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 1,549 records · Page 86Linked to original sources

PubMed2002

Data mining the protein data bank: residue interactions.

The protein databank contains a vast wealth of structural and functional information. The analysis of this macromolecular information has been the subject of considerable work in order to advance knowledge beyond the collection of molecular coordinates. This article presents a method that determines local structural information within proteins using mathematical data mining techniques. The mine program described returns many known configurations of residues such as the catalytic triad, metal binding sites and the N-linked glycosylation site; as well as many other multiple residue interactions not previously categorized. Because mathematical constructs are used as targets, this method can identify new information not previously known, and also provide unbiased results of typical structure and their expected deviations. Because the results are defined mathematically, they cannot indicate the biological implications of the results. Therefore two support programs are described that provide insight into the biological context for the mine results. The first allows a weighted RMSD search between a template set of coordinates and a list of PDB files, and the second allows the labeling of a protein with the template results from mining to aid in the classification of this protein.

Algorithms↗

PubMed1990

Self-organization of associative memory and pattern classification: recurrent signal processing on topological feature maps.

We extend the neural concepts of topological feature maps towards self-organization of auto-associative memory and hierarchical pattern classification. As is well-known, topological maps for statistical data sets store information on the associated probability densities. To extract that information we introduce a recurrent dynamics of signal processing. We show that the dynamics converts a topological map into an auto-associative memory for real-valued feature vectors which is capable to perform a cluster analysis. The neural network scheme thus developed represents a generalization of non-linear matrix-type associative memories. The results naturally lead to the concept of a feature atlas and an associated scheme of self-organized, hierarchical pattern classification.

Algorithms↗

PubMed2006

Unsupervised classification of single particles by cluster tracking in multi-dimensional space.

In cryo-electron microscopy (cryo-EM) single-particle reconstruction, the heterogeneity of two-dimensional projection image data resulting from the co-existence of different conformational or ligand binding states of a macromolecular complex remains a major obstacle as it impairs the validity of reconstructed density maps and limits the progress toward higher resolution. Classification of cryo-EM data according to the different conformations is difficult because of the coexistence of multiple orientations in a single dataset. Here, we present an unsupervised classification method, termed cluster tracking, which utilizes the continuity in multi-dimensional space induced by angular adjacency of projections in large datasets. In a proof of concept, the testing of cluster tracking on simulated projection data, which were generated from multiple conformations and orientations of an existing volume, produced clusters that are consistent with the conformational identity of the data. The application of the method to experimental cryo-EM projection data is found to result in a partition similar to the one generated by supervised classification.

Algorithms↗

PubMed1990

Automatic computer detection of clustered calcifications in digital mammograms.

The automatic detection of clusters of calcifications in digital mammograms has been investigated using image analysis techniques. The calcifications were segmented from the background of normal breast structure in the mammogram using a local area thresholding process. This procedure also identified other breast structures and the digital image properties of all segmented objects were analysed to extract clusters of calcifications. Seventy five clinical mammograms were digitised. These were divided into training and test sets of 25 and 50 films respectively. The results for the test set of 50 complete clinical mammograms show that the computer system achieves a 25/25 true positive film classification (i.e. those containing clusters of calcifications) with false positive clusters detected in 4/50 films. There were no false negative film classifications.

Algorithms↗

PubMed2002

Language trees and zipping.

In this Letter we present a very general method for extracting information from a generic string of characters, e.g., a text, a DNA sequence, or a time series. Based on data-compression techniques, its key point is the computation of a suitable measure of the remoteness of two bodies of knowledge. We present the implementation of the method to linguistic motivated problems, featuring highly accurate results for language recognition, authorship attribution, and language classification.

Algorithms↗

PubMed2003

Uncertainty in the output of artificial neural networks.

