Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Classification Algorithms”

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 433 records · Page 24Linked to original sources
PubMed2006

Spectral analysis of two-signed microarray expression data.

We give a simple and informative derivation of a spectral algorithm for clustering and reordering complementary DNA microarray expression data. Here, expression levels of a set of genes are recorded simultaneously across a number of samples, with a positive weight reflecting up-regulation and a negative weight reflecting down-regulation. We give theoretical support for the algorithm based on a biologically justified hypothesis about the structure of the data, and illustrate its use on public domain data in the context of unsupervised tumour classification. The algorithm is derived by considering a discrete optimization problem and then relaxing to the continuous realm. We prove that in the case where the data have an inherent 'checkerboard' sign pattern, the algorithm will automatically reveal that pattern. Further, our derivation shows that the algorithm may be regarded as imposing a random graph model on the expression levels and then clustering from a maximum likelihood perspective. This indicates that the output will be tolerant to perturbations and will reveal 'near-checkerboard' patterns when these are present in the data. It is interesting to note that the checkerboard structure is revealed by the first (dominant) singular vectors--previous work on spectral methods has focussed on the case of nonnegative edge weights, where only the second and higher singular vectors are relevant. We illustrate the algorithm on real and synthetic data, and then use it in a tumour classification context on three different cancer data sets. Our results show that respecting the two-signed nature of the data (thereby distinguishing between up-regulation and down-regulation) reveals structures that cannot be gleaned from the absolute value data (where up- and down-regulation are both regarded as 'changes').

Algorithms↗
PubMed1991

[Experience with computer-assisted EMG analysis: reference values for four muscles].

The characteristics of a newly developed computerized EMG-analysis system are lined out. The analysis steps include segmentation and parametrization of digitized EMG segments followed by automatic classification of single potentials into motor units. Comparison of parameters is used rather than a template matching algorithm for classification. For each motor unit a set of parameters can be computed in addition to duration and amplitude. Normative data are given for four muscles. 60 normal volunteers (50% female, 50% below age 50) were recorded from for each muscle. The recording conditions required to move the electrode until a motor unit could be recorded with its individual maximum. Furthermore, it was attempted to attain the maximum discharge rate of a particular motor unit. A highly significant correlation of the amplitude of the motor unit potentials with age was found for all four muscles. For the duration of the potentials a correlation with age was found only for the M. ext. digitorum brevis. Significant sex differences were found for nearly all parameters and all muscles. Pairwise comparisons of the muscles revealed significant differences of all muscles. The analysis showed, that a muscle can be regarded pathological if more than two of a total of seven recorded motor units show pathological parameters.

Age Factors↗
PubMed1997

Automated computer-assisted detection of follicles in ultrasound images of ovary.

Monitoring the follicles in women's ovaries is especially important in human reproduction. Today, the monitoring of follicles is done with human interaction. Such monitoring can be very demanding and inaccurate, and in most cases signifies additional burdens for the experts. In this paper, a new algorithm for automated computer-assisted detection of follicles in the ultrasound images of the ovary is proposed. It has a typical object recognition scheme (preprocessing, segmentation, and classification). The algorithm is based on the following idea: first, the ovary is estimated (coarsely) and then follicles are searched for. The methods used are known from literature (despeckle filter, Kirsch's operator, optimal thresholding, thinning, shape descriptions, classification), and the majority of our work was done experimenting with these methods and selecting the appropriate thresholds. The algorithm's computational complexity is of order of O(n2), which means about 6 min of processing time per an ultrasound image of dimensions of 768 x 576 pixels on HP 715 machines. It has been tested on a set of 20 real ultrasound images of the ovary. The recognition rate of follicles with these procedures was around 62%. The algorithm is not perfect, but it will be further modified and improved, as indicated in our conclusions.

Algorithms↗
PubMed2006

A genetic algorithm design for microcalcification detection and classification in digital mammograms.

