Moving window kernel PCA for adaptive monitoring of nonlinear processes

This paper discusses the monitoring of complex nonlinear and time-varying processes. Kernel principal component analysis (KPCA) has gained significant attention as a monitoring tool for nonlinear systems in recent years but relies on a fixed model that cannot be employed for time-varying systems. The contribution of this article is the development of a numerically efficient and memory saving moving window KPCA (MWKPCA) monitoring approach. The proposed technique incorporates an up- and downdating procedure to adapt (i) the data mean and covariance matrix in the feature space and (ii) approximates the eigenvalues and eigenvectors of the Gram matrix. The article shows that the proposed MWKPCA algorithm has a computation complexity of O(N2), whilst batch techniques, e.g. the Lanczos method, are of O(N3). Including the adaptation of the number of retained components and an l-step ahead application of the MWKPCA monitoring model, the paper finally demonstrates the utility of the proposed technique using a simulated nonlinear time-varying system and recorded data from an industrial distillation column.

Kids’ Life and Times: using an Internet survey to measure children’s health-related quality of life

Objective
To examine the psychometric properties of an internet version of a children and young person's quality of life measure originally designed as a paper questionnaire.

Methods
Participants were 3,440 10 and 11 year old children in Northern Ireland who completed the KIDSCREEN-27 online as part of a general attitudinal survey. The questionnaire was animated using cartoon characters that are familiar to most children and the questions appeared on screen and were read aloud by actors.

Results
Exploratory principal component analysis of the online version of the questionnaire supported the existence of five components in line with the paper version. The items loaded on the components that would be expected based on previous findings with five domains - physical well-being,psychological well-being, autonomy and parents, social support and peers and school environment.Internal consistency reliability of the five domains was measured using Cronbach's alpha and the results suggested that the scale scores were reliable. The domain scores were similar to those reported in the literature for the paper version.

Conclusions
These results suggest that the factor structure and internal consistency reliability scores of the KIDSCREEN-27 embedded within an online survey are comparable to those reported in the literature for the paper version.

Live-Cell Tracking Using SIFT Features in DIC Microscopic Videos

In this paper, a novel motion-tracking scheme using scale-invariant features is proposed for automatic cell motility analysis in gray-scale microscopic videos, particularly for the live-cell tracking in low-contrast differential interference contrast (DIC) microscopy. In the proposed approach, scale-invariant feature transform (SIFT) points around live cells in the microscopic image are detected, and a structure locality preservation (SLP) scheme using Laplacian Eigenmap is proposed to track the SIFT feature points along successive frames of low-contrast DIC videos. Experiments on low-contrast DIC microscopic videos of various live-cell lines shows that in comparison with principal component analysis (PCA) based SIFT tracking, the proposed Laplacian-SIFT can significantly reduce the error rate of SIFT feature tracking. With this enhancement, further experimental results demonstrate that the proposed scheme is a robust and accurate approach to tackling the challenge of live-cell tracking in DIC microscopy.

Metabolic profile changes in the testes of mice with streptozotocin-induced type 1 diabetes mellitus

Contrary to the traditional view, recent studies suggest that diabetes mellitus has an adverse influence on male reproductive function. Our aim was to determine the effect of diabetes on the testicular environment by identifying and then assessing perturbations in small molecule metabolites. Testes were obtained from control and streptozotocin-induced diabetic C57BL/6 mice, 2, 4 and 8 weeks post-treatment. Diabetic status was confirmed by glycated haemoglobin, non-fasting blood glucose, physiological condition and body weight. A novel extraction procedure was utilized to obtain protein free, low-molecular weight, water soluble extracts which were then assessed using H-1 nuclear magnetic resonance spectroscopy. Principal component analysis of the derived profiles was used to classify any variations, and specific metabolites were identified based on their spectral pattern. Characteristic metabolite profiles were identified for control and type 1 diabetic animals with the most distinctive being from mice with the largest physical deterioration and loss of body weight. Eight streptozotocin-treated animals did not develop diabetes and displayed profiles similar to controls. Diabetic mice had decreases in creatine, choline and carnitine and increases in lactate, alanine and myo-inositol. Betaine levels were found to be increased in the majority of diabetic mice but decreased in a few animals with severe loss of body weight and physical condition. The association between perturbations in a number of small molecule metabolites known to be influential in sperm function, with diabetic status and physiological condition, adds further impetus to the proposal that diabetes influences important spermatogenic pathways and mechanisms in a subtle and previously unrecognized manner.

