Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Classification Algorithms”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 325 records · Page 18Linked to original sources
PubMed2005

[Proteomic analysis of prostate cancer using surface enhanced laser desorption/ionization mass spectrometry].

OBJECTIVE: To identify the serum biomarkers of prostate cancer by using protein chip and bioinformatics. METHODS: Eighty three prostate cancer (PCA) patients and ninety five healthy people from mass screen in Changchun were detected by surface-enhanced laser desorption/ionization mass spectrometry (SELDI-MS). The data of spectra were analyzed by bioinformatics tools-Biomarker Wizard and Biomarker Pattern. RESULTS: Compared with the spectra of healthy people, there were 18 potential markers detected in the spectra of the PCA patients, the protein expression was high in 4 of which and low in the 10 of which. The softwares Biomarkerwizard and Biomarker Pattern automatically, under given conditions, selected 8 biomarker proteins to be used to establish a five layer decision tree differentiate to diagnose PCA and differentiate PCA from healthy people with a specificity of 92.632% and a sensitivity of 96.386%. CONCLUSION: New serum biomarkers of PCA have been identified, and this SELDI mass spectrometry coupled with decision tree classification algorithm will provide a highly accurate and innovative approach for the early diagnosis of PCA.

Aged↗
PubMed2006

The state-of-the-art in predictive toxicogenomics.

Predictive toxicogenomics, ie, the acquisition of advanced knowledge of the safety profile of a compound using genomic biomarkers, is a technology that provides much optimism for improving early drug discovery decisions. Toxicogenomics creates an opportunity to shift attrition to earlier stages in drug development to a point where course-corrective action can be taken with relatively lower financial costs, thus improving the efficiency of the drug development process. This review will survey the current state-of-the-art in toxicogenomics for predicting toxicity, both in vivo and in vitro, with emphasis on the use of classification algorithms and the importance of toxicogenomic databases for biomarker discovery and validation.

Algorithms↗
PubMed2006

Plasma proteomic profiling: search for lung cancer diagnostic and early detection markers.

Environmental and occupational exposure to asbestos is among the established risk factors for lung cancer, the leading cause of cancer-related deaths in the United States. This link between exposure to asbestos and the excessive death rate from lung cancer was evident in a study of former workers of an asbestos pipe insulation manufacturing plant in Tyler, TX. We performed comparative proteomic profiling of plasma samples that were collected from nine patients within 12 months before death and their age-, race- and exposure-matched disease-free controls on strong anion exchange chips using surface-enhanced laser desorption ionization time-of-flight mass spectrometry. A distance-dependent K-nearest neighbor (KNN) classification algorithm identified spectral features of m/z values 7558.9 and 15103.0 that were able to distinguish lung cancer patients from disease-free individuals with high sensitivity and specificity. The high correlation between the intensities of these two peaks (r=0.987) strongly suggests that they are the doubly and singly charged ions of the same protein product. Examination of these proteomic markers in the plasma samples of subjects from >5 years before death from lung cancer suggested that they are related to the early development of lung cancer. Validation of these biomarkers would have significant implications for the early detection of lung cancer and better management of high-risk patients.

Adult↗
PubMed2006

[Applying patial least-squares discriminant analysis on autofluorescence spectra to identify gastric cancer].

Measurement of fluorescence intensity was performed at excitation wavelength of 308 nm and emission wavelength in the range of 328-596 nm. The partial least-squares (PLS) method was used to analyze autofluorescence spectra of gastric cancer. The 42 normal samples and 42 cancer samples were taken from 42 gastric cancer patients. The normalized and centralized spectra of two kinds of samples showed similar but divergent patterns. PLS classification algorithm could differentiate cancer tissues from normal tissues with a sensitivity of 83.3%, a specificity of 95.2%, and a positive predictive value of 94.6%. We concluded therefore that the PLS method was a fast, effective choice for identification of gastric cancer.

Adult↗
PubMed1992

Automated prescreening of conventionally prepared cervical smears: a feasibility study.

The feasibility of an automated screening device that will analyze conventionally prepared cervical smears was examined. Consecutive normal and abnormal cervical smears were selected retrospectively and analyzed on a customized microscope imaging workstation system. Specialized image processing, feature extraction, and object classification algorithms were integrated on the imaging workstation to provide an Analysis Score for each slide analyzed. A threshold was set for the Analysis Score, and all slides with a score less than the threshold were classified as normal. Examination of the distribution of the Analysis Score for normal and abnormal cervical smears provided an estimate that 50% of all slides will be sorted as normal, and consequently need no cytologist review. With the Analysis Score threshold set for a 50% sort rate, zero false-negative slides were found in this slide set. This study has shown that a large percentage of slides can be correctly classified as normal while maintaining a high sensitivity to abnormal slides. Given the current workload problems in cytology laboratories, reducing the screening workload by one half will be beneficial. This study has shown that it is now feasible with current technology to provide a clinically useful device to automate screening of conventional cervical smears.