In this paper, we propose a genetic algorithm design to automatically classify and detect micocalcification clusters in digital mammograms. The proposed GA technique is characterised by transforming input images into a feature domain, where each pixel is represented by its mean and standard deviation inside a surrounding window of size 9 x 9 pixel. In the feature domain, chromosomes are constructed to populate the initial generation and further features are extracted to enable the proposed GA to search for optimised classification and detection of microcalcification clusters via regions of 128 x 128 pixels. Extensive experiments show that the proposed GA design is able to achieve high performances in microcalcification classification and detection, which are measured by ROC curves, sensitivity against specificity, areas under ROC curves and benchmarked by existing representative techniques.

Algorithms↗
PubMed2006

Short-term prediction of mortality in patients with systemic lupus erythematosus: classification of outcomes using random forests.

OBJECTIVE: To identify demographic and clinical characteristics that classify patients with systemic lupus erythematosus (SLE) at risk for in-hospital mortality. METHODS: Patients hospitalized in California from 1996 to 2000 with a principal diagnosis of SLE (N = 3,839) were identified from a state hospitalization database. As candidate predictors of mortality, we used patient demographic characteristics; the presence or absence of 40 different clinical conditions listed among the discharge diagnoses; and 2 summary indexes derived from the discharge diagnoses, the Charlson Index and the SLE Comorbidity Index. Predictors of patients at increased risk of mortality were identified and validated using random forests, a statistical procedure that is a generalization of single classification trees. Random forests use bootstrapped samples of patients and randomly selected subsets of predictors to create individual classification trees, and this process is repeated to generate multiple trees (a forest). Classification is then done by majority vote across all trees. RESULTS: Of the 3,839 patients, 109 died during hospitalization. Selecting from all available predictors, the random forests had excellent predictive accuracy for classification of death. The mean classification error rate, averaged over 10 forests of 500 trees each, was 11.9%. The most important predictors were the Charlson Index, respiratory failure, SLE Comorbidity Index, age, sepsis, nephritis, and thrombocytopenia. CONCLUSION: Information on clinical diagnoses can be used to accurately predict mortality among hospitalized patients with SLE. Random forests represent a useful technique to identify the most important predictors from a larger (often much larger) number and to validate the classification.

Adult↗
PubMed2001

ECG beat classification by a novel hybrid neural network.

This paper presents a novel hybrid neural network structure for the classification of the electrocardiogram (ECG) beats. Two feature extraction methods: Fourier and wavelet analyses for ECG beat classification are comparatively investigated in eight-dimensional feature space. ECG features are determined by dynamic programming according to the divergence value. Classification performance, training time and the number of nodes of the multi-layer perceptron (MLP), restricted Coulomb energy (RCE) and a novel hybrid neural network are comparatively presented. In order to increase the classification performance and to decrease the number of nodes, the novel hybrid structure is trained by the genetic algorithms (GAs). Ten types of ECG beats obtained from the MIT-BIH database and from a real-time ECG measurement system are classified with a success of 96% by using the hybrid structure.

Algorithms↗
PubMed2001

Classifying vertical facial deformity using supervised and unsupervised learning.

OBJECTIVES: To evaluate the potential for machine learning techniques to identify objective criteria for classifying vertical facial deformity. METHODS: 19 parameters were determined from 131 lateral skull radiographs. Classifications were induced from raw data with simple visualisation, C5.0 and Kohonen feature maps; and using a Point Distribution Model (PDM) of shape templates comprising points taken from digitised radiographs. RESULTS: The induced decision trees enable a direct comparison of clinicians' idiosyncrasies in classification. Unsupervised algorithms induce models that are potentially more objective, but their blackbox nature makes them unsuitable for clinical application. The PDM methodology gives dramatic visualisations of two modes separating horizontal and vertical facial growth. Kohonen feature maps favour one clinician and PDM the other. Clinical response suggests that while Clinician 1 places greater weight on 5 of 6 parameters, Clinician 2 relies on more parameters that capture facial shape. CONCLUSIONS: While machine learning and statistical analyses classify subjects for vertical facial height, they have limited application in their present form. The supervised learning algorithm C5.0 is effective for generating rules for individual clinicians but its inherent bias invalidates its use for objective classification of facial form for research purposes. On the other hand, promising results from unsupervised strategies (especially the PDM) suggest a potential use for objective classification and further identification and analysis of ambiguous cases. At present, such methodologies may be unsuitable for clinical application because of the invisibility of their underlying processes. Further study is required with additional patient data and a wider group of clinicians.