Application of Nonlinear PCA for Fault Detection in Polymer Extrusion Processes

This paper describes the application of an improved nonlinear principal component analysis (PCA) to the detection of faults in polymer extrusion processes. Since the processes are complex in nature and nonlinear relationships exist between the recorded variables, an improved nonlinear PCA, which incorporates the radial basis function (RBF) networks and principal curves, is proposed. This algorithm comprises two stages. The first stage involves the use of the serial principal curve to obtain the nonlinear scores and approximated data. The second stage is to construct two RBF networks using a fast recursive algorithm to solve the topology problem in traditional nonlinear PCA. The benefits of this improvement are demonstrated in the practical application to a polymer extrusion process.

Improved Nonlinear PCA for Process Monitoring Using Support Vector Data Description

Nonlinear principal component analysis (PCA) based on neural networks has drawn significant attention as a monitoring tool for complex nonlinear processes, but there remains a difficulty with determining the optimal network topology. This paper exploits the advantages of the Fast Recursive Algorithm, where the number of nodes, the location of centres, and the weights between the hidden layer and the output layer can be identified simultaneously for the radial basis function (RBF) networks. The topology problem for the nonlinear PCA based on neural networks can thus be solved. Another problem with nonlinear PCA is that the derived nonlinear scores may not be statistically independent or follow a simple parametric distribution. This hinders its applications in process monitoring since the simplicity of applying predetermined probability distribution functions is lost. This paper proposes the use of a support vector data description and shows that transforming the nonlinear principal components into a feature space allows a simple statistical inference. Results from both simulated and industrial data confirm the efficacy of the proposed method for solving nonlinear principal component problems, compared with linear PCA and kernel PCA.

Dietary patterns and bone mineral status in young adults: the Northern Ireland Young Hearts Project

Studies of individual nutrients or foods have revealed much about dietary influences on bone. Multiple food or nutrient approaches, such as dietary pattern analysis, could offer further insight but research is limited and largely confined to older adults. We examined the relationship between dietary patterns, obtained by a posteriori and a priori methods, and bone mineral status (BMS; collective term for bone mineral content (BMC) and bone mineral density (BMD)) in young adults (20-25 years; n 489). Diet was assessed by 7 d diet history and BMD and BMC were determined at the lumbar spine and femoral neck (FN). A posteriori dietary patterns were derived using principal component analysis (PCA) and three a priori dietary quality scores were applied (dietary diversity score (DDS), nutritional risk score and Mediterranean diet score). For the PCA-derived dietary patterns, women in the top compared to the bottom fifth of the 'Nuts and Meat' pattern had greater FN BMD by 0.074 g/cm(2) (P=0.049) and FN BMC by 0.40 g (P=0.034) after adjustment for confounders. Similarly, men in the top compared to the bottom fifth of the 'Refined' pattern had lower FN BMC by 0.41 g (P-0.049). For the a priori DDS, women in the top compared to the bottom third had lower FN BMD by 0.05 g/cm(2) after adjustments (P=0.052), but no other relationships with BMS were identified. In conclusion, adherence to a 'Nuts and Meat' dietary pattern may be associated with greater BMS in young women and a 'Refined' dietary pattern may be detrimental for bone health in young men.

2D silhouette and 3D skeletal models for human detection and tracking

In this paper we propose a statistical model for detection and tracking of human silhouette and the corresponding 3D skeletal structure in gait sequences. We follow a point distribution model (PDM) approach using a Principal Component Analysis (PCA). The problem of non-lineal PCA is partially resolved by applying a different PDM depending of pose estimation; frontal, lateral and diagonal, estimated by Fisher's linear discriminant. Additionally, the fitting is carried out by selecting the closest allowable shape from the training set by means of a nearest neighbor classifier. To improve the performance of the model we develop a human gait analysis to take into account temporal dynamic to track the human body. The incorporation of temporal constraints on the model increase reliability and robustness.

Organizational assessment in paediatric primary care:Development and initial validation of the primary care organizational questionnaire

Primary care in the United States is undergoing many changes. Reliable and valid instruments are needed to assess the effects of these changes. The Primary Care Organizational Questionnaire (PCOQ), a 56-item 5-point Likert scale survey that evaluates interactions among members of the clinic/practice and job-related attributes, was administered to clinicians and staff in 36 primary care practices serving paediatric populations in Connecticut. A priori scales were reliable (Cronbach alpha =0.7). Analysis of variance (ANOVA) showed greater heterogeneity across clinics than within clinics for 13 of 15 a priori scales, which were then included in a principal component factor analysis with varimax rotation. Eigenvalue analysis showed nine significant factors, largely similar to the a priori scales, indicating concurrent construct validity. Further research will ascertain the utility of the PCOQ in predicting the effectiveness of primary care practices in implementing disease management programmes. © 2006 Royal Society of Medicine Press.

