PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Principal Component Analysis”

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 451 records · Page 25Linked to original sources
PubMed2001

Evaluation of principal components analysis with high-performance liquid chromatography and photodiode array detection for the forensic differentiation of ballpoint pen inks.

Inks from seven black and eight blue ballpoint pens were separated by a high-performance liquid chromatography (HPLC) method utilizing a photodiode array detection (PDA). A classifier flowchart was designed for the chromatographic data based on the presence or absence of certain peaks at different wavelengths to qualitatively discriminate between the inks. The same data were quantitatively classified by principal components analysis (PCA) to estimate the separation between a pair of classes of ink samples. It was found that the black ballpoint pen inks were discriminated satisfactorily utilizing two-dimensional data of the peak areas and retention times at the optimum wavelengths. The blue pens were discriminated by analyzing the chromatographic data at four different wavelengths simultaneously with a cross-validated PCA. The results of this study indicated that HPLC-PDA coupled with chemometrics could make a powerful discriminating tool for the forensic chemist, especially when analyzing extensive and/or complex data.

Journal Article↗
PubMedDate not supplied

Application of principal components analysis to 1H-NMR data obtained from propolis samples of different geographical origin.

Propolis is a widely used natural remedy and a range of biological activities have been attributed to it. The chemical composition of propolis is highly variable and its quality is often controlled on the basis of one or two marker compounds. In order to progress towards a method for the quality control of this complex material, HPLC and 1H-NMR approaches as methods of quality control have been compared. HPLC analyses of 43 samples of propolis were carried out and six marker compounds were quantified in each sample. The same samples were analysed using 1H-NMR and the spectra were then converted into their first derivative forms and digitised using the software application MestRe-C. The digitised data were subjected to principal component analysis using the software application Simca-P. It was found that the chemical composition of propolis mapped well according to the geographical origins of the samples studied when the first three principal components were used to display them. In addition, each sample was assessed for anti-oxidant activity, and the results were then overlaid onto the sample groupings according to 1H-NMR data. It was observed that anti-oxidant properties also mapped quite well according to geographical origin.

Geography↗
PubMed2002

Modeling of gas absorption cross sections by use of principal-component-analysis model parameters.

Monitoring the amount of gaseous species in the atmosphere and exhaust gases by remote infrared spectroscopic methods calls for the use of a compilation of spectral data, which can be used to match spectra measured in a practical application. Model spectra are based on time-consuming line-by-line calculations of absorption cross sections in databases by use of temperature as input combined with path length and partial and total pressure. It is demonstrated that principal component analysis (PCA) can be used to compress the spectrum of absorption cross sections, which depend strongly on temperature, into a reduced representation of score values and loading vectors. The temperature range from 300 to 1000 K is studied. This range is divided into two subranges (300-650 K and 650-1000K), and separate PCA models are constructed for each. The relationship between the scores and the temperature values is highly nonlinear. It is shown, however, that because the score-temperature relationships are smooth and continuous, they can be modeled by polynomials of varying degrees. The accuracy of the data compression method is validated with line-by-line-calculated absorption data of carbon monoxide and water vapor. Relative deviations between the absorption cross sections reconstructed from the PCA model parameters and the line-by-line-calculated values are found to be smaller than 0.15% for cross sections exceeding 1.27 x 10(-21) cm(-1) atm(-1) (CO) and 0.20% for cross sections exceeding 4.03 x 10(-21) cm(-1) atm(-1) (H2O). The computing time is reduced by a factor of 10(4).

Journal Article↗
PubMed2003

Thermoanalytical, chemical and principal component analysis of plant drugs.