Algorithms↗
PubMed2005

Identification and quantification of industrial grade glycerol adulteration in red wine with fourier transform infrared spectroscopy using chemometrics and artificial neural networks.

Fourier transform infrared (FT-IR) single bounce micro-attenuated total reflectance (mATR) spectroscopy, combined with multivariate and artificial neural network (ANN) data analysis, was used to determine the adulteration of industrial grade glycerol in selected red wines. Red wine samples were artificially adulterated with industrial grade glycerol over the concentration range from 0.1 to 15% and calibration models were developed and validated. Single bounce infrared spectra of glycerol adulterated wine samples were recorded in the fingerprint mid-infrared region, 900-1500 cm(-1). Partial least squares (PLS) and PLS first derivatives were used for quantitative analysis (r2 = 0.945 to 0.998), while linear discriminant analysis (LDA) and canonical variate analysis (CVA) were used for classification and discrimination. The standard error of prediction (SEP) in the validation set was between 1.44 and 2.25%. Classification of glycerol adulterants in the different brands of red wine using CVA resulted in a classification accuracy in the range between 94 and 98%. Artificial neural network analysis based on the quick back propagation network (BPN) and the radial basis function network (RBFN) algorithms had classification success rates of 93% using BPN and 100% using RBFN. The genetic algorithm network was able to predict the concentrations of glycerol in wine up to an accuracy of r2 = 0.998.

Algorithms↗
PubMed2007

Computer algorithm for automated work group classification from free text: the DREAM technique.

OBJECTIVE: This study developed and tested a computer method to automatically assign subjects to aggregate work groups based on their free text work descriptions. METHODS: The Double Root Extended Automated Matcher (DREAM) algorithm classifies individuals based on pairs of subjects' free text word roots in common with those of standard classification systems and several explicitly defined linkages between term roots and aggregates. RESULTS: DREAM effectively analyzed free text from 5887 participants in a multisite chronic obstructive pulmonary disease prevention study (Lung Health Study). For a test set of 533 cases, DREAMs classifications compared favorably with those of a four-human panel. The humans rated the accuracy of DREAM as good or better in 80% of the test cases. CONCLUSIONS: Automated text interpretation is a promising tool for analyzing large data sets for applications in data mining, research, and surveillance. Work descriptive information is most useful when it can link an individual to aggregate entities that have occupational health relevance. Determining the appropriate group requires considerable expertise. This article describes a new method for making such assignments using a computer algorithm to reduce dependence on the limited number of occupational health experts. In addition, computer algorithms foster consistency of assignments.

Algorithms↗
PubMed1994

Growing a classification tree using the apparent misclassification rate.

A method to determine the size of a classification tree is proposed. This method is based on the change of the apparent misclassification rate (AMR) of the tree at each growing stage. The method is simple and fast compared to the other classification tree methods, which are based on minimizing a cost complexity function. To test the method, it was used to classify species of fungi, and the results are in good agreement with those obtained by linear discriminant analysis. Also, 21 proteins with known structures and functions were classified using the proposed method. For this purpose the coefficient of variation for several properties of the secondary structures of these proteins has been used. Again, the results were in good agreement with the classification obtained previously using dynamic programming.

Algorithms↗
PubMed1999

Time-frequency MEG-MUSIC algorithm.

We propose a method that incorporates the time-frequency characteristics of neural sources into magnetoencephalographic (MEG) source estimation. The method is based on the multiple-signal-classification (MUSIC) algorithm and it calculates a time--frequency matrix in which diagonal and off-diagonal terms are the auto and crosstime--frequency distributions of multichannel MEG recordings, respectively. The method averages this time-frequency matrix over the time--frequency region of interest. The locations of neural sources are then estimated by checking the orthogonality between the noise subspace of this averaged matrix and the sensor lead field. Accordingly, the method allows us to estimate the locations of neural sources from each time--frequency component. A computer simulation was performed to test the validity of the proposed method, and the results demonstrate its effectiveness.

