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 1,585 records · Page 88Linked to original sources

PubMed2005

Applying pattern recognition methods and structure property correlations to determine drug carrier potential of nicotinic acid and analogize to dihydropyridine.

Multivariate methods are utilized to compare nicotinic acid and dihydropyridine as a drug carrier. Nicotinic acid and dihydropyridine form ester groups on 10 beta-lactam antibiotics with an oxymethyl group forming a linkage between the antibiotic and the drug carrier (nicotinic acid or dihydropyridine). Calculated molecular properties are analyzed by self-organizing tree algorithm (SOTA), bivariate regression, cluster analysis, factor analysis, discriminant analysis, hierarchical classification, and principal coordinates analysis. Ten important pharmacological properties for each of the nicotinic acid and dihydropyridine antibiotic derivatives are numerically similar and highly correlated. Calculated molecular properties include molar refractivity, molar volume, parachor, index of refraction, partition coefficient (log P), polarizability, and polar surface area. Dermal permeability coefficients (Kp) for nicotinic acid derivatives are similar to values for dihydropyridine derivatives. Dermal permeabliity coefficients analyzed by hierarchical classification and SOTA analysis were shown to be closely interrelated and highly correlated. Ten properties of the nicotinic acid and dihydropyridine were compared by Passing-Bablok regression analysis and shown to be highly correlated (r=0.9879). Box plot analysis of 10 properties, inclusive of both groups of derivatives, showed narrow ranges in values. Cluster analysis of derivative properties showed the nicotinic acid derivatives to be highly similar to the dihydropyridine derivatives of the same antibiotics. Cluster analysis was performed by single linkage, complete linkage, and centroid linkage. Factor analysis showed the nicotinic acid derivatives to be interrelated and similar to the dihydropyridine derivatives. Discriminant analysis performed on all derivatives formed a single highly cohesive and non-differentiated cluster, demonstrating strong similarity among nicotinic acid and dihydropyridine derivatives. Principal coordinates analysis (determines similarity) of Kp values showed high similarity between the nicotinic acid and dihydropyridine antibiotic derivatives.

Dihydropyridines↗

PubMed1995

A new system for classification of cardiac death as arrhythmic, ischemic, or due to myocardial pump failure.

Existing classifications of cardiac death fail to incorporate current understanding of the pathophysiology of sudden cardiac death. We developed a new scheme for classifying cardiac death that defines 3 categories of underlying mechanism: primary arrhythmia, acute myocardial ischemia/infarction, and myocardial pump failure. Using this new system, we classified the mechanism of 106 definite cardiac deaths from the Recurrent Coronary Prevention Project. Fifty deaths (47%) were classified as arrhythmic, 46 (43%) as ischemic, and 9 (8%) as due to myocardial pump failure (1 death was not classifiable). All 36 witnessed arrhythmic deaths were sudden and 8 of 9 witnessed myocardial pump failure deaths were nonsudden. The 38 witnessed ischemic deaths were split evenly between sudden and nonsudden. Interrater agreement for the classification of mechanism was 100%. This classification scheme, if validated in subsequent studies, will provide a useful algorithm for classifying deaths by underlying mechanism.

Arrhythmias, Cardiac↗

PubMed2001

Magnetic resonance spectroscopy of the malignant prostate gland after radiotherapy: a histopathologic study of diagnostic validity.

