Smart-phone based electrocardiogram wavelet decomposition and neural network classification

This paper discusses ECG classification after parametrizing the ECG waveforms in the wavelet domain. The aim of the work is to develop an accurate classification algorithm that can be used to diagnose cardiac beat abnormalities detected using a mobile platform such as smart-phones. Continuous time recurrent neural network classifiers are considered for this task. Records from the European ST-T Database are decomposed in the wavelet domain using discrete wavelet transform (DWT) filter banks and the resulting DWT coefficients are filtered and used as inputs for training the neural network classifier. Advantages of the proposed methodology are the reduced memory requirement for the signals which is of relevance to mobile applications as well as an improvement in the ability of the neural network in its generalization ability due to the more parsimonious representation of the signal to its inputs.

On the utility of input selection and pruning for financial distress prediction models

Analyzes the use of linear and neural network models for financial distress classification, with emphasis on the issues of input variable selection and model pruning. A data-driven method for selecting input variables (financial ratios, in this case) is proposed. A case study involving 60 British firms in the period 1997-2000 is used for illustration. It is shown that the use of the Optimal Brain Damage pruning technique can considerably improve the generalization ability of a neural model. Moreover, the set of financial ratios obtained with the proposed selection procedure is shown to be an appropriate alternative to the ratios usually employed by practitioners.

A kernel-based two-class classifier for imbalanced data sets

Many kernel classifier construction algorithms adopt classification accuracy as performance metrics in model evaluation. Moreover, equal weighting is often applied to each data sample in parameter estimation. These modeling practices often become problematic if the data sets are imbalanced. We present a kernel classifier construction algorithm using orthogonal forward selection (OFS) in order to optimize the model generalization for imbalanced two-class data sets. This kernel classifier identification algorithm is based on a new regularized orthogonal weighted least squares (ROWLS) estimator and the model selection criterion of maximal leave-one-out area under curve (LOO-AUC) of the receiver operating characteristics (ROCs). It is shown that, owing to the orthogonalization procedure, the LOO-AUC can be calculated via an analytic formula based on the new regularized orthogonal weighted least squares parameter estimator, without actually splitting the estimation data set. The proposed algorithm can achieve minimal computational expense via a set of forward recursive updating formula in searching model terms with maximal incremental LOO-AUC value. Numerical examples are used to demonstrate the efficacy of the algorithm.

A fast linear-in-the-parameters classifier construction algorithm using orthogonal forward selection to minimize leave-one-out misclassification rate

We propose a simple and computationally efficient construction algorithm for two class linear-in-the-parameters classifiers. In order to optimize model generalization, a forward orthogonal selection (OFS) procedure is used for minimizing the leave-one-out (LOO) misclassification rate directly. An analytic formula and a set of forward recursive updating formula of the LOO misclassification rate are developed and applied in the proposed algorithm. Numerical examples are used to demonstrate that the proposed algorithm is an excellent alternative approach to construct sparse two class classifiers in terms of performance and computational efficiency.

An LDA and probability-based classifier for the diagnosis of Alzheimer's Disease from structural MRI

In this paper a custom classification algorithm based on linear discriminant analysis and probability-based weights is implemented and applied to the hippocampus measurements of structural magnetic resonance images from healthy subjects and Alzheimer’s Disease sufferers; and then attempts to diagnose them as accurately as possible. The classifier works by classifying each measurement of a hippocampal volume as healthy controlsized or Alzheimer’s Disease-sized, these new features are then weighted and used to classify the subject as a healthy control or suffering from Alzheimer’s Disease. The preliminary results obtained reach an accuracy of 85.8% and this is a similar accuracy to state-of-the-art methods such as a Naive Bayes classifier and a Support Vector Machine. An advantage of the method proposed in this paper over the aforementioned state of the art classifiers is the descriptive ability of the classifications it produces. The descriptive model can be of great help to aid a doctor in the diagnosis of Alzheimer’s Disease, or even further the understand of how Alzheimer’s Disease affects the hippocampus.

