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 667 records · Page 37Linked to original sources
PubMed2004

A school-based case identification process for identifying inner city children with asthma: the Breathmobile program.

BACKGROUND: Striking increases in the prevalence and morbidity of asthma among inner city children have been documented. OBJECTIVE: To establish and evaluate a large-scale, school-based case-detection process designed to efficiently and reliably identify inner city children with asthma. METHODS: A bilingual, seven-question, self-administered, parental asthma screening survey was developed. Clinical validation was achieved in a sample of 675 consecutive parents bringing a child to the school-based Breathmobile Program for initial evaluation, using a comprehensive evaluation by a physician specialist (ie, allergist) as the standard. Survey response patterns were used to construct a novel seven-model, tiered scoring algorithm and an abbreviated algorithm that predict the probability of a child being clinically classified as "yes asthma" or "no asthma." A systematic survey distribution process administered by a single coordinator was developed, and the impact of a classroom-oriented incentive offering a 25 dollars school supply gift certificate for survey return rates of >or= 80% was evaluated. RESULTS: A total of 636 parents provided one or more survey responses and information sufficient for clinical classification. The scoring algorithm correctly identified children with asthma (>or= 80% probability) with a sensitivity of 86.5%, a specificity of 83.6%, and a misclassification rate of 14.3% (91 of 636 children). The sensitivity for identifying persistent asthma was 91.3%. Asthma prevalence estimates derived using survey results from a larger sampling of the general population were similar to rates previously reported for comparable populations. The inclusion of an inexpensive incentive increased the median survey return rates from 35.3 to 65% (z= -11.9; p <.001). The screening process has been used to conduct 27,526 surveys at inner city schools. CONCLUSIONS: The Breathmobile case-detection process offers a validated, comprehensive, large-scale method with which to identify children with asthma at their school sites.

Adolescent↗
PubMed1999

Chronic wrist pain: diagnosis and management. Development and use of a new algorithm.

OBJECTIVE: Chronic wrist pain can be difficult to manage and the differential diagnosis is extensive. To provide guidelines for assessment of the painful wrist an algorithm was developed to encourage a structured approach to the diagnosis and management of these patients. METHODS: A review of the literature on causes of chronic wrist pain was undertaken; history taking, physical examination and imaging studies were evaluated systematically to determine which of the many potential conditions was the cause of the wrist pain. Chronic wrist pain was subdivided into pain of probable intra-articular or extra-articular origin. By means of this classification a clinical algorithm was developed to establish a diagnosis and its clinical usefulness was tested in a prospective study of 84 patients presenting to our outpatient clinic. RESULTS: A definite diagnosis could be established in 59% (49 of 84) of the cases by careful history taking, extensive physical examination, plain radiographs, ultrasound examination and bone scintigraphy. In 19% of the cases (16 of 84) a probable diagnosis was made resulting in a total figure 78% (65 of 84). Additional imaging studies (arthrography, magnetic resonance imaging and computed tomography) increased the definite diagnoses to 70% (59 of 84). CONCLUSION: The algorithm proved easy to use and by the use of careful history taking, thorough physical examination and simple imaging techniques (ultrasonography and scintigraphy) a diagnosis was made in 78% of cases.

Adolescent↗
PubMedDate not supplied

An inductive algorithm approach to knowledge acquisition for expert system development. A pilot study.