Algorithms↗
PubMed1996

Methodology of serial ECG classification using an adaptation of the NOVACODE for Q wave myocardial infarction in the Bypass Angioplasty Revascularization Investigation (BARI).

Serial electrocardiographic (ECG) changes are a critical component of the diagnostic algorithm for classification of myocardial ischemic events in large-scale clinical trials. This study describes a computerized serial ECG classification program developed at the St. Louis University Core ECG Laboratory for use in the Bypass Angioplasty Revascularization Investigation (BARI) trial, in which patients with multivessel coronary artery disease were randomized to receive either coronary artery bypass grafting or percutaneous transluminal coronary angioplasty. The St. Louis University program detects and codes serial changes in Q, ST, and T wave items according to Minnesota code (MC) criteria using a modified NOVACODE hierarchical classification system. Measurements using a seven-power calibrated coding loupe are used to generate the MC from a customized software program. Significant minor or major changes are detected by the serial comparison program and referred to a physician coder for verification. Serial comparison coding rules are used to adjust for weaknesses in the standard MC classification system resulting from instability at decision boundaries. Of 4,244 BARI randomized and registry study participants with follow-up ECGs received at the Core ECG Laboratory as of March 1995, a grade 2 MC Q wave progression was noted in 568 participants (13.4%) using MC criteria alone, as compared with 367 (8.6%) after the St. Louis University coding rules were applied. The incidence of grade 1 MC Q wave progressions was 16.4% (697/4,244) versus 6.1% (259/4,244) when the St. Louis University program was applied. Intraobserver variability for grade 2 Q wave progression codes determined from a sample of 812 serial.

Algorithms↗
PubMed2004

Neural spike classification using parallel selection of all algorithm parameters.

The Forster-Handwerker template-matching algorithm (J. Neurosci. Methods 31 (1990) 109) classifies neuronal spikes according to three parameters selected by the experimenter prior to running the algorithm. Thousands of different combinations of these parameter values are possible producing hundreds of different classifications for each input file. Using a 40-processor Linux-based parallel computing cluster, we ran their algorithm with an effective sampling of all combinations of parameter values in order to generate a list of the classifications that can be generated by the algorithm. A distance measure was used to quantify the similarity between classifications and then to create a distance table containing entries for the distances between all pairs of classifications. Using a self-organizing neural network (SON) and the distance table we group the classifications by similarity and select the best representative classifications that the Forster-Handwerker algorithm can produce.

Action Potentials↗
PubMed2026

A cfDNA fragmentomics classifier for noninvasive differentiation of benign and malignant renal masses.

Noninvasive differentiation of malignant and benign renal masses remains a major clinical challenge, particularly for radiologically indeterminate lesions. Here, we developed and validated a plasma cell-free DNA (cfDNA) fragmentomics-based machine learning classifier for renal mass characterization. The model was trained on 331 participants (171 cancer, 160 benign) and independently validated on 144 participants (73 cancer, 71 benign). Three cfDNA fragmentation features, including copy number variation (CNV), fragmentation-based methylation (FRAGMA), and nucleosome footprint (NF), derived from low-pass whole-genome sequencing, were integrated into an ensemble framework. The model achieved strong discriminative performance, with area under the curve (AUC) values of 0.956 in the training cohort and 0.946 in the validation cohort, outperforming individual feature-based models. At a predefined operating threshold corresponding to 90% sensitivity, specificity reached 0.90 and 0.87, respectively. Notably, most cancer samples exhibited low tumor fraction (TF&#x2009;<&#x2009;3%), yet the model maintained robust performance in low-TF samples (AUCs: 0.952 and 0.941, respectively). Performance remained consistent across tumor stage, grade, and histological subtypes. The classifier also demonstrated potential clinical utility in diagnostically challenging settings, including lipid-poor angiomyolipoma and oncocytoma, with 12 of 13 oncocytoma samples correctly classified in an independent cohort. In addition, the model correctly identified 85.3% of benign masses&#x2009;>&#x2009;4&#xa0;cm, for which surgical intervention is more commonly considered, and 84.6% of malignant tumors&#x2009;&#x2264;&#x2009;4&#xa0;cm, for which management can be challenging. Collectively, these findings support cfDNA fragmentomics as a promising noninvasive liquid biopsy approach for renal mass evaluation and clinical decision-making.

