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 433 records · Page 24Linked to original sources

PubMed1994

Adaptive classification of myocardial electrogram waveforms.

The shape of myocardial electrogram complexes can change gradually in response to electrical and physiological transients. These changes could affect the reliability of morphologic-based electrogram classifiers proposed for use in implantable cardioverters. In this report, we present a method of detecting gradual changes in the shape of electrogram complexes and evaluate the method by incorporating it into a simple adaptive classification scheme. Of the six subjects recruited to take part in a previous comparative study of myocardial electrogram features, we observed extensive morphologic drift of normal sinus beats in two subjects. Our results indicate that the adaptive classification scheme proposed here can reduce observed classification error rates compared to rates obtained without adaptation.

Algorithms↗

PubMed1993

Self-organized acoustic feature map in detection of fricative-vowel coarticulation.

The self-organizing map, a neural network algorithm of Kohonen, was used for the detection of coarticulatory variation of fricative [s] preceding vowels [a:], [i:], and [u:]. The results were compared with the psychoacoustic classification of the same samples to find out whether the map had extracted perceptually meaningful features of [s]. The map distinguished samples of [s] in front of [u:] from those in front of [a:] or [i:] throughout the fricative duration. Samples of [s] preceding [a:] and [i:] were distinguished from each other only just before (about 40 ms) the vowel onset. The results agreed with the perceptual classifications. Most judgments (82%) of [s] in front of [u:] were correct, and this variant of [s] was recognized from the first and second halves of segmented fricatives equally well. Samples of [s] in front of [a:] and [i:] were distinguished from each other less accurately. When halves of segmented [s] were perceptually judged, the differentiation between the following [a] and [i] was possible only on the basis of the second half of the fricative. The results demonstrate that the self-organizing map is a useful tool for the extraction of intersubject regularities in speech spectra. The map also provides an easily understandable, on-line, visualization of speech that can be used as feedback in therapy and education.

Acoustic Stimulation↗

PubMed1992

Treatment of lower extremity infections in diabetics.

The infected diabetic lower extremity has enjoyed a surge in popularity in the medical literature. There have been numerous papers outlining classification systems for ulcer depth, surgical approaches, and microbiology. Discussions on antibiotic use have usually been directed toward therapy of the "diabetic foot infections" as a group, without regard to differences in severity and location of these infections. These infections can vary from the most superficial of processes to a severe life- and limb-threatening sepsis. The author presents a review of the processes involved in the diabetic lower extremity infection and suggests a classification system for selection of empiric antibiotic therapy based on the severity of the infection.

Algorithms↗

PubMed1993

Computationally efficient cluster representation in molecular sequence megaclassification.

Molecular sequence megaclassification is a technique for automated protein sequence analysis and annotation. Implementation of the method has been limited by the need to store and randomly access a database of all the sequence pair similarities. More than 80,000 protein sequences are now present in the public databases, and the pair similarity data table for the full protein sequence database requires over 1 gigabyte of storage. In this paper we present a computationally efficient representation of groups based on a graph theory approach where sequence clusters are described by a minimal spanning tree of highest scoring similarity pairs. This representation allows a classification of N proteins to be stored in order(N) memory. The use of this minimal spanning tree representation simplifies analysis of groups, the description of group characteristics and the manual correction of artifacts resulting from false hits. The new tree representation also introduces new possibilities for artifact generation in sequence classification. Methods for detecting and removing these artifacts are discussed.

Algorithms↗

PubMed1995

Size estimation and magnification error in radiographic imaging: implications for classification of arteriovenous malformations.

PURPOSE: To assess magnification error in digital subtraction angiography as it pertains to arteriovenous malformation (AVM) size. METHODS: A rectangular grid phantom with equally spaced markers mounted in a stereotactic frame was imaged with digital angiographic equipment. The location and orientation of the grid was altered relative to the central plane of the phantom. Both linear and area measurements were made according to the perceived location of phantom markers using a standard catheter calibration technique and compared with stereotactically derived estimates. Finally, a single case example of an angiographically imaged rolandic AVM was used to compare linear dimensions obtained with both described techniques. RESULTS: The determination of location and size with standard angiographic imaging is subject to error because of the divergent geometry of the incident x-ray beam. The resulting nonconstant geometric magnification causes errors in linear measurements of 10% to 13% at depths of 7 cm from the calibration plane. Errors in area measurements at the same position increase by 20% to 25%. Measurements of maximum diameter or cross-sectional area may have an additional error when nonspherical objects are inclined to the viewing direction (40% at 45 degrees inclination). These errors are reduced to less than 1 mm using the stereotactic technique. Some commercial angiographic systems have internal software to enable a spatial calibration based on known distances in the image or on the diameter of a catheter. The catheter technique was accurate in the calibration direction (perpendicular to the catheter axis) but had a 12% error in the direction parallel to the catheter because of a nonunity aspect ratio in the video system. Measurement of the dimensions of a rolandic AVM using the catheter calibration technique had an error that ranged from -3% to +26% (standard error, 20%) with respect to the stereotactic technique. CONCLUSIONS: Numerous nonstereotactic referential systems for determining linear distances are inherently erroneous by varying degrees compared with the stereotactic technique. Area and volume determinations naturally increase this error further. To the extent that no standardized method for determining linear distances exists, significant variations in estimation of AVM size result. Classification schemes for AVMs have been hampered by this technical error.

