Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “CLASSIFICATION”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

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

At least 1,801 records · Page 100Linked to original sources
PubMed2006

Feature extraction for systolic heart murmur classification.

Heart murmurs are often the first signs of pathological changes of the heart valves, and they are usually found during auscultation in the primary health care. Distinguishing a pathological murmur from a physiological murmur is however difficult, why an "intelligent stethoscope" with decision support abilities would be of great value. Phonocardiographic signals were acquired from 36 patients with aortic valve stenosis, mitral insufficiency or physiological murmurs, and the data were analyzed with the aim to find a suitable feature subset for automatic classification of heart murmurs. Techniques such as Shannon energy, wavelets, fractal dimensions and recurrence quantification analysis were used to extract 207 features. 157 of these features have not previously been used in heart murmur classification. A multi-domain subset consisting of 14, both old and new, features was derived using Pudil's sequential floating forward selection (SFFS) method. This subset was compared with several single domain feature sets. Using neural network classification, the selected multi-domain subset gave the best results; 86% correct classifications compared to 68% for the first runner-up. In conclusion, the derived feature set was superior to the comparative sets, and seems rather robust to noisy data.

Aged↗
PubMed2005

Using aerial video to train the supervised classification of Landsat TM imagery for coral reef habitats mapping.

Management of coral reef resources is a challenging task, in many cases, because of the scarcity or inexistence of accurate sources of information and maps. Remote sensing is a not intrusive, but powerful tool, which has been successfully used for the assessment and mapping of natural resources in coral reef areas. In this study we utilized GIS to combine Landsat TM imagery, aerial photography, aerial video and a digital bathymetric model, to assess and to map submerged habitats for Alacranes reef, Yucatán, México. Our main goal was testing the potential of aerial video as the source of data to produce training areas for the supervised classification of Landsat TM imagery. Submerged habitats were ecologically characterized by using a hierarchical classification of field data. Habitats were identified on an overlaid image, consisting of the three types of remote sensing products and the bathymetric model. Pixels representing those habitats were selected as training areas by using GIS tools. Training areas were used to classify the Landsat TM bands 1, 2 and 3 and the bathymetric model by using a maximum likelihood algorithm. The resulting thematic map was compared against field data classification to improve habitats definition. Contextual editing and reclassification were used to obtain the final thematic map with an overall accuracy of 77%. Analysis of aerial video by a specialist in coral reef ecology was found to be a suitable source of information to produce training areas for the supervised classification of Landsat TM imagery in coral reefs at a coarse scale.

Animals↗
PubMed2006

Permutation entropy improves fetal behavioural state classification based on heart rate analysis from biomagnetic recordings in near term fetuses.

The relevance of the complexity of fetal heart rate fluctuations with regard to the classification of fetal behavioural states has not been satisfyingly clarified so far. Because of the short behavioural states, the permutation entropy provides an advantageous complexity estimation leading to the Kullback-Leibler entropy (KLE). We test the hypothesis that parameters derived from KLE can improve the classification of fetal behaviour states based on classical heart rate fluctuation parameters (SDNN, RMSSD, ln(LF), ln(HF)). From measured heartbeat sequences (35 healthy fetuses at a gestational age between 35 and 40 completed weeks) representative intervals of 256 heartbeats were visually preclassified into fetal behavioural states. Employing discriminant analysis to separate the states 1F, 2F and 4F, the best classification result by classical parameters was 80.0% (SDNN). After additionally considering KLE parameters it was improved significantly (p<0.0005) to 94.3% (ln(LF), KLE_Mean). It could be confirmed that KLE can improve the state classification. This might reflect the consideration of different physiological aspects by classical and complexity measures.

Cardiotocography↗
PubMed2006

Classification of breast masses in mammograms using genetic programming and feature selection.

Mammography is a widely used screening tool and is the gold standard for the early detection of breast cancer. The classification of breast masses into the benign and malignant categories is an important problem in the area of computer-aided diagnosis of breast cancer. A small dataset of 57 breast mass images, each with 22 features computed, was used in this investigation; the same dataset has been previously used in other studies. The extracted features relate to edge-sharpness, shape, and texture. The novelty of this paper is the adaptation and application of the classification technique called genetic programming (GP), which possesses feature selection implicitly. To refine the pool of features available to the GP classifier, we used feature-selection methods, including the introduction of three statistical measures--Student's t test, Kolmogorov-Smirnov test, and Kullback-Leibler divergence. Both the training and test accuracies obtained were high: above 99.5% for training and typically above 98% for test experiments. A leave-one-out experiment showed 97.3% success in the classification of benign masses and 95.0% success in the classification of malignant tumors. A shape feature known as fractional concavity was found to be the most important among those tested, since it was automatically selected by the GP classifier in almost every experiment.

