Information Extraction And Quantitative Analysis Of Positive Signals In Tomographic Chip

The fluorescent signal detection of chromatography chip is the key of the disease diagnosis. Immunochromatography technology is one of most appealing fast detection techniques in 1990s. Its basic principle is the interactions between antigen and antibody. The detection of fluorescent signal in chromatography chip is an important task of diagnosing disease because the fluorescent information presents the degree of interaction between antigen and antibody. The disease can be detected through the fluorescent information. However, nowadays, most immune chromatography experimental results are qualitative instead of quantitative, which restricts the application of chromatographic chip.Toward this issue, the paper proposes two algorithms to auto-extract the fluorescent information in the chromatography chip and quantitative analysis is taken to detect diseases. The main work and innovation in this paper are: (1)Aiming at two types of immune chromatography chip image– collaurum chip image and chip image marked with quantum dots, two algorithms are proposed. (2) Realize the quantitative analysis on fluorescent information.A new algorithm is introduced for the automatic identification of fluorescent signal which applies to the collaurum chip image. Based on the features of chromatographic chips, mathematic morphology in RGB color space is used to filter and enhance the images, then pyramid connection is used to segment the areas of fluorescent signal. After that, Gaussian Mixture Model is leveraged for detecting the fluorescent signal. At last, the average fluorescent intensity is calculated from fluorescent areas. It proved that the algorithm has a good effect on segmenting the fluorescent areas and it is able to detect the fluorescent signal quickly and accurately to achieve the quantitative detection of chromatographic chip through experimental data analysis.For immunochromatography chip image marked with quantum dots, we firstly use SMQT grayscale image enhancement algorithm to process gray image in the HSI space whose fluorescent information is very fuzzy. The enhancement method can overcome the disadvantage of traditional enhancement algorithm and retain the details of the image information. It can improve the identification of fuzzy image information effectively. Then we use statistical iteration to segment the image to acquire the fluorescent information based on its histogram and scattergram. At last, the two algorithms are successfully applied into the image processing model of the immunochromatography analyzer system. The experimental results show that the proposed algorithms have a significant improvement in both accuracy and automation of image extraction. Its prospect is promising.