Robustness of classification rules that incorporate additional information

The discrimination problem for two normal populations with the same covariance matrix when additional information on the population is available is considered. A study of the robustness properties against training sample contamination of classification rules that incorporate this additional information is performed. These rules have received recently attention where their total misclassification probability (TMP) is proved to be lower than Fisher's linear discriminant rule. The results of a simulation study on the TMP which compares the behaviour of the new rules against Fisher's rule and some of its robustified versions under different types of contamination are presented. These results show that the rules that incorporate the additional information not only have lower TMP, but they also prevent against some types of contamination. In order to achieve prevention from all types of contamination a robustifed version of these rules is recommended.     

Evaluation of the completion of influenza vaccination

OBJECTIVE: To find the reasons which determine failures to comply with anti-flu vaccinations, so that these can be corrected and the coverage of this preventive action be increased. DESIGN: Observational crossover study, done by means of a telephone survey of people over 65. A questionnaire with closed questions, composed after a pilot study and validated by Cronbach's alpha. SETTING: Primary Care Centre (PCC). PATIENTS: We calculated a population sample for qualitative variables (_ = 0.05; p = 0.60; e = 0.05) of 294 people over 65, chosen from the PCC records, by means of random sampling (K = 4) stratified for age and discounting the telephone selection bias. MEASUREMENTS AND RESULTS: The proportion of vaccinated patients (60.9%) obtained in our study did not significantly differ from that in the general population. The percentage of patients included in the programme for the first time was 14%. Level of satisfaction among those vaccinated was 89.4%, with 8.9% of problems detected being light. Main causes of non-vaccination were: thinking that they didn't need it (63.5%), ignorance of the campaign (35.7%), fear of the reaction (24.3%), forgetting (10.4%). The main form of access to the campaign information was from the PCC, both through individuals and posters. Lack of information was statistically significant (p < 0.00001) as a determinant of non-vaccination, without other factors (age, sex, associated pathologies...) explaining these differences. CONCLUSIONS: Individualised and on-going health education by the PCC is fundamental. This would enable the identification of the group not vaccinated due to their express refusal and the recovery of non-vaccinated patients.     

Improving the testability of switched-capacitor filters

A design-for-test methodology for SC filters is presented, based on an architecture using some additional circuitry and providing extra capabilities for both off- and on-line tests. The approach uses a comparison (voting) mechanism to indicate whether or not two copies of a filter element (a biquad, for instance) have a similar response during their actual operation. The design and implementation of a few filter examples are included to assess the potential usefulness of this new approach.     

Methamphetamine Blocks Adenosine A2A Receptor Activation via Sigma 1 and Cannabinoid CB1 Receptors

Methamphetamine is, worldwide, one of the most consumed drugs of abuse. One important side effect is neurodegeneration leading to a decrease in life expectancy. The aim of this paper was to check whether the drug affects one of the receptors involved in neurodegeneration/neuroprotection events, namely the adenosine A_2A receptor (A_2AR). First, we noticed that methamphetamine does not affect A_2A functionality if the receptor is expressed in a heterologous system. However, A_2AR becomes sensitive to the drug upon complexes formation with the cannabinoid CB₁ receptor (CB₁R) and the sigma 1 receptor (σ₁R). Signaling via both adenosine A_2AR and cannabinoid CB₁R was affected by methamphetamine in cells co-expressing the two receptors. In striatal primary cultures, the A_2AR–CB₁R heteromer complex was detected and methamphetamine not only altered its expression but completely blocked the A_2AR- and the CB₁R-mediated activation of the mitogen activated protein kinase (MAPK) pathway. In conclusion, methamphetamine, with the participation of σ₁R, alters the expression and function of two interacting receptors, A_2AR, which is a therapeutic target for neuroprotection, and CB₁R, which is the most abundant G protein-coupled receptor (GPCR) in the brain.     

Numerical characterization and experimental verification of an in-plane MEMS-actuator with thin-film aluminum heater

In this paper, a novel concept of a thermo-mechanical MEMS actuator using aluminum thin-film heaters on a thermal oxide for electrical insulation is presented. The actuator is part of an universal tensile testing platform for thermo-mechanical material characterization of one dimensional materials on a micro- and nano-scopic scale under different environmental conditions, as varying temperatures, pressure, moisture or even vacuum and is realised in BDRIE technology. It is shown, that the actuator concept fulfills the requirements for the use in a tensile loading stage along with heterogeneously integrated nanofunctional elements, following a specimen centered approach in line with bottom-up self-assembly processes. Simulation and experiment agree very well in the thermal and mechanical domain and allow subsequent optimisation of the actuator performance.     

Effect of ethanol administration on fatty acid desaturation

The effect of ethanol on the fatty acid desaturation by rat liver has been studied using liquid diets of different composition. Acute ethanol administration increased triacylglycerols of total liver lipids, but did not modify significantly the lipidic composition of microsomes. The Δ6 and Δ5 desaturases were inhibited by ethanol whereas the Δ9 desaturase and fatty acid synthetase were apparently modified only by diet composition. NADH-cytochrome (cyt.) c reductase was partially inhibited, whereas NADH-cyt. b₅ reductase remained practically unaltered and NADPH-cyt. c reductase activity was enhanced. Decreased electrons supplied by the microsomal cyt. b₅ electron transport chain would not be the reason for the inhibition of Δ6 and Δ5 desaturases by ethanol.     

Effect of cold environment on hepatic microsomal Δ6 and Δ9 desaturase activity of male rats

Male rats maintained at 24 C and then shifted to 5 C for 5 days increased food intake and decreased in growth rate and food conversion. No modification was observed in Δ6 desaturase activity, while Δ9 desaturase activity decreased after this period of time. These results were confirmed by liver microsomal and mitochondrial fatty acid composition. The phospholipid composition of liver microsomes was unaltered, whereas in mitochondria, phosphatidylcholine and sphingomyelin decreased and phosphatidylethanolamine increased due to the cold environment. The influence of food intake and weight changes on fatty acid metabolism was studied using (i) rats maintained at 5 C with restricted food intake to match the food intake of those kept at 24 C with food ad libitum and (ii) rats maintained at 24 C whose food intake was also restricted so that their growth rate would be the same as that of rats maintained at 5 C with food ad libitum, respectively. These results indicate that the negative metabolic balance state of these cold conditions is not an active factor modifier of Δ6 desaturase activity, whereas it decreases Δ9 desaturase activity, reflecting the lipogenic characteristics of the latter enzyme.     

NEURAL-NETWORK-AIDED DESIGN OF AUTOMOBILE EXHAUST CATALYSTS

A priori design of catalysts is not yet possible. Such task would demand unavailable scientific knowledge of the correlations among synthesis parameters and resulting solid state and surface structures, on the one hand, and among those atomic-level structural details and their catalytic functions, on the other hand. To avoid testing every possible combination, therefore, the applied chemist or chemical engineer must identify empirical correlations underlying the existing experimental data base.

The ability of artificial neural networks to identify complex correlations and to predict the result of experiments has recently generated considerable interest in various areas of science and engineering. In this paper, neural networks are used to identify composition-performance relationships in automobile exhaust catalysts.

This work employs an artificial neural network technique to do a sensitivity analysis of the conversions of pollutant gases as a function of the catalyst composition and the operating conditions. This approach converges on the optimum catalyst composition and operating condition in order to produce specified conversions of carbon monoxide, hydrocarbons and nitrogen oxides, to carbon dioxide, water and di-nitrogen respectively.