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 397 records · Page 22Linked to original sources

PubMed1991

Data selection and treatment of chemicals tested for genotoxicity and carcinogenicity.

A database containing qualitative and quantitative results of experimental studies in the fields of genotoxicity and carcinogenicity has been developed. By analyzing results of the studies performed by the U.S. National Toxicology Program, or by a similar program developed in Japan, or reported in the scientific literature, as well performed by private organizations, information has been collected relating to 3389 chemicals, identified by their CAS number. The studies considered for the database include three genotoxicity/mutagenicity short-term test (STTs), namely, two in vitro (Salmonella, gene mutation assay, and mammalian cells/human lymphocytes chromosome aberration assay) and one in vivo, the rodent bone marrow micronucleus assay. To investigate the possible predictive value of these STT assays for carcinogenicity, the results of animal long-term bioassays have also been collected. We have re-evaluated all the genotoxicity studies and the majority of those cases studied in different laboratories with contrasting results has been resolved; a small proportion of questionable cases is, however, still present in the database. In total, 2898 (85.5%) of the chemicals have been tested in the Salmonella assay; 1399 (41.3%) have been tested in the in vitro chromosome aberration assay; 319 (9.4%) have been tested in the in vivo rodent bone marrow cell micronucleus assay; 716 (21.2%) of the chemicals have been tested in the in vivo animal long-term bioassay. For 1118 chemicals tested in the Salmonella assay, 30,650 quantitative studies have been included in the database, thus allowing a possible classification of mutagenic chemicals according to their mutagenic potency.(ABSTRACT TRUNCATED AT 250 WORDS)

Algorithms↗

PubMed1992

Application of artificial neural network to computer-aided diagnosis of coronary artery disease in myocardial SPECT bull's-eye images.

We have developed a computerized system that can aid in the radiologist's diagnosis in the detection and classification of coronary artery diseases. The technique employs a neural network to analyze 201Tl myocardial SPECT bull's-eye images. This multi-layer feed-forward neural network with a backpropagation algorithm has 256 input units (pattern: compressed 16 x 16-matrix images), 5-140 units in a single hidden layer, and eight output units (diagnosis: one normal and seven different types of abnormalities). The neural network was taught using pairs of training (learning) input data (bull's-eye "EXTENT" image) and desired output data ("correct" diagnosis). The effects of the numbers of hidden units and learning iterations in the network on the recognition performance were examined. In our initial stage, the results show that the recognition performance of the neural network is better than that of the radiology resident but worse than that of the experienced radiologist. Our study also demonstrates that the result produced in the neural network depends on the variety of the training examples used. The preliminary study suggests that the neural network approach is useful for the computer-aided diagnosis of coronary artery diseases in myocardial SPECT bull's-eye images.

Coronary Disease↗

PubMed1990

[Adaptative procedures for pre-processing of evoked potentials].

The investigation of evoked potentials requires suitable consideration of physiological and pathophysiological characteristics of spontaneous and evoked electrical activity of the brain. For this purpose a preprocessing strategy based on adaptive recursive estimation of statistical parameters was developed. In this way, artifact handling, classification, filtering and further preprocessing of spontaneous EEG and evoked potentials can be improved.

Algorithms↗

PubMed2026

Improving insurance deduction identification: a hybrid artificial intelligence model using machine learning and expert systems.

