Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “classifier”

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 1,243 records · Page 69Linked to original sources
PubMed2005

Study of probabilistic neural networks to classify the active compounds in medicinal plants.

Probabilistic neural networks (PNNs) were utilized for the classifications of 102 active compounds from diverse medicinal plants with anticancer activity against human rhinopharyngocele cell line KB. Molecular descriptors calculated from structure alone were used to represent molecular structures. A subset of the calculated descriptors selected using factor correlation analysis and forward stepwise regression was used to construct the prediction models. Linear discriminant analysis (LDA) was also utilized to construct the classification model to compare the results with those obtained by PNNs. The accuracy of the training set, the cross-validation set, and the test set given by PNNs and LDA were 100, 92.3, 90.9% and 71.8, 92.3, 54.5%, respectively, which indicated that the results obtained by PNNs agree well with the experimental values of these compounds and also revealed the superiority of PNNs over LDA approach for the classification of anticancer activities of compounds. The models built in this work would be of potential help in the design of novel and more potent anticancer agents.

Algorithms↗
PubMed2006

Augmentation of near infrared diffuse reflectance and transmittance spectral data for the development of robust PLSBC models for classifying double blind clinical trial tablets.

The water content of clinical trial tablets can be different between and within different tablet batches, depending on the relative humidity conditions during their production, packaging, storage and analysis. These water variations lead to important spectral variations in the near infrared spectral region which can lead to a wrong identification if the classification model was based on unrepresentative data towards the water content. As model development for clinical trial studies needs to be extremely fast - within one working day - with generally only one batch available, the principle of data augmentation has to be applied to render more robust classification models. Therefore, tablets available for constructing the model are being processed in order to increase or decrease their water content and to make them more representative for tablets to be tested in the future. The inclusion of a deliberate water variation is the most efficient way to develop a model, for which no additional model redevelopment will be required to pass the system suitability tests and to obtain a correct identification.

Chemistry, Pharmaceutical↗
PubMed2006

Pathological features of renal neoplasms classified by size and symptomatology.

PURPOSE: We examined the relationship between tumor size and pathological findings in a contemporary series of surgical renal lesions and we characterized the relationship of incidental and symptomatic tumors to pathological findings. MATERIALS AND METHODS: We retrospectively reviewed the records of patients treated surgically for renal lesions suspicious for malignancy between March 2000 and May 2005. Specimens were examined for a gross and microscopic description. Statistical analysis was used to determine the correlation of size and symptomatology. RESULTS: A total of 349 renal masses from 331 patients were identified. Of the 349 renal masses 56 (16.0%) were benign, 289 (82.8%) were renal cell carcinoma and 4 (1.1%) were other malignancies. The percent of malignant tumors increased from 72.1% for those less than 2 cm to 93.7% for those greater than 7 cm (OR 1.39, 95% CI 1.17 to 1.65). Of the 349 renal masses 258 (73.9%) were discovered incidentally and 91 (26.1%) were symptomatic. Mean size of incidental and symptomatic tumors was 3.7 and 6.2 cm, respectively (p < 0.001). When comparing T1 incidental and symptomatic tumors, there was no significant difference in the overall frequency of malignancy. When comparing T2 incidental and symptomatic tumors, the groups had similar malignancy rates (90.9% and 100%, respectively, p = 0.16). However, symptomatic lesions showed an increased incidence of high grade malignancy (78.4% vs 40.9%, p = 0.012). CONCLUSIONS: Smaller renal tumors are more likely to be benign or be a lower grade of malignancy. T1 renal tumors are more likely to be detected incidentally than T2 tumors. When T1 incidental and symptomatic tumors were compared, there was no difference between the malignancy rates. However, when T2 incidental and symptomatic tumors were compared, symptomatic tumors were more likely to be high grade malignancy.

Adenoma, Oxyphilic↗
PubMed2006

Support vector classifiers via gradient systems with discontinuous righthand sides.

Gradient dynamical systems with discontinuous righthand sides are designed using Persidskii-type nonsmooth Lyapunov functions to work as support vector machines (SVMs) for the discrimination of nonseparable classes. The gradient systems are obtained from an exact penalty method applied to the constrained quadratic optimization problems, which are formulations of two well known SVMs. Global convergence of the trajectories of the gradient dynamical systems to the solution of the corresponding constrained problems is shown to be independent of the penalty parameters and of the parameters of the SVMs. The proposed gradient systems can be implemented as simple analog circuits as well as using standard software for integration of ODEs, and in order to use efficient integration methods with adaptive stepsize selection, the discontinuous terms are smoothed around a neighborhood of the discontinuity surface by means of the boundary layer technique. The scalability of the proposed gradient systems is also shown by means of an implementation using parallel computers, resulting in smaller processing times when compared with traditional SVM packages.

Algorithms↗
PubMed2005

Classifying spatial patterns of brain activity with machine learning methods: application to lie detection.