Analysis of the performance of artificial neural networks (ANNs) is usually based on aggregate results on a population of cases. In this paper, we analyze ANN output corresponding to the individual case. We show variability in the outputs of multiple ANNs that are trained and "optimized" from a common set of training cases. We predict this variability from a theoretical standpoint on the basis that multiple ANNs can be optimized to achieve similar overall performance on a population of cases, but produce different outputs for the same individual case because the ANNs use different weights. We use simulations to show that the average standard deviation in the ANN output can be two orders of magnitude higher than the standard deviation in the ANN overall performance measured by the Az value. We further show this variability using an example in mammography where the ANNs are used to classify clustered microcalcifications as malignant or benign based on image features extracted from mammograms. This variability in the ANN output is generally not recognized because a trained individual ANN becomes a deterministic model. Recognition of this variability and the deterministic view of the ANN present a fundamental contradiction. The implication of this variability to the classification task warrants additional study.

Algorithms↗

PubMed1998

Automatic quantification of immunohistochemically stained cell nuclei using unsupervised image analysis.

A method for quantification of images of immunohistochemically stained cell nuclei by computing area proportions is presented. The image is transformed by a principal component transform. The resulting first component image is used to segment the objects from the background using dynamic thresholding of the P2/A-histogram, where P2/A is a global roundness measure. Then the image is transformed into principal component hue, defined as the angle around the first principal component. This image is used to segment positive and negative objects. The method is fully automatic and the principal component approach makes it robust with respect to illumination and focus settings. An independent test set consisting of images grabbed with different focus and illumination for each field of view was used to test the method, and the proposed method showed less variation than the intraoperator variation using supervised Maximum Likelihood classification.

Algorithms↗

PubMed2003

Ensemble machine learning on gene expression data for cancer classification.

Whole genome RNA expression studies permit systematic approaches to understanding the correlation between gene expression profiles to disease states or different developmental stages of a cell. Microarray analysis provides quantitative information about the complete transcription profile of cells that facilitate drug and therapeutics development, disease diagnosis, and understanding in the basic cell biology. One of the challenges in microarray analysis, especially in cancerous gene expression profiles, is to identify genes or groups of genes that are highly expressed in tumour cells but not in normal cells and vice versa. Previously, we have shown that ensemble machine learning consistently performs well in classifying biological data. In this paper, we focus on three different supervised machine learning techniques in cancer classification, namely C4.5 decision tree, and bagged and boosted decision trees. We have performed classification tasks on seven publicly available cancerous microarray data and compared the classification/prediction performance of these methods. We have observed that ensemble learning (bagged and boosted decision trees) often performs better than single decision trees in this classification task.

Algorithms↗

PubMed2006

Binary tree-based generic demosaicking algorithm for multispectral filter arrays.

In this paper, we extend the idea of using mosaicked color filter array (CFA) in color imaging, which has been widely adopted in the digital color camera industry, to the use of multispectral filter array (MSFA) in multispectral imaging. The filter array technique can help reduce the cost, achieve exact registration, and improve the robustness of the imaging system. However, the extension from CFA to MSFA is not straightforward. First, most CFAs only deal with a few bands (3 or 4) within the narrow visual spectral region, while the design of MSFA needs to handle the arrangement of multiple bands (more than 3) across a much wider spectral range. Second, most existing CFA demosaicking algorithms assume the fixed Bayer CFA and are confined to properties only existed in the color domain. Therefore, they cannot be directly applied to multispectral demosaicking. The main challenges faced in multispectral demosaicking is how to design a generic algorithm that can handle the more diversified MSFA patterns, and how to improve performance with a coarser spatial resolution and a less degree of spectral correlation. In this paper, we present a binary tree based generic demosaicking method. Two metrics are used to evaluate the generic algorithm, including the root mean-square error (RMSE) for reconstruction performance and the classification accuracy for target discrimination performance. Experimental results show that the demosaicked images present low RMSE (less than 7) and comparable classification performance as original images. These results support that MSFA technique can be applied to multispectral imaging with unique advantages.

Algorithms↗

PubMed1994

Digital signal processing chip implementation for detection and analysis of intracardiac electrograms.