PURPOSE: Financial challenges in healthcare systems worldwide, especially in low- and middle-income countries like Iran, have increased hospitals' reliance on insurance reimbursements. Unrecognized insurance deductions often cause severe financial shortages, making efficient deduction management crucial. This study aimed to design a hybrid intelligent system for identifying and predicting insurance deductions by combining machine learning and expert system frameworks. DESIGN/METHODOLOGY/APPROACH: A mixed-methods design was applied in four stages. First, a scoping review identified the causes and patterns of insurance deductions. Second, interviews with 15 insurance experts produced a validated checklist and a dataset from inpatient billing records. Third, using the CRISP-DM methodology, machine learning algorithms were developed and tested in SPSS Modeler alongside a fuzzy expert system developed in MATLAB. Finally, the model was validated using the holdout method. FINDINGS: Four categories of deduction drivers were identified: service provision, registration errors, document submission issues, and revenue conversion processes. The CHAID decision tree outperformed other algorithms with a 99% precision rate and the lowest Mean Absolute Error (9.43). A brief assessment of potential overfitting was conducted to ensure that the CHAID model's high accuracy was interpreted cautiously and supported by the validation results. The fuzzy expert system with validated rules was adaptable for deduction classification, especially for cases unsuitable for quantitative modeling. ORIGINALITY/VALUE: The hybrid model improves detection and prevention of deductions, offering actionable insights for hospital administrators, insurers, and policymakers. Its implementation can enhance hospital information systems, streamline claims processing, and optimize revenue management amid financial constraints.

Machine Learning↗

PubMed2025

Support vector machine classification of 18F-FDG PET scans across subtypes of amyotrophic lateral sclerosis.

PURPOSE: While 18F-FDG PET imaging has demonstrated diagnostic value in people with Amyotrophic Lateral Sclerosis (PwALS) and group-level differences were identified between different disease subtypes (e.g., genetic and clinical variants), refining and validating a machine-learning-based subject-level diagnostic algorithm may improve the general applicability and reliability of 18F-FDG PET as a diagnostic tool in ALS. In this study, we employed support vector machines (SVM) to further explore the diagnostic potential of 18F-FDG PET in ALS, alongside its ability to classify between different genetic subtypes or clinical phenotypes. METHODS: 18F-FDG PET data of 36 healthy volunteers (HV), 25 people with ALS-mimicking diseases (Mimics), and 167 PwALS, grouped by genetic status (e.g., sporadic (sALS) or carrying a C9orf72 hexanucleotide repeat expansion (ALSC9orf72RE) and onset (bulbar or spinal) type, acquired with Biograph 'TruePoint' PET/CT scanner, were included in the study (Dataset 1). A second dataset of 183 PwALS and 31 Mimics acquired with Biograph 'HiRez' scanner was included as an independent cross-validation set (Dataset 2). PET images were spatially normalised to MNI space to fit linear SVMs with cross-validation. Only age-matched groups were considered to eliminate age-related effects. RESULTS: For Dataset 1, the linear SVM resulted in an average accuracy of 0.86 for the classification of ALS vs. HV, 0.53 for ALS vs. Mimics, 0.83 for ALSC9orf72RE vs. sALS, and 0.58 for bulbar vs. spinal onset. These findings were corroborated with Dataset2, with an accuracy of up to 0.76 for ALSC9orf72RE vs. sALS, and 0.59 for bulbar vs. spinal. CONCLUSION: 18F-FDG brain PET imaging, combined with SVM and age-matching, can distinguish between ALSC9orf72RE and sALS with good accuracy, but lacks sufficient discriminative power to differentiate between ALS and Mimics and between different sites of onset.

Humans↗

PubMed1995

Feasibility study for identifying adverse events attributable to vaccination by record linkage.

To investigate the feasibility of using a record linkage method for identifying vaccine attributable adverse events, computerized hospital admissions and vaccination records from South East Kent district were linked and checked for accuracy. Records for 90% of children under 2 years of age admitted to hospital over a 2-year period were matched with vaccination records using a computer algorithm based on name, date of birth, sex, and post-code supplemented by visual inspection. Relative to this gold standard, matching on date of birth, sex and postcode alone had a sensitivity of 60% and an incorrect match rate of 0.2% after matches to more than one vaccine recipient were excluded. Manual checking of a sample of admissions showed that only 4% had been assigned incorrect International Classification of Disease (ICD) codes. Routine record linkage of ICD admission codes to vaccination records therefore yields data of good quality which may be used for surveillance purposes.

Adverse Drug Reaction Reporting Systems↗

PubMed1989

Fully automated measurements by light microscopy of tissue sections using a cellular array computer.