Knowledge acquisition, which consists of knowledge elicitation and knowledge representation, often is considered the weakest link in the design of expert systems. Systems frequently are built on the knowledge of one expert and require extensive use of knowledge engineering techniques to elicit this knowledge from the expert. Inductive algorithms are a potential alternative method of knowledge acquisition for expert system development. The aim of this pilot study was to examine the feasibility of applying machine learning techniques, specifically, inductive algorithms, to an existing research database as a method for knowledge elicitation and knowledge representation for expert system development. Two inductive algorithms (C4 and Classification and Regression Trees [CART]) that generate decision trees were selected for the analysis using a data set of 201 patients hospitalized for Pneumocystis carinii pneumonia. Neither C4 nor CART produced trees with an accuracy that was significantly better than the baseline accuracy (71.3%) for prediction of outcome in the data set. The mean accuracy of the C4 decision trees was below baseline and the mean accuracy of CART decision trees was 74.6%. The experts found both algorithms comprehensible, but not adequate, and identified important missing predictor variables. The study findings suggest that additional research is needed to examine the appropriate use of inductive algorithms in the transformation of nursing data and information into nursing knowledge.

Algorithms↗
PubMedDate not supplied

[Analysis of intracranial pressure signals using artificial neural networks].

Intracranial pressure (ICP) is influenced by an array of predictable and unpredictable factors. Statistical modelling of this signal has only limited applicability because of the significant load of stochastic components. We tested the efficiency of an alternative approach, based on the methodology of artificial neural networks (ANNs) in the on-line prediction of future values of ICP and in the classification of signal properties. Satisfactory accuracy of forecasting was achieved with the ANNs for a 3-minute prediction horizon, while the prediction quality with autoregressive models of statistical origin was proved unsatisfactory. The results obtained with the ANNs were further improved when signal pre-processing with wavelet transform was employed. Nevertheless, even with the ANN methodology, no sudden breakdowns in the ICP signal (which in this respect might be compared to a "catastrophe") can be forecast with any practical applicability. We therefore applied two ANN algorithms, oriented at classification and discrimination of the global properties of the ICP signal. The neural network was expected to discriminate those sets of signal properties, which were assumed to correspond to certain clinical conditions of the patient. In a "dynamic pattern classification" the network was presented with several sections of ICP records. This was combined with information about the assignment of a given record to one of four arbitrary classes of danger. In this mode no data pre-processing was carried out, in contrast to our second approach, in which the signal was pre-processed with statistical analyses and only these intermediate coefficients were fed to the ANN classifier. The results obtained with both classification methods at their present stage of training were similar and approximated to a 70% rate of judgements consistent with expert scoring. Nevertheless, the method based on the assessment of global parameters of the ICP record seems more promising, because it leaves the possibility of extending the set of training data by information from other diagnostic modalities. The study aims towards the development of a pseudo-intelligent computer expert system, which has would be taught salient links between data extracted from the ICP signal and higher- order data, which contributed to the expert score. Hence the system would be able to make decisions on the basis of a reduced set of input information, available from a standard monitoring modality.

Algorithms↗
PubMed2004

An optimal structure-discriminative amino acid index for protein fold recognition.

Identifying the fold class of a protein sequence of unknown structure is a fundamental problem in modern biology. We apply a supervised learning algorithm to the classification of protein sequences with low sequence identity from a library of 174 structural classes created with the Combinatorial Extension structural alignment methodology. A class of rules is considered that assigns test sequences to structural classes based on the closest match of an amino acid index profile of the test sequence to a profile centroid for each class. A mathematical optimization procedure is applied to determine an amino acid index of maximal structural discriminatory power by maximizing the ratio of between-class to within-class profile variation. The optimal index is computed as the solution to a generalized eigenvalue problem, and its performance for fold classification is compared to that of other published indices. The optimal index has significantly more structural discriminatory power than all currently known indices, including average surrounding hydrophobicity, which it most closely resembles. It demonstrates >70% classification accuracy over all folds and nearly 100% accuracy on several folds with distinctive conserved structural features. Finally, there is a compelling universality to the optimal index in that it does not appear to depend strongly on the specific structural classes used in its computation.

Algorithms↗
PubMed1996

The application of local grey level histograms to organelle classification in histological images.

The performance of a histogram matching algorithm was evaluated for the classification of organelles in histological images. The test set used for this investigation comprised several digital images of sheep pituitary gland captured using a CCD camera linked to a Tunnelling Electron Microscope. Results show that the normalized image histogram is a good feature to choose as the basis of object recognition in biological images.

