NMR spectral quantitation by principal component analysis.

The use of principal component analysis (PCA) for simultaneous spectral quantitation of a single resonant peak across a series of spectra has gained popularity among the NMR community. The approach is fast, requires no assumptions regarding the peak lineshape and provides quantitation even for peaks with very low signal-to-noise ratio. PCA produces estimates of all peak parameters: area, frequency, phase and linewidth. If desired, these estimates can be used to correct the original data so that the peak in all spectra has the same lineshape. This ability makes PCA useful not only for direct peak quantitation, but also for processing spectral data prior to application of pattern recognition/classification techniques. This article briefly reviews the theoretical basis of PCA for spectral quantitation, addresses issues of data processing prior to PCA, describes suitable and unsuitable datasets for PCA applications and summarizes the developments and the limitations of the method.     

Classification of macular and optic nerve disease by principal component analysis

In this study, pattern electroretinography (PERG) signals were obtained by electrophysiological testing devices from 70 subjects. The group consisted of optic nerve and macular diseases subjects. Characterization and interpretation of the physiological PERG signal was done by principal component analysis (PCA). While the first principal component of data matrix acquired from optic nerve patients represents 67.24% of total variance, the first principal component of the macular patients data matrix represents 76.81% of total variance. The basic differences between the two patient groups were obtained with first principal component, obviously. In addition, the graphic of second principal component vs. first principal component of optic nerve and macular subjects was analyzed. The two patient groups were separated clearly from each other without any hesitation. This research developed an auxiliary system for the interpretation of the PERG signals. The stated results show that the use of PCA of physiological waveforms is presented as a powerful method likely to be incorporated in future medical signal processing.     

Distinguishing normal and abnormal tracheal breathing sounds by principal component analysis

Expired and inspired tracheal breathing sounds (BS) were recorded from 10 normal subjects and 8 patients with respiratory diseases, including bronchial asthma, sarcoidosis, fibrosing lung disease, chronic bronchitis, and radiation pneumonitis. Frequency spectra were generated using Fast Fourier Transform (FFT), and we observed considerable differences between BS spectra of normal subjects and patients. The frequency of peak amplitude and mean frequency of the BS spectra of patients were significantly higher than those of normal subjects. Spectral features were extracted by dividing each spectra into equal frequency bands--each feature being the mean amplitude of each FFT element within a frequency band. We used Principal Component Analysis to compare spectral feature sets and found a clear separation between normal and abnormal tracheal BS for 10, 20, and 40 features/spectra. We conclude that Principal Component Analysis of BS could become a new method of diagnosing respiratory disease in an automated fashion.     

Recovering dipole sources from scalp-recorded event-related-potentials using component analysis: principal component analysis and independent component analysis.

Principal component analysis (PCA) and independent component analysis (ICA) were examined in their ability to recover dipole sources from simulated data. Datasets of EEG segments were generated that contained cortical sources that were temporally overlapping or non-overlapping, and dipole sources with varying degree of spatial orthogonality. For temporal overlapping sources, both PCA and ICA resulted in components that required multiple-source equivalent current dipole models. The spatially overlapping sources affected the PCA method more than ICA, resulting in single PCA components in which all non-orthogonal sources were represented. For both PCA and ICA, dipole models with fixed-location dipoles successfully accounted for most of the variance in the component weights, even when the spatial or temporal overlap of the generating sources required multiple-dipole models.     

Longitudinal principal component analysis of somatic and motor characteristics

Longitudinal principal component analysis was applied to the study of growth patterns of several somatic and motor characteristics. The data are from the “Leuven Growth Study of Belgian Boys,” in which six annual observations were made. Peak growth velocity and age at peak growth velocity for each variable were estimated. The results for height show three components sufficient to provide an adequate representation of the original information. The first component can be interpreted as a growth distance component, which characterizes the general position of an individual growth curve relative to the average growth curve during the period analyzed. Components 2 and 3 reflect fluctuations in percentile level during the age period studied and can be conceived as indices of stability. Component 2 provides additional information about the age at peak height velocity; component 3 characterizes variation in peak height velocity. These interpretations can, in general, be assumed for the other characteristics analyzed.     

