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 37 records · Page 2Linked to original sources
PubMed2005

Hands-free human computer interaction via an electromyogram-based classification algorithm.

A four-electrode system for hands-free computer cursor control, based on the digital processing of Electromyogram (EMG) signals is proposed. The electrodes are located over the right frontalis, the procerus, the left temporalis and the right temporalis muscles in the head. This system is meant to enable individuals paralyzed from the neck down (e.g., due to Spinal Cord Injury) to interact with computers using point-and-click graphic interfaces. The intention is to translate electromyograms derived from muscle contractions associated with specific facial movements into five cursor actions, namely: Left, Right, Up, Down and Left-click. This translation is accomplished by a digital signal processing classification algorithm that takes advantage of the divergent spectral nature of the EMG signals produced by the frontalis, temporalis, and procerus muscles, respectively. The effectiveness of the algorithm is evaluated by comparing its performance to that of a previously developed three-electrode EMG-based algorithm, using Matlab simulations. The results indicate that the algorithm classifies with great accuracy and provides a marked improvement over the previous three-electrode system.

Algorithms↗
PubMedDate not supplied

Comparison of linear and nonlinear classification algorithms for the prediction of drug and chemical metabolism by human UDP-glucuronosyltransferase isoforms.

Partial least squares discriminant analysis (PLSDA), Bayesian regularized artificial neural network (BRANN), and support vector machine (SVM) methodologies were compared by their ability to classify substrates and nonsubstrates of 12 isoforms of human UDP-glucuronosyltransferase (UGT), an enzyme "superfamily" involved in the metabolism of drugs, nondrug xenobiotics, and endogenous compounds. Simple two-dimensional descriptors were used to capture chemical information. For each data set, 70% of the data were used for training, and the remainder were used to assess the generalization performance. In general, the SVM methodology was able to produce models with the best predictive performance, followed by BRANN and then PLSDA. However, a small number of data sets showed either equivalent or better predictability using PLSDA, which may indicate relatively linear relationships in these data sets. All SVM models showed predictive ability (>60% of test set predicted correctly) and five out of the 12 test sets showed excellent prediction (>80% prediction accuracy). These models represent the first use of pattern recognition methods to discriminate between substrates and nonsubstrates of human drug metabolizing enzymes and the first thorough assessment of three classification algorithms using multiple metabolic data sets.

Algorithms↗
PubMed2001

Analysis of pain-related somatosensory evoked magnetic fields using the MUSIC (multiple signal classification) algorithm for magnetoencephalography.

We evaluated the effectiveness of the Multiple Signal Classification (MUSIC) algorithm by analysing pain-related somatosensory-evoked magnetic fields (SEFs) by 148-channel whole-head-type magnetoencephalography. MUSIC peaks of middle latency components were located around the primary somatosensory cortex (SI), contralateral to the stimulated finger. Long latency components were located around the bilateral secondary somatosensory cortices (SII) and cingulate gyri. Peaks at the SII and cingulate gyri were more prominent on very painful and moderately painful stimulation than on weak stimulation. The results were in very good agreement with results from single dipole estimation. These findings suggest that the MUSIC algorithm could be a useful tool for analysis of pain-related SEFs.

Adult↗
PubMed2006

An examination of the reliability of a classification algorithm for subgrouping patients with low back pain.

STUDY DESIGN: Test-retest design to examine interrater reliability. OBJECTIVE: Examine the interrater reliability of individual examination items and a classification decision-making algorithm using physical therapists with varying levels of experience. SUMMARY OF BACKGROUND DATA: Classifying patients based on clusters of examination findings has shown promise for improving outcomes. Examining the reliability of examination items and the classification decision-making algorithm may improve the reproducibility of classification methods. METHODS: Patients with low back pain less than 90 days in duration participating in a randomized trial were examined on separate days by different examiners. Interrater reliability of individual examination items important for classification was examined in clinically stable patients using kappa coefficients and intraclass correlation coefficients. The findings from the first examination were used to classify each patient using the decision-making algorithm by clinicians with varying amounts of experience. The reliability of the classification algorithm was examined with kappa coefficients. RESULTS: A total of 123 patients participated (mean age 37.7 [+/-10.7] years, 44% female), 60 (49%) remained stable between examinations. Reliability of range of motion, centralization/peripheralization judgments with flexion and extension, and the instability test were moderate to excellent. Reliability of centralization/peripheralization judgments with repeated or sustained extension or aberrant movement judgments were fair to poor. Overall agreement on classification decisions was 76% (kappa = 0.60, 95% confidence interval 0.56, 0.64), with no significant differences based on level of experience. CONCLUSION: Reliability of the classification algorithm was good. Further research is needed to identify sources of disagreements and improve reproducibility.

