Application Of Group Testing Method In Biomedicine

To strengthen the detection and control of infectious diseases is an important task of the health departments in all countries. For some infectious diseases, such as AIDS, syphilis, hepatitis B, and etc, If the infected patients can be detected early and the source of infection can be controlled, we will be able to prevent the spread of diseases in a wide range.How to detect the samples efficiently, accurately and economically is a problem worthy of study. We usually use the blood or body fluid samples to do the test. For a large number of samples to be tested, it is a very economical and convenient method to detect the samples by the method of grouping. Because of its low detection limit, high specificity, simple operation, small sample size and low cost, packet detection method is widely used. This paper briefly introduces the background, the basic concepts and methods of the grouping detection method, and analyzes the impact of various factors on the results of the packet detection, such as the dilution effect, the variation of sample number in the group, the existence of detection errors and etc, when the blood samples are divided into groups to do the test. Because of the high cost of gene detection, grouping detection method is also widely used in the field of genetic testing.The main content of this paper is divided into two parts. First, it assumes that the overall result only relevant to the individual state, the biomarker concentration of the pool is the arithmetic average of the biomarker concentrations of the individual specimens. Secondly, the estimation of the parameters is studied, when the test results of each group are not directly obtained, but also affected by the detection error. For example, overall antibody concentration is reflected by the optical density. Then the optical density readings will be affected by many other factors. Then the models under the two cases is used to estimate the parameters by means of EM algorithm, and simulation calculation is carried out. In the second part. Pfeiffer in his article points out that when using the gene model which he proposed to estimate the genotype probability. Using the maximum likelihood estimation method to estimate the result is not ideal, so we propose a new method to estimate the model parameter and give out the property of this estimator, we also do some simulation to show the good performance of our new method. Finally, we proposed the conclusion.