Gene expression data classification with Kernel principal component analysis

One important feature of the gene expression data is that the number of genes M far exceeds the number of samples N. Standard statistical methods do not work well when N < M. Development of new methodologies or modification of existing methodologies is needed for the analysis of the microarray data. In this paper, we propose a novel analysis procedure for classifying the gene expression data. This procedure involves dimension reduction using kernel principal component analysis (KPCA) and classification with logistic regression (discrimination). KPCA is a generalization and nonlinear version of principal component analysis. The proposed algorithm was applied to five different gene expression datasets involving human tumor samples. Comparison with other popular classification methods such as support vector machines and neural networks shows that our algorithm is very promising in classifying gene expression data.     

Feature selection of stabilometric parameters based on principal component analysis

This study addresses the challenge of identifying the features of the Centre of pressure (COP) trajectory that are most sensitive to postural performance, with the aim of avoiding redundancy and allowing a straightforward interpretation of the results. Postural sway in 50 young, healthy subjects was measured by a force platform. Thirty-seven stabilometric parameters were computed from the one-dimensional and two-dimensional COP time series. After normalisation to the relevant biomechanical factors, by means of multiple regression models, a feature selection process was performed based on principal component analysis. Results suggest that COP two-dimensional time series can be primarily characterised by four parameters, describing the size of the COP path over the support surface; the principal sway direction; and the shape and bandwidth of the power spectral density plot. COP one-dimensional time series (antero-posterior (AP) and medio-lateral (ML)) can be characterised by six parameters describing COP dispersion along the AP direction; mean velocity along the ML and AP directions; the contrast between ML and AP regulatory activity; and two parameters describing the spectral characteristics of the COP along the AP direction. On the basis of the results obtained, some guidelines are suggested for the choice of stabilometric parameters to use, with the aim of promoting standardisation in quantitative posturography.     

Interpretation of laboratory results using multidimensional scaling and principal component analysis

Principal component analysis (PCA) and multidimensional scaling (MDS) are a set of mathematical techniques which uncover the underlying structure of data by examining the relationships between variables. Both MDS and PCA use proximity measures such as correlation coefficients or Euclidean distances to generate a spatial configuration (map) of points where distances between points reflect the relationship between individuals with their underlying set of data. Multidimensional scaling, when compared to PCA, gives more readily interpretable solutions of lower dimensionality and does not depend on the assumption of a linear relationship between variables. Both MDS and PCA were applied to electrolyte profiles of patients with acute renal failure and patients without apparent disease. The MDS was superior to PCA in separating renal patients from normal patients. The one-dimensional and two-dimensional solutions of MDS and PCA were compared.     

Hydroxy-fatty acid profiles of Legionella species: diagnostic usefulness assessed by principal component analysis.

Twenty-nine species (76 strains) of members of the genus Legionella were analyzed for their cellular hydroxylated fatty acids (OH-FAs). The individual patterns were unusually complex and included both monohydroxylated and dihydroxylated chains of unbranched or branched (iso and anteiso) types. Comparison of the strain profiles by SIMCA (Soft Independent Modelling of Class Analogy) principal component analysis revealed four main groups. Group 1 included Legionella pneumophila plus L. israelensis strains, and group 2 included L. micdadei and L. maceacherneii strains. These two closely related groups were characterized by the occurrence of di-OH-FAs and differed mainly in the amounts of 3-OH-a21:0, 3-OH-n21:0, 3-OH-n22:0, and 3-OH-a23:0. Group 3 (13 species) was distinguished by i14:0 at less than 3%, 3-OH-3-OH-n14:0 at greater than 5%, 3-OH-n15:0 at greater than 2%, and minute amounts of OH-FAs with chains longer than 21:0. Group 4 (12 species) was heterogeneous. Its main characteristics were the presence of 3-OH-n12:0 and 3-OH-n13:0, 3-OH-i14:0 at greater than 5%, as well as significant amounts of 3-OH-a21:0 and 3-OH-n21:0. The groupings obtained by OH-FA profiles were found to reflect DNA-DNA homology groupings reasonably well, and the profiles appear to be useful for differentiation of Legionella species.     