Software was developed for the acquisition, segmentation and analysis of microscopic OD-images on a VICOM digital image processor, extended with a VISIOMORPH morphoprocessor board. The delineation algorithms for peroxisomes, lysosomes, and nuclei in liver, kidney, and adrenal gland sections start by thresholding the difference between the original image and a low pass filtered version. The resulting binary mask is then processed by morphological operations in order to produce an object overlay. The efficiency of the programs is evaluated by comparing delineated objects at different OD-levels, created by varying the stain or by multiplying the original pixel values with constant factors. Manual delineation on some images is also used as a reference. More complex algorithms are used for the delineation of muscle fibres in ATP-ase-stained sections and immunocytochemically labelled cells in monolayer preparations. Muscle images from parallel sections with different stainings are matched with a coordinate transform, enabling the transfer of the object mask from a single delineated image to the unprocessed images and thus obtain all necessary information for fibre classification. After segmentation, the OD-images and their object overlays are fed into a data extraction program, measuring for each delineated object user-selected features. Data are sent to a VAX for statistical interpretation.

Adrenal Cortex↗

PubMed1995

Debris compensation of DNA histograms and its effect on S-phase analysis.

Debris compensation is an important variable affecting S-phase fraction (SPF) analysis in flow cytometric DNA histograms. The SPF was estimated in fresh frozen breast carcinomas using the following four debris subtraction algorithms: modeling debris as an exponential curve (EXP); the incorporation of nuclei cut a single time into the exponential moel (EXP-SC); the random cutting of nuclei into multiple pieces of varying sizes (MC); and a combination of both nuclear cutting models (SC-MC). Comparison of SPF estimates indicated that the various debris subtraction models yielded differences in SPF, with SPF values obtained using the exponential model having considerable variation compared to SPF estimates from histograms where debris was modeled by algorithms based on nuclear slicing and fragmentation. However, SPF estimates could be affected by initial placement of the nuclear debris boundaries, the coefficient of variation of the G0/1 peak, and the relative amount of debris. Using the ratio of the height of the G0/1 peak to the height of the debris between the chicken red blood cells (CRBC) and G0/1 peaks as an objective measurement of nuclear debris, debris compensation was necessary in diploid DNA histograms where this ratio was as low as 1.5%. Taken in the context of SPF prognostic cutoff levels, variation in debris models and boundaries can change the classification of cases with borderline SPF levels into the poor prognostic high SPF categories, thereby making the comparison of SPF values derived from different studies difficult.

Adenocarcinoma↗

PubMed1994

The way to EEG-classification: transition from language of patterns to language of systems.

An approach is proposed based on identification of systems underlying EEG to reduce the number of descriptors required. Five-minute EEG records were collected from 39 normal subjects in 3 states: eyes closed, fingering beads, eyes open. The algorithm included a spectral analysis, separation of rhythms by a finite memory filter, and autocorrelation analysis of the oscillations and their envelope. Only three oscillatory systems were identified reliably. All of them pertain to the alpha-band: independent frontal and occipital alpha-rhythms, and a mu-rhythm. A division of the subjects into 5 groups has been based on the presence or absence of these systems. Across groups, the mean frequency of the alpha-system was 10.2 Hz, mu-system--10.0 Hz. The mean damping decrement equalled .24 for the occipital alpha system, .30 for mu. The assumption that rhythms in the alpha-band can be created due to the physiological equivalent of narrow-band filters with stochastic neuronal activity at their input has been confirmed in the majority of cases.

Adolescent↗

PubMed1991

Microcomputer software applied to corneal stromal biometry.

Here we report a new method for image analysis of the corneal stroma. To obtain the biometric characteristics of a given area, we perform three different treatments of the same image. These automated predictions closely match each parameter (length, surface area, etc.), as measured manually by a "blind investigator." Furthermore, to obtain quantifiable, average values, we have increased the number of successive image measurements, which has led to the development of a series of programs designed to optimize automated data handling. Finally, the acquired, calculated parameters are summarized in the form of intermediate tables for each series of images, and as a final summary table incorporating t-test values and permitting comparison between two stromas (e.g., normal and pathological). Multivariate analysis and an ascending hierarchical classification demonstrate the main trends in differences of pathological vs. normal stroma.