Algorithms↗
PubMed1990

[Cutaneous manifestations of tuberculosis].

Tuberculous lesions of the skin occur rarely nowadays. We therefore have mostly lost the knowledge to recognize this entity. Furthermore, the great variability of the clinical and particularly dermatologic manifestations of the disease can tax the most astute clinician. The challenge is even greater, when the patient has an intercurrent condition such as a malignancy or AIDS. In order to make understanding easier, we replace all descriptive terms, some carried over from the last century, by a useful pathogenetic and clinical, algorithmic classification. Finally the necessity of biopsy and specific cultures for proper diagnosis and treatment of any skin lesion consistent with skin tuberculosis is emphasized.

Antitubercular Agents↗
PubMed1985

[Quantitative technics in ultrasonic diagnosis].

This paper presents techniques for computer-aided analysis of medical ultrasonic images. Methods of image processing, feature extraction and a classification algorithm for tissue characterization are described. Based on these techniques a pattern recognition system for semi-automatic analysis of ultrasonic images has been developed and proved within three clinical studies, which aimed to improve the ultrasound diagnosis of the prostate, the thyroid and the female breast. The results show that computer analysis of ultrasonic images is capable of providing a quantitative diagnosis as well as of differentiating tumor lesions by sonography.

Breast Neoplasms↗
PubMed1987

[Problem-specific grading in disease course assessment and calculation of the likelihood of transition of chronic cardiovascular diseases].

For the course rating of cardiovascular caused drop in efficiency a problem-specific grading is introduced, the characterization of which is composed of simple, non-invasive parameters. The grading is specified for 4 clinical diagnostic classes (hypertension, coronary heart diseases, pulmonary heart, "other heart and vessel diseases") and takes into consideration the frequent coincidence or connection of several diagnoses of heart and vessel diseases with each other. The choice of parameter and connection have to fulfil the demand, which is essential for the course rating of heart and vessel diseases, like practicability proportionate with the frequency of heart and vessel diseases, adequate registration of changes in severe degrees, statement about the patient and not about several parameters. Furthermore it has to make it possible to calculate provisional probabilities under variable conditions as well as to make possible a valuation of prognosis. For that purpose the construction of suitable matrixes is shown and an integral projection is represented, which is used for the interpretation of courses dependent on severe degrees. The conception of grading has the function to modify necessary parameters for inclusion in the hierarchy according to medical starting position (for instance primary mode of registration). The model which is applicable to level of population makes possible an electronic data - processing - based evaluation including algorithmic classification.

Cardiovascular Diseases↗
PubMed1993

New developments in speech pattern element hearing aids for the profoundly deaf.

Two new developments in speech pattern processing hearing aids will be described. The first development is the use of compound speech pattern coding. Speech information which is invisible to the lipreader was encoded in terms of three acoustic speech factors; the voice fundamental frequency pattern, coded as a sinusoid, the presence of aperiodic excitation, coded as a low-frequency noise, and the wide-band amplitude envelope, coded by amplitude modulation of the sinusoid and noise signals. Each element of the compound stimulus was individually matched in frequency and intensity to the listener's receptive range. Audio-visual speech receptive assessments in five profoundly hearing-impaired listeners were performed to examine the contributions of adding voiceless and amplitude information to the voice fundamental frequency pattern, and to compare these codings to amplified speech. In both consonant recognition and connected discourse tracking (CDT), all five subjects showed an advantage from the addition of amplitude information to the fundamental frequency pattern. In consonant identification, all five subjects showed further improvements in performance when voiceless speech excitation was additionally encoded together with amplitude information, but this effect was not found in CDT. The addition of voiceless information to voice fundamental frequency information did not improve performance in the absence of amplitude information. Three of the subjects performed significantly better in at least one of the compound speech pattern conditions than with amplified speech, while the other two performed similarly with amplified speech and the best compound speech pattern condition. The three speech pattern elements encoded here may represent a near-optimal basis for an acoustic aid to lipreading for this group of listeners. The second development is the use of a trained multi-layer-perceptron (MLP) pattern classification algorithm as the basis for a robust real-time voice fundamental frequency extractor. This algorithm runs on a low-power digital signal processor which can be incorporated in a wearable hearing aid. Aided lipreading for speech in noise was assessed in the same five profoundly hearing-impaired listeners to compare the benefits of conventional hearing aids with those of an aid which provided MLP-based fundamental frequency information together with speech+noise amplitude information. The MLP-based pattern element aid gave significantly better performance in the reception of consonantal voicing contrasts from speech in pink noise than that achieved with conventional amplification and consequently, it also gave better overall performance in audio-visual consonant identification.(ABSTRACT TRUNCATED AT 400 WORDS)

Algorithms↗
PubMed1997

Neural net-based identification of cells expressing the p300 tumor-related antigen using fluorescence image analysis.