Characterization of premotor interneurones by their input patterns--application of principal component analysis to cat cervical interneurones

Principal component analysis of input patterns of cat C6-C8 interneurones (300 cells) revealed that identified premotor interneurones (11 cells) activated from skin afferents and projecting to T1 motoneurones possessed a special input pattern, characterized by restricted distribution on the plane of the first (Prin 1) versus second (Prin 2) principal component (high positive values of both components). These premotor neurones were located mostly in laminae V-VI. Among other laminae V-VI cells descending in the lateral funiculus to T1 similar to such premotor neurones, there were cells distributed similarly on the Prin 1-2 plane. Further, a majority of interneurones antidromically activated from the T1 motor nucleus at low thresholds also showed a distribution on the plane similar to the premotor neurones. We suggest that premotor neurones of this input pattern constitute a major group among laminae V-VI premotor neurones projecting to T1.     

一种基于小波域主成分分析的心电压缩算法

目的为降低心电信号存储和传输的数据量,并克服传统心电压缩方法只利用导联内相关性的劣势,本文提出一种基于小波域主成分分析和分层编码(w PCA_LC)的压缩方法。方法首先通过心电电极获取12通道心电数据,对所有通道的心电信号做小波变换,每个尺度下的小波系数组成小波系数矩阵,在每个系数矩阵上做主成分分析(principal component analysis,PCA),之后对小波系数小的主成分做[位置增量,数据]的编码方式,其他主成分采用霍夫曼编码,最后使用本文算法压缩圣彼得堡心率失常数据库。结果实验表明,在均方根误差为5.2%时,本文算法的压缩比为71,远高于基于稀疏分解的方法和基于小波变换阈值选择的方法。结论基于小波域主成分分析的心电压缩算法对多导联心电信号具有较好的压缩性能。     

Effects of habitual physical exercise on physiological age in men aged 20–85 years as estimated using principal component analysis

A population of 221 healthy adult men (aged 20–85 years) was studied to determine whether those who exercised regularly were in good biological condition, and also whether those who were in a state of high physical fitness were in a good state biologically, in terms of physiological age (PA) and physical fitness age (FA) as estimated by principal component analysis. A group of 17 physiological function tests and 5 physical fitness tests were employed to estimate PA and FA, respectively. The results of this study indicated that those who maintained high physical fitness at all age decade groups from 20 to 79 years had a trend towards maintaining a relatively lower PA (physiologicallyyounger). Mean PA and FA of the trained group were younger by 4.7 and 7.3 years, respectively than those of the untrained group. In addition, the slope of regression line of PA on chronological age was more gentle in the trained group than that in the untrained group. These results would suggest that those who are in a state of high physical fitness maintain a relatively good physiological condition, and that regular physical exercise may delay physiological changes normally seen with aging, and consequently may increase the life span.     

Optical pathology using oral tissue fluorescence spectra: classification by principal component analysis and k-means nearest neighbor analysis

The spectral analysis and classification for discrimination of pulsed laser-induced autofluorescence spectra of pathologically certified normal, premalignant, and malignant oral tissues recorded at a 325-nm excitation are carried out using MATLAB@R6-based principal component analysis (PCA) and k-means nearest neighbor (k-NN) analysis separately on the same set of spectral data. Six features such as mean, median, maximum intensity, energy, spectral residuals, and standard deviation are extracted from each spectrum of the 60 training samples (spectra) belonging to the normal, premalignant, and malignant groups and they are used to perform PCA on the reference database. Standard calibration models of normal, premalignant, and malignant samples are made using cluster analysis. We show that a feature vector of length 6 could be reduced to three components using the PCA technique. After performing PCA on the feature space, the first three principal component (PC) scores, which contain all the diagnostic information, are retained and the remaining scores containing only noise are discarded. The new feature space is thus constructed using three PC scores only and is used as input database for the k-NN classification. Using this transformed feature space, the centroids for normal, premalignant, and malignant samples are computed and the efficient classification for different classes of oral samples is achieved. A performance evaluation of k-NN classification results is made by calculating the statistical parameters specificity, sensitivity, and accuracy and they are found to be 100, 94.5, and 96.17%, respectively.     