Patterns of brain activity during deception have recently been characterized with fMRI on the multi-subject average group level. The clinical value of fMRI in lie detection will be determined by the ability to detect deception in individual subjects, rather than group averages. High-dimensional non-linear pattern classification methods applied to functional magnetic resonance (fMRI) images were used to discriminate between the spatial patterns of brain activity associated with lie and truth. In 22 participants performing a forced-choice deception task, 99% of the true and false responses were discriminated correctly. Predictive accuracy, assessed by cross-validation in participants not included in training, was 88%. The results demonstrate the potential of non-linear machine learning techniques in lie detection and other possible clinical applications of fMRI in individual subjects, and indicate that accurate clinical tests could be based on measurements of brain function with fMRI.

Adult↗
PubMed2006

Classification of fMRI independent components using IC-fingerprints and support vector machine classifiers.

We present a general method for the classification of independent components (ICs) extracted from functional MRI (fMRI) data sets. The method consists of two steps. In the first step, each fMRI-IC is associated with an IC-fingerprint, i.e., a representation of the component in a multidimensional space of parameters. These parameters are post hoc estimates of global properties of the ICs and are largely independent of a specific experimental design and stimulus timing. In the second step a machine learning algorithm automatically separates the IC-fingerprints into six general classes after preliminary training performed on a small subset of expert-labeled components. We illustrate this approach in a multisubject fMRI study employing visual structure-from-motion stimuli encoding faces and control random shapes. We show that: (1) IC-fingerprints are a valuable tool for the inspection, characterization and selection of fMRI-ICs and (2) automatic classifications of fMRI-ICs in new subjects present a high correspondence with those obtained by expert visual inspection of the components. Importantly, our classification procedure highlights several neurophysiologically interesting processes. The most intriguing of which is reflected, with high intra- and inter-subject reproducibility, in one IC exhibiting a transiently task-related activation in the 'face' region of the primary sensorimotor cortex. This suggests that in addition to or as part of the mirror system, somatotopic regions of the sensorimotor cortex are involved in disambiguating the perception of a moving body part. Finally, we show that the same classification algorithm can be successfully applied, without re-training, to fMRI collected using acquisition parameters, stimulation modality and timing considerably different from those used for training.

Algorithms↗
PubMed2005

Improved attainment of NKF classified lipid target levels after conversion from cyclosporine to tacrolimus in renal graft recipients.

In an open, prospective, multicenter study, stable renal graft recipients were converted to tacrolimus because of cyclosporine-related side effects. Seventy-five patients were switched primarily because of hyperlipidemia. After the switch to tacrolimus, mean total cholesterol was reduced by 15% at month 6. One hundred seventy-seven additional patients were switched primarily for other indications: hypertrichosis, gingival hyperplasia, and arterial hypertension, and these symptoms also improved after the switch. In this analysis, serum lipid levels were categorized according to a modified standard classification of lipid parameters for renal transplant patients (published by the NKF Work Group). The aim was to estimate the proportion of patients reaching normal lipid levels after the conversion to tacrolimus therapy. In patients with primary indication hyperlipidemia, the proportion with normal cholesterol levels increased significantly from 5.6% at baseline to 37.5% at month 6 (P < .05). For LDL cholesterol, the increase was from 54.1% at baseline to 64.9% at month 6, and for triglycerides the improvement was from 25.4% to 33.8%. HDL cholesterol levels remained stable. Similar changes of lipid parameters were also observed in the subgroups of patients converted to tacrolimus primarily because of other indications. After conversion from cyclosporine to tacrolimus, a significantly higher proportion of stable renal graft recipients reached normal total cholesterol levels. For LDL cholesterol and triglycerides, a trend for normalization was observed. Thus, the improvement of serum lipid levels resulted for many patients in a change to a better level class and improved or normalized their cardiovascular risk parameters.

Cholesterol, LDL↗
PubMed2004

Ecdysozoan phylogeny and Bayesian inference: first use of nearly complete 28S and 18S rRNA gene sequences to classify the arthropods and their kin.

Relationships among the ecdysozoans, or molting animals, have been difficult to resolve. Here, we use nearly complete 28S+18S ribosomal RNA gene sequences to estimate the relations of 35 ecdysozoan taxa, including newly obtained 28S sequences from 25 of these. The tree-building algorithms were likelihood-based Bayesian inference and minimum-evolution analysis of LogDet-transformed distances, and hypotheses were tested wth parametric bootstrapping. Better taxonomic resolution and recovery of established taxa were obtained here, especially with Bayesian inference, than in previous parsimony-based studies that used 18S rRNA sequences (or 18S plus small parts of 28S). In our gene trees, priapulan worms represent the basal ecdysozoans, followed by nematomorphs, or nematomorphs plus nematodes, followed by Panarthropoda. Panarthropoda was monophyletic with high support, although the relationships among its three phyla (arthropods, onychophorans, tardigrades) remain uncertain. The four groups of arthropods-hexapods (insects and related forms), crustaceans, chelicerates (spiders, scorpions, horseshoe crabs), and myriapods (centipedes, millipedes, and relatives)-formed two well-supported clades: Hexapoda in a paraphyletic crustacea (Pancrustacea), and 'Chelicerata+Myriapoda' (a clade that we name 'Paradoxopoda'). Pycnogonids (sea spiders) were either chelicerates or part of the 'chelicerate+myriapod' clade, but not basal arthropods. Certain clades derived from morphological taxonomy, such as Mandibulata, Atelocerata, Schizoramia, Maxillopoda and Cycloneuralia, are inconsistent with these rRNA data. The 28S gene contained more signal than the 18S gene, and contributed to the improved phylogenetic resolution. Our findings are similar to those obtained from mitochondrial and nuclear (e.g., elongation factor, RNA polymerase, Hox) protein-encoding genes, and should revive interest in using rRNA genes to study arthropod and ecdysozoan relationships.