The adoption of digital signal processing (DSP) microchips for detection and analysis of electrocardiographic signals offers a means for increased computational speed and the opportunity for design of customized architecture to address real-time requirements. A system using the Motorola 56001 DSP chip has been designed to realize cycle-by-cycle detection (triggering) and waveform analysis using a time-domain template matching technique, correlation waveform analysis (CWA). The system digitally samples an electrocardiographic signal at 1000 Hz, incorporates an adaptive trigger for detection of cardiac events, and classifies each waveform as normal or abnormal. Ten paired sets of single-chamber bipolar intracardiac electrograms (1-500 Hz) were processed with each pair containing a sinus rhythm (SR) passage and a corresponding arrhythmia segment from the same patient. Four of ten paired sets contained intraatrial electrograms that exhibited retrograde atrial conduction during ventricular pacing; the remaining six paired sets of intraventricular electrograms consisted of either ventricular tachycardia (4) or paced ventricular rhythm (2). Of 2,978 depolarizations in the test set, the adaptive trigger failed to detect 6 (99.8% detection sensitivity) and had 11 false triggers (99.6% specificity). Using patient dependent thresholds for CWA to classify waveforms, the program correctly identified 1,175 of 1,197 (98.2% specificity) sinus rhythm depolarizations and 1,771 of 1,781 (99.4% sensitivity) abnormal depolarizations. From the results, the algorithm appears to hold potential for applications such as real-time monitoring of electrophysiology studies or detection and classification of tachycardias in implantable antitachycardia devices.

Algorithms↗

PubMed2004

[Implication of probability analysis for diagnosis of stable coronary diseases. Comparison of non-invasive diagnostic tests].

UNLABELLED: A prospective collection of clinical (history), electrocardiography (ECG), exercise electrocardiography (EE), dobutamine stress echocardiography (SE) and catheterization data was performed in 551 patients with chest pain regarded as definite or probable stable angina pectoris. All patients (65% male) without prior history of myocardial infarction undergone clinical evaluation: chest pain classification according to Diamond, ECG, EE (regarded as positive on the basis of > or = 1 mm ST-depression), SE (ischemia was defined as new or worsening wall motion abnormalities using a 16-segment model) and coronary angiography (CA):CAD was defined as > or = 50% narrowing of at least one major vessel. Sensitivity and specificity of ECG, EE, SE was calculated: and results are as follows respectively: 23% and 87%, 93% and 21%, 85% and 69%. On the basis of these results with implication of probability analysis two algorithms were developed by computer program. They incorporated pretest variables: age, gender, chest pain classification, ECG and results of one or two non-invasive tests: EE and (or) SE. The sensitivity and specificity of the algorithm is 96% and 44%. CONCLUSIONS: 1. Chest pain characteristics remains an effective tool for estimating probability of CAD. 2. The diagnostic value of resting electrocardiography in stable CAD is low. 3. SE has comparably sensitivity but significantly higher specificity than EE. 4. We demonstrated that when the likelihood of CAD is high on the basis of initial evaluation, diagnostic non-invasive testing is not indicated before CA. 5. When the probability is intermediate or low, implementation of first choice test should be different in women (SE) and men (EE).

Adult↗

PubMed1992

Case-mix groups for VA hospital-based home care.

The purpose of this study is to group hospital-based home care (HBHC) patients homogeneously by their characteristics with respect to cost of care to develop alternative case mix methods for management and reimbursement (allocation) purposes. Six Veterans Affairs (VA) HBHC programs in Fiscal Year (FY) 1986 that maximized patient, program, and regional variation were selected, all of which agreed to participate. All HBHC patients active in each program on October 1, 1987, in addition to all new admissions through September 30, 1988 (FY88), comprised the sample of 874 unique patients. Statistical methods include the use of classification and regression trees (CART software: Statistical Software; Lafayette, CA), analysis of variance, and multiple linear regression techniques. The resulting algorithm is a three-factor model that explains 20% of the cost variance (R2 = 20%, with a cross validation R2 of 12%). Similar classifications such as the RUG-II, which is utilized for VA nursing home and intermediate care, the VA outpatient resource allocation model, and the RUG-HHC, utilized in some states for reimbursing home health care in the private sector, explained less of the cost variance and, therefore, are less adequate for VA home care resource allocation.

