LIBS Imaging Study Of CoP Particles Endocytosis Of Hepatocellular Carcinoma Cells And Lung Cancer Tissues

In recent years,as people’s material conditions are abundant and living conditions are improving,various types of cancer have become more frequent.Laser-Induced Breakdown Spectroscopy(LIBS)is an emerging tissue analysis technology that can identify cancer tissues.As far as we know,there is no relevant research on lung cancer tissues based on LIBS,and research on the cell level is rarely reported.Therefore,this article has carried out research on the endocytosis of liver cancer He PG2 cells and LIBS imaging analysis of lung cancer tissues.This article first introduces the application of LIBS technology at home and abroad in biological tissue element imaging and cancer discriminative diagnosis,and puts forward the main research content of this article.Secondly,it introduces the relevant principles of LIBS,introduces the data mining algorithms and related model evaluation indexes and image processing algorithms used in this article.Thirdly,based on the established LIBS system,a Nd:YAG laser with a frequency of 4Hz and a wavelength of 1064 nm is used to carry out the research on cell and tissue samples.The detection delay of the laser is optimized for Hep G2 cell samples,the element distribution map of Co is established,and the relationship between cell density and Co element intensity is analyzed.The confidence ellipse interval with positive correlation is established.For lung tissue samples to optimize the experimental parameters,including laser and sample distance,the atmosphere gas flow,and laser detection delay.The LIBS spectra of lung large cell carcinoma and common lung cancer samples are obtained.Combined with the NIST spectral database,we establish the element distribution images of K,Ca,Na,Mg,Fe,C,H,O,N and analyze the correlation between the element map and the staining map.Finally,the cluster analysis of lung cancer tissue based on LIBS is carried out.Based on principal component analysis and cluster analysis method,adopting different principal component combinations combined with fuzzy C-means clustering,k-means clustering and k-medoid clustering methods.We obtain different clustering renderings by combining the stained image and the element distribution images.Registration realizes the division between cancer cell and normal cell regions,and evaluates the accuracy of the clustering effect.The study indicates that K-means clustering has the best effect,with the highest accuracy,sensitivity,specificity and accuracy of 98.68%,98.02%,99.38% and 99.40%,respectively.Compared with the previous work,the accuracy has increased by 14 percentage points,realizing automatic pathological diagnosis of lung cancer tissue.The correlation between the element diagrams of K,Ca,Na and Mg and the cluster results is analyzed.It is found that Na element has the best correlation with the distribution of cancer.The correlation coefficients are all above 0.8,followed by Mg element and Ca element,and K element has the worst correlation.