PURPOSE: Accurate spatial representation of tumor clearance after conformal radiotherapy is an endpoint of clinical importance. Magnetic resonance spectroscopy (MRS) can diagnose malignancy in the untreated prostate gland through measurements of cellular metabolites. In this study we sought to describe spectral metabolic changes in prostatic tissue after radiotherapy and validate a multivariate analytic strategy (based on MRS) that could identify viable tumor. METHODS AND MATERIALS: Transrectal ultrasound-guided prostate biopsies from 35 patients were obtained 18-36 months after external beam radiotherapy. One hundred sixteen tissue specimens were subjected to 1H MRS, submitted to histopathology, and analyzed for correlation with a multivariate strategy specifically developed for biomedical spectra. RESULTS: The sensitivity and specificity of MRS in identifying a malignant biopsy were 88.9% and 92% respectively, with an overall classification accuracy of 91.4%. The diagnostic spectral regions identified by our algorithm included those due to choline, creatine, glutamine, and lipid. Citrate, an important discriminating resonance in the untreated prostate gland, was invisible in all spectra, regardless of histology. CONCLUSIONS: Although the spectral features of prostate tissue markedly change after radiotherapy, MRS combined with multivariate methods of analysis can accurately identify histologically malignant biopsies. MRS shows promise as a modality that could integrate three-dimensional measures of tumor response.

Aged↗

PubMed1998

Preoperative intensive care unit admission and hemodynamic monitoring in patients scheduled for major elective noncardiac surgery: a retrospective review of 95 patients.

OBJECTIVE: To review experience with preoperative intensive care unit (ICU) admission and hemodynamic monitoring to determine which patients benefited and how. DESIGN: Retrospective review over 32-month period (1991 to 1994). SETTING: Surgical ICU of a university teaching hospital. PARTICIPANTS: Ninety-five patients admitted to the surgical ICU before a major elective noncardiac, nonthoracic surgical procedure. INTERVENTIONS: All patients underwent hemodynamic monitoring with a pulmonary artery catheter (PAC). Interventions were made at the discretion of the ICU attending and attending surgeon, based on a general algorithm. Patients were categorized based on history or hemodynamics. The historic classification was as follows: group I, patients with cardiac disease documented by history and cardiac imaging, n = 37; group II, patients with cardiac disease documented by history, but not cardiac imaging, n = 24; group III, patients without documented cardiac disease, but with other significant medical problems, n = 34. Hemodynamic classification considered patients to have subnormal parameters if the cardiac index was < 2.5 L/min/m2, the mixed venous oxygen saturation was < 65%, or the oxygen delivery index was < 350 mL/min/m2 (n = 45), and normal parameters if greater than these (n = 50). MAIN RESULTS: There were no differences in APACHE II scores. Group I patients had greater Goldman Cardiac Risk Indices than group III patients (7.4 +/- 4.8 v 5.0 +/- 3.0). Patients in group I had a significantly greater incidence of subnormal initial hemodynamic values (63%) than patients in group II (47%) or group III (32%). The incidence of postoperative cardiovascular complications among groups was not different. Fifty patients (52%) had normal hemodynamics initially; two (4%) developed postoperative cardiovascular complications compared with 10 patients (22%) of the 45 with subnormal initial hemodynamic values. Of these 45 patients, 24 (52%) had their hemodynamic parameters corrected preoperatively with crystalloids, packed red blood cells, inotropes, and/or afterload reduction. Two of these 24 patients (8%) experienced postoperative cardiovascular complications, compared with 8 of the remaining 21 patients who had no attempt to normalize their hemodynamic values preoperatively other than maintaining a normal pulmonary artery occlusion pressure. CONCLUSIONS: Patients who had normal initial preoperative hemodynamic parameters or abnormal initial parameters that were normalized preoperatively experienced significantly fewer perioperative cardiovascular complications than those with abnormal initial values that were not normalized preoperatively. These results suggest that there may be benefit to the practice of preoperative ICU admission, hemodynamic monitoring with a PAC, and "optimization" of cardiac function in selected patients undergoing major elective noncardiac surgery. Further studies are needed to better delineate the most appropriate patient populations and effective therapeutic protocol.

Adult↗

PubMed2000

Synovial, chondropathic, depositional: the radiological categories of arthritis. A review.

