SEARCH · PubMed Health

Results for “Classification Algorithms”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

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

At least 955 records · Page 53Linked to original sources

PubMed2003

Diagnostic decision rule for support in clinical assessment of the need for surgical intervention in horses with acute abdominal pain.

A prospective survey of horses with colic referred to a university hospital was undertaken to elaborate on a simple clinical decision support system capable of predicting whether or not horses require surgical intervention. Cases were classified as requiring surgical intervention or not on the basis of intraoperative findings or necropsy reports. Logistic regression analysis was applied to identify predictors with the strongest association with treatment needed. The classification and regression tree (CART) methodology was used to combine the variables in a simple classification system. The performance of the elaborated algorithms, as diagnostic instruments, was recorded as test sensitivity and specificity. The CART method generated 5 different classification trees with a similar basic structure consisting of: degree of pain, peritoneal fluid colour, and rectal temperature. The tree, constructed at a prevalence of 15% surgical cases, appeared to be the best proposal made by CART. In this classification tree, further discrimination of cases was obtained by including the findings of rectal examination and packed cell volume. When regarded as a test system, the sensitivity and specificity was 52% and 95%, respectively, corresponding to positive and negative predictive values of 68% and 91%. The variables examined in the present study did not provide a safe clinical decision rule. The classification tree constructed at 15% surgical cases was considered feasible, the proportion of horses incorrectly predicted to be without need of immediate surgery (false negatives) was small, whereas the proportion of horses incorrectly predicted to be in need of immediate surgery (false positives) was large. Some of the false positive horses were amenable to surgical treatment, although these cases did not conform to the strict definition of a surgical case. A less rigorous definition of a surgical case than that used in the present study would lower the percentage of false positives.

Abdomen, Acute↗

PubMed2005

Congenital scalp and calvarial deficiencies: principles for classification and surgical management.

Congenital defects of the scalp and skull present a challenge for care providers because of a combination of their rarity and the magnitude of potential morbidity. Recent advancements in autogenous and alloplastic cranioplasty and scalp reconstruction techniques argue for a comprehensive consideration of this problem. This article (1) reviews the causes of congenital scalp and calvarial defects; (2) proposes a classification system based on defect type, similar to the tumor-node-metastasis classification, in that defect location, defect size, and extent of neuromeningeal involvement are the critical variables; and (3) presents algorithms for care based on the defect classification. A set of management principles on which treatment plans can be based for these unique problems is provided.

Algorithms↗

PubMed2005

Support-vector-machine classification of linear functional motifs in proteins.

Our algorithm predicts short linear functional motifs in proteins using only sequence information. Statistical models for short linear functional motifs in proteins are built using the database of short sequence fragments taken from proteins in the current release of the Swiss-Prot database. Those segments are confirmed by experiments to have single-residue post-translational modification. The sensitivities of the classification for various types of short linear motifs are in the range of 70%. The query protein sequence is dissected into short overlapping fragments. All segments are represented as vectors. Each vector is then classified by a machine learning algorithm (Support Vector Machine) as potentially modifiable or not. The resulting list of plausible post-translational sites in the query protein is returned to the user. We also present a study of the human protein kinase C family as a biological application of our method.

Databases, Genetic↗

PubMed1982

Diagnosis of brain death by common carotid artery velocity waveform analysis.

In addition to neurologic criteria, some test of cessation of brain blood flow is recommended before diagnosing "brain death." Cerebral arteriography, radionuclide scintigraphy, and contrast computed axial tomography, though reliable, possess significant practical limitations. Analysis of the dimensions and contours of common carotid Doppler velocity tracings of brain-dead patients has identified qualitative and quantitative differences not only from normal subjects, but also from patients with complete atheromatous carotid occlusion and from those unconscious after brain injury. Though accurate separation was initially made using computer-assisted classification function analysis, a simplified algorithmic approach using only three velocity waveform variables has been developed without loss of accuracy. The speed, simplicity, portability, and inexpensiveness of this approach commend its use.

