| Aim:Presently,the incidence and mortality of lung cancer have surpassed other cancers worldwide,becoming the highest death factor for cancer patients.The pathological type of peripheral lung cancer is more common in non-small cell adenocarcinoma,which can induce pleural effusion.Pleural effusion is a kind of body fluid produced in the chest cavity,the cells and biomarkers in its effusion can assist in the diagnosis of some related diseases.As research reports,there are many types of components in pleural effusion,not only single tumor cells but also tumor cell clusters,and the proportion of tumor clusters is related to the severity of the metastasis.So the detection of tumor clusters in the pleural effusion is closely related to the diagnosis,tumor metastasis and the prognosis of lung cancer.But,in the process of clinical examination,it is a challenge to detect single lung cancer cells or differentiate tumor cell clusters from large pleural effusion.Microfluidic technology has the ability of the micron level control,and a great plasticity character.By adding physical,chemical or immune antibody on its material,this technique can greatly extend the analysis throughput.The histopathological examination had always been treated as the "gold standard" of disease diagnosis.The analysis of the target cell's shape or other feature characters is the foundation of discrimination.And computer-based image recognition technology can detect the difference of features when human eyes failed to identify.This study was to explore the combination of microfluidic chip technology and computer image recognition technology,to analyze the composition of pleural effusion,in order to lay the foundation for further analysis a large amount of pleural effusion and diagnosis the tumor cells in the future.Methods:1.Use the Solid Works software to design the microfluidic sorting chip.2.The samples selected from the A549 lung cancer cell lines and SV40 virus immortalized Me-5A mesothelium cell line as a source of tumor cell group and testing group,after being centrifuged,using phosphate buffer as the solvent to prepare a certain concentration of tumor cell groups.The white blood cells were extracted from healthy volunteers by red blood cell lysis as the negative control group.3.After loading the sample on the injection pump,set a series of velocity ratio in each entrance,to obtain the chip's filling rate,recovery rate,distribution of the chip rate,dynamic characteristics,etc.The fluid characteristics of the chip were observed by using the ANSYS Fluent fluid simulation software.4.After retrieving the cell clusters,nonspecifically stain them with AO fluorescent solution,capture the images of cells by CCD camera.5.Using the Matlab software to analyze the R and G channels of each image respectively,to obtain the feature values of tumor cells and clusters.Results:1.The maximum recovery rate is about 80% when Va=10ml/h,and Vb=8.5ml/h.2.Large clusters(>10)only appeared in outlet IV and V of the microfluidic chip.3.When analyzing the characteristics of AO fluorescence staining,the tumor cells and non-tumor cells were distinguished by contrast and correlation features.Conclusion:1.Under the velocity of inlet A and B 10 ml/h and 8.5 ml/h,the recovery rate will reach about 80%,tumor cell clusters were collected in channel III,IV and V,which is suitable for pleural effusion analysis.2.The features of the two selected cell lines can be distinguished to a certain extent,which can be used for the identification of a variety of tumor cells.3.The sorting and identification of the system can be combined freely,which can improve the efficiency of each part by optimizing each module.4.Because there is no statistical difference between the control group and the testing group in viability examination,thus other downstream analysis can be conducted to assist the identification of the cell types. |
PDF Full Text Request