Algorithms↗
PubMed1991

A method for quantitative autoradiography over stained sections of tumors exposed in vivo to radiolabeled antibodies.

Quantitative autoradiography can be an important element in calculating dose patterns delivered to tumors and normal tissues of animals injected with radiolabeled antibodies. In this article we develop a computer algorithm that permits quantitative evaluation of grain density over large areas of stained tissue with minimal grain counting. This algorithm, using spectral classification of brightfield and darkfield images, image ratios, and manual counting of small sections of an autoradiograph, yields images defined in terms of grain density, independent of staining artifacts. These quantitative images give a precise indication of intratumor radioimmunoglobulin targeting, and can serve as source functions for dosimetric calculations.

Algorithms↗
PubMed2002

Prediction of food protein allergenicity: a bioinformatic learning systems approach.

Food hypersensitivity is constantly increasing in Western societies with a prevalence of about 1-2% in Europe and in the USA. Among children, the incidence is even higher. Because of the introduction of foods derived from genetically modified crops on the marketplace, the scientific community, regulatory bodies and international associations have intensified discussions on risk assessment procedures to identify potential food allergenicity of the newly introduced proteins. In this work, we present a novel biocomputational methodology for the classification of amino acid sequences with regard to food allergenicity and non-allergenicity. This method relies on a computerised learning system trained using selected excerpts of amino acid sequences. One example of such a successful learning system is presented which consists of feature extraction from sequence alignments performed with the FASTA3 algorithm (employing the BLOSUM50 substitution matrix) combined with the k-Nearest-Neighbour (kNN) classification algorithm. Briefly, the two features extracted are the alignment score and the alignment length and the kNN algorithm assigns the pair of extracted features from an unknown sequence to the prevalent class among its k nearest neighbours in the training (prototype) set available. 91 food allergens from several specialised public repositories of food allergy and the SWALL database were identified, pre-processed, and stored, yielding one of the most extensively characterised repositories of allergenic sequences known today. All allergenic sequences were classified using a standard one-leave-out cross validation procedure yielding about 81% correctly classified allergens and the classification of 367 non-allergens in an independent test set resulted in about 98% correct classifications. The biocomputational approach presented should be regarded as a significant extension and refinement of earlier attempts suggested for in silico food safety assessment. Our results show that the framework described here is powerful enough to become useful as part of a multiple-procedure test scheme that also depicts other evaluation approaches such as solid phase immunoassay and tests for stability to digestions.

Algorithms↗
PubMed1995

Diagnostic reliability of dermoscopic criteria for detecting malignant melanoma.

BACKGROUND: Recently a new standardized terminology in dermoscopy has been provided by a Consensus Meeting held by the Committee on Analytical Morphology of the Arbeitsgemeinschaft Dermatologische Forschung in Hamburg in order to be applied to further studies in this field. OBJECTIVE: In this study on 159 pigmented skin tumors including 65 melanomas, the validity of various dermoscopic criteria proposed by the Consensus Meeting for detecting melanoma was evaluated. METHODS: In each lesion, a detailed clinical and dermoscopic examination, photographic documentation of the clinical and dermoscopic appearance, surgical excision and histopathologic evaluation were performed. Statistical analyses including chi 2 statistics, logistic regression analysis and CART (classification and regression tree) analysis were applied. RESULTS: The diagnosis of melanoma using dermoscopy could be obtained easily, if combinations of the following criteria were observed: whitish veil, pigment network alterations (e.g. irregular pigment network, narrow pigment network, broad pigment network), irregular extensions, black dots and gray-blue areas. Interestingly, however, clinical and dermoscopic examination to detect melanomas yielded the same results, namely a sensitivity of 94%. The combined use of both methods led to an increase in the diagnostic sensitivity of 95%, whereas the combination of clinical and dermoscopic examination with logistic regression analysis of dermoscopic criteria enabled us to detect all 65 melanomas in our data set, thus providing a sensitivity of 100%. Furthermore, an algorithm for the classification of melanoma based on the evaluation of dermoscopic criteria has been constructed by CART analysis showing that the presence of a whitish veil in combination with a pigment network inevitably indicates melanoma. CONCLUSION: The validity of the various dermoscopic criteria set forth by the Consensus Meeting could be demonstrated as some criteria (e.g. whitish veil, irregular pigment network, irregular extensions) were observed with a significant higher frequency in melanomas.

