Functional magnetic resonance imaging of swine brain during change in thiopental anesthesia into EEG burst-suppression level--a preliminary study

Deepening anesthesia produces well known changes in electroencephalogram (EEG) and evoked potentials, differing in pathological and normal brain. Yet, it is not known how the T2*-weighted signal changes in the healthy brain during deepening anesthesia. We studied the effect of thiopental bolus on functional magnetic resonance imaging (fMRI) in the healthy brain using porcine model. In five pigs (2–3 months, 20–25 kg), the control bolus prior to fMRI resulted in a change into burst-suppression. After the recovery of continuous EEG, fMRI (4 min) was performed with a single bolus of thiopental (11.4-17.1 mg/kg) administered 1 min after the onset of imaging. This was repeated in four of five pigs. Positive (6-8%) or negative (-3 to -8%) signal intensity changes correlated to the thiopental bolus injection were seen in the group average fMRI response. Positive response was 1.6% and negative response 2.3% of the total brain region of interest (ROI) voxels. Responding voxels were distributed more prominently in the thalamic ROI (4.5%) than in the cortical ROI (2.2%). The group average of unthresholded voxel time courses showed that the net effect of thiopental bolus was a small (0.5%) but a clear positive change in the thalamic region, while variance changed in the global level. In conclusion, this study is the first to show that significant signal intensity changes occur in fMRI response during the sudden deepening of thiopental anesthesia. However, these responses are neither anatomically constant nor global in the healthy swine brain.     

Hydroperoxide formation during autoxidation of conjugated linoleic acid methyl ester

The aim of this study was to investigate whether hydroperoxides are formed in the autoxidation of conjugated linoleic acid (CLA) methyl ester both in the presence and absence of α‐tocopherol. The existence of hydroperoxide protons was confirmed by D₂O exchange and by chemoselective reduction of the hydroperoxide groups into hydroxyl groups using NaBH₄. These experiments were followed by nuclear magnetic resonance (NMR) spectroscopy. The ¹³C and ¹HNMR spectra of a mixture of 9‐hydroper‐oxy‐10‐trans,12‐cis‐octadecadienoic acid methyl ester (9‐OOH) and 13‐hydroperoxy‐9‐cis, 11‐trans‐octadecadienoic acid methyl ester (13‐OOH), which are formed during the autoxidation of methyl linoleate, were studied in detail to allow the comparison between the two linoleate hydroperoxides and the CLA methyl ester hydroperoxides. The ¹³CNMR spectra of samples enriched with one of the two linoleate hydroperoxide isomers were assigned using 2D NMR techniques, namely Correlated Spectroscopy (COSY), gradient Heteronuclear Multiple Bond Correlation (gHMBC), and gradient Heteronuclear Single Quantum Correlation (gHSQC). The ¹³C and ¹H NMR experiments performed in this study show that hydroperoxides are formed during the autoxidation of CLA methyl ester both in the presence and absence of α‐tocopherol and that the major isomers of CLA methyl ester hydroperoxides have a conjugated monohydroperoxydiene structure similar to that in linoleate hydroperoxides.     

Use of Neural Networks in the Simulation and Optimization of Pressure Swing Adsorption Processes

In this work a method for the simulation and optimization of a pressure swing adsorption process for the separation of nitrogen from air by using neural networks was developed. The model is used to obtain a prediction for the process performance, namely, the specific product and yield, over a wide range of operating conditions. These results are compared with the predictions from a mass tranfer model, and a very good agreement is found. The network developed is also used to minimize a cost objective function, and it is shown that it can easily be used in process optimization and/or control.     

Endotoxins associated with cyanobacteria and their removal during drinking water treatment

The aim of this study was to investigate endotoxin concentrations in cyanobacterial water blooms and strains, and to assess the removal of endotoxins during drinking water treatment. Endotoxin concentrations were measured from 151 hepatotoxic, neurotoxic and non-toxic cyanobacterial water blooms by using Limulus amebocyte lysate (LAL) assay, and the results were compared to bacterial data. Endotoxin activities ranged from 20 to 3.8 x 10(4) endotoxin units (EU) per ml. Endotoxicity of the samples correlated with phycobiliprotein concentration that was used to assess cyanobacterial abundance, heterotrophic plate count, and Aeromonas spp. but it did not correlate with the number of coliforms or streptococci. The high endotoxin concentrations occasionally detected in the water bloom samples were probably due to Gram negative bacteria that existed together with cyanobacteria since the 26 axenic cyanobacterial strains from different genera that were studied showed very low endotoxin activity. No differences in endotoxin activity were detected between hepatotoxic, neurotoxic and non-toxic strains. Removal of endotoxins during drinking water treatment was studied at nine waterworks that previously had been associated with high numbers of cyanobacteria and that used different processes for water purification. Endotoxin concentration in raw waters ranged from 18 to 356 EU ml(-1). The treatment processes reduced 59-97% of the endotoxin activity; in the treated water the concentration ranged from 3 to 15 EU ml(-1). The most significant reduction occurred at the early stages of water treatment, during coagulation, settling and sand filtration. Activated carbon filtration either increased or had no effect on endotoxin concentration. Ozonation and chlorination had little effect on the endotoxin concentrations.     