Wavelets and the generalization of the variogram

The experimental variogram computed in the usual way by the method of moments and the Haar wavelet transform are similar in that they filter data and yield informative summaries that may be interpreted. The variogram filters out constant values; wavelets can filter variation at several spatial scales and thereby provide a richer repertoire for analysis and demand no assumptions other than that of finite variance. This paper compares the two functions, identifying that part of the Haar wavelet transform that gives it its advantages. It goes on to show that the generalized variogram of order k=1, 2, and 3 filters linear, quadratic, and cubic polynomials from the data, respectively, which correspond with more complex wavelets in Daubechies's family. The additional filter coefficients of the latter can reveal features of the data that are not evident in its usual form. Three examples in which data recorded at regular intervals on transects are analyzed illustrate the extended form of the variogram. The apparent periodicity of gilgais in Australia seems to be accentuated as filter coefficients are added, but otherwise the analysis provides no new insight. Analysis of hyerpsectral data with a strong linear trend showed that the wavelet-based variograms filtered it out. Adding filter coefficients in the analysis of the topsoil across the Jurassic scarplands of England changed the upper bound of the variogram; it then resembled the within-class variogram computed by the method of moments. To elucidate these results, we simulated several series of data to represent a random process with values fluctuating about a mean, data with long-range linear trend, data with local trend, and data with stepped transitions. The results suggest that the wavelet variogram can filter out the effects of long-range trend, but not local trend, and of transitions from one class to another, as across boundaries.

Independent evaluation of the ability of spaceborne radar and lidar to retrieve the microphysical and radiative properties of ice clouds

The combination of radar and lidar in space offers the unique potential to retrieve vertical profiles of ice water content and particle size globally, and two algorithms developed recently claim to have overcome the principal difficulty with this approach-that of correcting the lidar signal for extinction. In this paper "blind tests" of these algorithms are carried out, using realistic 94-GHz radar and 355-nm lidar backscatter profiles simulated from aircraft-measured size spectra, and including the effects of molecular scattering, multiple scattering, and instrument noise. Radiation calculations are performed on the true and retrieved microphysical profiles to estimate the accuracy with which radiative flux profiles could be inferred remotely. It is found that the visible extinction profile can be retrieved independent of assumptions on the nature of the size distribution, the habit of the particles, the mean extinction-to-backscatter ratio, or errors in instrument calibration. Local errors in retrieved extinction can occur in proportion to local fluctuations in the extinction-to-backscatter ratio, but down to 400 m above the height of the lowest lidar return, optical depth is typically retrieved to better than 0.2. Retrieval uncertainties are greater at the far end of the profile, and errors in total optical depth can exceed 1, which changes the shortwave radiative effect of the cloud by around 20%. Longwave fluxes are much less sensitive to errors in total optical depth, and may generally be calculated to better than 2 W m(-2) throughout the profile. It is important for retrieval algorithms to account for the effects of lidar multiple scattering, because if this is neglected, then optical depth is underestimated by approximately 35%, resulting in cloud radiative effects being underestimated by around 30% in the shortwave and 15% in the longwave. Unlike the extinction coefficient, the inferred ice water content and particle size can vary by 30%, depending on the assumed mass-size relationship (a problem common to all remote retrieval algorithms). However, radiative fluxes are almost completely determined by the extinction profile, and if this is correct, then errors in these other parameters have only a small effect in the shortwave (around 6%, compared to that of clear sky) and a negligible effect in the longwave.

Foods derived from animals: the impact of animal nutrition on their nutritive value and ability to sustain long-term health

Foods derived from domestic animals are a significant source of nutrients in the UK diet. However, certain aspects of some animal-derived foods, notably levels of saturated fatty acids, have given rise to concerns that these foods may contribute to the risk of cardiovascular disease, the metabolic syndrome and other conditions. However, the composition of the many animal-derived foods is not constant and can often be enhanced by manipulating the nutrition of the animal. This paper reviews these possibilities with particular attention to lipids, and draws attention to the fact that milk in particular, contains a number of compounds which may, for example, exert anti-carcinogenic effects. It is clear that the role of animal nutrition in creating foods closer to the optimum composition for long-term human health will not only be more relevant in the future, but will be vital in attempts to improve the health of the human population.

Successful long-term ultra dry storage of seed of 15 species of Brassicaceae in a genebank: variation in ability to germinate over 40 years and dormancy

Seed of 15 species of Brassicaceae were stored hermetically in a genebank (at -5 degrees C to -10 degrees C with c. 3% moisture content) for 40 years. Samples were withdrawn at intervals for germination tests. Many accessions showed an increase in ability to germinate over this period. due to loss in dormancy. Nevertheless, some dormancy remained after 40 years' storage and was broken by pre-applied gibberellic acid. The poorest seed survival occurred in Hormatophylla spinosa. Even in this accession the ability to germinate declined by only 7% between 1966 and 2006. Comparison of seeds from 1966 stored for 40 years with those collected anew in 2006 from the original sampling sites, where possible, showed few differences, other than a tendency (7 of 9 accessions) for the latter to show greater dormancy. These results for hermetic storage at sub-zero temperatures and low moisture contents confirm that long-term seed storage can provide a successful technology for ex situ plant biodiversity conservation.