Adult↗

PubMed2000

Development and application of a virtual environment for reconstructive surgery.

OBJECTIVE: This paper details the development and application of a Virtual Environment for Reconstructive Surgery (VERS). It addresses the technical and user-interface challenges in developing such a system, and the lessons learned during application of the system in the case of a 17-year-old boy with a severe facial defect arising from the removal of a soft-tissue sarcoma. MATERIALS AND METHODS: Computed tomography (CT) scans were segmented into bone and soft-tissue classifications using traditional and novel algorithms, a surface mesh was generated, and imaging artifacts were removed, yielding a mesh suitable for visualization. This patient-specific mesh was then used in a virtual environment by the surgeons for preoperative visualization of the defect, planning of the surgery, and production of a custom surgical template to aid in repairing the defect. RESULTS: This system was successfully used to plan the surgery of the patient and to produce a custom, patient-specific template that was used to harvest bone from a donor site in order to reconstruct the defect. CONCLUSION: Despite technical challenges, virtual-environment surgical planning is useful as a clinical tool for preoperative visualization, cephalometric analysis, and surgical intervention. It can provide a more precise surgical result than would otherwise be realized using traditional methods.

Adolescent↗

PubMed1998

A Generalized Estimate of the SLR B Polynomial Ripples for RF Pulse Generation.

The nonlinearity of the parameter relations for the Shinnar-Le Roux RF pulse design algorithm has induced to performa classification based on the features of the slice profile dueto the RF pulse. In the present paper a generalization ofthe relation between the ripple amplitudes of the SLR B polynomial and those of the slice profile is given. It allows generation of RF pulses with better slice profiles and slightly reduced energy, avoiding any a priori classification. The effect of our estimation has been shown by generating several pulses by generalized estimation of B polynomial ripples. In addition, their behavior has been compared to that of analogous pulses generated by means of the classification just mentioned. Copyright 1998 Academic Press.

Journal Article↗

PubMedDate not supplied

The ratio of AIDS to non-AIDS Medicaid medical costs from 1992 to 2000.

Our research objective was to calculate and forecast the monthly increase in medical and prescription costs for Medicaid patients with acquired immunodeficiency syndrome (AIDS) and compare these values with costs for non-AIDS patients. A retrospective analysis of AIDS patients and a control group of Georgia Medicaid beneficiaries was conducted between January 1, 1988, and December 31, 1991. AIDS patients were defined using the Keyes algorithm of combinations of International Classification of Diseases, 9th Revision, Clinical Modification codes. The AIDS patient group was matched demographically to a group of non-AIDS patients. Data were adjusted to account for eligibility status, and the ratio of AIDS costs to non-AIDS costs was modeled with an econometric time series procedure. A total of 1966 AIDS patients were identified from 900,000 Medicaid recipients in the study period; 58.0% were male and 59.8% were black. Age was bimodal at < or = 1 year and 33 years. The best fit for the medical cost ratios produced a significant regression coefficient of .37. The initial ratio of AIDS to non-AIDS forecast was 4.25 in January 1992. The January 2000 forecast of this ratio increased to 42.56. This increase equates to an additional $8510.19 per AIDS patient-month for January 2000 in 1991 dollars. The outpatient prescription ratio for AIDS versus non-AIDS patients was not predictable. However, the greatest observed discrepancies were attributed to the expense for antihemophilia products. Overall, the most important finding was the accelerating medical costs for treating AIDS patients compared with costs for treating non-AIDS patients. These results may, in part, reflect additional costs for treating intravenous drug users and pediatric AIDS patients.

Acquired Immunodeficiency Syndrome↗

PubMed2006

Identification of primary tumors of brain metastases by SIMCA classification of IR spectroscopic images.

