PubMed Health⌕ Search

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 451 records · Page 25Linked to original sources
PubMed1985

Double-judgment psychophysics: problems and solutions.

Many paradigms for comparing identification thresholds with detection thresholds require the observer to make double judgments. We show that these paradigms can produce misleading results because of response biases and attentional shifts. For example, the subject's response bias plus correlated noise can mimic inhibition between channels. Some of these same problems can affect single-judgment paradigms. A detailed analysis of the double-judgment forced-choice paradigm reveals that there are a multiplicity of optimal strategies, some of which enhance identification over detection. Several improved analysis techniques for minimizing the effects of cognitive factors are proposed for both the double-judgment forced-choice paradigm and the double-judgment rating-scale paradigm. A classification scheme for distinguishing different types of interactions and correlations is developed. When the new rating-scale algorithm is applied to the detection of well-separated spatial frequencies, substantial masking but negligible inhibition is found. The rating-scale paradigm is shown to be useful in revealing not only the sensitivity and the interactions of the underlying mechanisms but also the observer's information-processing strategies.

Attention↗
PubMed1995

Texture analysis in quantitative MR imaging. Tissue characterisation of normal brain and intracranial tumours at 1.5 T.

The diagnostic potential of texture analysis in quantitative tissue characterisation by MR imaging at 1.5 T was evaluated in the brain of 6 healthy volunteers and in 88 patients with intracranial tumours. Texture images were computed from calculated T1 and T2 parameter images by applying groups of common first-order and second-order grey level statistics. Tissue differentiation in the images was estimated by the presence or absence of significant differences between tissue types. A fine discrimination was obtained between white matter, cortical grey matter, and cerebrospinal fluid in the normal brain, and white matter was readily separated from the tumour lesions. Moreover, separation of solid tumour tissue and peritumoural oedema was suggested for some tumour types. Mutual comparison of all tumour types revealed extensive differences, and even specific tumour differentiation turned out to be successful in some cases of clinical importance. However, no discrimination between benign and malignant tumour growth was possible. Much texture information seems to be contained in MR images, which may prove useful for classification and image segmentation.

Algorithms↗
PubMed1995

A simulation algorithm for ultrasound liver backscattered signals.

In this study, we present a simulation algorithm for the backscattered ultrasound signal from liver tissue. The algorithm simulates backscattered signals from normal liver and three different liver abnormalities. The performance of the algorithm has been tested by statistically comparing the simulated signals with corresponding signals obtained from a previous in vivo study. To verify that the simulated signals can be classified correctly we have applied a classification technique based on an artificial neural network. The acoustic features extracted from the spectrum over a 2.5 MHz bandwidth are the attenuation coefficient and the change of speed of sound with frequency (dispersion). Our results show that the algorithm performs satisfactorily. Further testing of the algorithm is conducted by the use of a data acquisition and analysis system designed by the authors, where several simulated signals are stored in memory chips and classified according to their abnormalities.

Acoustics↗
PubMed2026

Automated Classification of Lymphoma Subtypes From Histopathological Images Using a U-Net Deep Learning Model: Comparative Evaluation Study.

BACKGROUND: Accurate classification and grading of lymphoma subtypes are essential for treatment planning. Traditional diagnostic methods face challenges of subjectivity and inefficiency, highlighting the need for automated solutions based on deep learning techniques. OBJECTIVE: This study aimed to investigate the application of deep learning technology, specifically the U-Net model, in classifying and grading lymphoma subtypes to enhance diagnostic precision and efficiency. METHODS: In this study, the U-Net model was used as the primary tool for image segmentation integrated with attention mechanisms and residual networks for feature extraction and classification. A total of 620 high-quality histopathological images representing 3 major lymphoma subtypes were collected from The Cancer Genome Atlas and the Cancer Imaging Archive. All images underwent standardized preprocessing, including Gaussian filtering for noise reduction, histogram equalization, and normalization. Data augmentation techniques such as rotation, flipping, and scaling were applied to improve the model's generalization capability. The dataset was divided into training (70%), validation (15%), and test (15%) subsets. Five-fold cross-validation was used to assess model robustness. Performance was benchmarked against mainstream convolutional neural network architectures, including fully convolutional network, SegNet, and DeepLabv3+. RESULTS: The U-Net model achieved high segmentation accuracy, effectively delineating lesion regions and improving the quality of input for classification and grading. The incorporation of attention mechanisms further improved the model's ability to extract key features, whereas the residual structure of the residual network enhanced classification accuracy for complex images. In the test set (N=1250), the proposed fusion model achieved an accuracy of 92% (1150/1250), a sensitivity of 91.04% (1138/1250), a specificity of 89.04% (1113/1250), and an F1-score of 90% (1125/1250) for the classification of the 3 lymphoma subtypes, with an area under the receiver operating characteristic curve of 0.95 (95% CI 0.93-0.97). The high sensitivity and specificity of the model indicate strong clinical applicability, particularly as an assistive diagnostic tool. CONCLUSIONS: Deep learning techniques based on the U-Net architecture offer considerable advantages in the automated classification and grading of lymphoma subtypes. The proposed model significantly improved diagnostic accuracy and accelerated pathological evaluation, providing efficient and precise support for clinical decision-making. Future work may focus on enhancing model robustness through integration with advanced algorithms and validating performance across multicenter clinical datasets. The model also holds promise for deployment in digital pathology platforms and artificial intelligence-assisted diagnostic workflows, improving screening efficiency and promoting consistency in pathological classification.