Algorithms↗

PubMed1993

Structural classification of the P450 superfamily based on consensus sequence comparison.

The novel method for classification of P450s has been proposed basing on consecutive multiple alignments of consensus sequences. A code can be assigned to each P450 amino acid sequence reflecting its relative distance from the corresponding consensus sequences at three hierarchical levels. Thus, only the principle of sequence similarity is chosen for classification without involving such criteria as enzymatic properties, phylogenetic assignment or chronology.

Algorithms↗

PubMedDate not supplied

[Automated classification system (ACS) for detection of high risk groups exposed to radiation].

The paper presents the results of using an experimental version of the automatic classifying system (ACS) based on unstatistical methods of recognizing the images with estimate calculation algorithms. According to 21 blood parameters obtained on an automatic analyser, ACS could divide, at higher than 90% significance, groups of persons living in the polluted areas of the towns of Klintsy and Zlynka who have accumulated mean radiation dose of 15 and 63 mZv, respectively. The results suggest that ACS may be used to divide subjects into risk groups in terms of exposures to unfavorable environmental factors.

Algorithms↗

PubMed2026

AniAnn's: alignment-free annotation of tandem repeat arrays using fast average nucleotide identity estimates.

MOTIVATION: Satellite DNA has long posed challenges for genome assembly and analysis due to its low sequence complexity and poor mappability. These large heterochromatic arrays of tandem repeats are ubiquitous across eukaryotic genomes, yet remain understudied. Current methods for annotating satellite regions, and other classes of tandem repeat arrays, are limited in their ability to annotate divergent or novel sequences. RESULTS: In this work, we introduce AniAnn's, an algorithm for annotating large blocks of tandemly repeating DNAs. AniAnn's exploits the high Average Nucleotide Identity (ANI) shared between repeat units of the same array to quickly and accurately infer the boundaries of such arrays. We show that AniAnn's improves the annotation of satellites and other tandem repeats within a variety of plant and animal genomes, while requiring only a fraction of the runtime compared to previous approaches. We conclude by exploring several use cases of AniAnn's as a lightweight method for masking repeats prior to whole-genome alignment as well as the de novo annotation and classification of satellite repeats. AVAILABILITY: AniAnn's is open source software and available at github.com/marbl/anianns.

Algorithms↗

PubMed1988

Comparison of the performances of an automated arrhythmia detector working on original and virtual ECG tracings.

Two approaches can be taken to improve the performance of an automatic arrhythmia detector: perfecting the detection algorithms or improving the quality of the investigated traces by preprocessing the original traces. This paper reports on the results of a data preprocessing approach. Preprocessing consists in constructing new traces, which we call virtual. They are mathematically obtained from the original traces and referred to the dominant cardiac electric axis. The classifications obtained with an arrhythmia detector using both virtual and original traces are presented and discussed. By comparing the performance indices obtained under the two different conditions, it can be seen that a diagnosis based on the virtual traces is as acceptable as one based on the original traces. This result should be judged as favorable, since the algorithm was not adjusted or calibrated to the virtual traces, while those who developed it had certainly calibrated the parameters to the original traces.

Algorithms↗

PubMed1985

Tests of homogeneity and trend with medians.

Algorithms for tests of homogeneity and trend with medians are presented. Under certain conditions the tests are more efficient than many standard nonparametric tests on one-way classification designs. Computationally the tests are very simple and could be performed by hand or with the help of a calculator. Application to certain types of data derived in cytogenetics is pointed out with an example.

Animals↗

PubMed1990

Cases of alleged asbestos-related disease: a radiologic re-evaluation.

