ALESSANDRA NOGUEIRA SANTOS,ALEXANDRE CARLOS BRANDÃO RAMOS,DEMÉTRIO ARTUR WERNER SOARES,EDSON GIULIANI RAMOS FERNANDES,JOSÉ RENATO GARCIA BRAGA,NIRTON CRISTI SILVA VIEIRA,ÁLVARO ANTONIO ALENCAR DE QUEI
申请号：
BRPI1016277
公开号：
BRPI1016277B1
申请日:
2010.12.09
申请国别(地区):
BR
年份:
2018
代理人:
摘要：
The device biossensor integrated with artificial neural network for simultaneous monitoring of hemometabólitos.The invention describes the development process of a system bioeletroquímico multienzimático controlled and automated by an artificial neural network for the simultaneous determination of hemometabóLitos cholesterol, urea and glucose.Dendrímeros of polyglycerol.Polyglycerol arborescent or polyglycerol with a high density of branch (pgld) were used as a platform nanoestruturada to obtain a bioconjugado multienzimático with propriedaDes hemometabólitos sensing of glucose, cholesterol and urea.Bioeletroquímicos electrodes were prepared after immobilization of the enzymes of cholesterol oxidase (CoOx)Glucose oxidase (GOx) and urease (UR) in dendrímeros of polyglycerol of generation 2 (pgldg) using carbodiimide derivative.The applicability of the systems monoenzimáticos pgldg2 - CoOx, pgldg2 - GOx, pgldg2 - ur and multienzimático or pgldg2 - CoOx \/ GOx \/ ur in quantitative determination, and at the same time, hemometabólitos cholesterol.Glucose, urea was tested in in vitro conditions that simulate physiological conditions of the human body.An artificial neural network (RNA) was used as the main tool for data classification and pattern recognition in the sign of the biossensor for the prediction of electrical behavior ofThe device multienzimático.A neural network type multilayer perceptron was used.And the results showed that the application of the algorithm to learning of retropropagação of érro (backpropagation) led to a high predictive quality.Since 95% of the coefficients of determination R2 were above 0.98.The use of neural network led to a precise adjustment of the curves of electrical signal of the device biossensor compared to polynomial regression technique due to the característicThe nonlinear of these curves.The results obtained with the biossensor nanoestruturado controlled and alitomatizado by artificial neural network showed an excellent efficienc