Humans↗
PubMed2025

GeomeTRe: accurate calculation of geometrical descriptors of tandem repeat proteins.

MOTIVATION: Structured tandem repeat proteins (STRPs) are characterized by preserved structural motifs arranged in a modular way. The structural and functional diversity of STRPs makes them particularly important for studying evolution and novel structure-function relationships, and ultimately for designing new synthetic proteins with specific functions. One crucial aspect of their classification is the estimation of geometrical parameters, which can provide better insight into their properties and the relationship between the spatial arrangement of repeated units and protein function. Calculating geometric descriptors for STRPs is challenging because naturally occurring repeats are not "perfect" and often contain insertions and deletions. Existing tools for predicting structural symmetry work well on simple cases but often fail for most natural proteins. RESULTS: Here, we present GeomeTRe, an algorithm that calculates geometrical descriptors such as curvature (yaw), twist (roll), and pitch for a protein structure with known repeat unit positions. The algorithm simulates the movement of consecutive units, identifies rotational axes, and calculates the corresponding Tait-Bryan angles. GeomeTRe's parameters can enhance STRP annotation and classification by identifying variations in geometric arrangements among different functional groups. The package is fast and suitable for processing large protein structure datasets when repeat region information (e.g. from RepeatsDB) is available. AVAILABILITY AND IMPLEMENTATION: GeomeTRe is available as a Python package; source code and documentation can be found at https://github.com/BioComputingUP/GeomeTRe.

Algorithms↗
PubMed1993

Color constancy. II. Results for two-stage linear recovery of spectral descriptions for lights and surfaces.

Our analysis of color constancy in a companion paper [J. Opt. Soc. Am A 10, 2148 (1993)] provided an algorithm that lets one test how well linear color constancy schemes work. Here we present the results of applying the algorithm to a large parametric class of color constancy problems involving bilinear models that relate photoreceptoral spectral sensitivities, surface reflectance functions, and illuminant spectral power distributions. These results, supported by simulation and further analysis, provide a detailed classification of two-stage linear methods for recovering the spectral properties of reflectances and illuminants from reflected lights.

Algorithms↗
PubMed1995

Discriminant analysis algorithm based on a distance function and on a Bayesian decision.

We propose a new algorithm for the allocation of an individual to one of several possible groups or populations. The algorithm enables us to define a finite partition over the sample space, based on distance function. This partition is used, jointly with the application of a standard Bayesian decision rule, to allocate individuals to the populations. The algorithm also provides a measure of the allocation confidence for each individual, in a similar manner to that of logistic regression. The error rates for classification are also computed using the leave-one-out method. Results are compared with those obtained with other discriminant analysis techniques previously reported: Fisher's linear discriminant function, the quadratic discriminant function, logistic discrimination, and others.

Adolescent↗
PubMed1977

Computer interpreted fetal electroencephalogram: sharp wave detection and classification of infants for one year neurological outcome.

The presence of visually discernible sharp waves (SWs) in the fetal electroencephalogram (FEEG) has been found to be associated with abnormal neurological infant outcome, but no method of programmed SW detection for FEEG was available. In order to develop an algorithm for SW detection, the first and second derivatives for visually identified SWs and non-SWs were examined and five random variables chosen for discriminant function analysis (DFA). The resulting equation, incorporated into program logic along with logic for artifact rejection, produced classifications from 85% to 89% consistent with visual identifications, suggesting that the number of SWs/epoch (NSW) corresponds with visually identified SWs. In addition, in 61 cases using a threshold for NSW derived by DFA, computer recognized SWs were found to be significantly related to the overall visual interpretation of the tracings (P less than 0.005). Finally, NSW alone produced correct classification of 65.5% of infants for 1 year neurological outcome. The overall consistency was increased to as high as 80% using additional FEEG and neonatal data. These findings imply that some forms of brain damage are present before birth and can be detected during labor using FEEG.