Evaluation of quality of life and priorities of patients with glaucoma

PURPOSE: To investigate the quality of life and priorities of patients with glaucoma.

METHODS: Patients diagnosed with glaucoma and no other ocular comorbidity were consecutively recruited. Clinical information was collected. Participants were asked to complete three questionnaires: EuroQuol (EQ-5D), time tradeoff (TTO), and choice-based conjoint analysis. The latter used five-attribute outcomes: (1) reading and seeing detail, (2) peripheral vision, (3) darkness and glare, (4) household chores, and (5) outdoor mobility. Visual field loss was estimated by using binocular integrated visual fields (IVFs).

RESULTS: Of 84 patients invited to participate, 72 were enrolled in the study. The conjoint utilities showed that the two main priorities were "reading and seeing detail" and "outdoor mobility." This rank order was stable across all segmentations of the data by demographic or visual state. However, the relative emphasis of these priorities changed with increasing visual field loss, with concerns for central vision increasing, whereas those for outdoor mobility decreased. Two subgroups of patients with differing priorities on the two main attributes were identified. Only 17% of patients (those with poorer visual acuity) were prepared to consider TTO. A principal component analysis revealed relatively independent components (i.e., low correlations) between the three different methodologies for assessing quality of life.

CONCLUSIONS: Assessments of quality of life using different methodologies have been shown to produce different outcomes with low intercorrelations between them. Only a minority of patients were prepared to trade time for a return to normal vision. Conjoint analysis showed two subgroups with different priorities. Severity of glaucoma influenced the relative importance of priorities.

Fault Diagnostics for Internal Combustion Engines – Current and Future Techniques

The adoption of each new level of automotive emissions legislation often requires the introduction of additional emissions reduction techniques or the development of existing emissions control systems. This, in turn, usually requires the implementation of new sensors and hardware which must subsequently be monitored by the on-board fault detection systems. The reliable detection and diagnosis of faults in these systems or sensors, which result in the tailpipe emissions rising above the progressively lower failure thresholds, provides enormous challenges for OBD engineers. This paper gives a review of the field of fault detection and diagnostics as used in the automotive industry. Previous work is discussed and particular emphasis is placed on the various strategies and techniques employed. Methodologies such as state estimation, parity equations and parameter estimation are explained with their application within a physical model diagnostic structure. The utilization of symptoms and residuals in the diagnostic process is also discussed. These traditional physical model based diagnostics are investigated in terms of their limitations. The requirements from the OBD legislation are also addressed. Additionally, novel diagnostic techniques, such as principal component analysis (PCA) are also presented as a potential method of achieving the monitoring requirements of current and future OBD legislation.

Behavioural and physiological responses of heifer calves to acute stressors: long-term consistency and relationship with adult reactivity to milking

The present study investigated the long-term consistency of individual differences in dairy cattles’ responses in tests of behavioural and hypothalamo–pituitary–adrenocortical (HPA) axis reactivity, as well as the relationship between responsiveness in behavioural tests and the reaction to first milking. Two cohorts of heifer calves, Cohorts 1 (N = 25) and 2 (N = 16), respectively, were examined longitudinally from the rearing period until adulthood. Cohort 1 heifers were subjected to open field (OF), novel object (NO), restraint, and response to a human tests at 7 months of age, and were again observed in an OF test during first pregnancy between 22 and 24 months of age. Subsequently, inhibition of milk ejection and stepping and kicking behaviours were recorded in Cohort 1 heifers during their first machine milking. Cohort 2 heifers were individually subjected to OF and NO tests as well as two HPA axis reactivity tests (determining ACTH and/or cortisol response profiles after administration of exogenous CRH and ACTH, respectively) at 6 months of age and during first lactation at approximately 29 months of age. Principal component analysis (PCA) was used to condense correlated response measures (to behavioural tests and to milking) within ages into independent dimensions underlying heifers’ reactivity. Heifers demonstrated consistent individual differences in locomotion and vocalisation during an OF test from rearing to first pregnancy (Cohort 1) or first lactation (Cohort 2). Individual differences in struggling in a restraint test at 7 months of age reliably predicted those in OF locomotion during first pregnancy in Cohort 1 heifers. Cohort 2 animals with high cortisol responses to OF and NO tests and high avoidance of the novel object at 6 months of age also exhibited enhanced cortisol responses to OF and NO tests at 29 months of age. Measures of HPA axis reactivity, locomotion, vocalisation and adrenocortical and behavioural responses to novelty were largely uncorrelated, supporting the idea that stress responsiveness in dairy cows is mediated by multiple independent underlying traits. Inhibition of milk ejection and stepping and kicking behaviours during first machine milking were not related to earlier struggling during restraint, locomotor responses to OF and NO tests, or the behavioural interaction with a novel object. Heifers with high rates of OF and NO vocalisation and short latencies to first contact with the human at 7 months of age exhibited better milk ejection during first machine milking. This suggests that low underlying sociality might be implicated in the inhibition of milk ejection at the beginning of lactation in heifers.

