RESUMO

The quantification of target analytes in complex matrices requires special calibration approaches to compensate for additional capacity or activity in the matrix samples. The standard addition is one of the most important calibration procedures for quantification of analytes in such matrices. However, this technique requires a great number of reagents and material, and it consumes a considerable amount of time throughout the analysis. In this work, a new calibration procedure to analyze biological samples is proposed. The proposed calibration, called the addition calibration technique, was used for the determination of zinc (Zn) in blood serum and erythrocyte samples. The results obtained were compared with those obtained using conventional calibration techniques (standard addition and standard calibration). The proposed addition calibration was validated by recovery tests using blood samples spiked with Zn. The range of recovery for blood serum and erythrocyte samples were 90-132% and 76-112%, respectively. Statistical studies among results obtained by the addition technique and conventional techniques, using a paired two-tailed Student's t-test and linear regression, demonstrated good agreement among them.

Assuntos

RESUMO

An on-line preconcentration system for zinc determination in 24-h urine, blood plasma and erythrocyte matrices by flame atomic absorption spectrometry (FAAS) was used. This procedure was based on adsorption of Zn(II) ions onto a minicolumn filled with silica gel, chemically modified with niobium(V) oxide (Nb(2)O(5)-SiO(2)). The determination of the optimum conditions for Zn(II) preconcentration was done using two-level full factorial and Doehlert designs. In the optimization procedure, four variables (sample pH, eluent concentration, sample flow rate and eluent flow rate) were investigated. The results obtained from the full factorial design demonstrated that the sample pH and sample flow rate variables, and their interactions, were statistically significant. A Doehlert matrix was used in order to determine the optimum conditions for the sample pH and sample flow rate. The optimized conditions for sample pH and flow rate sampling were 6.6 and 7.1 ml min(-1), respectively, to obtain the maximum Zn(II) preconcentration and determination in the biological samples studied. Parameters of analytical curve, precision, effect of other ions in the proposed system and accuracy were achieved to assess the proposed method. The accuracy was confirmed by analysis of certified reference materials (urine Seronorm Trace Elements) and recovery tests in blood plasma and erythrocyte samples. Detection limit (3sigma/S) of 0.77 microg l(-1), precision (calculated as relative standard deviation) of 1.5% for Zn(II) concentration of 10 microg l(-1) (n=7) and a sampling frequency of 27 samples/h were obtained from the proposed system.

RESUMO

Apesar de estarmos numa época de tecnologia e de avanços em estudos e diagnósticos em relação à doença arterial coronária (DAC), a realidade é que tal patologia ainda representa a princiapal causa de mortalidade e morbidade no mundo ocidental. Neste trabalho foi explorada a potencialidade das redes neurais e Lógica Fuzzy no estado de fatores de risco de doenças arteriais coronarianas. Na sua forma mais geral, uma rede neural, é uma máquina que é projetada para modelar a maneira como o cérebro realiza uma tarefa particular ou função de interesse, por meio da sua propriedade de aprender a partir de um conjunto de dados de entrada. Participaram deste estudo 33 indivíduos funcionários da Universidade Federal de Santa Catarina (UFSC) no mês de maio de 2001. A partir de dados clínicos e laboratoriais obtidos desses pacientes (idade, sexo, colesterol total, HDL colesterol, LDL colesterol, triglicerideos, tabagismo, Hipertensão Arterial Sistêmica,Diabetes Melittus, massa corporal, estatura, perímetro da cintura e do abdomen) se calculou o índice de massa corporal, indice cintura quadril, relação colesterol HDL colestetol e também o risco para a doença arterial coronariana. Neste estudo, aplicando a Teoria de Fuzzy Sets (Conjuntos Difusos ou Nebulosos) aos princípios da lógica clássica e resultados de rede neural sobre este conjunto de dados, foi possível determinar um valor percentual pertinente ao grau de risco de doença coronária de cada indivíduo. Concluiu-se que o sistema proposto mostrou maior potencialidade em relação à forma tradicional de avaliação para.

Assuntos