Thermal decomposition and elemental content of commercial raw plant materials used in medicine-roots, rhizomes and bark originating from different medicinal plant species were analyzed. The thermal decomposition was performed using the derivatograph. The content of non-metallic (N, P, S, Cl, I and B) and metallic (Ca, Mg, Fe, Mn, Cu and Zn) elements was determined by spectrophotometric techniques after previous mineralization of samples. In order to obtain more clear classification of the analyzed plant materials principal component analysis (PCA) was applied. Interpretation of PCA results for three databases (thermoanalytical, non-metals and metals data sets) allows the statement that samples of roots, rhizomes and bark from the same plant species in majority of cases are characterized by similar elemental composition and similar course of their thermal decomposition. In this way the differences in general chemical composition of medicinal plants raw materials can be determined.

Differential Thermal Analysis↗
PubMed1995

Principal component analysis and the scaled subprofile model compared to intersubject averaging and statistical parametric mapping: I. "Functional connectivity" of the human motor system studied with [15O]water PET.

Using [15O]water PET and a previously well studied motor activation task, repetitive finger-to-thumb opposition, we compared the spatial activation patterns produced by (1) global normalization and intersubject averaging of paired-image subtractions, (2) the mean differences of ANCOVA-adjusted voxels in Statistical Parametric Mapping, (3) ANCOVA-adjusted voxels followed by principal component analysis (PCA), (4) ANCOVA-adjustment of mean image volumes (mean over subjects at each time point) followed by F-masking and PCA, and (5) PCA with Scaled Subprofile Model pre- and postprocessing. All data analysis techniques identified large positive focal activations in the contralateral sensorimotor cortex and ipsilateral cerebellar cortex, with varying levels of activation in other parts of the motor system, e.g., supplementary motor area, thalamus, putamen; techniques 1-4 also produced extensive negative areas. The activation signal of interest constitutes a very small fraction of the total nonrandom signal in the original dataset, and the exact choice of data preprocessing steps together with a particular analysis procedure have a significant impact on the identification and relative levels of activated regions. The challenge for the future is to identify those preprocessing algorithms and data analysis models that reproducibly optimize the identification and quantification of higher-order sensorimotor and cognitive responses.

Adult↗
PubMed2000

Classification of chili powders by thin-layer chromatography and principal component analysis.

Silica gel, aluminium oxide, diatomaceous earth, polyamide, cyano, diol and amino plates have been tested for their capacity to separate the color pigments of six chili powders of different origin by both adsorption and reversed-phase thin-layer chromatography. The plates were evaluated at 340 and 440 nm wavelengths. Best separation of color pigments was obtained on impregnated diatomaceous earth layer using acetone-water 17:3 v/v eluent. It was found that the pigment composition of chili powders showed marked differences. Principal component analysis employed for the classification of the chili powders according to their pigment composition indicated that these differences can be used for the determination of the similarity or dissimilarity of the chili powders.

Capsicum↗
PubMed1992

Depression in Benin: an assessment using the Comprehensive Psychopathological Rating Scale and the principal component analysis.

Ninety two major depressed outpatients were rated with the Comprehensive Psychopathological Rating Scale (CPRS) in Cotonou, in Benin (West-Africa). Patients satisfied modified DSM III major depression criteria and were French-speaking. Men, civil servants, and city dwellers were over-represented in the population sample. The examination of item frequency yielded rather 'western-like' clinical features of depression: with differences described previously: a lower frequency of suicidal thoughts and guilt feelings, a higher frequency of somatic complaints and ideas of persecution. Principal component analysis reinforced 'western-like' aspects. The relationship between the so-called 'western culture-bound symptoms' and the so-called 'African ones' is discussed.

Adult↗
PubMed1982

Principal component analysis of evoked responses and the effects of alcohol on the geniculo-striate system of the monkey.