Aftercare↗

PubMed2006

Prevalence-dependent decision limits for the early detection of type 2 diabetes mellitus in venous blood, venous plasma and capillary blood during glucose challenge.

BACKGROUND: The glycemia decision limits recommended by WHO/ADA for type 2 diabetes detection are derived from clinical signs in advanced stages of the disease. Since insulin secretion patterns and sensitivitity are impaired at the beginning of type 2 diabetes, this stage may be better suited to identify decision limits with higher diagnostic efficiency than those currently applied. METHODS: Oral glucose tolerance tests were performed in 300 subjects. Glucose concentrations were measured at 30-min intervals in venous plasma, venous blood and capillary blood. Insulin concentrations in venous plasma, an insulin sensitivity index and body mass index were used to indicate a type 2 diabetic state. A multiple logistic regression procedure was "trained" using only subjects "clearly" considered to be non-diseased or diseased based on an oral glucose tolerance test according to WHO criteria. This insulin algorithm was applied to the whole study group, leading to definitive classification into the non-diseased or the diseased group. This a posteriori classification was used to identify cutoff values with the highest diagnostic efficiency. RESULTS: The diagnostic efficiency was significantly higher when decision limits lower than the WHO recommendations for glucose concentrations were applied in a preselected subpopulation and in all three sample systems tested, e.g., 9.49 mmol/L (171 mg/dL) for venous plasma and 8.94 mmol/L (161 mg/dL) for capillary blood in the 2-h post-load state. The optimized and WHO 2-h cutoff values corresponded to a disease prevalence of 28% and approximately 5% (20% in the fasting state), respectively. Diagnostic efficiency was higher in the 2-h post-load than in the fasting state. Combining fasting values with 2-h post-load values did not further improve the diagnostic efficiency. Glucose concentrations determined from capillary blood were as efficient as those from venous blood or plasma. The number of diabetic subjects detected differed considerably between capillary blood and venous plasma for the WHO/ADA cutoff values, but not for the optimized cutoff values. CONCLUSIONS: The efficiency of type 2 diabetes diagnosis can be improved by optimizing cutoff values according to disease prevalence. Unexpectedly, the optimized 2-h post-load cutoff was lower for capillary blood than for venous plasma. It is proposed to identify a risk group e.g., by characteristics of the metabolic syndrome in which the 2-h post-challenge concentration is determined using lower cut-off values than presently recommended.

Adult↗

PubMed2005

Gene selection from microarray data for cancer classification--a machine learning approach.

A DNA microarray can track the expression levels of thousands of genes simultaneously. Previous research has demonstrated that this technology can be useful in the classification of cancers. Cancer microarray data normally contains a small number of samples which have a large number of gene expression levels as features. To select relevant genes involved in different types of cancer remains a challenge. In order to extract useful gene information from cancer microarray data and reduce dimensionality, feature selection algorithms were systematically investigated in this study. Using a correlation-based feature selector combined with machine learning algorithms such as decision trees, naïve Bayes and support vector machines, we show that classification performance at least as good as published results can be obtained on acute leukemia and diffuse large B-cell lymphoma microarray data sets. We also demonstrate that a combined use of different classification and feature selection approaches makes it possible to select relevant genes with high confidence. This is also the first paper which discusses both computational and biological evidence for the involvement of zyxin in leukaemogenesis.

Algorithms↗

PubMed2004

Automated linking of free-text complaints to reason-for-visit categories and International Classification of Diseases diagnoses in emergency department patient record databases.