Brain Diseases↗
PubMed1992

Data processing for multi-channel optical recording: action potential detection by neural network.

Using a neural network, we have developed a program for fast and precise detection of action potentials (AP) in raw multi-channel optical recording data. The AP detection was performed in two steps: first, peaks were detected in raw optical data, and, second, the peaks were classified by the neural network into APs, noise and undecided peaks. The network was optimized and trained by the backpropagation learning algorithm, employing some thousands of manually classified peaks. The performance of the optimized network was found to be not completely satisfactory, although it was better than the classification by template matching and nearest-neighbor rules. The addition of a signal-to-noise ratio (SNR) of a peak to the network classification improved the classification performance: in comparison with the manual classification results, 96% of manually classified APs were detected. The causes of classification errors were discussed. In spite of the fact that the program required a slight amount of human intervention for undecided peaks, the program could allow mostly automatic AP detection.

Action Potentials↗
PubMed1990

A hybrid classifier for automated radiologic diagnosis: preliminary results and clinical applications.

We describe the design, implementation, and preliminary evaluation of a computer system to aid clinicians in the interpretation of cranial magnetic-resonance (MR) images. The system classifies normal and pathologic tissues in a test set of MR scans with high accuracy. It also provides a simple, rapid means whereby an unassisted expert may reliably label an image with his best judgment of its histologic composition, yielding a gold-standard image; this step facilitates objective evaluation of classifier performance. This system consists of a preprocessing module; a semiautomatic, reliable procedure for obtaining objective estimates of an expert's opinion of an image's tissue composition; a classification module based on a combination of the maximum-likelihood (ML) classifier and the isodata unsupervised-clustering algorithm; and an evaluation module based on confusion-matrix generation. The algorithms for classifier evaluation and gold-standard acquisition are advances over previous methods. Furthermore, the combination of a clustering algorithm and a statistical classifier provides advantages not found in systems using either method alone.

Algorithms↗
PubMed1996

Patient classification system: an optimization approach.

A patient classification system was developed integrating a patient acuity instrument with a computerized nursing distribution method based on an optimization algorithm. The objective was to minimize the total number of nursing personnel used by optimally assigning the nursing staff to meet the acuity needs of the various clinical units in the hospital. The validity of the model to assign staff was established by a 30-day comparison with the existing manual method of assignment.

Algorithms↗
PubMedDate not supplied

An empirical classification of drinking patterns among alcoholics: binge, episodic, sporadic, and steady.

Steady (daily, continuous) versus nonsteady (binge, episodic, bout, intermittent) drinking pattern have been influential Jellinek's (1960) formulation of delta and gamma drinkers, and are used as variables in various typological systems and drinker profiles. However, definitions of drinking patterns vary widely across studies, and most studies rely on one self-report item to establish a subject's pattern. To systematize and empirically test drinking-pattern schemas, we developed detailed definitions of binge, episodic, sporadic, and steady drinking patterns. A computer algorithm was written in SAS to classify 94 male alcoholics participating in outpatient conjoint therapy, using 6-month pretreatment drinking data from the Timeline Followback Interview. The final classification was: 3 (3%) binge, 33 (35%) episodic, 12 (13%) sporadic, and 40 (43%) steady drinkers. Six (6%) were unclassifiable (due to too few drinking days or too many interruptions to the pattern) by the computer. Episodic, sporadic, and steady drinkers did not differ in demographics, alcohol-related consequences, global psychological distress, or marital satisfaction. Steady drinking was associated with later onset of drinking problems (> 25), while episodic and sporadic drinking were associated with earlier onset. These results are contrary to current use of "binge drinking" as a variable associated with Type 1 alcoholism. Predictive validity analyses indicated that steady drinkers continued to drink more frequently than episodic and sporadic drinkers during treatment and 6 months posttreatment. Also, preliminary data indicate that pretreatment drinking pattern may be predictive of similar within-treatment urge-to-drink patterns. Implications for research and treatment are discussed.

Adult↗
PubMed1994

Performance analysis of manual and automated systemized nomenclature of medicine (SNOMED) coding.