Algorithms↗
PubMed2005

Image segmentation feature selection and pattern classification for mammographic microcalcifications.

Since microcalcifications in X-ray mammograms are the primary indicator of breast cancer, detection of microcalcifications is central to the development of an effective diagnostic system. This paper proposes a two-stage detection procedure. In the first stage, a data driven, closed form mathematical model is used to calculate the location and shape of suspected microcalcifications. When tested on the Nijmegen University Hospital (Netherlands) database, data analysis shows that the proposed model can effectively detect the occurrence of microcalcifications. The proposed mathematical model not only eliminates the need for system training, but also provides information on the borders of suspected microcalcifications for further feature extraction. In the second stage, 61 features are extracted for each suspected microcalcification, representing texture, the spatial domain and the spectral domain. From these features, a sequential forward search (SFS) algorithm selects the classification input vector, which consists of features sensitive only to microcalcifications. Two types of classifiers-a general regression neural network (GRNN) and a support vector machine (SVM)--are applied, and their classification performance is compared using the Az value of the Receiver Operating Characteristic curve. For all 61 features used as input vectors, the test data set yielded Az values of 97.01% for the SVM and 96.00% for the GRNN. With input features selected by SFS, the corresponding Az values were 98.00% for the SVM and 97.80% for the GRNN. The SVM outperformed the GRNN, whether or not the input vectors first underwent SFS feature selection. In both cases, feature selection dramatically reduced the dimension of the input vectors (82% for the SVM and 59% for the GRNN). Moreover, SFS feature selection improved the classification performance, increasing the Az value from 97.01 to 98.00% for the SVM and from 96.00 to 97.80% for the GRNN.

Breast Neoplasms↗
PubMed2000

Feasibility of computer-assisted recognition of different dental hard tissues.

For pattern recognition, fuzzy set theory has been proven to be highly useful. The aim of the present investigation was to combine the fuzzy set with an ultrasonic scaler, to test its suitability for automatic detection of different tooth substances. An experienced operator placed the tip of a piezoceramic ultrasonic scaler on a tooth, thereby inducing oscillations in the contact area around the ultrasonic tip. Each surface showed a characteristic oscillatory behavior in the immediate vicinity of the tip. The oscillations were then re-transmitted to the scaler tip and recorded by the measurement of variations in current and voltage. Because the ultrasonic scaler is driven by piezoceramics, it can be used as both an oscillatory excitation and a sensor system. The data were processed with Fast Fourier transformation and analyzed by means of a fuzzy pattern recognition algorithm. Re-classification of the different measurements was carried out by the experienced operator's assessment. With a combination of six features (frequencies), re-classification was correct for 99% of all surfaces. The diagnostic reliability of the system was tested by the assessment of 50 teeth for which no learning data had previously been recorded. The unknown samples were correctly classified to 100%. The excellent results of these experiments suggest promising possibilities for the implementation of new diagnostic and therapeutic instruments in periodontology practice.

Algorithms↗
PubMed1987

[Various approaches to optimizing the hemodynamics of myocardial infarct patients].

