Study Of Reducing Temperature Effect On Gold Immunochromotographic Assay Strip Quantitative Detection Based On Artificial Neural Network

Recently, gold immunochromotograhic assay strip (GICAS) has been wildly applied in clinical diagnosis and has become one of developmental biochemical test in the 21^st century because it is effective, simple, contamination-free, stable and adapted to single sample detection. Therefore, GICAS quantitative analysis has gained great attention. It has significant and more practical value to enhancing accuracy of GICAS quantitative analysis. In order to improve the accuracy quantitative analysis of GICAS, temperature effect on its accuracy was analyzed and studied deeply, and a resolution was put forward. The main research work in this paper is as follows: 1) Current development of GICAS and application of intelligence technology to GICAS was presented. 2) The principle of GICA strip and its representative quantitative detection system were introduced. Effect factors for GICAS were analyzed in detail, especially temperature effect on the GICAS quantitative analysis. Relative experiments showed that study on temperature effects on the accuracy of quantitative analysis was essential. 3) BP neural network ,which was widely used, and the improved BP algorithm were introduced briefly. In order to reduce temperature effect and improve accuracy of GICAS, a method of three-dimensional working curves based on artificial neural network was suggested. The reseach results indicated that it was much better to use three-dimensional working curves than two-dimensional curves. 4) A method by the control of temperature in order to reduce temperature effect in practice was put forward. According to the requirement of the GICA strip quantitative detection system, the fuzzy neural networks PID is used by combining Fuzzy logic and artificial neural network. The simulation results indicated the method would obtain good results. Based on this method, a constant temperature realization design which based on DSP chip and used thermo-electric module was presented in the paper. Obviously, this study will help the improvement of GICAS study and instrumentdesign.