The accepted optimum logic frame for complex diagnostic problems is sequential hypothesis testing. The X-ray diagnosis of arthritis has now become sufficiently complex to make this the procedure of choice, but standard classifications of arthritis lack the discriminatory power needed for its effective deployment. This obstacle can be overcome if these standard classifications are replaced by a more discriminant classification based on radiologically identifiable discriminators. This classification divides arthritis into three groups: synovial, chondropathic and depositional. The initial categorization is usually made fairly simply from two or three markers, of which the most important is the site of any erosions present. Once categorized, the subsequent analysis can proceed in an approximately binary manner using the discriminators appropriate for that category. The use of this radiological classification simplifies the diagnostic approach, reduces the workload, and provides the algorithm needed for sequential hypothesis testing.

Algorithms↗

PubMed1997

Classifying rehabilitation inpatients by expected functional gain.

OBJECTIVES: To create a more suitable payment system for medical rehabilitation, the authors developed a companion classification system to the original functional independence measure-function-related groups (FIM-FRGs), which classify patients having similar lengths of stay in a rehabilitation hospital or inpatient unit. The companion system presented here groups patients according to their gains in functional status during the rehabilitation stay. METHODS: Data from 84,492 patients discharged from 252 rehabilitation facilities in 1992 were provided by the Uniform Data System for Medical Rehabilitation. Classification rules were formed using clinical judgment and a recursive partitioning algorithm. The gain-FRGs system used four predictor variables: (1) diagnosis leading to disability, admission scores on the (2) motor and (3) cognitive subscales of the FIM, and (4) patient age. RESULTS: The gain-FRGs system contained 74 patient groups and explained 21% of the variation in functional gain for patients in a different set of records withheld for validation. CONCLUSIONS: The gain-FRGs system should be considered for prospective payment systems because it gives the provider an incentive to improve patient outcomes, which is missing in a payment system based on FIM-FRGs alone.

Activities of Daily Living↗

PubMedDate not supplied

Strengthening sociometric prediction: scientific advances in the assessment of children's peer relations.

This study assessed the strength of sociometric classification in the prediction of concurrent sociobehavioral adjustment. Differential adjustment for subgroups of unclassified children were also examined. Participants were 881 fifth graders (ages 9 to 12). Classification strength (CS) and unclassified subgroups were determined through newly developed algorithms. CS added significantly to the prediction of all areas of adjustment. For example, highly rejected children were at extreme risk for victimization whereas highly controversial children were most likely to be bullies and relationally aggressive. Unclassified subgroups were found to exhibit adjustment problems mirroring those of their extreme status group counterparts. Findings support that increasing the sensitivity of sociometric measurement results in both greater predictive strength and enhanced understanding of underlying social processes.

Adjustment Disorders↗

PubMed2005

The use of verbal autopsy in attributing cause of death from epilepsy.

Verbal autopsy (VA) is a method used increasingly in low-income countries to ascertain the likely cause of death. Originally developed for use in pediatrics and infectious disease, it is also useful in situations with poor data on adult mortality. Using information on cause of death ascertained through interviews with relatives or other associates of the deceased, assessment of the likely cause of death is performed by expert clinicians. Methodological issues related to the VA method include robustness when dealing with less-frequent causes of death; use of open or closed mortality classification; and possible selection bias in assessing cause of death. However, alternatives, including death certification, share many sources of imprecision. The clear advantage of the VA method is that it does not depend on access to clinic services, although clinical experts are used for the classification process. Even this may be challenged by the development of algorithms for the automatic coding of cause of death. VA has not been systematically applied to studies of causes of death in developed countries where physicians are available to complete death certificates. However, in studies of sudden unexplained death in developed countries, techniques similar to VA are used to ascertain circumstances surrounding death to be able to potentially classify it as sudden and unexplained. Thus, family and friends of the deceased are potentially valuable contributors, using the VA methodology: VA is useful in low-income countries to estimate the burden of epilepsy on all-cause mortality and, in industrialized nations it offers insights into sudden death from epilepsy.

Autopsy↗

PubMed2002

A parallel mixture of SVMs for very large scale problems.