Many pathology departments rely on the accuracy of computer-generated diagnostic coding for surgical specimens. At present, there are no published guidelines to assure the quality of coding devices. To assess the performance of systemized nomenclature of medicine (SNOMED) coding software, manual coding was compared with automated coding in 9353 consecutive surgical pathology reports at the Baltimore Veterans Affairs Medical Center. Manual SNOMED coding produced 13,454 morphologic codes comprising 519 distinct codes; 209 were unique codes (assigned to only one report apiece). Automated coding obtained 23,744 morphologic codes comprising 498 distinct codes, of which 129 were unique codes. Only 44 (.5%) instances were found in which automated coding missed key diagnoses on surgical case reports. Thus, automated coding compared favorably with manual coding. To achieve the maximum performance, departments should monitor the output from automatic coders. Modifications in reporting style, code dictionaries, and coding algorithms can lead to improved coding performance.

Classification↗
PubMed1993

Computational waveform analysis and classification of auditory brainstem evoked potentials.

The widely used quantitative descriptors of amplitude and latency of evoked potentials, for peaks and troughs along the waveform, relate to only a limited number of points along the waveform, ignoring the interposed data. Moreover, these descriptors are typically determined manually, rendering them susceptible to user bias. We propose and demonstrate a machine-scoring algorithm for the identification and measurement of Auditory Brainstem Evoked Potentials (ABEP) peaks I, III and V. We further introduce an algorithm for the quantitative analysis of ABEP by waveform, and for clustering records according to waveform characteristics. The results of computerized peak identification and measurement, without user intervention, were correlated with manual measurements of the same peaks in a large number of waveforms. The waveform analysis and classification procedure differentiated waveforms to monaural left, monaural right and binaural stimulation, as well as according to the recording montage. These results underscore the advantages of using information in the waveform of ABEP, which has so far been overlooked. The automated algorithms for evaluation of ABEP by waveform hold the promise of a more comprehensive and consistent evaluation, and hence improved sensitivity.

Adult↗
PubMed1985

[Automatic and semiautomatic contour finding of the left ventricle in the 2 dimensional echocardiogram. In vitro studies in formalin-fixed swine hearts].

In order to test semiautomatic and automatic contour finding procedures in 2-dimensional echocardiograms we determined endocardial borders in 42 short-axis slices of post-mortem animal hearts after interactive image enhancement such as scaling, normalisation and linearisation of grey levels semiautomatically and automatically by a complex contour finding algorithm. The areas calculated on the basis of these semiautomatic and automatic procedures were compared with "true" anatomic areas derived from planimetry. The complex computer algorithm is based on the detection and analysis of grey level gradients. The algorithm first creates a raw contour which still contains intra- and extracavitary artefacts as well as interrupted endocardial strings. Using a statistical iterative classification procedure and a least-square polynomial approximation the endocardial strings were structured, completed and smoothed and the artefacts eliminated. We were able to determine endocardial contours by semiautomatic methods in 33 (79%) and by automatic procedures in 30 (73%) of the echocardiograms. The correlation between semiautomatic and "true" contours was r = 0.97; y = 1.01x-0.46; standard error of the estimate (SEE) 0.51 cm2, between automatic and "true" contours r = 0.98; y = 0.99x - 0.31; SEE 0.43 cm2; the correlation parameters between the anatomic "true" areas and the areas calculated on the basis of manually derived borders were r = 0.98; y = 0.97x - 0.05; SEE 0.45 cm2. From our studies we conclude that left ventricular endocardium in short axis slices of postmortem animal hearts could reliably and reproducibly be detected by semiautomatic as well as by automatic procedures using a contour finding algorithm.

Animals↗
PubMed1990

Plosive/fricative distinction: the voiceless case.

Using only three measures of the waveform, the zero-crossing rate, the logarithm of the root-mean-square (rms) energy, and the derivative of the log rms energy with respect to time [termed rate of rise (ROR)], voiceless plosives (including affricates) can be distinguished from voiceless fricatives in word-initial, medial, and final positions. Peaks in the ROR contour are considered for significance to the plosive/fricative distinction by examining the log rms energy and zero-crossing rate. Then, the magnitude of the first significant peak in the ROR contour is used as the primary classifier. The algorithm was tested on 1364 tokens (720 word-initial tokens produced by four female and four male speakers; 360 word-medial tokens produced by two males and two females; 320 word-final tokens produced by two males and two females). Data from two male and two female speakers (360 word-initial tokens) were used as a training set, and the remaining data were used as a test set. The overall rate of correct classification was 96.8%. Implications of this result are discussed.

Algorithms↗
PubMed1985

Maximum likelihood estimation of variance components for a multivariate mixed model with equal design matrices.

An algorithm is described for estimating variance and covariance components by restricted maximum likelihood for a multivariate mixed two-way classification with equal design matrices. The procedure involves a transformation to canonical scale, effectively reducing a q-variate analysis to q corresponding univariate analyses. A small numerical example is given as well as a large-scale practical application.

Analysis of Variance↗