Optimization of CCT Equations Using Calculated Grain Boundary Soluble Compositions for the Simulation of Austenite Decomposition of Steels

Metallurgical and Materials Transactions B - New CCT equations have been developed and optimized to simulate the start temperatures of the austenite decomposition process in low-alloyed steels...     

Exopolysaccharides produced by probiotic strains modify the adhesion of probiotics and enteropathogens to human intestinal mucus

Exopolysaccharides (EPSs) are exocellular polymers present in the surface of many bacteria, including Lactobacillus and Bifidobacterium. The genome sequence of several strains revealed the presence of EPS-encoding genes. However, the physiological role that EPSs play in the bacterial ecology still remains uncertain. In this study, we have assessed the effect of EPSs produced by Lactobacillus rhamnosus GG, Bifidobacterium longum NB667, and Bifidobacterium animalis IPLA-R1 on the adhesion of probiotic and enteropathogen strains to human intestinal mucus. The EPS fraction GG had no significant effect on the adhesion of L. rhamnosus GG and B. animalis IPLA-R1. However, the EPS fractions NB667 and IPLA-R1 significantly reduced the adherence of both probiotic strains. In contrast, the three EPS fractions increased the adhesion of Enterobacter sakazakii ATCC 29544 and Escherichia coli NCTC 8603. Higher adherence of Salmonella enterica serovar Typhimurium ATCC 29631 and Clostridium difficile ATCC 9689 was detected in the presence of the EPS fractions GG and NB667. In general, these effects were obtained at EPS concentrations of up to 5 mg/ml, and they were EPS dose dependent. The competitive exclusion of probiotics in the presence of EPS could suggest the involvement of these biopolymers in the adhesion to mucus. The increase in the adherence of enteropathogens could be explained if components of the pathogen surface are able to bind to specific EPSs and the bound EPSs are able to adhere to mucus. To the best of our knowledge, this is the first work reporting the effect of EPSs from probiotics on bacterial adhesion properties.     

Adhesion of selected Bifidobacterium strains to human intestinal mucus and the role of adhesion in enteropathogen exclusion

The ability of potential probiotic strains to adhere to the intestinal mucosa and exclude and displace pathogens is of utmost importance for therapeutic manipulation of the enteric microbiota. The ability of seven selected human bifidobacterial strains and five human enteropathogenic strains to adhere to human intestinal mucus was analyzed and compared with that of four strains isolated from chicken intestines. The adhesion of the bifidobacterial strains ranged from 3 to 16% depending on the strain. Bifidobacterium strains of animal origin adhered significantly better than did strains of human origin. Of the pathogenic bacteria, Escherichia coli NCTC 8603 had the highest adhesion value (20%), Salmonella Typhimurium ATCC 29631, Enterobacter sakazakii ATCC 29544, and Clostridium difficile ATCC 9689 had adhesion values ranging from 10 to 15%, and Listeria monocytogenes ATCC 15313 had the lowest adhesive value (3%). The ability of these bifidobacteria to inhibit pathogen adhesion and to displace pathogens previously adhering to mucus was also tested. The inhibition of pathogens adhesion by these bifidobacterial strains was variable and clearly strain dependent. In general, bifidobacterial strains of animal origin were better able to inhibit and displace pathogens than were human strains. Preliminary characterization of bacterial adhesion was accomplished using different pretreatments to explore adhesion mechanisms. The results indicate that different molecules are implicated in the adhesion of bifidobacteria to the human intestinal mucus, constituting a multifactorial process.     

So near and yet so far: New insight into properties of some well-known classifier paradigms

This article provides some new insight into the properties of four well-established classifier paradigms, namely support vector machines (SVM), classifiers based on mixture density models (CMM), fuzzy classifiers (FCL), and radial basis function neural networks (RBF). It will be shown that these classifiers can be formulated in a way such that they are functionally equivalent or at least highly similar. The interpretation of a specific classifier as being an SVM, CMM, FCL, or RBF then only depends on the objective function and the optimization algorithm used to adjust the parameters. The properties of these four paradigms, however, are very different: a discriminative classifier such as an SVM is expected to have optimal generalization capabilities on new data, a generative classifier such as a CMM also aims at modeling the processes from which the observed data originate, and a comprehensible classifier such as an FCL is intended to be parameterized and understood by human domain experts. We will discuss the advantages and disadvantages of these properties and show how they can be measured numerically in order to compare these classifiers. In such a way, the article aims at supporting a practitioner in assessing the properties of classifier paradigms and in selecting or combining certain paradigms for a given application problem.