【作者】 陈瑶；

【导师】 付神贺； 朱思祁；

【作者基本信息】 暨南大学 ， 光学工程（专业学位）， 2020， 硕士

【摘要】 癌症具有成因复杂,致死率高、治愈率低等特点,是人类目前尚未攻克的一个医学难题。从微观层面上对癌细胞进行病理学研究,有助于科研人员更加深入地了解癌症的成因及治疗方法。为了快速识别组织中癌细胞并了解其分布情况,本文利用高光谱显微成像技术,对组织切片中的正常细胞和癌细胞进行光谱采集,并依靠其光谱信息和空间信息建立数学模型,对细胞种类进行鉴别,进而了解癌细胞的分布情况并实现癌细胞的精准快速检测。医学研究表明癌细胞因不受限制的复制和增殖,会导致DNA异倍体出现,细胞核增大且畸变,DNA和RNA含量比正常细胞显著增高。针对这些差异,本文分别提出了采用H&E染色剂和DAPI染色剂对细胞进行染色,使得这些差异能够通过染色后的光谱展示出来,从而实现对正常细胞和癌细胞的鉴别。本文首先利用癌细胞DNA和RNA含量升高引起的p H值变化会导致H&E染色变化的特点,通过建立癌细胞和正常细胞的透射光谱差异,结合支持向量机自适应算法,实现对癌细胞的鉴别。考虑到细胞核内染色质不均匀会导致各处p H值不一样,本文在癌细胞鉴别实验中首次引入了核内DNA分布均匀度作为修正参数,使癌细胞鉴别的灵敏性和特异性分别达到99%和98%。为了进一步优化,本研究采用DAPI染色剂替代H&E染色剂对组织染色并使用高光谱技术对癌细胞鉴别,DAPI对DNA浓度和DNA/RNA比例极其敏感,因此其荧光光谱能够有效反应出癌细胞核和正常细胞核的差异,本研究表明利用此方法对100组癌细胞进行鉴别的灵敏性和特异性均为100%。更多还原

【Abstract】 Cancer was characterized by complex causes,high mortality and cure rate,and cancer was a medical problem that has not been solved yet.Pathological studies of cancer cells at the microscopic level could help researchers better understand the causes and treatment of cancer.In order to quickly identify cancer cells and understand the distribution of cancer cells in the tissue,in this paper,we used hyperspectral imaging technology to collect the spectrum of normal cells and cancer cells from tissue section.And then we relied on the spectral information and spatial information of cells to establish a mathematical model to identify cell types,so as to understand the distribution of cancer cells and realize accurate and rapid detection of cancer cells.Medical studies have shown that the unrestricted replication and proliferation of cancer cells lead to the appearance of DNA aneuploidy,the enlargement and distortion of the nucleus,and the DNA and RNA content were significantly higher than that of normal cells.In view of the characteristics of these differences,in this article,we proposed to use H&E stain and DAPI stain to stain the cells respectively,so that these differences could be shown by the staining spectrum,so as to achieve the identification of normal cells and cancer cells.In this paper,firstly,the p H change caused by the increase of DNA and RNA content of cancer cells would lead to the change of H&E staining.By establishing the transmission spectrum difference between cancer cells and normal cells and combining with the SVM adaptive algorithm,the identification of cancer cells was realized.We considered that the heterogeneous chromatin in the nucleus would lead to different p H values,in this paper,the uniformity of DNA distribution in the nucleus was introduced for the first time as a modified parameter in the cancer cell identification experiment,so that the sensitivity and specificity of cancer cell identification could reach 99% and 98% respectively.To further optimize,in this study,DAPI stain was used to replace H&E stain for tissue staining,and we used hyperspectral imaging technique to identify cancer cells.DAPI was extremely sensitive to DNA concentration and DNA/RNA ratio,so the fluorescence spectrum of cells stained by DAPI could effectively reflect the differences between cancer cells and normal cells.This study demonstrated that the sensitivity and specificity of this method for the identification of 100 groups of cancer cells were 100%.更多还原