STUDY OBJECTIVE: The use of the International Classification of Diseases system to describe emergency department (ED) case mix has disadvantages. We therefore developed computer algorithms that recognize a combination of words, word fragments, and word patterns to link free-text complaint fields to 20 reason-for-visit categories. We examine the feasibility and reliability of applying these reason-for-visit categories to ED patient-visit databases. METHODS: We analyzed a database (containing complaints and International Classification of Diseases diagnoses for 1 year's visits to a single ED) using a 3-step process (create initial terms, maximize sensitivity, maximize specificity) to define inclusion and exclusion terms for 20 reason-for-visit categories. To assess the reliability of the reason-for-visit assignment algorithm, we repeated the final 2 steps on a second database, composed of visits sampled from 21 EDs. For each database, we determined the prevalence of complaints that link to each reason-for-visit category and the distributions of International Classification of Diseases, Ninth Revision diagnoses that resulted for all patients and patients stratified by age. RESULTS: The 20 reason-for-visit categories capture 77% of all patients in database 1 (mean age 33.5 years) and 67% of all patients in database 2 (mean age 38.9 years). The percentage of visits captured by the 20 reason-for-visit categories, by age range, for databases 1 and 2 are (respectively) 0 to 2 years (84% and 76%), 3 to 10 years (82% and 74%), 11 to 65 years (76% and 68%), and 66 years or older (69% and 60%). The proportions of all complaints that link to each reason-for-visit category are largely similar between databases. Every complaint field that is linked to each reason-for-visit category includes at least 1 term that relates it to the category title, and the most frequently assigned diagnoses in each reason-for-visit category are those that one would expect to be associated with the reason-for-visit category complaints. CONCLUSION: The method by which free-text complaint fields are parsed into reason-for-visit categories is feasible and reasonably reliable; the finalized database 1 reason-for-visit category inclusion/exclusion terms lists required only modest changes to work well in database 2. The reason-for-visit categories used here are broadly defined to maximize the proportion of visits that they capture; more narrowly defined reason-for-visit categories will require more extensive revision of their inclusion/exclusion terms lists when used in different databases. A prospective, reason-for-visit-based ED classification system could have several useful applications (including syndromic surveillance), although content validity analysis will be necessary to investigate this hypothesis.

Adolescent↗

PubMed2003

Selection of fusion levels in scoliosis surgery.

Fusion levels are selected with the goal of having a balanced physiologic spine on the postoperative films. We have proposed an algorithm for selecting fusion levels based on the King-Moe scoliosis classification scheme and the flexibility of a compensatory curve. We propose the acronym TEAMS as a pneumonic. Curve type defines the region of the spine to be fused. The ends of the Cobb measurement around the region are the temporary endpoints. The endpoints are extended to avoid an apex due to the sagittal or frontal planes. The mobile disc is defined at the caudal end of the fusion area. Finally, the bending films, traction films, or intraoperative films are used to define the stable zone of Harrington that must contain the ends of the fusion mass. This scheme yields a simple, easily taught algorithm that results in reproducible endpoints for scoliosis fusion.

Adolescent↗

PubMed2004

A new scoring function and associated statistical significance for structure alignment by CE.

A new scoring function for assessing the statistical significance of protein structure alignment has been developed. The new scores were tested empirically using the combinatorial extension (CE) algorithm. The significance of a given score was given a p-value by curve-fitting the distribution of the scores generated by a random comparison of proteins taken from the PDB_SELECT database and the structural classification of proteins (SCOP) database. Although the scoring function was developed based on the CE algorithm, it is portable to any other protein structure alignment algorithm. The new scoring function is examined by sensitivity, specificity, and ROC curves.

Algorithms↗

PubMed1999

Patients with newly diagnosed carcinoma of the breast: validation of a claim-based identification algorithm.

The objectives of this study were to validate a claims-based algorithm for identification of patients with newly diagnosed carcinoma of the breast and to optimize the algorithm. Claims data from all females aged 21 years or older who enrolled in a large California health maintenance organization during the study period from October 1, 1994 through March 31, 1996 were analyzed. Medical records of the patients identified through the claims-based algorithm were reviewed to determine whether the patients were correctly identified. The initial algorithm had a positive predictive value of 84% which was similar to the previous study. The percentages of correct identification significantly increased with the patient's age at diagnosis. Other patient demographic characteristics and facility characteristics were not related to the accuracy of the identification. Using a classification tree procedure and additional information from the false-positive cases, the initial algorithm was modified for improvement. The best-modified algorithm had a positive predictive value of 92% while only 0.5% (4/837) of the true-positive cases were excluded. The results once again demonstrated that patients with newly diagnosed carcinomas of the breast can be identified using claims data. These databases provide an efficient and effective tool for performing health services studies on large patient populations.

Adult↗