Algorithms↗
PubMed1989

A new method for the classification of subvigil stages, using the Fourier transform, and its application to sleep apnea.

An online classification for two-second segments of the EEG is developed. Based on the calculation of a norm and three characteristic frequencies the automatic classification shows satisfactory agreement with visual classification. The influence of the Fourier phase spectra is discussed. Trajectories in the parameter space are introduced. The automatic classification was applied to patients with a pronounced sleep-wake disorder and proved to be valuable in the diagnosis of a disturbed sleep structure.

Electroencephalography↗
PubMed1995

A new drug classification for computer systems: the ATC extension code.

During the testing of the Read Clinical Codes in general practice medical records in Australia, it became apparent that the pharmaceutical section of the codes was not applicable in a country with different brand names, pack sizes and forms. For pharmacoepidemiological studies, structured classification of both morbidity and pharmaceuticals is required for meaningful analysis. The search for a suitable pharmaceutical classification proved fruitless. While the Australian Government has recently adopted the Anatomical Therapeutic Chemical (ATC) Classification as the national standard, this only classifies drugs to the generic level. None of the extended coding systems used in hospital pharmacies, by community pharmacists, or by Government are hierarchically structured. The extension code we have developed, is an analytical algorithm comprising independent fields for: dosage; strength; manufacturer and brand; and pack size. The codes within each field are also structured in a hierarchical manner. The result is an extension code of 21 digits, each digit or group of digits having a meaning. The structure of this classification will allow analysis of any aspect of the drug prescribed. This system is designed for computerised entry of text and transparent coding of the data--not for manual coding on paper nor manual code entry to the computer.

Algorithms↗
PubMed1995

Surgical classification of obstetric fistulas.

OBJECTIVE: To develop a surgical classification for obstetric fistulas in order to compare surgical techniques and results. METHODS: Based on a retrospective analysis of 775 consecutive fistula patients, the following classification is presented: (type I) fistulas not involving the urethral closing mechanism; (type II) fistulas involving the urethral closing mechanism; and (type III) ureter and other exceptional fistulas. Type II fistulas can be further divided into: (A) without (sub)total urethra involvement, and (B) with (sub)total urethra involvement; and (a) without a circumferential defect, and (b) with a circumferential defect. This classification was applied prospectively in over 2700 consecutive fistula patients. RESULTS: The surgical technique becomes progressively more complicated from type I through type IIBb. The results of closure and continence worsen progressively from type I through type IIBb. Personal experience in the case of type III fistulas is very limited. CONCLUSION: This classification enables a systematic comparison of different surgical techniques and an objective evaluation of results from different centers.

Adult↗
PubMed1986

Observer variation in the classification of mammographic parenchymal patterns.

Wolfe has described different cancer risks associated with a classification of four patterns of the breast parenchyma on mammography, but there is however little information available on the ability of radiologists to agree on the classification of the different patterns. We have assessed inter-rater agreement on the assignment of films to one of the four mammographic patterns described by Wolfe. One hundred xeromammograms were selected, copied and distributed to 10 radiologists who were experts in mammography. Films were classified according to the presence or absence of several radiological signs, according to diagnosis and recommendation, and according to mammographic pattern. Agreement was assessed after correction for agreement expected by chance, using the Kappa statistic. In general, high levels of agreement were found for the classification of mammographic pattern. Agreement on the classification of mammographic pattern was substantially greater than agreement for any other feature of mammographic interpretation, including diagnosis and recommendation.

Breast Neoplasms↗
PubMed1987

Clinical evaluation of children's classification behavior.

Classification skills among 88 unimpaired school children (22 each at grades k, 2, 4, and 6) were evaluated using the Iconic-Symbolic (IS) test of the Muma Assessment Program (Muma and Muma, 1979). IS stimuli are 18 test plates, each of which contains pictures of three objects. The child is required to select any two of these pictures. Evaluation of response patterns is intended to permit judgement of subject classification strategy and changes therein across the middle childhood years. In the current study, children, after exposure to standard IS test procedures, were interviewed in order to determine the rationale they used in forming their paired-object classes. Results revealed little isomorphism between the rationale that children, in fact, employed in the task and the classification criteria presumed, a priori, by the IS procedures. Minimal differences were found in the degree of response-rationale isomorphism across the child sample. Implications of these results are discussed as they relate to the evaluation of classification skills among unimpaired and clinical populations.