Support vector machines (SVMs) are the state-of-the-art models for many classification problems, but they suffer from the complexity of their training algorithm, which is at least quadratic with respect to the number of examples. Hence, it is hopeless to try to solve real-life problems having more than a few hundred thousand examples with SVMs. This article proposes a new mixture of SVMs that can be easily implemented in parallel and where each SVM is trained on a small subset of the whole data set. Experiments on a large benchmark data set (Forest) yielded significant time improvement (time complexity appears empirically to locally grow linearly with the number of examples). In addition, and surprisingly, a significant improvement in generalization was observed.

Algorithms↗

PubMed2003

An open Internet platform to distributed image processing applied to dermoscopy.

Proprietary systems for dermoscopy images analysis are available to improve the diagnosis and follow-up of the pigmented skin lesions. Their performance seems comparable with that of a human expert. Progress in computer-aided classification of melanocytic lesions depends notably on judicious choices of the algorithms dedicated to the extraction of signs from the dermoscopy images and of the method which combines these signs to classify the lesions. To allow the researcher's community to benefit from their large set of elementary algorithms already available for dermoscopy, we set up a system accessible through the Internet which: allows the engineers to register their algorithms while preserving their secrecy: their programs run on their own server; lets a user to define its own sequence of image analysis and to apply it to its images: the system automatically calls the appropriate remote programs; makes possible and stimulates the synergy of worldwide researchers in order to validate algorithms of images analysis best suited to achieve the correct diagnosis and to track the malignant melanoma; makes these techniques available to the greatest number of users through the Web and thus to support a mass screening; reduces the maintenance of the system to the minimum: it requires users only an Internet browser and engineers to follow a simple widespread standardised interface for distributed programs. Various problems should be addressed: the lack of standardisation of images acquisition: algorithms based on relative colours are best suited to this system; the copyrights on images and algorithms; charging the use of remote computer resources. This system allows for an international collaborative work in the fight against the malignant melanoma by offering a conceptual and technical platform of teledermoscopy. It is intended to support synergy between the engineers and the users implied in the diagnosis and teaching of dermoscopy.

Algorithms↗

PubMed2004

[The functional state classification and evaluation of the stability level in mental loads based on the factor structure of heart rate variability parameters].

In the paper, outcomes of the researches devoted to factor analysis of heart rate variability parameters and definition of the most informative parameters for diagnostics of functional states and an evaluation of level of stability to mental loads, are presented. The factor structure of parameters, which unclude integral level of heart rate variability (1), balance between activity of vagus and brain cortical-limbic systems (2), integrated level of cardiovascular system functioning (3), is substantiated. Factor analysis outcomes have been used for construction of functional state classification, for their differential diagnostics, and for development and check of algorithm for evaluation of the stability level in mental loads.

Adult↗

PubMed1991

[Plastic surgery measures in combined soft tissue defects of the lower leg].

Isolated or combined soft tissue defects of the lower leg are still a challenge as regard diagnostic and differential treatment. We distinguish acute and chronic isolated or combined soft tissue defect. Acute trauma within a multiple injury pattern is a special form of the acute soft tissue defect. For diagnostic purposes we use a preoperative and intraoperative standardized diagnostic programme. A new classification oriented towards the therapeutic procedure is presented, as is therapy algorithm which has been created.

Follow-Up Studies↗

PubMed2005

Computer-aided detection of lung nodules: false positive reduction using a 3D gradient field method and 3D ellipsoid fitting.