Algorithms↗

PubMedDate not supplied

A classification of Scottish infants using latent class analysis.

This paper illustrates the use of latent class analysis to classify 50,000 infants into a small number of classes or case types, as a preliminary to a study of the allocation of neonatal hospital resources throughout Scotland. Information, extracted from a detailed neonatal discharge record, was summarized by 11 clinical and diagnostic catagorical variables. Statistical models incorporating 1 to 6 latent classes were then estimated using the EM algorithm. The 4 class model was chosen because it provided a good description of the data and the resulting classes had a medical interpretation. The factors influencing the choice of model are discussed and goodness of fit tests are presented. The stability of the classes was also investigated using random halves of the data and an earlier comparable data set.

Classification↗

PubMed1990

Comparison of pattern recognition methods for computer-assisted classification of spectra of heart sounds in patients with a porcine bioprosthetic valve implanted in the mitral position.

The diagnostic performance of two pattern recognition methods (or classifiers) to detect valvular degeneration was evaluated in 48 patients with a porcine bioprosthetic heart valve inserted in the mitral position. Twenty patients had a normal porcine bioprosthetic valve and 28 patients had a degenerated bioprosthetic valve. One method was based on the Gaussian-Bayes model and the second on the "nearest neighbor" algorithm using three distance measurements. Eighteen diagnostic features were extracted from the sound spectrum of each patient and, for each method, a two-class supervised learning approach was used to determine the most discriminant diagnostic patterns composed of 6 features or less. The probability of error of the classifiers was estimated with the leave-one-out approach. The performance of each method to discriminate between normal and degenerated bioprosthetic valves was verified by clinical evaluation of the valves. The best performance in evaluation of the sound spectrum (98% correct classifications) was obtained with the Bayes classifier and two patterns of six features each. The percentage of false positive classifications of valve degeneration was 0% and the percentage of false negative classifications was 4%. Sensitivity for the detection of valve degeneration was 96%, specificity was 100%, positive predictive value was 100%, and negative predictive value was 95%. The best performance of the nearest neighbor method (94% correct classifications) was obtained by using the Mahalanobis distance and five patterns composed of three, four, five, or six diagnostic features. Using a pattern composed of only three features, the percentage of false positive classifications for degeneration was 10% and the percentage of false negative classifications was 4%.(ABSTRACT TRUNCATED AT 250 WORDS)

Bayes Theorem↗

PubMed1990

System for flow sorting chromosomes on the basis of pulse shape.

Sorting on the basis of the complex features resolved by chromosome slit-scan analysis requires rapid and flexible pulse shape acquisition and processing for determining sort decisions before droplet breakoff. Fluorescence scans of chromosome morphology contain centromeric index and banding information suitable for chromosome classification, but these scans are often characterized by variability in length and height and require sophisticated data processing procedures for identification. Setting sort criteria on such complex morphological data requires digitization and subsequent computation by an algorithm tolerant of variations in overall pulse shape. We demonstrate here the capability to sort individual chromosomes based on their morphological features measured by slit-scan flow cytometry. To do this we have constructed a sort controller capable of acquiring an 128 byte chromosome waveform and executing a series of numerical computations resulting in an area-based centromeric index sort decision in less than 2 ms. The system is configured in a NOVIX microprocessor, programmed in FORTH, and interfaced to a slit-scan flow cytometer data acquisition system. An advantage of this configuration is direct control over the machine state during program execution for minimal processing time. Examples of flow sorted chromosomes are shown with their corresponding fluorescence pulse shapes.

Animals↗

PubMedDate not supplied

Image processing for the rest of us: the potential utility of inexpensive computerized image analysis in clinical pathology and radiology.