Attention↗
PubMed1989

Site classification for the osseointegrated implant.

Descriptions of jaw anatomy and bone quality have been based on total jaw resorption classifications. These classifications have not been specific for treatment planning for the osseointegrated implant. The proposed classification describes specific sites by bone quantity and quality and proximity to vital structures. This classification is suggested as an aid in assigning a prognostic value to implants and for the purpose of clarity and communication between the various dental disciplines.

Alveolar Process↗
PubMed1993

Thymoma: a study of the pathologic classification of 71 cases with evaluation of the Muller-Hermelink system.

A pathologic study of 71 consecutive cases of thymoma was conducted to determine the value of histologic classification for predicting the invasive potential of thymomas. The traditional and newly proposed Muller-Hermelink classification systems were studied for correlation with thymomas of the three clinicopathologic stages identified as noninvasive thymoma, microinvasive thymoma, and macroinvasive thymoma. None of the histologic types classified by the traditional system were correlated specifically with invasive growth. In contrast, the cortical-type thymoma of the Muller-Hermelink system correlated strongly with both microinvasive thymoma and macroinvasive thymoma (P < .001 for both). Although not all cortical thymomas manifested invasive growth and not all invasive thymomas were of the cortical type, the statistical data suggested that cortical thymomas should be regarded as potentially malignant tumors with the propensity for invasive growth. Spindle-cell thymoma of the traditional system or medullary thymoma of the Muller-Hermelink system, on the other hand, was not associated with invasiveness and behaved as a benign tumor. All other thymomas that were not purely spindle cell or medullary cell in nature had a potential for invasion and aggressive behavior, and this potential seemed to be inversely proportional to the percentage of fusiform cells in such neoplasms. Nevertheless, the results of this study support the usefulness of the Muller-Hermelink classification system in predicting the aggressive potential of thymomas. A group of microinvasive thymomas was identified as a separate group for analysis in this study. Although no recurrences were observed during the follow-up period, microinvasive thymomas deserve further study as a separate group because they correlated strongly with cortical thymoma. The association of thymoma with myasthenia gravis was rather high (63.4%) in this study. Both the mixed-type thymoma of the traditional classification system and the cortical-type thymoma of the Muller-Hermelink system showed statistically significant correlation with myasthenia gravis (P = .002 and .001, respectively).

Adult↗
PubMed1983

The postoperative classification for uterine cervical cancer and its clinical evaluation.

A postoperative classification for uterine cervical cancer has been made in consideration of the spatial spreading of cancer, biological malignancy of 120 cases which were treated with radical hysterectomy and pelvic lymphadnectomy. This classification corresponds extremely well to prognosis. The 5-year survival of the cases with prognostic index (P.I.) 9 or less was 96.1%, while those with P.I. above 10 showed 31.8%. In the Shinshu University School of Medicine clinic, this classification has become indispensable for decision of postoperative irradiation, selection of irradiation methods, and chemotherapeutic agents. Using this classification for individualized therapy, the survival rate was elevated in advanced cancer, class III or IV, with lymph node metastasis and P.I. above 10.

Adenocarcinoma↗
PubMed1985

Classification of hearts with overriding aortic and pulmonary valves.

Despite the clarity of the sequential segmential segmental approach to complex congenital heart malformations, the classification of hearts with overriding arterial valves remains contentious. A series of 67 hearts, all with overriding arterial valves, has therefore been studied in an attempt to provide clear and unambiguous criteria for their classification. There were 51 hearts with an overriding aortic valve, 13 hearts with an overriding pulmonary valve and 3 specimens with overriding of both valves. In each of these categories the degree of override and the underlying morphology varied considerably. The options to classify these hearts are limited. Using the "50 per cent rule" as a device to catalogue the type of connexion - irrespective of the morphology - all hearts were described in unambiguous fashion. A comparison with synonyms, as frequently used for purposes of classification, revealed that the latter are often insufficient properly to classify the basic abnormality. By classifying the type of connexion and describing the morphology separately, no basic problem remains in distinguishing between hearts with double outlet right ventricle and subpulmonary ventricular septal defect and hearts with the morphology of Fallot's tetralogy with an aorta almost exclusively arising from the right ventricle. Similarly, the classification of hearts with complete transposition and subpulmonary defect in the setting of the so-called Taussig-Bing heart is brought back to its proper perspective. Proper and consistent application of the sequential segmental approach leaves no room for ambiguity, even in complicated hearts with overriding arterial valves.

Aortic Valve↗
PubMed1993

A comparison of two classification systems for hemifacial microsomia.

