Study On Abnormal Detection Method Of Bone Tissue Engineering Cells Based On Neural Network

Cell detection is an important field in biomedicine.Both the biomedical field and the clinical medicine field play an irreplaceable role.Nowadays,for the problem of degenerative and traumatic articular cartilage damage,the medical field has begun to apply bone tissue engineering technology to it.Bone tissue engineering technology includes 3D printing technology,in vitro cell culture,and cell-directed induction differentiation.Cell detection also plays an indispensable role.In order to ensure the reliability of cell culture,traditional staining needs to sacrifice a batch of experimental cells and the experimental method is more complicated.With the rapid development of deep learning,it also provides new ideas for cell abnormality detection,and at the same time enables deep learning to serve related research in the field of biomedicine.The target detection algorithm has obvious advantages,improves the detection efficiency,greatly saves the cost,and has good detection accuracy compared to the cumbersome manual detection method.This article proposes a set of abnormal detection methods for rat bone marrow mesenchymal stem cells,and verifies the relationship with the degree of cell differentiation.This method mainly includes image acquisition system,image preprocessing method,anomaly detection model and verification of the degree of cell differentiation.The main research contents of this paper are as follows:(1)Considering the huge demand of deep learning for data sets,this paper made an study of image sensors and Open CV,and built a set of microscope-based stem cell image acquisition system,which could meet the relevant acquisition of images required by the experiment.In addition,about 10,000 cell images were collected through this medium.(2)Considering the particularity of BMSCs,the annotation of abnormal cell images should have high accuracy,reliability,timeliness and other characteristics.In this paper,an improved algorithm of edge detection based on Canny operator is constructed so that a set of image preprocessing method of BMSCs,solved the biomedical field dependence on abnormal tagging of BMSCs,the guidance of professionals from now on for ectomesenchymal abnormal rat bone marrow stem cells achieved only using image preprocessing method can realize precise annotations.(3)In this paper,on the basis of fully absorbing the previous research on target detection,we deeply analyzed the compatibility between target detection model and biomedicine,selected the target detection model that is most suitable for BMSCs,and completed the relevant optimization and adjustment,making its abnormal detection accuracy in rat bone marrow mesenchymal stem cells reach 84.3%.The reliability and feasibility of this algorithm were verified by trypan blue staining,cell fluorescence staining and automatic cell counting apparatus.A set of effective abnormal detection methods of BMSCs are obtained.(4)The relationship between the abnormal rate of rat bone marrow mesenchymal stem cells and the degree of cell differentiation was proposed,and the need and economy of verifying the relationship between the two were analyzed.In this paper,chondroblast differentiated cells were obtained by directional induction of BMSCs,and an experiment was designed to detect abnormal cells before the experiment,and the degree of cell differentiation was calculated by immunofluorescence staining for the cells after the experiment.It was concluded that the rate of cell abnormality was inversely correlated with the degree of cell differentiation.It will pave the way for bone tissue engineering technology.