patterns of genetic diversity due to the differing likelihood of gene flow via aquatic corridors. To determine whether phylogenetic patterns were consistent with these hypotheses, we sequenced portions of the cytochrome oxidase I and 12S rRNA genes of the mitochondrial genome from 96 individuals belonging to 10 species in the genus Desmognathus. In addition, we combined our dataset with an earlier published dataset for the 12S rRNA genes. The order of species divergence is consistent with aquatic ancestors having displaced taxa into more terrestrial habitats, but the major lineages within the genus Desmognathus arose suddenly, and therefore, the specific sequence of events is not well resolved. The phylogenetic analyses among species suggest that direct-development and a terrestrial lifestyle are ancestral in the genus Desmognathus, but the degree of adult terrestriallity is labile, with some species having re-invaded terrestrial habitats. We present evidence of a clade of Desmognathus quadramaculatus from North Carolina that is distinct from the D. quadramaculatus/Desmognathus marmoratus clade. Within species, estimates of Tajima's D and Fu and Li's statistics suggest the species experienced population expansions at different times in the past. Current levels of sequence diversity in northern populations, therefore, reflect different arrival times, and hence, differences in the opportunity for among population divergence. The recent arrival of most species over large portions of their geographic ranges suggests that most extant communities have been assembled, a posteriori, by the recent assortment of species along the aquatic to terrestrial gradient according to their ecologies.

Animals↗
PubMed2003

A phylogeny of Chinese species in the genus Phrynocephalus (Agamidae) inferred from mitochondrial DNA sequences.

We investigated the phylogenetic relationships among most Chinese species of lizards in the genus Phrynocephalus (118 individuals collected from 56 populations of 14 well-defined species and several unidentified specimens) using four mitochondrial gene fragments (12S rRNA, 16S rRNA, cytochrome b, and ND4-tRNA(LEU)). The partition-homogeneity tests indicated that the combined dataset was homogeneous, and maximum-parsimony (MP), neighbor-joining (NJ), maximum-likelihood (ML) and Bayesian (BI) analyses were performed on this combined dataset (49 haplotypes including outgroups for 2058bp in total). The maximum-parsimony analysis resulted in 24 equally parsimonious trees, and their strict consensus tree shows that there are two major clades representing the Chinese Phrynocephalus species: the viviparous group (Clade A) and the oviparous group (Clade B). The trees derived from Bayesian, ML, and NJ analyses were topologically identical to the MP analysis except for the position of P. mystaceus. All analyses left the nodes for the oviparous group, the most basal clade within the oviparous group, and P. mystaceus unresolved. The phylogenies further suggest that the monophyly of the viviparous species may have resulted from vicariance, while recent dispersal may have been important in generating the pattern of variation among the oviparous species.

Animals↗