Analysis of diffusion tensor magnetic resonance imaging data using principal component analysis

An analysis method for diffusion tensor (DT) magnetic resonance imaging data is described, which, contrary to the standard method (multivariate fitting), does not require a specific functional model for diffusion-weighted (DW) signals. The method uses principal component analysis (PCA) under the assumption of a single fibre per pixel. PCA and the standard method were compared using simulations and human brain data. The two methods were equivalent in determining fibre orientation. PCA-derived fractional anisotropy and DT relative anisotropy had similar signal-to-noise ratio (SNR) and dependence on fibre shape. PCA-derived mean diffusivity had similar SNR to the respective DT scalar, and it depended on fibre anisotropy. Appropriate scaling of the PCA measures resulted in very good agreement between PCA and DT maps. In conclusion, the assumption of a specific functional model for DW signals is not necessary for characterization of anisotropic diffusion in a single fibre.     

Statistical analysis of mitochondrial pathologies in childhood: identification of deficiencies using principal component analysis

Mitochondrial pathologies are a heterogeneous group of metabolic disorders that are frequently characterized by anomalies of oxidative phosphorylation, especially in the respiratory chain. The identification of these anomalies may involve many investigations, and biochemistry is a main tool. However, considering the whole set of biochemical data, the interpretation of the results by the traditionally used statistical methods remains complex and does not always lead to an unequivocal conclusion about the presence or absence of a respiratory chain defect. This arises from three main problems: (a) the absence of an a priori-defined control population, because the determination of the control values are derived from the whole set of investigated patients, (b) the small size of the population studied, (c) the large number of variables collected, each of which creates a wide variability. To cope with these problems, the principal component analysis (PCA) has been applied to the biochemical data obtained from 35 muscle biopsies of children suspected of having a mitochondrial disease. This analysis makes it possible for each respiratory chain complex to distinguish between different subsets within the whole population (normal, deficient, and, in between, borderline subgroups of patients) and to detect the most discriminating variables. PCA of the data of all complexes together showed that mitochondrial diseases in this population were mainly caused by multiple deficits in respiratory chain complexes. This analysis allows the definition of a new subgroup of newborns, which have high respiratory chain complex activity values. Our results show that the PCA method, which simultaneously takes into account all of the concerned variables, allows the separation of patients into subgroups, which may help clinicians make their diagnoses.     

Phenotypes of comorbidity in OSAS patients: combining categorical principal component analysis with cluster analysis

Phenotyping obstructive sleep apnea syndrome's comorbidity has been attempted for the first time only recently. The aim of our study was to determine phenotypes of comorbidity in obstructive sleep apnea syndrome patients employing a data‐driven approach. Data from 1472 consecutive patient records were recovered from our hospital's database. Categorical principal component analysis and two‐step clustering were employed to detect distinct clusters in the data. Univariate comparisons between clusters included one‐way analysis of variance with Bonferroni correction and chi‐square tests. Predictors of pairwise cluster membership were determined via a binary logistic regression model. The analyses revealed six distinct clusters: A, ‘healthy, reporting sleeping related symptoms’; B, ‘mild obstructive sleep apnea syndrome without significant comorbidities’; C₁: ‘moderate obstructive sleep apnea syndrome, obesity, without significant comorbidities’; C₂: ‘moderate obstructive sleep apnea syndrome with severe comorbidity, obesity and the exclusive inclusion of stroke’; D₁: ‘severe obstructive sleep apnea syndrome and obesity without comorbidity and a 33.8% prevalence of hypertension’; and D₂: ‘severe obstructive sleep apnea syndrome with severe comorbidities, along with the highest Epworth Sleepiness Scale score and highest body mass index’. Clusters differed significantly in apnea–hypopnea index, oxygen desaturation index; arousal index; age, body mass index, minimum oxygen saturation and daytime oxygen saturation (one‐way analysis of variance P < 0.0001). Binary logistic regression indicated that older age, greater body mass index, lower daytime oxygen saturation and hypertension were associated independently with an increased risk of belonging in a comorbid cluster. Six distinct phenotypes of obstructive sleep apnea syndrome and its comorbidities were identified. Mapping the heterogeneity of the obstructive sleep apnea syndrome may help the early identification of at‐risk groups. Finally, determining predictors of comorbidity for the moderate and severe strata of these phenotypes implies a need to take these factors into account when considering obstructive sleep apnea syndrome treatment options.