Brain metastases are secondary intracranial lesions which occur more frequently than primary brain tumors. The four most abundant types of brain metastasis originate from primary tumors of lung cancer, colorectal cancer, breast cancer and renal cell carcinoma. As metastatic cells contain the molecular information of the primary tissue cells and IR spectroscopy probes the molecular fingerprint of cells, IR spectroscopy based methods constitute a new approach to determine the origin of brain metastases. IR spectroscopic images of 4 by 4 mm2 tissue areas were recorded in transmission mode by a FTIR imaging spectrometer coupled to a focal plane array detector. Unsupervised cluster analysis revealed variances within each cryosection. Selected clusters of five IR images with known diagnoses trained a supervised classification model based on the algorithm soft independent modeling of class analogies (SIMCA). This model was applied to distinguish normal brain tissue from brain metastases and to identify the primary tumor of brain metastases in 15 independent IR images. All specimens were assigned to the correct tissue class. This proof-of-concept study demonstrates that IR spectroscopy can complement established methods such as histopathology or immunohistochemistry for diagnosis.

Adult↗

PubMed2006

Specimen-specific beam models for fast and accurate prediction of human trabecular bone mechanical properties.

Direct assessment of bone competence in vivo is not possible, hence, it is inevitable to predict it using appropriate simulation techniques. Although accurate estimates of bone competence can be obtained from micro-finite element models (muFE), it is at the expense of large computer efforts. In this study, we investigated the application of structural idealizations to represent individual trabeculae by single elements. The objective was to implement and validate this technique. We scanned 42 human vertebral bone samples (10 mm height, 8 mm diameter) with micro-computed tomography using a 20 microm resolution. After scanning, direct mechanical testing was performed. Topological classification and dilation-based algorithms were used to identify individual rods and plates. Two FE models were created for each specimen. In the first one, each rod-like trabecula was modeled with one thickness-matched beam; each plate-like trabecula was modeled with several beams. From a simulated compression test, assuming one isotropic tissue modulus for all elements, the apparent stiffness was calculated. After reducing the voxel size to 40 microm, a second FE model was created using a standard voxel conversion technique. Again, one tissue modulus was assumed for all elements in all models, and a compression test was simulated. Bone volume fraction ranged from 3.7% to 19.5%; Young's moduli from 43 MPa to 649 MPa. Both models predicted measured apparent moduli equally well (R2 = 0.85), and were in excellent agreement with each other (R2 = 0.97). Tissue modulus was estimated at 9.0 GPa and 10.7 GPa for the beam FE and voxel FE models, respectively. On average, the beam models were solved in 219 s, reducing CPU usage up to 1150-fold as compared to 40 microm voxel FE models. Relative to 20 microm voxel models 10,000-fold reductions can be expected. The presented beam FE model is an abstraction of the intricate real trabecular structure using simple cylindrical beam elements. Nevertheless, it enabled an accurate prediction of global mechanical properties of microstructural bone. The strong reduction in CPU time provides the means to increase throughput, to analyze multiple loading configuration and to increase sample size, without increasing computational costs. With upcoming in vivo high-resolution imaging systems, this model has the potential to become a standard for mechanical characterization of bone.

Biomechanical Phenomena↗

PubMed1999

Long-term oral platelet glycoprotein IIb/IIIa receptor antagonism with sibrafiban after acute coronary syndromes: study design of the sibrafiban versus aspirin to yield maximum protection from ischemic heart events post-acute coronary syndromes (SYMPHONY) trial. Symphony Steering Committee.