This study was designed to test the effects of alcohol on visual evoked potentials in nonhuman primates performing a cognitive task. Flash evoked potentials were recorded from monkeys involved in a delayed matching-to-sample (DMS) paradigm in which the flash served as an alerting signal before each trial. Event-related potentials were recorded from the lateral geniculate nucleus and homolateral striate cortex before, during, and after intravenous administration of saline or ethanol (0.25, 0.5, 1.0, and 2.0 g/kg). Average evoked potentials (AEPs) were computed. Residual waveforms were obtained by subtracting the predrug AEP from postdrug AEPs. A principal component analysis was employed to define the alcohol alterations on the evoked responses. In the analysis each AEP was represented by 40 time points spaced 12 msec apart. These reduced representations of the AEP were entered in the variance-covariance matrix calculations. The first five eigenvectors were computed and plotted. Alcohol produced the greatest variance in the AEPs at the two highest dose levels. So the data were grouped together into three experimental categories: saline, low-dose (0.25-0.5 g/kg) and high-dose (1.0-2.0 g/kg). A correlation template, representing each category, was computed by correlating individual eigenvectors with each sequential average composed of 10 individual evoked potentials in the 200 trials of an experimental session. Alcohol affected the state vector from the brain by loading the correlation coefficient in the opposite direction following alcohol administration in two principal components. One or two of the eigenvectors significantly (P less than 0.01) shifted in geniculate nucleus, indicating that either the nucleus or a previous station was affected by alcohol. In comparison, three or more eigenvectors from striate cortex were shifted significantly following alcohol injection. This difference may be explained by the effect of alcohol on multisynaptic brain structures, including the brain-stem reticular formation, which in turn influenced the cortex.

Animals↗
PubMed2005

Distinctiveness, typicality, and recollective experience in face recognition: a principal components analysis.

In this study, participants rated previously unseen faces on six dimensions: familiarity, distinctiveness, attractiveness, memorability, typicality, and resemblance to a familiar person. The faces were then presented again in a recognition test in which participants assigned their positive recognition decisions to either remember (R), know (K), or guess categories. On all dimensions except typicality, faces that were categorized as R responses were associated with significantly higher ratings than were faces categorized as K responses. Study ratings for R and K responses were then subjected to a principal components analysis. The factor loadings suggested that R responses were influenced primarily by the distinctiveness of faces, but K responses were influenced by moderate ratings on all six dimensions. These findings indicate that the structural features of a face influence the subjective experience of recognition.

Face↗
PubMed2006

Principal component analysis of urine metabolites detected by NMR and DESI-MS in patients with inborn errors of metabolism.

Urine metabolic profiles of patients with inborn errors of metabolism were examined with nuclear magnetic resonance (NMR) and desorption electrospray ionization mass spectrometry (DESI-MS) methods. Spectra obtained from the study of urine samples from individual patients with argininosuccinic aciduria (ASA), classic homocystinuria (HCY), classic methylmalonic acidemia (MMA), maple syrup urine disease (MSUD), phenylketonuria (PKU) and type II tyrosinemia (TYRO) were compared with six control patient urine samples using principal component analysis (PCA). Target molecule spectra were identified from the loading plots of PCA output and compared with known metabolic profiles from the literature and metabolite databases. Results obtained from the two techniques were then correlated to obtain a common list of molecules associated with the different diseases and metabolic pathways. The combined approach discussed here may prove useful in the rapid screening of biological fluids from sick patients and may help to improve the understanding of these rare diseases.

Case-Control Studies↗
PubMedDate not supplied

Data reduction of multichannel fields: global field power and principal component analysis.

Electroencephalographic data recorded for topographical analysis constitute multidimensional observations, and the present paper illustrates methods of data analysis of multichannel recordings where components of evoked brain activity are identified quantitatively. The computation of potential field strength (Global Field Power, GFP) is used for component latency determination. Multivariate statistical methods like Principal Component Analysis (PCA) may be applied to the topographical distribution of potential values. The analysis of statistically defined components of visually elicited brain activity is illustrated with data sets stemming from different experiments. With spatial PCA the dimensionality of multichannel data is reduced to only three components that account for more than 90% of the variance. The results of spatial PCA relate to experimental conditions in a meaningful way, and this method may also be used for time segmentation of topographic potential maps series.

