Study And Implementation Of Nucleic Acid Body Fittings Based On Fluorescence Image Segmentation

With the global tumor incidence and mortality increasing year by year,tumor has become one of the major diseases endangering human health.Malignant tumors are incurable.Early detection and early treatment can greatly improve the cure rate of tumors.However,early-stage tumors are so poorly characterized that they are often difficult to detect with common biochemical assays.It is of great significance for the effective prevention and treatment of tumors to develop effective early diagnosis methods for tumors.Tumor is a genetic disease,and when genes are abnormal,differences in protein expression can result in tumor formation.Functionalized nucleic acid aptamers can use differences in protein expression to find specific markers of cancer cells,which are very promising for early diagnosis and treatment of cancer.The screening of nucleic acid aptamers requires repeated rounds of experiments,which is cumbersome,time-consuming,and prone to aerosol contamination,which can easily lead to the failure of screening experiments.Therefore,it is necessary to verify the specific performance of nucleic acid aptamers during screening to determine whether the screening experiment is effective and to monitor the progress of the screening experiment.In order to solve the above problems,our research group has developed an automated nucleic acid aptamer screening instrument and a fluorescence imaging verification device,which can complete the automated screening and performance verification of nucleic acid aptamers.In this paper,following the requirements of nucleic acid aptamer fluorescence imaging verification instruments,an aptamer performance verification system software based on fluorescence imaging and image segmentation is developed.The software can calculate the average fluorescence intensity of a single cell by collecting images of nucleic acid aptamers binding to target cells,so as to determine the performance of aptamers and monitor the progress of screening experiments.The application value of this paper is to provide a new method for the performance verification of nucleic acid aptamers and to provide support for the research group to develop fully automated nucleic acid aptamer screening and verification integrated instruments.In this paper,according to the characteristics of the cell microscopic images collected under white light,the images are denoised by bilateral filtering,histogram equalization and image enhancement algorithm based on wavelet are used to adjust the uneven illumination of the image.Moreover,a cell segmentation method based on the combination of Unet++ network and concave centroid morphological method is designed to perform cell segmentation on the preprocessed images.The Unet++ neural network,which is suitable for small datasets and medical image segmentation,is used as the image segmentation network.In order to enhance the segmentation of cell boundaries and improve the problem of unbalanced positive and negative samples caused by uneven cell growth,a loss function combining cross-entropy with boundaries and Dice coefficients is introduced.Compared with other methods,it is shown that the segmentation method has better segmentation effect for the images collected in this paper.It is difficult to count the problem that the cells still stick together after the segmentation.By analyzing the adhesion situation,we divides the types of adhesion into three categories,and proposes a segmentation method based on concave points combined with the center of gravity to further segment the adhesion area in the image after network segmentation.Experiments show that the method can effectively segment the sticky region.For fluorescence images,the method of image registration is adopted in this study.The fluorescence image is registered with the cell segmentation image to extract the cell fluorescence area,which effectively removes the interference of background fluorescence and uneluted fluorescent particles.In order to verify the effectiveness of the algorithm in this study,we designed the cell membrane staining experiment and the 5-8F nasopharyngeal carcinoma screening experiment.The results were compared with flow cytometry and PCR amplification methods,it is indicated that the proposed algorithm based on fluorescence imaging and image segmentation can effectively verify the specificity of nucleic acid aptamers and monitor the process of nucleic acid aptamer screening.Finally,in order to meet the needs of experimenters to verify the performance of nucleic acid aptamers and monitor the screening process in the experiment,a software interface system for the specific verification algorithm of nucleic acid aptamers based on fluorescent images is developed.