Humans↗
PubMed2004

A constrained-syntax genetic programming system for discovering classification rules: application to medical data sets.

This paper proposes a new constrained-syntax genetic programming (GP) algorithm for discovering classification rules in medical data sets. The proposed GP contains several syntactic constraints to be enforced by the system using a disjunctive normal form representation, so that individuals represent valid rule sets that are easy to interpret. The GP is compared with C4.5, a well-known decision-tree-building algorithm, and with another GP that uses Boolean inputs (BGP), in five medical data sets: chest pain, Ljubljana breast cancer, dermatology, Wisconsin breast cancer, and pediatric adrenocortical tumor. For this last data set a new preprocessing step was devised for survival prediction. Computational experiments show that, overall, the GP algorithm obtained good results with respect to predictive accuracy and rule comprehensibility, by comparison with C4.5 and BGP.

Algorithms↗
PubMed2006

Detecting potential labeling errors in microarrays by data perturbation.

MOTIVATION: Classification is widely used in medical applications. However, the quality of the classifier depends critically on the accurate labeling of the training data. But for many medical applications, labeling a sample or grading a biopsy can be subjective. Existing studies confirm this phenomenon and show that even a very small number of mislabeled samples could deeply degrade the performance of the obtained classifier, particularly when the sample size is small. The problem we address in this paper is to develop a method for automatically detecting samples that are possibly mislabeled. RESULTS: We propose two algorithms, a classification-stability algorithm and a leave-one-out-error-sensitivity algorithm for detecting possibly mislabeled samples. For both algorithms, the key structure is the computation of the leave-one-out perturbation matrix. The classification-stability algorithm is based on measuring the stability of the label of a sample with respect to label changes of other samples and the version of this algorithm based on the support vector machine appears to be quite accurate for three real datasets. The suspect list produced by the version is of high quality. Furthermore, when human intervention is not available, the correction heuristic appears to be beneficial.

Artifacts↗
PubMed1997

Noise covariance incorporated MEG-MUSIC algorithm: a method for multiple-dipole estimation tolerant of the influence of background brain activity.

This paper proposes a method of localizing multiple current dipoles from spatio-temporal biomagnetic data. The method is based on the multiple signal classification (MUSIC) algorithm and is tolerant of the influence of background brain activity. In this method, the noise covariance matrix is estimated using a portion of the data that contains noise, but does not contain any signal information. Then, a modified noise subspace projector is formed using the generalized eigenvectors of the noise and measured-data covariance matrices. The MUSIC localizer is calculated using this noise subspace projector and the noise covariance matrix. The results from a computer simulation have verified the effectiveness of the method. The method was then applied to source estimation for auditory-evoked fields elicited by syllable speech sounds. The results strongly suggest the method's effectiveness in removing the influence of background activity.

Algorithms↗
PubMed1996

Detecting cortical activities from fMRI time-course data using the MUSIC algorithm with forward and backward covariance averaging.

A method is proposed for processing time-course fMRI data taken with successive single-shot echo-planar imaging. The proposed method uses a two-dimensional version of the multiple signal classification (MUSIC) algorithm and the technique called covariance averaging, both of which were developed in the field of sensor-array processing. The proposed method consists of four steps: calculate the averaged data covariance matrix, determine the number of activities using this covariance matrix, estimate the locations of the activities, and estimate their time evolution curves. Computer simulation results showed that a nearly fourfold improvement in the spatial resolution can be attained due to the method's super-resolution capability.

Algorithms↗