We are developing a computer-aided detection system to assist radiologists in the detection of lung nodules on thoracic computed tomography (CT) images. The purpose of this study was to improve the false-positive (FP) reduction stage of our algorithm by developing features that extract three-dimensional (3D) shape information from volumes of interest identified in the prescreening stage. We formulated 3D gradient field descriptors, and derived 19 gradient field features from their statistics. Six ellipsoid features were obtained by computing the lengths and the length ratios of the principal axes of an ellipsoid fitted to a segmented object. Both the gradient field features and the ellipsoid features were designed to distinguish spherical objects such as lung nodules from elongated objects such as vessels. The FP reduction performance in this new 25-dimensional feature space was compared to the performance in a 19-dimensional space that consisted of features extracted using previously developed methods. The performance in the 44-dimensional combined feature space was also evaluated. Linear discriminant analysis with stepwise feature selection was used for classification. The parameters used for feature selection were optimized using the simplex algorithm. Training and testing were performed using a leave-one-patient-out scheme. The FP reduction performances in different feature spaces were evaluated by using the area Az under the receiver operating characteristic curve and the number of FPs per CT section at a given sensitivity as accuracy measures. Our data set consisted of 82 CT scans (3551 axial sections) from 56 patients with section thickness ranging from 1.0 to 2.5 mm. Our prescreening algorithm detected 111 of the 116 solid nodules (nodule size: 3.0-30.6 mm) marked by experienced thoracic radiologists. The test Az values were 0.95 +/- 0.01, 0.88 +/- 0.02, and 0.94 +/- 0.01 in the new, previous, and combined feature spaces, respectively. The number of FPs per section at 80% sensitivity in these three feature spaces were 0.37, 1.61, and 0.34, respectively. The improvement in the test Az with the 25 new features was statistically significant (p<0.0001) compared to that with the previous 19 features alone.

Algorithms↗

PubMed1999

Epidemiological classification of acute myocardial infarction: time for a change?

AIMS: The classification of an acute ischaemic cardiac event is traditionally based on cardiac enzymes, electrocardiography (ECG) and clinical symptoms. The impact of new specific cardiac markers on the diagnostic classification of suspected acute myocardial infarction remains poorly studied. We therefore set out to compare the diagnostic and prognostic information provided by the MONICA code and a patient classification based on the maximal level of creatine kinase MB isoenzyme. The significance of typical pain and various ECG algorithms were separately analysed. METHODS AND RESULTS: The study population consisted of 311 consecutive patients who were evaluated for suspected acute myocardial infarction in a regional referral hospital. Patients were retrospectively classified according to the MONICA criteria, by a simplified code combining symptoms and creatine kinase MB, and solely using the maximal creatine kinase MB concentration. Total mortality was followed for 1 and 5 years. The creatine kinase MB based classification was shown to be the strongest predictor of mortality (OR=2.8-3.7, p<0.001) for outcome both at 1 and 5 years. Typical pain and a positive Minnesota ECG had no prognostic relevance. However, an analysis algorithm of the admission ECG was predictive of 1- and 5-year survival. CONCLUSIONS: The epidemiological classification of suspected acute myocardial infarction could be based solely on a specific cardiac marker, such as creatine kinase MB mass. This approach contains prognostic information and is accurate enough for the structured diagnosis of acute myocardial infarction. Other outcome predictors could be used to identify patient subgroups and assess therapy.

Age Distribution↗

PubMed2005

Data-partitioning using the Hilbert space filling curves: effect on the speed of convergence of Fuzzy ARTMAP for large database problems.

The Fuzzy ARTMAP algorithm has been proven to be one of the premier neural network architectures for classification problems. One of the properties of Fuzzy ARTMAP, which can be both an asset and a liability, is its capacity to produce new nodes (templates) on demand to represent classification categories. This property allows Fuzzy ARTMAP to automatically adapt to the database without having to a priori specify its network size. On the other hand, it has the undesirable side effect that large databases might produce a large network size (node proliferation) that can dramatically slow down the training speed of the algorithm. To address the slow convergence speed of Fuzzy ARTMAP for large database problems, we propose the use of space-filling curves, specifically the Hilbert space-filling curves (HSFC). Hilbert space-filling curves allow us to divide the problem into smaller sub-problems, each focusing on a smaller than the original dataset. For learning each partition of data, a different Fuzzy ARTMAP network is used. Through this divide-and-conquer approach we are avoiding the node proliferation problem, and consequently we speedup Fuzzy ARTMAP's training. Results have been produced for a two-class, 16-dimensional Gaussian data, and on the Forest database, available at the UCI repository. Our results indicate that the Hilbert space-filling curve approach reduces the time that it takes to train Fuzzy ARTMAP without affecting the generalization performance attained by Fuzzy ARTMAP trained on the original large dataset. Given that the resulting smaller datasets that the HSFC approach produces can independently be learned by different Fuzzy ARTMAP networks, we have also implemented and tested a parallel implementation of this approach on a Beowulf cluster of workstations that further speeds up Fuzzy ARTMAP's convergence to a solution for large database problems.