The classification of hemifacial microsomia (HFM) aids in diagnosis, treatment planning, prognostic predictions and data evaluation. The aetiological and phenotypic heterogeneities of HFM, however, make its classification problematic. This study used data from 50 patients to examine the classification of HFM and to compare two systems: OMENS and SAT. The results were concordant with current literature and demonstrated the phenotypic heterogeneity of HFM. Essentially, both classifications embody the major craniofacial defects, but the OMENS system appears to be further refined by its differentiation between soft tissue and nerve defects, and between orbital and mandibular defects. Neither system, however, records deafness or grades auricular tags, although tags occurred in 34% of cases and two patients with otherwise 'normal' ears had tags. Therefore, it is suggested that auricular tags be graded as minor ear malformations. Furthermore, the OMENS system could be strengthened by the addition of an asterisk to the acronym in cases with serious non-craniofacial anomalies, for example OMENS*. This adds little complexity to the acronym, but immediately indicates when a patient's features lie towards the more generalised oculoauriculovertebral end of the phenotypic spectrum.

Abnormalities, Multiple↗
PubMed1988

Classification of acute leukaemias.

Acute leukaemias have traditionally been classified according to the nature of the predominating cells as judged by cytomorphology and cytochemistry. A codification of this classification into L1-L3 subdivisions for lymphoblastic cases (ALL) and M1-M7 for myeloid cases (AML), proposed by the FAB group of haematologists, has been used extensively in the past decade. Some criticisms of this codification are presented and other approaches to classification are discussed. Among these are included the potential importance of multiple lineage expression, measurements of cell differentiation using cytochemical criteria, the relevance of isoenzyme biochemistry of leukaemic cell populations, their surface antigenic differences as demonstrated by the use of monoclonal antibodies and the growing contribution of cytogenetics. The development and complexity of combined classifications which attempt to incorporate most of these features is illustrated. A simplified classification into broad groups is proposed, ALL being divided by surface antigenic and cytogenetic criteria and AML by morphology and cytochemistry. These groups may be further elaborated as appropriate.

Acute Disease↗
PubMed1996

Analysis and classification of interictal spike discharges in benign partial epilepsy of childhood on the basis of the Hilbert transformation.

The spatial distribution of instantaneous power during the occurrence of rolandic spikes (computed via Hilbert transformation) can be utilized for classification of topographically different spike types. By visual analysis of the instantaneous power maps we found seven characteristic spike classes which correspond to the results of visual analysis of potential distribution. A neural network (NN) was trained with representatives of these classes. Single instantaneous power maps of 55/56 visually selected rolandic spikes, as well as map sequences of averaged spike segments (on-line) of 17 patients, were correctly classified by means of NN. The sensitivity of the NN for spikes from unknown patients was 98%. The classification scheme enables an objective classification of (multi-) regional spikes for selective averaging and for studies dealing with syndrome classification.

Child↗
PubMedDate not supplied

Elemental classification in multi-detector stem images using image analysis clustering techniques.

The multi-detector STEM images form a multi-dimensional measurement space where picture elements of morphologically distinct regions cluster and can be separated for classification using image processing clustering techniques. These images are highly correlated and improve the classification only at the high cost of adding dimensionality. Data reduction techniques which take the class separation into account can be used to compress the useful information carried by these images into a few components to which clustering techniques can be successfully applied. At high magnification, a slight displacement is sometimes observed, which can slightly impair the classification result. The redundancy of the information in the quadrant images suggests their use for phase retrieval - which, added to an energy loss data channel, may greatly improve the classification.

Animals↗
PubMed1984

Multivariate statistical classification of noisy images (randomly oriented biological macromolecules).

Multivariate Statistical Analysis (MSA) methods have recently been introduced for analyzing images of biological macromolecules [Van Heel and Frank, Ultramicroscopy 6 (1981) 187]. With these techniques, the significant characteristics of each molecular image can be expressed in merely 2 to 8 factorial coordinate values rather than in the typical 64 X 64 = 4096 pixel grey values that originally described the image. This very large reduction in total amount of data facilitates the understanding of the general behavior of a set of molecular images in terms of classes or of general trends in the data set. The (artificial) intelligence of the procedure, however, lies in the decision-making or classification phase. The theory and philosophy of multivariate statistical classification are reviewed using generalized metrics. Problem-dependent classification rationales are proposed. A set of computer-generated "randomly oriented molecular images" are used to test the classification schemes. This model experiment is a step towards 3D structure analysis of macromolecules based on large numbers of (noisy) electron microscopical images of randomly oriented biological macromolecules.

Animals↗

Refine your search to explore more results.