Common variants of apolipoprotein A-IV differ in their ability to inhibit low density lipoprotein oxidation

Apolipoprotein A-IV (apoA-IV) inhibits lipid peroxidation, thus demonstrating potential anti-atherogenic properties. The aim of this study was to investigate how the inhibition of low density lipoprotein (LDL) oxidation was influenced by common apoA-IV isoforms. Recombinant wild type apoA-IV (100 mu g/ml) significantly inhibited the oxidation of LDL (50 mu g protein/ml) by 5 mu M CuSO4 (P < 0.005), but not by 100 mu M CuSO4, suggesting that it may act by binding copper ions. ApoA-IV also inhibited the oxidation of LDL by the water-soluble free-radical generator 2,2'-azobis(amidinopropane) dihydrochloride (AAPH; I mM), as shown by the two-fold increase in the time for half maximal conjugated diene formation (T-1/2; P < 0.05) suggesting it can also scavenge free radicals in the aqueous phase. Compared to wild type apoA-IV, apoA-IV-S347 decreased T-1/2 by 15% (P = 0.036) and apoA-IV-H360 increased T-1/2 by 18% (P = 0.046). All apoA-IV isoforms increased the relative electrophoretic mobility of native LDL, suggesting apoA-IV can bind to LDL and acts as a site-specific antioxidant. The reduced inhibition of LDL oxidation by apoA-IV-S347 compared to wild type apoA-IV may account for the previous association of the APOA4 S347 variant with increased CHD risk and oxidative stress. (c) 2006 Elsevier Ireland Ltd. All rights reserved.

Analysing the ability to retain sidechain hydrogen-bonds in mutant proteins

Motivation: Hydrogen bonds are one of the most important inter-atomic interactions in biology. Previous experimental, theoretical and bioinformatics analyses have shown that the hydrogen bonding potential of amino acids is generally satisfied and that buried unsatisfied hydrogen-bond-capable residues are destabilizing. When studying mutant proteins, or introducing mutations to residues involved in hydrogen bonding, one needs to know whether a hydrogen bond can be maintained. Our aim, therefore, was to develop a rapid method to evaluate whether a sidechain can form a hydrogen-bond. Results: A novel knowledge-based approach was developed in which the conformations accessible to the residues involved are taken into account. Residues involved in hydrogen bonds in a set of high resolution crystal structures were analyzed and this analysis is then applied to a given protein. The program was applied to assess mutations in the tumour-suppressor protein, p53. This raised the number of distinct mutations identified as disrupting sidechain-sidechain hydrogen bonding from 181 in our previous analysis to 202 in this analysis.

Variation in the ability of human influenza A viruses to induce and inhibit the IFN-beta pathway

We investigated the ability of a selection of human influenza A viruses, including recent clinical isolates, to induce IFN-beta production in cultured cell lines. In contrast to the well-characterized laboratory strain A/PR/8/34, several, but not all, recent isolates of H3N2 viruses resulted in moderate IFN-beta stimulation. Through the generation of recombinant viruses, we were able to show that this is not due to a loss of the ability of the NS1 genes to suppress IFN-beta induction; indeed, the NS1 genes behaved similarly with respect to their abilities to block dsRNA signaling. Interestingly, replication of A/Sydney/5/97 virus was less Susceptible to pre-treatment with IFN-alpha than the other viruses. In contrast to the universal effect on dsRNA signaling, we noted differences in the effect of NS1 proteins on expression of interferon stimulated genes and also genes induced by a distinct pathway. The majority of NS1 proteins blocked expression From both IFN-dependent and TNF-dependent promoters by an apparent post-transcriptional mechanism. The NS1 gene of A/PR/8/34 NS1 did not confer these blocks. We noted striking differences in the Cellular localization of different influenza A virus NS1 proteins during infection, which might explain differences in biological activity. (C) 2005 Elsevier Inc. All rights reserved.

The glycosylation pattern of baculovirus expressed envelope protein E2 affects its ability to prevent infection with bovine viral diarrhoea virus

We have investigated the role of glycosylation of the envelope glycoprotein E2 of bovine viral diarrhoea virus (BVDV), produced in insect cells, in BVDV infection. When amino acids predicated to code for the C-terminal N-linked glycosylation site were mutated the resulting protein was less efficient than wild type protein at preventing infection of susceptible cells with BVDV. In addition, mutational analysis showed that a further two predicted N-terminal N-linked glycosylation sites of E2 are required for efficient production of recombinant protein. (c) 2005 Elsevier B.V. All rights reserved.