Recent progress in computer technology in both hardware and software, combined with marked cost reductions, have placed quantitatively accurate video densitometry systems within the reach of the individual clinician, biomedical researcher, and community hospital. While much of the attention generated by advances in image processing has focussed on larger scale procedures, such as CAT, chemical shift, and positron emission tomography, important applications can be found for considerably more modest systems. In this article, we discuss three such applications of DUMAS, a personal computer-based imaging system developed by the Image Processing Center at Drexel University. A potential technique for quantifying numbers of estrogen receptors in tumorous breast tissue samples as a predictor of patient responsiveness to hormonal therapy is described first, along with possible sources of error. The second application, also related to clinical pathology and cancer, outlines methods for relating changes in nuclear and cell morphology to the diagnosis of Sezary Cell Syndrome. The utility of binary image filtering methods in the classification of cell types is discussed. The third application involves the development of a semi-automatic procedure for the determination of vessel diameter in arteriograms. A detailed description of the optimization and curve-fitting algorithms is provided along with preliminary test results comparing various approaches. The need for user demand to fuel research and development in small-scale imaging systems is also discussed.

Algorithms↗

PubMed1992

Histopathological prediction of liver metastasis after curative resection of colorectal cancer.

To estimate the risk of liver metastasis after curative resection of colorectal cancer, resected specimens from 290 patients (45 with metachronous liver metastasis) were examined and the relationships between 10 histopathological variables and liver metastasis were analysed using our application of the Akaike information criterion (AIC). Of the 10 variables examined, the depth of venous invasion (Vd) had the greatest prognostic value for metastasis, followed by the number of venous invasions, the number of lymphovascular invasions, lymph node metastasis and type of infiltration. The prediction of liver metastasis was further improved by combining Vd with lymphocyte infiltration, mucinous production, interstitial fibrosis or depth of penetration, although these four variables per se were minimally informative for metastasis. We conclude that the prediction of liver metastasis is best achieved by combining Vd with other variables. Our risk group classification, and the estimated probability of liver metastasis for each group, are shown.

Adenocarcinoma↗

PubMedDate not supplied

Staging of maxillary cancer. Which classification?

Of the many proposed classifications for staging maxillary sinus cancer, none has been adopted universally and none is known to be superior to the others. This study identified the best of six currently used classifications using data from 53 previously untreated patients with squamous cell carcinoma of the maxillary sinus. Analysis of each classification's ability to stage the majority of patients, produce a balanced distribution of T stages, and correlate T stage with treatment and prognosis revealed Harrison's classification to be the best. Harrison's classification should be adopted worldwide as the classification of choice for staging squamous cell carcinoma of the maxillary sinus.

Algorithms↗

PubMed1994

DNA sequence analysis using hierarchical ART-based Classification Networks.

Adaptive resonance theory (ART) describes a class of artificial neural network architectures that act as classification tools which self-organize, work in realtime, and require no retraining to classify novel sequences. We have adapted ART networks to provide support to scientists attempting to categorize tandem repeat DNA fragments from Onchocerca volvulus. In this approach, sequences of DNA fragments are presented to multiple ART-based networks which are linked together into two (or more) tiers; the first provides coarse sequence classification while the subsequent tiers refine the classifications as needed. The overall rating of the resulting classification of fragments is measured using statistical techniques based on those introduced by Zimmerman, et al. (1994) to validate results from traditional phylogenetic analysis. Tests of the Hierarchical ART-based Classification Network, or HABclass network, indicate its value as a fast, easy-to-use classification tool which adapts to new data without retraining on previously classified data.

Algorithms↗

PubMed1989

Classification and evaluation of the obesities.

Obesity is a condition of multifactorial etiology that can be associated with important health and functional consequences. Suggestions for the proper evaluation of obese patients have been presented along with brief descriptions of the rationales for their use. The evaluation protocol has been summarized in two algorithms to aid in the performance of a complete and organized work-up.

Adipose Tissue↗