We report on preliminary investigations of the use of an image analysis system to perform preliminary algorithmic classification of images of fluorochrome-labeled cells followed by capture of gray-level images of potentially abnormal cells for analysis by a neural network. Cells were labeled with an antibody against a bladder cancer tumor-associated antigen, and the neural net was used to distinguish true-positive cells from negative cells, false-positive cells (autofluorescent or nonspecific labeling), and cell-sized artifacts. Gray-level cell images were digitized and processed for analysis by a feed-forward neural network using back-propagation. The network was trained and tested with two independent image sets. Various network configurations and activation functions were investigated, including a sinusoidal activation function. At high power, the network agreed completely with the human observer's classification. At low power, a strong clustering of cells classified by the network with expert classification was seen, while the neural network showed roughly 75% concordance with the human observer. In addition, a set of four features extracted from raw cell images were investigated. The features were: shape factor, texture, area, and average pixel intensity. A network trained with these features performed better than one operating with gray-level images. We conclude that using neural networks to recognize and classify images captured by an image analysis microscope is feasible.

Antigens, Neoplasm↗
PubMedDate not supplied

Critique of arrhythmia detectors based on heuristic rules.

Every arrhythmia detector employs a beat classifier to discriminate between normal (N) and ventricular (V) beats. In most of these beat-classification algorithms, a set of rules is employed to distinguish between N and V beats using a common set of features extracted from the real-time ECG signal and/or correlation of QRS complexes with the dominant QRS template. A common set of these features includes: beat area, beat width, beat amplitude, beat polarity, and R-to-R interval. Heuristic methods are commonly used to adapt the rules to particular databases. These classifiers are rule-based classifiers that employ AND-OR binary structures and hand-tuned thresholds for making decisions in the feature space. The complexity of the feature space increases as the number of features increases. For k features, a k-dimensional space is required. Thus, the separation between N and V space distributions becomes more difficult, especially since these distributions overlap. When AND-OR binary structures with hand-tuned thresholds or linear-separation techniques are used to separate N and V distributions in a k-dimensional feature space, errors are guaranteed, because these distributions are not linearly separable. As a results, these algorithms have limited dynamic ranges. This means that the sensitivity for a certain class of beats (N or V) will grow only at the expense of positive predictivity for that class, and vice versa.

Algorithms↗
PubMed1998

The analysis of exfoliated cervical cells by infrared microscopy.

The screening technique developed by Dr. Papanicolau for cervical cancer has significantly increased opportunities for early detection and treatment of this disease. Recently there has been a great deal of concern related to the accuracy of the Pap smear screening technique for cervical cancer and the need for highly skilled technologists to reliably examine the prepared samples. The use of infrared spectroscopy as a technique to discriminate between normal and malignant cervical cell samples has been reported by several research groups. Samples of cervical cells can be prepared for spectroscopic diagnosis by centrifuging cells obtained by the normal Pap smear technique and applying them to an infrared transparent window. A major concern with diagnosis by infrared spectroscopy is the homogeneity of the sample and whether the spectral features used in the analysis are caused by localized groups of cells or can be attributed to the majority of the cells in the sample. In this paper we report on the use of automated infrared microscopic mapping techniques to measure the infrared spectra at fixed intervals across the sample covering a 5 mm spot. Various post processing techniques have been applied to the spectral results to create images revealing specific features of the sample. The techniques applied to the data include: 1) Baseline corrected peak height; 2) Band ratioing to compensate for thickness; 3) Correlation metrics; 4) Similarity matching, and 5) Mahalanobis distance classification algorithm.

Algorithms↗
PubMedDate not supplied

Source contribution of lead in house dust from a lead mining waste superfund site.

The relative contribution to household dust of lead particles from a mining waste superfund site and lead-based paint is investigated. Automated individual particle analysis (IPA) based on scanning electron microscopy (SEM) and X-ray energy spectroscopy (EDX) is used to develop a classification algorithm for determining lead particle source contribution in household dust vacuum bags. On a volume basis the proportion derived from the mining waste is found to be 26%, the proportion derived from a paint source is 16%, and the proportion from soil is 37%. In 15% of the lead particles identified a specific originating source could not be determined. Using a weighting method accounting for the lead concentration per particle rather than volume the contributions were similar for mining waste and paint, 21% and 23%, respectively, but the soil contribution was reduced to 8%, and the source for 29% of the lead could not be identified. These results suggested that the contribution of waste piles to the lead present in household dust is at least as important a source as paint. There is evidence to suggest that a large percentage of lead in the soil also originated from the waste piles and the overall contribution, therefore, of the waste piles may be greater than the contribution from paint.