Adolescent↗
PubMed2006

Analysis of magnetic source localization of P300 using the multiple signal classification algorithm.

The authors studied the localization of P300 magnetic sources using the multiple signal classification (MUSIC) algorithm. Six healthy subjects (aged 24-34 years old) were investigated with 148-channel whole-head type magnetencephalography using an auditory oddball paradigm in passive mode. The authors also compared six stimulus combinations in order to find the optimal stimulus parameters for P300 magnetic field (P300m) in passive mode. Bilateral MUSIC peaks were located on the mesial temporal, superior temporal and parietal lobes. Interestingly, all MUSIC peaks in these regions emerged earlier in the right hemisphere than in the left hemisphere, suggesting that the right hemisphere has predominance over the left in the processing activity associated with P300m. There were no significant differences among the six stimulus combinations in evoking those P300m sources. The results of the present study suggest that the MUSIC algorithm could be a useful tool for analysis of the time-course of P300m.

Adult↗
PubMed2006

A comparison of wavelet, ridgelet, and curvelet-based texture classification algorithms in computed tomography.

The research presented in this article is aimed at the development of an automated imaging system for classification of normal tissues in medical images obtained from computed tomography (CT) scans. This article focuses on comparing the discriminating power of several multi-resolution texture analysis techniques using wavelet, ridgelet, and curvelet-based texture descriptors. The approach consists of two steps: automatic extraction of the most discriminative texture features of regions of interest and creation of a classifier that automatically identifies the various tissues. The algorithms are extensively tested and results are compared with standard texture classification algorithms. Tests indicate that using curvelet-based texture features significantly improves the classification of normal tissues in CT scans.

Algorithms↗
PubMed1996

Limitations of tachycardia confirmation and rate classification algorithms in a third-generation implantable cardioverter defibrillator.

Newer ICDs provide antitachycardia (ATP) and bradycardia pacing and cardioversion and defibrillation shocks based on sensed interval criteria. The objectives of this investigation were to determine the algorithm related errors in tachycardia confirmation and rate classification that occurred in patients with a third-generation, noncommitted, tiered ICD therapy. Forty-three consecutive patients with the Guardian ATP 4210 ICD, which uses an X out of Y sensed interval counting algorithm for tachycardia detection, confirmation, and classification were studied. Surface ECGs, intracardiac electrograms, stored data logs, and sense histories were reviewed to diagnose errors due to these algorithms that resulted in delivery of inappropriate therapy or inhibition of appropriate therapy. Sixty-eight classification or confirmation algorithm errors from 7,610 tachycardia detections (< 1%) were diagnosed in 23 (53%) of 43 patients. Three types of errors not related to device or sensing lead malfunction or programming mistakes were seen. In 26 episodes, the confirmation algorithm failed to detect late tachycardia reversion of nonsustained tachyarrhythmias, on the last or next to last sensed interval, and did not inhibit ATP (n = 17) or shocks (n = 9). In 28 episodes, inaccurate classification of tachycardia rate resulted in inappropriate ATP (n = 23) or shock (n = 5) therapy. In 14 episodes, the posttherapy reconformation algorithm produced inhibition of VVI pacing and prolonged asystole following shock therapy. These errors in tachycardia confirmation and rate classification were due to the inherent limitations of the X out of Y counting algorithm.

Adult↗
PubMed2001

A comparison of classification algorithms to automatically identify chest X-ray reports that support pneumonia.

We compared the performance of expert-crafted rules, a Bayesian network, and a decision tree at automatically identifying chest X-ray reports that support acute bacterial pneumonia. We randomly selected 292 chest X-ray reports, 75 (25%) of which were from patients with a hospital discharge diagnosis of bacterial pneumonia. The reports were encoded by our natural language processor and then manually corrected for mistakes. The encoded observations were analyzed by three expert systems to determine whether the reports supported pneumonia. The reference standard for radiologic support of pneumonia was the majority vote of three physicians. We compared (a) the performance of the expert systems against each other and (b) the performance of the expert systems against that of four physicians who were not part of the gold standard. Output from the expert systems and the physicians was transformed so that comparisons could be made with both binary and probabilistic output. Metrics of comparison for binary output were sensitivity (sens), precision (prec), and specificity (spec). The metric of comparison for probabilistic output was the area under the receiver operator characteristic (ROC) curve. We used McNemar's test to determine statistical significance for binary output and univariate z-tests for probabilistic output. Measures of performance of the expert systems for binary (probabilistic) output were as follows: Rules--sens, 0.92; prec, 0.80; spec, 0.86 (Az, 0.960); Bayesian network--sens, 0.90; prec, 0.72; spec, 0.78 (Az, 0.945); decision tree--sens, 0.86; prec, 0.85; spec, 0.91 (Az, 0.940). Comparisons of the expert systems against each other using binary output showed a significant difference between the rules and the Bayesian network and between the decision tree and the Bayesian network. Comparisons of expert systems using probabilistic output showed no significant differences. Comparisons of binary output against physicians showed differences between the Bayesian network and two physicians. Comparisons of probabilistic output against physicians showed a difference between the decision tree and one physician. The expert systems performed similarly for the probabilistic output but differed in measures of sensitivity, precision, and specificity produced by the binary output. All three expert systems performed similarly to physicians.