Combined investigation of systemic and intracardiac hemodynamic parameters (radiocardiography and Fick's procedure, transvenous catheterization, and right-ventricular and pulmonary arterial blood pressure measurement) was carried out in 428 myocardial infarction patients admitted to hospital within the early hours after the attack; changes in necrotic area were monitored for 7 days (precordial ECG cartography and serial serum CPK assays). Computer-assisted combined data analysis made use of image identification techniques, the images being based on a general approximation algorithm and informative coefficient. The algorithm permitted a classification of patients' condition by left-ventricular insufficiency (a 77.4% reliability) and by the spread of necrotic area (a 72.5% reliability). The prognosis chart, constructed on the basis of comparison between actual hemodynamic values and their standard vectors, brought out optimum regimens for every parameter associated with the least risk of left-ventricular insufficiency and necrotic spread.

Adult↗
PubMed1994

Accurate segmentation of respiration waveforms from infants enabling identification and classification of irregular breathing patterns.

Although the precise causes of Sudden Infant Death Syndrome (SIDS) are still unclear there is evidence to suggest that hypoxaemia may be a contributory factor. Transcutaneous oxygen monitors can be used but are unsatisfactory for young babies in the home. As an alternative approach, respiratory patterns can be studied but attempts at classification of individual breaths are often unsuccessful particularly during periods of extraneous noise or movement artefact. We have developed a robust algorithm which provides accurate segmentation and classification of breaths even in the presence of noise or movement. This improves on previous techniques by deferring the decision on an uncertain candidate breath until more information is available; yet the delay incurred is two breaths at most. The use of look-up tables and decision trees means that computational requirements are kept to a minimum and implementation in a simple home monitor is quite possible.

Algorithms↗
PubMed2002

A novel approach to remote homology detection: jumping alignments.

We describe a new algorithm for protein classification and the detection of remote homologs. The rationale is to exploit both vertical and horizontal information of a multiple alignment in a well-balanced manner. This is in contrast to established methods such as profiles and profile hidden Markov models which focus on vertical information as they model the columns of the alignment independently and to family pairwise search which focuses on horizontal information as it treats given sequences separately. In our setting, we want to select from a given database of "candidate sequences" those proteins that belong to a given superfamily. In order to do so, each candidate sequence is separately tested against a multiple alignment of the known members of the superfamily by means of a new jumping alignment algorithm. This algorithm is an extension of the Smith-Waterman algorithm and computes a local alignment of a single sequence and a multiple alignment. In contrast to traditional methods, however, this alignment is not based on a summary of the individual columns of the multiple alignment. Rather, the candidate sequence is at each position aligned to one sequence of the multiple alignment, called the "reference sequence." In addition, the reference sequence may change within the alignment, while each such jump is penalized. To evaluate the discriminative quality of the jumping alignment algorithm, we compare it to profiles, profile hidden Markov models, and family pairwise search on a subset of the SCOP database of protein domains. The discriminative quality is assessed by median false positive counts (med-FP-counts). For moderate med-FP-counts, the number of successful searches with our method is considerably higher than with the competing methods.

Algorithms↗
PubMed1997

Simultaneous spike detection and topographic classification in pediatric surface EEGs.

In this study, an algorithm is introduced for the automatic detection and simultaneous topographic classification of interictal regional spike activity in pediatric surface EEG records. The algorithm is based on the classification of the topographic distribution of instantaneous power by means of a 'group' trained classifier. The results of automatic spike analysis were compared with the decisions of two experienced electroencephalographers. Four routine EEG records exhibiting (multi)regional spikes were examined. The mean selectivity for the automatic spike detector was 84.6% (mean sensitivity 88.1%, mean specificity 89.3%) and for the electroencephalographers 85.3%. All spikes detected by the algorithm were simultaneously classified according to their topographic characteristics. The results of automatic spike classification (lateralization/localization) corresponded to the results of visual analysis.

Algorithms↗
PubMed2006

Simulation and analysis of optical skin lesion images.