Profiling the ionome of rice and its use in discriminating geographical origins at the regional scale, China

Element profile was investigated for their use to trace the geographical origin of rice (Oryza sativa L.) samples. The concentrations of 13 elements (calcium (Ca), potassium (K), magnesium (Mg), phosphorus (P), boron (B), manganese (Mn), iron (Fe), nickel (Ni), copper (Cu), arsenic (As), selenium (Se), molybdenum (Mo), and cadmium (Cd)) were determined in the rice samples by inductively coupled plasma optical emission and mass spectrometry. Most of the essential elements for human health in rice were within normal ranges except for Mo and Se. Mo concentrations were twice as high as those in rice from Vietnam and Spain. Meanwhile, Se concentrations were three times lower in the whole province compared to the Chinese average level of 0.088 mg/kg. About 12% of the rice samples failed the Chinese national food safety standard of 0.2 mg/kg for Cd. Combined with the multi-elemental profile in rice, the principal component analysis (PCA), discriminant function analysis (DFA) and Fibonacci index analysis (FIA) were applied to discriminate geographical origins of the samples. Results indicated that the FIA method could achieve a more effective geographical origin classification compared with PCA and DFA, due to its efficiency in making the grouping even when the elemental variability was so high that PCA and DFA showed little discriminatory power. Furthermore, some elements were identified as the most powerful indicators of geographical origin: Ca, Ni, Fe and Cd. This suggests that the newly established methodology of FIA based on the ionome profile can be applied to determine the geographical origin of rice.

Application of factor analysis and geostatistics to identify processes controlling the distribution and bioaccessibility of trace elements in Northern Ireland soils

Geologic and environmental factors acting over varying spatial scales can control
trace element distribution and mobility in soils. In turn, the mobility of an element in soil will affect its oral bioaccessibility. Geostatistics, kriging and principal component analysis (PCA) were used to explore factors and spatial ranges of influence over a suite of 8 element oxides, soil organic carbon (SOC), pH, and the trace elements nickel (Ni), vanadium (V) and zinc (Zn). Bioaccessibility testing was carried out previously using the Unified BARGE Method on a sub-set of 91 soil samples from the Northern Ireland Tellus1 soil archive. Initial spatial mapping of total Ni, V and Zn concentrations shows their distributions are correlated spatially with local geologic formations, and prior correlation analyses showed that statistically significant controls were exerted over trace element bioaccessibility by the 8 oxides, SOC and pH. PCA applied to the geochemistry parameters of the bioaccessibility sample set yielded three principal components accounting for 77% of cumulative variance in the data
set. Geostatistical analysis of oxide, trace element, SOC and pH distributions using 6862 sample locations also identified distinct spatial ranges of influence for these variables, concluded to arise from geologic forming processes, weathering processes, and localised soil chemistry factors. Kriging was used to conduct a spatial PCA of Ni, V and Zn distributions which identified two factors comprising the majority of distribution variance. This was spatially accounted for firstly by basalt rock types, with the second component associated with sandstone and limestone in the region. The results suggest trace element bioaccessibility and distribution is controlled by chemical and geologic processes which occur over variable spatial ranges of influence.

Chemometric techniques in multivariate statistical modelling of process plant

The techniques of principal component analysis (PCA) and partial least squares (PLS) are introduced from the point of view of providing a multivariate statistical method for modelling process plants. The advantages and limitations of PCA and PLS are discussed from the perspective of the type of data and problems that might be encountered in this application area. These concepts are exemplified by two case studies dealing first with data from a continuous stirred tank reactor (CSTR) simulation and second a literature source describing a low-density polyethylene (LDPE) reactor simulation.