The Study Of Fluorescence Immunochromatographic Quantitative Detection System Based On Artificial Neural Network

Fluorescence immunochromatographic quantitative detection is a commonly used and effective method of quantitative concentration detecting. Combining the principle of immunoassay and the technique of fluorescence analysis, It is characterized by convenient testing process, good sensitivity and high stability. Therefore, to design and implement a quantitative detection system based on the fluorescence immunochromatographic technique will have broad application prospects in biochemical detection and medical diagnosis and other related fields.In order to realize the complete process of fluorescence immunochromatographic quantitative detection for practical application, a quantitative detection system needs to integrate various functioning modules, mainly including motion controlling, signal conditioning and sampling, fluorescence quantitative information processing, curve fitting, test time management, test id input and results database. Therefore, in this paper, based on a summary of the project background and research status, the specific content is divided into the following sections:(1) According to the principle of fluorescence immunochromatographic quantitative detection technology, select the best matching excitation light source and photoelectric sensor based on the fluorescence spectra of test samples. Then design and simulate optical model, building up practical optical measuring system, develop the fluorescence signal conditioning and sampling circuit. Finally realize the detection sub module of fluorescence information and verifying it.(2) Complete the hardware and software design and debugging of the main circuit of fluorescence immunochromatographic quantitative detection system. The main circuit adopts C8051F020 microcontroller as the core. Hardware design includes overall design and detailed design of circuit modules such as power supply, serial port expansion module; System is equipped with the SMALL-RTOS51 operating system. Software design including driven design, overall design, detailed design and the underlying file system according to the work characteristics of FLASH memory.(3) According to the characteristics of the fluorescence signal, achieve effective denoising and peak extraction through the signal preprocessing and feature extraction. On the basis of this, establish a calibration curve to implement quantitative detection process by the principle of BP neural network. With reference to foreign similar detection instrument, do comparative experiments to verify our system and test the limit of detection, detection reproducibility and quantitative accuracy.The experiment results show that the fluorescence immunochromatographic quantitative detection system designed in this paper can realize quantitative detection. The system can measure the concentration in the range of 200ng/mL-15000ng/mL. Repeatability error is less than 1.5%. Quantitative detection error is less than 2%. Since it provides theoretical basis and experimental analysis for the actual application of fluorescence immunochromatographic technique, this quantitative detection system has important clinical value and practical significance.