Algorithms↗

PubMed2002

A robust method for extraction and automatic segmentation of brain images.

A new protocol is introduced for brain extraction and automatic tissue segmentation of MR images. For the brain extraction algorithm, proton density and T2-weighted images are used to generate a brain mask encompassing the full intracranial cavity. Segmentation of brain tissues into gray matter (GM), white matter (WM), and cerebral spinal fluid (CSF) is accomplished on a T1-weighted image after applying the brain mask. The fully automatic segmentation algorithm is histogram-based and uses the Expectation Maximization algorithm to model a four-Gaussian mixture for both global and local histograms. The means of the local Gaussians for GM, WM, and CSF are used to set local thresholds for tissue classification. Reproducibility of the extraction procedure was excellent, with average variation in intracranial capacity (TIC) of 0.13 and 0.66% TIC in 12 healthy normal and 33 Alzheimer brains, respectively. Repeatability of the segmentation algorithm, tested on healthy normal images, indicated scan-rescan differences in global tissue volumes of less than 0.30% TIC. Reproducibility at the regional level was established by comparing segmentation results within the 12 major Talairach subdivisions. Accuracy of the algorithm was tested on a digital brain phantom, and errors were less than 1% of the phantom volume. Maximal Type I and Type II classification errors were low, ranging between 2.2 and 4.3% of phantom volume. The algorithm was also insensitive to variation in parameter initialization values. The protocol is robust, fast, and its success in segmenting normal as well as diseased brains makes it an attractive clinical application.

Adult↗

PubMed2005

A machine learning approach for automated recognition of movement patterns using basic, kinetic and kinematic gait data.

This paper investigated application of a machine learning approach (Support vector machine, SVM) for the automatic recognition of gait changes due to ageing using three types of gait measures: basic temporal/spatial, kinetic and kinematic. The gaits of 12 young and 12 elderly participants were recorded and analysed using a synchronized PEAK motion analysis system and a force platform during normal walking. Altogether, 24 gait features describing the three types of gait characteristics were extracted for developing gait recognition models and later testing of generalization performance. Test results indicated an overall accuracy of 91.7% by the SVM in its capacity to distinguish the two gait patterns. The classification ability of the SVM was found to be unaffected across six kernel functions (linear, polynomial, radial basis, exponential radial basis, multi-layer perceptron and spline). Gait recognition rate improved when features were selected from different gait data type. A feature selection algorithm demonstrated that as little as three gait features, one selected from each data type, could effectively distinguish the age groups with 100% accuracy. These results demonstrate considerable potential in applying SVMs in gait classification for many applications.

Aged↗

PubMed1999

Using neural networks and genetic algorithms to enhance performance in an electronic nose.

Sensitivity, repeatability, and discernment are three major issues in any classification problem. In this study, an electronic nose with an array of 32 sensors was used to classify a range of odorous substances. The collective time response of the sensor array was first partitioned into four time segments, using four smooth time-windowing functions. The dimension of the data associated with each time segment was then reduced by applying the Karhunen-Loéve (truncated) expansion (KLE). An ensemble of the reduced data patterns was then used to train a neural network (NN) using the Levenberg-Marquardt (LM) learning method. A genetic algorithm (GA)-based evolutionary computation method was used to devise the appropriate NN training parameters, as well as the effective database partitions/features. Finally, it was shown that a GA-supervised NN system (GANN) outperforms the NN-only classifier, for the classes of the odorants investigated in this study (fragrances, hog farm air, and soft beverages).

Algorithms↗