BACKGROUND/PURPOSE: In order to properly analyse the effectiveness of methods for optically differentiating malignant from benign skin lesions, it is necessary to have a set of images for which the ground truth is known. However, aspects of the ground truth of clinical images such as true lesion boundary position are unknown or not known precisely. Therefore, a skin/lesion image simulation with known features including boundary location, skin pattern and lesion colour is needed to enable accurate assessment of feature estimation algorithms for lesion classification. METHODS: In this paper, monochrome and colour skin/lesion images are synthesised with known characteristics including boundary, colour and skin pattern. Skin pattern is simulated with segmented lines with variations in length, orientation and intensity. Skin and lesion textures are modelled by an auto-regressive (AR) process. Monochrome skin lesion images are obtained by combining monochrome skin and lesion textures under the control of a known lesion shape with the addition of skin pattern. Colour skin lesion images are generated by mixing coloured skin and lesion textures. Finally, an inflammation area and image artefacts such as hair and specular reflection are added. RESULTS: The synthesised images provide the image set for evaluating image pre-processing, segmentation and skin pattern analysis. The pre-processing includes hair removal and specular reflection reduction. An AR model interpolation is suggested for hair removal, and multiple illumination processing is developed to decrease specular reflection. A fast snake algorithm is extended to detect the boundaries of skin lesion and inflammation areas. Skin line direction is detected as a feature to measure the disruption of skin pattern caused by lesion. CONCLUSIONS: Simulation of monochrome and colour skin/lesion image has been investigated, which is an alternative way to provide image set with known characteristics to validate image processing algorithms for image pre-processing, lesion/inflammation boundary detection and skin pattern analysis.

Colorimetry↗
PubMed2005

Multiclass cancer classification and biomarker discovery using GA-based algorithms.

MOTIVATION: The development of microarray-based high-throughput gene profiling has led to the hope that this technology could provide an efficient and accurate means of diagnosing and classifying tumors, as well as predicting prognoses and effective treatments. However, the large amount of data generated by microarrays requires effective reduction of discriminant gene features into reliable sets of tumor biomarkers for such multiclass tumor discrimination. The availability of reliable sets of biomarkers, especially serum biomarkers, should have a major impact on our understanding and treatment of cancer. RESULTS: We have combined genetic algorithm (GA) and all paired (AP) support vector machine (SVM) methods for multiclass cancer categorization. Predictive features can be automatically determined through iterative GA/SVM, leading to very compact sets of non-redundant cancer-relevant genes with the best classification performance reported to date. Interestingly, these different classifier sets harbor only modest overlapping gene features but have similar levels of accuracy in leave-one-out cross-validations (LOOCV). Further characterization of these optimal tumor discriminant features, including the use of nearest shrunken centroids (NSC), analysis of annotations and literature text mining, reveals previously unappreciated tumor subclasses and a series of genes that could be used as cancer biomarkers. With this approach, we believe that microarray-based multiclass molecular analysis can be an effective tool for cancer biomarker discovery and subsequent molecular cancer diagnosis.

Algorithms↗
PubMed2000

Nocturia.

PURPOSE: We review the current state of knowledge of nocturia and present algorithms for diagnosis, classification and treatment. MATERIALS AND METHODS: We reviewed the recent literature on nocturia, and state-of-the-art methods of diagnosis, classification and treatment. RESULTS: Nocturia, which is among the most bothersome of all urological symptoms, has heretofore been poorly classified and understood. Multiple factors may result in nocturia, including pathological conditions, such as cardiovascular disease, diabetes mellitus, lower urinary tract obstruction, anxiety or primary sleep disorders, and behavioral and environmental factors. Nocturia may be attributed to nocturnal polyuria (nocturnal urine overproduction) and/or diminished nocturnal bladder capacity. Distinction between these conditions is made by a simple arithmetic analysis of a 24-hour voiding diary. CONCLUSIONS: Nocturia has been poorly studied, and its etiology and pathogenesis have been classified only recently. We present a scheme for diagnosis and care of patients suffering from loss of sleep due to nocturnal voiding.

Algorithms↗