Acute Disease↗
PubMedDate not supplied

[Premature ejaculation: diagnosis, classification, algorithm of the patients' examination and results of their application].

We have analysed literature data and own experience with management of premature ejaculation (PE) in 3200 patients. This gave grounds for formulation of this phenomenon, creation of the classification and the examination algorithm. We applied the proposed methods in the treatment of 258 PE patients and made the conclusion about their adequacy and efficacy. PE is a polyetiological phenomenon and a multidisciplinary problem. Urological examination must be used as the first stage of the diagnostic process.

Algorithms↗
PubMed1994

An evaluation of computer assisted clinical classification algorithms.

The Mayo Clinic has a long tradition of indexing patient records in high resolution and volume. Several algorithms have been developed which promise to help human coders in the classification process. We evaluate variations on code browsers and free text indexing systems with respect to their speed and error rates in our production environment. The more sophisticated indexing systems save measurable time in the coding process, but suffer from incompleteness which requires a back-up system or human verification. Expert Network does the best job of rank ordering clinical text, potentially enabling the creation of thresholds for the pass through of computer coded data without human review.

Abstracting and Indexing↗
PubMed2004

Statistical evaluation of local alignment features predicting allergenicity using supervised classification algorithms.

BACKGROUND: Recently, two promising alignment-based features predicting food allergenicity using the k nearest neighbor (kNN) classifier were reported. These features are the alignment score and alignment length of the best local alignment obtained in a database of known allergen sequences. METHODS: In the work reported here a much more comprehensive statistical evaluation of the potential of these features was performed, this time for the prediction of allergenicity in general. The evaluation consisted of the following four key components. (1) A new high quality database consisting of 318 carefully selected, non-redundant allergens and 1,007 sequences carefully selected to be non-allergens. (2) Three different supervised algorithms: the kNN classifier, the Bayesian linear Gaussian classifier, and the Bayesian quadratic Gaussian classifier. (3) A large set of local alignment procedures defined using the FASTA3 alignment program by means of a wide range of different parameter settings. (4) Novel performance curves, alternative to conventional receiver-operating characteristic curves, to display not only average behaviors but also statistical variations due to small data sets. RESULTS: The linear Gaussian classifier proved most useful among the tested supervised machine learning algorithms, closely followed by the quadratic Gaussian equivalent and kNN. The overall best classification results were obtained with a novel feature vector consisting of the combined alignment scores derived from local alignment procedures using different substitution matrices. CONCLUSIONS: The models reported here should be useful as a part of an integrated assessment scheme for potential protein allergenicity and for future comparisons with alternative bioinformatic approaches.

Algorithms↗
PubMed2006

Binary tree of SVM: a new fast multiclass training and classification algorithm.

We present a new architecture named Binary Tree of support vector machine (SVM), or BTS, in order to achieve high classification efficiency for multiclass problems. BTS and its enhanced version, c-BTS, decrease the number of binary classifiers to the greatest extent without increasing the complexity of the original problem. In the training phase, BTS has N - 1 binary classifiers in the best situation (N is the number of classes), while it has log4/3 ((N + 3)/4) binary tests on average when making a decision. At the same time the upper bound of convergence complexity is determined. The experiments in this paper indicate that maintaining comparable accuracy, BTS is much faster to be trained than other methods. Especially in classification, due to its Log complexity, it is much faster than directed acyclic graph SVM (DAGSVM) and ECOC in problems that have big class number.

Algorithms↗
PubMed1995

MR image classification algorithms for neurosurgical and stereotactic radiosurgical trajectory planning and volumetric analysis.