Action Potentials↗
PubMed2005

Chromatographic classification and comparison of commercially available reversed-phase liquid chromatographic columns containing polar embedded groups/amino endcappings using principal component analysis.

Polar embedded phases have become increasingly popular in liquid chromatography (LC) analysis. These phases can produce diverse chromatographic selectivities as a result of their differing base silica, the type of polar embedded group (i.e. amide, urea, carbamate, ether or sulphonamide moieties) and the length of the alkyl ligand. Four column characterization protocols, using differing test probes, have been used to characterize 18 of these phases together with 17 alkyl phases (some of which contained novel polar endcapping, i.e. amino), which have been evaluated using principal component analysis (PCA). PCA provided graphical comparisons of the differences/similarities between these phases and between their corresponding C-alkyl, amino endcapped and enhanced polar selectivity phases.

Amines↗
PubMed1999

Disease-specific changes in equine ground reaction force data documented by use of principal component analysis.

OBJECTIVE: To assess the force plate as a diagnostic aid in equine locomotor abnormalities, particularly for abnormalities such as navicular disease that do not have specific diagnostic criteria. ANIMALS: 17 Thoroughbreds without observable locomotor abnormalities (group A), 6 Thoroughbreds with superficial digital flexor tendon injury (group B), and 8 Thoroughbreds with navicular disease (group C). PROCEDURE: Using a force plate, ground reaction force patterns were recorded at the trot. Peak limb vertical force and force/time curve parameters were derived from 4 identifiable points at the beginning and end of vertical and craniocaudal horizontal plots. Principal component analysis (PCA) of group-A data was undertaken on beginning and end of stride data, and the first 2 components were represented graphically. The PCA rotation matrices were applied to equivalent data for horses of groups B and C. RESULTS: Asymmetry of peak vertical force (PVF) could not be differentiated among groups A, B, and C. Values for group-B horses, however, were significantly outside mean group-A values on the PCA plot for beginning of stride phase variables. Group-B data were within the group-A range for end of stride phase variables. Values for group-C horses were significantly outside the group-A range for beginning of stride phase variables and were outside mean group-A values for end of stride phase variables. CONCLUSIONS: PCA of force/time data provides a sensitive method to evaluate the force/time curve associated with 2 specific injury/disease processes. CLINICAL RELEVANCE: Horses alter weight-bearing in biomechanically distinct ways, thus creating potential for the force plate to become an important diagnostic and prognostic tool.

Animals↗
PubMedDate not supplied

The symptoms of hyperglycaemia in people with insulin-treated diabetes: classification using principal components analysis.

BACKGROUND AND AIMS: People with insulin-treated diabetes commonly experience symptoms of hyperglycaemia, but the nature of these symptoms and their origins are poorly understood. The aims of this study were (1) to identify and classify the symptoms of hyperglycaemia experienced by people with insulin-treated diabetes and (2) to identify patient characteristics associated with intensity of, and glycaemic threshold for, glycaemic symptoms. METHODS: Common hyperglycaemic symptoms were identified from preliminary interviews. Eighteen symptoms were used in a questionnaire. Four hundred participants estimated the intensities with which they experienced these symptoms during hyperglycaemia. Principal components analysis (PCA) was used to examine correlations between symptoms. Associations between symptom intensity, glycaemic threshold, and other characteristics were examined with multiple regression. RESULTS: In total, 361 participants (90.2%) reported experiencing hyperglycaemic symptoms. PCA suggested four symptom groupings: (1) feeling tense, irritability, restlessness, poor concentration (agitation) (2) thirst, dry mouth, need to urinate, not feeling right, sweet/funny taste, weakness (osmotic) (3) dizziness, blurred vision, light-headedness, weakness (neurological) (4) headache, nausea (malaise). Mean symptom intensity was associated with younger age. The median (range) estimated blood glucose threshold for symptom onset was 15 (8-30) mmol/L; there was a weak tendency for this threshold to be elevated in people who had impaired hypoglycaemia awareness. CONCLUSIONS: People with insulin-treated diabetes commonly reported symptoms associated with hyperglycaemia. PCA separated these into four groups. Osmotic symptoms appear to be specific to hyperglycaemia; symptoms in the other groups may suggest underlying physiological mechanisms, but are relatively non-specific. Symptoms are more intense in younger people and may be reported at lower blood glucose concentrations in people with normal awareness of hypoglycaemia.