Air Pollution, Indoor↗
PubMed1980

An algorithm for the classification of untoward events in large scale clinical trials.

The algorithm described in this paper derives from a proposal by Karch and Lasagna and was developed for analysis of events reported in large scale clinical trials. It is presented in the form of a decision table which allows classification of events on a five-point scale according to the probability of their relationship with the drug under study. The degree of reproducibility of judgements made with the algorithm has been assessed and the decision table has been utilized for classifying all untoward events reported during an extended clinical trial with an anti-inflammatory drug.

Clinical Trials as Topic↗
PubMed2004

A double classification tree search algorithm for index SNP selection.

BACKGROUND: In population-based studies, it is generally recognized that single nucleotide polymorphism (SNP) markers are not independent. Rather, they are carried by haplotypes, groups of SNPs that tend to be coinherited. It is thus possible to choose a much smaller number of SNPs to use as indices for identifying haplotypes or haplotype blocks in genetic association studies. We refer to these characteristic SNPs as index SNPs. In order to reduce costs and work, a minimum number of index SNPs that can distinguish all SNP and haplotype patterns should be chosen. Unfortunately, this is an NP-complete problem, requiring brute force algorithms that are not feasible for large data sets. RESULTS: We have developed a double classification tree search algorithm to generate index SNPs that can distinguish all SNP and haplotype patterns. This algorithm runs very rapidly and generates very good, though not necessarily minimum, sets of index SNPs, as is to be expected for such NP-complete problems. CONCLUSIONS: A new algorithm for index SNP selection has been developed. A webserver for index SNP selection is available at http://cognia.cu-genome.org/cgi-bin/genome/snpIndex.cgi/

Algorithms↗
PubMed2000

Gaussian processes for classification: mean-field algorithms.

We derive a mean-field algorithm for binary classification with gaussian processes that is based on the TAP approach originally proposed in statistical physics of disordered systems. The theory also yields an approximate leave-one-out estimator for the generalization error, which is computed with no extra computational cost. We show that from the TAP approach, it is possible to derive both a simpler "naive" mean-field theory and support vector machines (SVMs) as limiting cases. For both mean-field algorithms and support vector machines, simulation results for three small benchmark data sets are presented. They show that one may get state-of-the-art performance by using the leave-one-out estimator for model selection and the built-in leave-one-out estimators are extremely precise when compared to the exact leave-one-out estimate. The second result is taken as strong support for the internal consistency of the mean-field approach.

Algorithms↗
PubMed2006

The classification and treatment algorithm for post-burn cervical contractures reconstructed with free flaps.

Neck contractures after burn produce restrictions in motion and unacceptable aesthetic outcomes. Proper planning and tissue selection is essential to minimize donor site morbidity while optimizing outcomes. A classification system and treatment algorithm aids in achieving this goal. Between December 1999 and January 2003, 40 burn patients underwent release and reconstruction with free perforator flaps. Neck extensibility and zone of injury were evaluated. Choice of reconstruction was based on available tissue, restriction degree and zones involved. Cervical territories were classified according to movement restrictions and amount of improvement. Reconstructive territories were classified as central above (CA), central below (CB), central above and below (CAB) and lateral (L). Single, split, double and preexpanded free flaps were used for the reconstructions. Maximal gain in motion was noted at 4 weeks and maintained for the average 11 months follow-up. Types of reconstructive territories showed significant effects on range of motion while etiology and time between injury and reconstruction showed no impact on the functional outcome. Classification of neck territories aids in improving outcomes while minimizing donor morbidity. The central above territory, when reconstructed with free flaps, yielded the most rewarding improvement. A classification and treatment algorithm aids in achieving significant improvements in range of neck motion while taking into consideration the donor sites.

Adolescent↗
PubMed2006

Algorithm for the classification of multi-modulating signals on the electrocardiogram.

This article discusses the algorithm to measure electrocardiogram (ECG) and respiration simultaneously and to have the diagnostic potentiality for sleep apnoea from ECG recordings. The algorithm is composed by the combination with the three particular scale transform of a(j)(t), u(j)(t), o(j)(a(j)) and the statistical Fourier transform (SFT). Time and magnitude scale transforms of a(j)(t), u(j)(t) change the source into the periodic signal and tau(j) = o(j)(a(j)) confines its harmonics into a few instantaneous components at tau(j) being a common instant on two scales between t and tau(j). As a result, the multi-modulating source is decomposed by the SFT and is reconstructed into ECG, respiration and the other signals by inverse transform. The algorithm is expected to get the partial ventilation and the heart rate variability from scale transforms among a(j)(t), a(j+1)(t) and u(j+1)(t) joining with each modulation. The algorithm has a high potentiality of the clinical checkup for the diagnosis of sleep apnoea from ECG recordings.

Algorithms↗