Neurosurgical and stereotactic radiosurgical trajectory planning involves a phase of image data acquisition followed by localization of tissues to be avoided or crossed during the procedure. Success of the procedure is often assessed during follow-up by performing volumetric analysis of tissues of interest. Both localization and volumetric analysis can be facilitated by classifying the input data into different tissue types. This study compares classification results based on four different input image data: (1) MR data obtained with one TR and TE parameter combination, (2) MR data acquired with eight combinations of TR and TE parameters, (3) input data compressed into four principal components, and (4) the input data (3) together with first- and second-order texture features. The algorithms (a) to reduce the dimensionality using principal components, (b) to generate texture features, (c) to determine feature usefulness in classification using the Wilk's lambda criterion, and (d) to classify the input set are described. Classification results are found to be poor for input data 1 and much improved (qualitatively and quantatively) for input data 2. Input data 3 produces classification results (quantitatively and qualitatively) similar to that of 2. Input data 4 is found to produce quantitative results similar to those of 2 and 3, but the qualitative results show enhanced classification of some tissues and distortion of others.

Algorithms↗
PubMed2006

Multi-class classification algorithm for optical diagnosis of oral cancer.

We report development of a direct multi-class spectroscopic diagnostic algorithm for discrimination of high-grade cancerous tissue sites from low-grade as well as precancerous and normal squamous tissue sites of human oral cavity. The algorithm was developed making use of the recently formulated theory of total principal component regression (TPCR). The in vivo autofluorescence spectral data acquired from patients screened for neoplasm of oral cavity at the Government Cancer Hospital, Indore, was used to train and validate the algorithm. The diagnostic algorithm based on TPCR was found to provide satisfactory performance in classifying the tissue sites in four different classes - high-grade squamous cell carcinoma, low-grade squamous cell carcinoma, leukoplakia, and normal squamous tissue. The classification accuracy for these four classes was observed to be approximately 94%, 100%, 100% and 91% for the training data set (based on leave-one-out cross-validation), and was approximately 90%, 90%, 85% and 88%, respectively for the corresponding classes for the independent validation data set.

Adult↗
PubMedDate not supplied

Results of a new classification algorithm combining K nearest neighbors and recursive partitioning.

We present results of a new computational learning algorithm combining favorable elements of two well-known techniques: K nearest neighbors and recursive partitioning. Like K nearest neighbors, the method provides an independent prediction for each test sample under consideration, while like recursive partitioning, it incorporates an automatic selection of important input variables for model construction. The new method is applied to the problem of correctly classifying a set of chemical data samples designated as being either active or inactive in a biological screen. Training is performed at varying levels of intrinsic model complexity, and classification performance is compared to that of both K nearest neighbor and recursive partitioning models trained using the identical protocol. We find that the cross-validated performance of the new method outperforms both of these standard techniques over a considerable range of user parameters. We discuss advantages and drawbacks of the new method, with particular emphasis on its parameter robustness, required training time, and performance with respect to chemical structural class.

Journal Article↗
PubMed2003

Analysis of raw microneurographic recordings based on wavelet de-noising technique and classification algorithm: wavelet analysis in microneurography.

We propose a new technique for analyzing the raw neurogram which enables the study of the discharge behavior of individual and group neurons. It utilizes an ideal bandpass filter, a modified wavelet de-noising procedure, an action potential detector, and a waveform classifier. We validated our approach with both simulated data generated from muscle sympathetic neurograms sampled at high rates in five healthy subjects and data recorded from seven healthy subjects during lower body negative pressure suction. The modified wavelet method was superior to the classical discriminator method and the regular wavelet de-noising procedure when applied to simulated neuronal signals. The detected spike rate and spike amplitude rate of the action potentials correlated strongly with number of bursts detected in the integrated neurogram (r = 0.79 and 0.89, respectively, p < 0.001). Eight major action potential waveform classes were found to describe more than 81% of all detected action potentials in all subjects. One class had characteristics similar in shape and in average discharge frequency (27.4 +/- 5.1 spikes/min during resting supine position) to those of reported single vasoconstrictor units. The newly proposed technique allows a precise estimate of sympathetic nerve activity and characterization of individual action potentials in multiunit records.

Action Potentials↗
PubMed2004

Application of the random forest classification algorithm to a SELDI-TOF proteomics study in the setting of a cancer prevention trial.

A thorough discussion of the random forest (RF) algorithm as it relates to a SELDI-TOF proteomics study is presented, with special emphasis on its application for cancer prevention: specifically, what makes it an efficient, yet reliable classifier, and what makes it optimal among the many available approaches. The main body of the paper treats the particulars of how to successfully apply the RF algorithm in a proteomics profiling study to construct a classifier and discover peak intensities most likely responsible for the separation between the classes.

Algorithms↗