Adolescent↗
PubMed1997

Principal Component Analysis of the Absorption Spectra of the Dye Thiacyanine in the Presence of the Surfactant AOT: Precise Identification of the Dye-Surfactant Aggregates

Spectral change of a cationic dye 3,3'-diethylthiacyanine iodide (THIA) in the presence of an anionic surfactant AOT has been presented. The THIA-AOT system exhibits blue-shifted metachromasia at AOT concentrations below its critical micellar concentration (CMC) and is thought to be due to the aggregation of electrostatically bound dye-AOT complex (DS). Metachromasia is gradually reversed to the monomeric band (peak at 425 nm) by AOT above its CMC. Principal component analysis (PCA) method has been applied for spectral analysis; the results show that the metachromatic peaks at 377 and 366 nm originate, respectively, from trimer and hexamer of the dye associated with AOT. From PCA, the molar absorption coefficient spectra of the individual absorbing components and the equilibrium constants for the systems, monomer right harpoon over left harpoon trimer and hexamer right harpoon over left harpoon monomer below and above CMC of AOT, respectively, have also been obtained. The micellar aggregation number of AOT obtained from PCA is found to be 16 which is in good agreement with the literature value.

Journal Article↗
PubMed2000

Principal component analysis and blind separation of sources for optical imaging of intrinsic signals.

The analysis of data sets from optical imaging of intrinsic signals requires the separation of signals, which accurately reflect stimulated neuronal activity (mapping signal), from signals related to background activity. Here we show that blind separation of sources by extended spatial decorrelation (ESD) is a powerful method for the extraction of the mapping signal from the total recorded signal. ESD is based on the assumptions (i) that each signal component varies smoothly across space and (ii) that every component has zero cross-correlation functions with the other components. In contrast to the standard analysis of optical imaging data, the proposed method (i) is applicable to nonorthogonal stimulus-conditions, (ii) can remove the global signal, blood-vessel patterns, and movement artifacts, (iii) works without ad hoc assumptions about the data structure in the frequency domain, and (iv) provides a confidence measure for the signals (Z score). We first demonstrate on orientation maps from cat and ferret visual cortex, that principal component analysis, which acts as a preprocessing step to ESD, can already remove global signals from image stacks, as long as data stacks for at least two-not necessarily orthogonal-stimulus conditions are available. We then show that the full ESD analysis can further reduce global signal components and-finally-concentrate the mapping signal within a single component both for differential image stacks and for image stacks recorded during presentation of a single stimulus.

Animals↗
PubMed1983

Principal component analysis of Motokawa's data on wavelength dependence of retinal processes.

Electric excitability of the human eye as determined by measuring the threshold for a sensation of phosphene in response to electric stimulation of the eye was found by Koiti Motokawa to increase temporarily after a brief illumination (J. Neurophysiol., 1949, 112, 475-488). While changing the wavelength of illuminating light widely, he found that the time course of the variation in the eye's electric excitability after illumination differed characteristically according to the wavelength. His data on this point (Tohoku J. exp. Med., 1949, 51, 197-205) were subjected to the principal component analysis. Three components were found necessary and sufficient for their linear combinations to reproduce time courses of the excitability enhancement after illumination with lights of varying wavelengths; one of the three components makes a great contribution to the excitability enhancement by green lights, the other to the one by red lights and the remainder to the one by blue lights. This is in support of Motokawa's view that his data are interpretable as summation effects of the three retinal processes which are excited preferentially by red, green and blue lights, respectively.

Electric Stimulation↗