Animals↗
PubMed2001

The Age-Related Eye Disease Study system for classifying age-related macular degeneration from stereoscopic color fundus photographs: the Age-Related Eye Disease Study Report Number 6.

PURPOSE: To describe the system for grading age-related macular degeneration from fundus photographs in the Age-Related Eye Disease Study. This is a prospective multicenter cohort study of the course of age-related macular degeneration and a placebo-controlled clinical trial of the effect of high-dose vitamin and mineral supplements on development of advanced age-related macular degeneration. METHODS: Standardized stereoscopic 30-degree color fundus photographs are taken at 11 clinical centers and evaluated by graders at a reading center for quality and age-related macular degeneration abnormalities. Macular subfields are defined by a grid, and presence and severity of various age-related macular degeneration abnormalities are graded using standard/example photographs and measuring circles. Advanced age-related macular degeneration abnormalities (presence/absence) include retinal pigment epithelial detachment, serous (or hemorrhagic) sensory retinal detachment, hard exudates, subretinal/subretinal pigment epithelial hemorrhage, subretinal fibrous tissue, and photocoagulation scars. Other abnormalities (multistep scales) include retinal pigment epithelial abnormalities (geographic atrophy, depigmentation, and increased pigment), and drusen (size, type [hard versus soft distinct or indistinct], and total area). Contemporaneous variability is tested by having different graders evaluate samples of photographs, and temporal variability by annual regrading of a dedicated subset. RESULTS: In a cumulative sample of 1230 eyes for contemporaneous reproducibility, agreement on advanced age-related macular degeneration was 96% (kappa = 0.88) and for four-step age-related macular degeneration level was 83% (kappa = 0.77). Agreement was moderate for pigment epithelial detachment (kappa = 0.54) and substantial for serous sensory retinal detachment, hard exudates, subretinal/subretinal pigment epithelial hemorrhage, subretinal fibrous tissue, and central geographic atrophy (kappa = 0.73-0.82). For pigment abnormalities, agreement was substantial for geographic atrophy (kappa = 0.63; kappa(weighted) = 0.71, 0.75 weight for one-step disagreements), and moderate for depigmentation (kappa = 0.41, kappa(weighted) = 0.51) and increased pigment (kappa = 0.54, kappa(weighted) = 0.71). For drusen, agreement was moderate to substantial for presence/maximum size (kappa = 0.50, kappa(weighted) = 0.68), type (kappa = 0.61, kappa(weighted) = 0.69), and area (kappa = 0.56, kappa(weighted) = 0.77). Six annual regrades of 119 eyes showed little temporal drift in grading of various age-related macular degeneration abnormalities, except for drusen type. CONCLUSIONS: The Age-Related Eye Disease Study has demonstrated satisfactory reliability for detecting onset of advanced age-related macular degeneration in a cohort, and moderate to substantial agreement on various abnormalities across the age-related macular degeneration spectrum. The Age-Related Eye Disease Study system for classification of age-related macular degeneration is suitable for longitudinal multicenter studies.

Cohort Studies↗
PubMed1999

How should interlobar pleural invasion be classified? Prognosis of resected T3 non-small cell lung cancer.

BACKGROUND: The results of surgical treatment for non-small cell lung cancer with interlobar pleural involvement and direct invasion of the other lobe have seldom been documented. METHODS: Of 1,130 consecutive patients who were operated on for primary bronchogenic carcinoma between 1984 and 1997, we studied 132 patients who had complete resection of T3 non-small cell carcinoma. RESULTS: The structures involved were as follows: parietal pleura, 49 patients; chest wall, 45; interlobar pleura, 19; main bronchus within 2 cm of the carina, 11; mediastinal pleura, 6; and diaphragm, 1. Patients with N2 disease had a significantly worse survival than those with N0 (p = 0.0054) and N1 disease (p = 0.0165). The survival of patients with involvement of the interlobar pleura was significantly worse than that of patients with T1 (p = 0.0001) or T2 disease (p = 0.0484), and was similar to that of patients with T3 disease (p = 0.9821). CONCLUSIONS: In patients with T3 disease, mediastinal lymph node involvement influenced survival significantly. Patients with involvement of the interlobar pleura should be regarded as having T3 lesions.

Adult↗