Chest radiographs were re-evaluated from 439 active and retired tireworkers previously designated as having a condition consistent with an asbestiform mineral exposure. The review was performed in an independent manner by three board-certified radiologists according to guidelines from an international classification system. The percentage of cases with abnormalities consistent with an asbestiform mineral exposure found separately by the three radiologists was 3.7, 3.0, and 2.7%. Application of an algorithm to form a consensus evaluation indicated that approximately 3.6% (16) of the subjects evaluated may have a condition consistent with an asbestos exposure. A more detailed review, however, revealed that only 11 workers, or 2.5% of the total, would have a reasonable likelihood of having such a condition. Most cases were normal and the majority of abnormalities present on the radiographs evaluated were nonoccupational in origin. Prevalent conditions identified included healed tuberculosis, histoplasmosis, emphysema, discoid atelectasis, effusions, healed rib fractures, scarring due to infection or old inflammatory disease, possible cancer, miscellaneous nonspecific linear markings consistent with cigarette smoking and aging, and heart and vascular system diseases--the latter evidenced by an abnormally large number of subjects with healed coronary artery bypass surgery and pacemaker implants. In summary, the best estimate from this study indicates that possibly 16 (3.6%), but more realistically 11 (2.5%), of the 439 tireworkers evaluated may have a condition consistent with exposure to an asbestiform mineral. This represents a 40-fold difference between the re-evaluation results and the original survey work.

Asbestosis↗

PubMed2025

Predicting host tropism in influenza a viruses: insights from multi-segment nucleotide signatures.

BACKGROUND: Influenza A virus (IAV) poses a significant public health threat due to its cross-species transmission and complex host adaptation mechanisms. This study integrated whole-genome data from avian, human, swine, and bovine IAV strains, using machine learning to predict viral host tropism based on nucleotide site features and to identify key sites driving host adaptation along with their synergistic effects. METHODS: A total of 64,000 IAV sequences from avian, human, swine, and bovine hosts were analyzed to build host-prediction models. A four-class classification framework (avian, human, swine, bovine) was constructed using nucleotide site features from all eight genomic segments (PB2, PB1, PA, HA, NP, NA, MP, NS). Eight machine learning algorithms (logistic regression, decision tree, random forest, SVM, KNN, gradient boosting, XGBoost, LightGBM) were benchmarked via 10-fold stratified cross-validation. Model performance was evaluated using accuracy, precision, recall, F1-score, AUPRC, and AUC. SHAP (SHapley Additive exPlanations) analysis prioritized critical nucleotide sites, while bivariate association tests identified synergistic/antagonistic interactions between sites. Nucleotide composition profiles were compared across host groups using hierarchical clustering and heatmap visualization. RESULTS: The XGBoost algorithm demonstrated the best and most stable performance, achieving an AUC value of over 0.95 in distinguishing human-derived sequences from non-human ones. SHAP analysis identified the top 20 critical nucleotide sites for each gene segment, such as sites 46 and 698 in the NS segment. Nucleotide composition analysis revealed high similarity between human and swine sequences in the HA and PB2 segments, and between avian and bovine sequences. The HA segment was particularly challenging in differentiating human from swine strains. Bivariate site association analysis uncovered significant synergistic or antagonistic effects between key sites within gene segments, forming complex networks. For instance, in the NS segment, a positive prediction contribution was observed when sites 371, 698, and 419 were all G. CONCLUSIONS: This study advances our mechanistic understanding of IAV host adaptation, identifies molecular determinants for zoonotic risk stratification, and establishes a scalable machine learning framework for predicting viral host tropism through nucleotide signature analysis, thereby enhancing surveillance strategies and informing preventive measures against emerging viral threats.

Influenza A virus↗

PubMed1994

Domains, constructs and the development of the breathing problems questionnaire.

Health-related quality of life (QoL) questionnaires are commonly developed as content valid instruments with conventionally defined domain subscales. We contrasted content valid domain subscales with construct valid construct subscales and developed a 13-domain QoL questionnaire, the Breathing Problems Questionnaire (BPQ), for patients with chronic bronchitis. In a series of studies, we examined the constructs relevant to COPD patients' experience of health. First, we provided psychometric evidence that items in the BPQ form two distinct groups: functional problems and negative evaluations. Second, we showed that problems were significantly more correlated with morbidity whereas negative evaluations were significantly more correlated with neuroticism. Third, we showed that negative evaluations correlated with neuroticism whereas positive evaluations (measured by the Satisfaction with Illness Scale) correlated with extraversion. Patients are more likely to make positive evaluations of their illness when they recognize that they are seriously ill. Most of the BPQ domains are subcategories of the construct of problems: both domains and construct classifications provide useful information.

Aged↗

PubMedDate not supplied

[Neuroradiographic diagnosis of degenerative disk disease--the Dusseldorf Diagnostic Disk program. II. Reproduction of pain caused by diskography].