BACKGROUND: Despite progress, atherosclerotic vascular disease remains a major cause of morbidity and mortality. Intravenous therapy with platelet glycoprotein (GP) IIb/IIIa receptor antagonists improves outcome in patients with acute coronary syndromes (ACS). Whether potent long-term antiplatelet therapy with oral GP IIb/IIIa antagonists will further improve outcome at a dose that is tolerable in long-term treatment is unknown. TRIAL DESIGN: SYMPHONY (Sibrafiban versus aspirin to Yield Maximum Protection from ischemic Heart events post-acute cOroNary syndromes) was a randomized, double-blind, aspirin-controlled trial with 2 concentration regimens of sibrafiban, an oral peptidomimetic GP IIb/IIIa antagonist, for long-term treatment instead of aspirin in patients after an ACS. Patients were eligible for SYMPHONY if they presented within 7 days of an ACS (>/=20 minutes of ischemic symptoms), had been clinically stable for at least 12 hours, and met one of the following inclusion criteria: ST-segment depression or elevation of at least 0.5 mm or new left bundle branch block with the ACS or elevated creatinine kinase MB more than the upper limit of normal and >3% of total creatine kinase or, if creatine kinase MB was not measured, an elevated level of troponin T or I. Approximately 9000 patients post ACS were randomized 1:1:1 to treatment with either aspirin (80 mg every 12 hours) or high-dose or low-dose sibrafiban every 12 hours. Assignment of tablet strength (3, 4.5, or 6 mg) within the sibrafiban arms was based on body weight and renal function to achieve a target steady-state plasma concentration. The duration of study drug therapy was 90 days. Patients who had intracoronary stenting during the course of the study initially received a blinded stent medication assignment for 2 to 4 weeks based on their initial randomization as follows: aspirin/ticlopidine 250 mg twice daily, low-dose sibrafiban/ticlopidine placebo twice daily, and high-dose sibrafiban/ticlopidine placebo twice daily. After the second interim safety assessment by the Data and Safety Monitoring Board the stent regimen for the low-dose group was modified to include ticlopidine 250 mg twice daily. END POINTS: The primary efficacy end point of SYMPHONY was the 90-day incidence of a composite of all-cause mortality, myocardial infarction or reinfarction, and severe recurrent ischemia. A clinical events classification committee was established to determine the end points of reinfarction and severe recurrent ischemia. The primary safety end points were the incidence of major bleeding or minor bleeding and the combined incidence of major and minor bleeding. Bleeding classification was done by computer algorithm. Tolerability was assessed by the rate of study drug discontinuation from bleeding.

Angina, Unstable↗

PubMed2003

Non-invasive mapping of connections between human thalamus and cortex using diffusion imaging.

Evidence concerning anatomical connectivities in the human brain is sparse and based largely on limited post-mortem observations. Diffusion tensor imaging has previously been used to define large white-matter tracts in the living human brain, but this technique has had limited success in tracing pathways into gray matter. Here we identified specific connections between human thalamus and cortex using a novel probabilistic tractography algorithm with diffusion imaging data. Classification of thalamic gray matter based on cortical connectivity patterns revealed distinct subregions whose locations correspond to nuclei described previously in histological studies. The connections that we found between thalamus and cortex were similar to those reported for non-human primates and were reproducible between individuals. Our results provide the first quantitative demonstration of reliable inference of anatomical connectivity between human gray matter structures using diffusion data and the first connectivity-based segmentation of gray matter.

Adult↗

PubMed2006

The role of single-photon emission computed tomography/computed tomography in benign and malignant bone disease.

Radiological (plain radiographs, computed tomography [CT], magnetic resonance imaging [MRI]) and nuclear medicine methods (bone scan, leukocyte scan) both provide unique information about the status of the skeleton. Both have typical strengths and weaknesses, which often lead to the sequential use of different procedures in daily routine. This use causes the unnecessary loss of time and sometimes money, if redundant information is obtained without establishing a final diagnosis. Recently, new devices for hybrid imaging (single-photon emission computed tomography/computed tomography [SPECT/CT], positron emission tomography/computed tomography [PET/CT]) were introduced, which allow for direct fusion of morphological (CT) and functional (SPECT, PET) data sets. With regard to skeletal abnormalities, this approach appears to be extremely useful because it combines the advantages of both techniques (high-resolution imaging of bone morphology and high sensitivity imaging of bone metabolism). By the accurate correlation of both, a new quality of bone imaging has now become accessible. Although researchers undertaking the initial studies exclusively used low-dose CT equipment, a new generation of SPECT/CT devices has emerged recently. By integrating high-resolution spiral CT, quality of bone imaging may improve once more. Ongoing prospective studies will have to show whether completely new diagnostic algorithms will come up for classification of bone disease as a consequence of this development. Besides, the role of ultrasonography and MRI for bone and soft-tissue imaging also will have to be re-evaluated. Looking at the final aim of all imaging techniques--to achieve correct diagnosis in a fast, noninvasive, comprehensive, and inexpensive way--we are now on the edge of a new era of multimodality imaging that will probably change the paths and structure of medicine in many ways. Presently, hybrid imaging using SPECT/CT has been proven to increase sensitivity and specificity of bone scintigraphy. This was mainly achieved by identifying benign bone conditions with increased bone turnover. Therefore, SPECT/CT should be applied whenever equivocal findings of planar bone imaging occur. It also helps to improve accuracy of leukocyte scanning to detect/exclude osteomyelitis and to define sites of inflammation. We therefore regard SPECT/CT as a valuable tool to optimize bone imaging, which might become even more important if new radiopharmaceuticals become available to image specific cell functions.