Pain reproduction after increasing pressure via an intradiscal injection was tested in three groups of patients, as a function of the stage of degeneration, the height of the disk space, and degenerative reactions of the upper plate. The purpose of the pain reproduction test is to classify pain due to disk pathology according to segments. The more advance the degeneration of the disk, the more accurate the pain reproduction. However, freedom from pain cannot always be demonstrated even in healthy subjects. While similar results are found with increasing thinning of the disks, these tests do not correlate with the upper plate reactions. The importance of this approach as compared to other imaging methods lies in the fact that it enables algogenic degenerative disk changes to be distinguished more clearly from those which are subclinical. However, the pain reproduction test is only one element in segmental classification as far as surgical treatment of pseudoradicular disk-related pathology is concerned.

Adult↗

PubMed1995

More about the difference between men and women: evidence from linear neural networks and the principal-component approach.

The ability of a statistical/neural network to classify faces by sex by means of a pixel-based representation has not been fully investigated. Simulations with pixel-based codes have provided sex-classification results that are less impressive than those reported for measurement-based codes. In no case, however, have the reported pixel-based simulations been optimized for the task of classifying faces by sex. A series of simulations is described in which four network models were applied to the same pixel-based face code. These simulations involved either a radial basis function network or a perceptron as a classifier, preceded or not by a preprocessing step of eigendecomposition. It is shown that performance comparable to that of the measurement-based models can be achieved with pixel-based input (90%) when the data are preprocessed. The effect of the eigendecomposition preprocessing of the faces is then compared with spatial-frequency analysis of face images and analyzed in terms of the perceptual information it captures. It is shown that such an examination may offer insight into the facial aspects important to the sex-classification process. Finally, the contribution of hair information to the performance of the model is evaluated. It is shown that, although the hair contributes to the sex-classification process, it is not the only important contributor.

Algorithms↗

PubMed1992

Automatic wave form classification of extracellular multineuron recordings.

A PC-based method for the reconstruction of individual spike trains from extracellular multineuron recordings is described. Starting with virtually no knowledge about the wave forms in a record, a fully automatic template-finding algorithm extracts templates using the entire data set. In a second step, individual spike trains are reconstructed.

Algorithms↗

PubMed1995

Neural-network-based classification of cognitively normal, demented, Alzheimer disease and vascular dementia from single photon emission with computed tomography image data from brain.

Single photon emission with computed tomography (SPECT) hexamethylphenylethyleneamineoxime technetium-99 images were analyzed by an optimal interpolative neural network (OINN) algorithm to determine whether the network could discriminate among clinically diagnosed groups of elderly normal, Alzheimer disease (AD), and vascular dementia (VD) subjects. After initial image preprocessing and registration, image features were obtained that were representative of the mean regional tissue uptake. These features were extracted from a given image by averaging the intensities over various regions defined by suitable masks. After training, the network classified independent trials of patients whose clinical diagnoses conformed to published criteria for probable AD or probable/possible VD. For the SPECT data used in the current tests, the OINN agreement was 80 and 86% for probable AD and probable/possible VD, respectively. These results suggest that artificial neural network methods offer potential in diagnoses from brain images and possibly in other areas of scientific research where complex patterns of data may have scientifically meaningful groupings that are not easily identifiable by the researcher.

Aged↗

PubMed1993

Numerical classification and identification of Acinetobacter genomic species.

A total of 211 Acinetobacter strains (representing all currently recognized genomic species) were tested for 329 biochemical characters. Overall similarities of all strains were determined for 145 characters by numerical taxonomic techniques, the UPGMA algorithm and the S(SM)) and the S(J) coefficients as measures of similarity. Seven clusters (two or more strains) and three unclustered strains were recovered at a similarity level of 80.0% (S(SM). At this level a complete correspondence between phenotypic cluster and genomic species was found only for genomic species 12 (Ac. radioresistens). At higher similarity levels (84.0% to 84.6% (S(SM)), however, several subclusters were found, each representing a single genomic species. An exception were the strains belonging to the genetically closely related species of the Acinetobacter calcoaceticus-baumannii complex. These were recovered scattered in several subclusters. The degree of genomic relatedness between some DNA groups correlated with phenotypic similarities, especially for DNA group 8 (Ac. Iwoffii) and 15 of Tjernberg and Ursing, and for DNA group 4 (Ac. haemolyticus) and 6. For the majority of genomic species, two identification matrices were constructed consisting of 22 and 10 diagnostic characters, respectively. The correct identification rates for the matrices were 98.0% (22 tests) and 90.8% (10 tests) taking a Willcox probability > 0.9. For unambiguous identification of some genomic species, however, additional methods (preferably DNA-DNA hybridization or ribotyping) should be used.

Acinetobacter↗