Bone Neoplasms↗

PubMed2004

Robust sparse hyperplane classifiers: application to uncertain molecular profiling data.

Molecular profiling studies can generate abundance measurements for thousands of transcripts, proteins, metabolites, or other species in, for example, normal and tumor tissue samples. Treating such measurements as features and the samples as labeled data points, sparse hyperplanes provide a statistical methodology for classifying data points into one of two categories (classification and prediction) and defining a small subset of discriminatory features (relevant feature identification). However, this and other extant classification methods address only implicitly the issue of observed data being a combination of underlying signals and noise. Recently, robust optimization has emerged as a powerful framework for handling uncertain data explicitly. Here, ideas from this field are exploited to develop robust sparse hyperplanes, i.e., classification and relevant feature identification algorithms that are resilient to variation in the data. Specifically, each data point is associated with an explicit data uncertainty model in the form of an ellipsoid parameterized by a center and covariance matrix. The task of learning a robust sparse hyperplane from such data is formulated as a second order cone program (SOCP). Gaussian and distribution-free data uncertainty models are shown to yield SOCPs that are equivalent to the SCOP based on ellipsoidal uncertainty. The real-world utility of robust sparse hyperplanes is demonstrated via retrospective analysis of breast cancer related transcript profiles. Data-dependent heuristics are used to compute the parameters of each ellipsoidal data uncertainty model. The generalization performance of a specific implementation, designated "robust LIKNON," is better than its nominal counterpart. Finally, the strengths and limitations of robust sparse hyperplanes are discussed.

Breast Neoplasms↗

PubMed2002

Conserved codon composition of ribosomal protein coding genes in Escherichia coli, Mycobacterium tuberculosis and Saccharomyces cerevisiae: lessons from supervised machine learning in functional genomics.

Genomics projects have resulted in a flood of sequence data. Functional annotation currently relies almost exclusively on inter-species sequence comparison and is restricted in cases of limited data from related species and widely divergent sequences with no known homologs. Here, we demonstrate that codon composition, a fusion of codon usage bias and amino acid composition signals, can accurately discriminate, in the absence of sequence homology information, cytoplasmic ribosomal protein genes from all other genes of known function in Saccharomyces cerevisiae, Escherichia coli and Mycobacterium tuberculosis using an implementation of support vector machines, SVM(light). Analysis of these codon composition signals is instructive in determining features that confer individuality to ribosomal protein genes. Each of the sets of positively charged, negatively charged and small hydrophobic residues, as well as codon bias, contribute to their distinctive codon composition profile. The representation of all these signals is sensitively detected, combined and augmented by the SVMs to perform an accurate classification. Of special mention is an obvious outlier, yeast gene RPL22B, highly homologous to RPL22A but employing very different codon usage, perhaps indicating a non-ribosomal function. Finally, we propose that codon composition be used in combination with other attributes in gene/protein classification by supervised machine learning algorithms.

Algorithms↗

PubMed1999

Automated model-based tissue classification of MR images of the brain.

We describe a fully automated method for model-based tissue classification of magnetic resonance (MR) images of the brain. The method interleaves classification with estimation of the model parameters, improving the classification at each iteration. The algorithm is able to segment single- and multispectral MR images, corrects for MR signal inhomogeneities, and incorporates contextual information by means of Markov random Fields (MRF's). A digital brain atlas containing prior expectations about the spatial location of tissue classes is used to initialize the algorithm. This makes the method fully automated and therefore it provides objective and reproducible segmentations. We have validated the technique on simulated as well as on real MR images of the brain.

Algorithms↗

PubMed2006

Using generalized additive models for construction of nonlinear classifiers in computer-aided diagnosis systems.

Several investigators have pointed out the possibility of using computer-aided diagnosis (CAD) schemes, as second readers, to help radiologists in the interpretation of images. One of the most important aspects to be considered when the diagnostic imaging systems are analyzed is the evaluation of their diagnostic performance. To perform this task, receiver operating characteristic curves are the method of choice. An important step in nearly all CAD systems is the reduction of false positives, as well as the classification of lesions, using different algorithms, such as neural networks or feature analysis, and several statistical methods. A statistical model more often employed is linear discriminant analysis (LDA). However, LDA implies several limitations in the type of variables that it can analyze. In this work, we have developed a novel approach, based on generalized additive models (GAMs), as an alternative to LDA, which can deal with a broad variety of variables, improving the results produced by using the LDA model. As an application, we have used GAM techniques for reducing the number of false detections in a computerized method to detect clustered microcalcifications, and we have compared this with the results obtained when LDA was applied. Employing LDA, the system achieved a sensitivity of 80.52% at a false-positive rate of 1.90 false detections per image. With the GAM, the sensitivity increased to 83.12% and 1.46 false positives per image.

Algorithms↗

PubMed2006

On weighting clustering.

Recent papers and patents in iterative unsupervised learning have emphasized a new trend in clustering. It basically consists of penalizing solutions via weights on the instance points, somehow making clustering move toward the hardest points to cluster. The motivations come principally from an analogy with powerful supervised classification methods known as boosting algorithms. However, interest in this analogy has so far been mainly borne out from experimental studies only. This paper is, to the best of our knowledge, the first attempt at its formalization. More precisely, we handle clustering as a constrained minimization of a Bregman divergence. Weight modifications rely on the local variations of the expected complete log-likelihoods. Theoretical results show benefits resembling those of boosting algorithms and bring modified (weighted) versions of clustering algorithms such as k-means, fuzzy c-means, Expectation Maximization (EM), and k-harmonic means. Experiments are provided for all these algorithms, with a readily available code. They display the advantages that subtle data reweighting may bring to clustering.

Algorithms↗

PubMed2006

Validating an asthma case detection instrument in a Head Start sample.

Although specific tests screen children in preschool programs for vision, hearing, and dental conditions, there are no published validated instruments to detect preschool-age children with asthma, one of the most common pediatric chronic conditions affecting children in economically disadvantaged communities of color. As part of an asthma education intervention, a 15-item Brief Respiratory Questionnaire (BRQ) was developed to detect children with probable asthma in Head Start and subsidized preschool settings in communities with high asthma prevalence and associated morbidity. Preschool personnel administered the BRQ to consenting parents of 419 enrolled children. Trained interviewers administered validation interviews (VALs) to parents of 149 case-positive children and 51 case-negative children. Three physicians independently assessed deidentified summaries of the VALs that captured responses about signs and symptoms of asthma, diagnosis and treatment, and use of medical services. The physicians' assessments of the summarized VALs were the validated standard to which the BRQ classifications were compared. A simple algorithm of 4 items was identified that can be administered and scored by nonmedical preschool personnel in less than 5 minutes. The chance-corrected agreement between these 4 items of the BRQ and the VAL was good: kappa, .73 (95% confidence interval, 0.62-0.84); specificity, 96%; sensitivity, 73%; and positive predictive value, 97%. The BRQ appears to be a valid instrument for detecting children with probable asthma in Head Start and other subsidized preschool settings in communities with high prevalence of asthma.

Allergens↗