定量研究尼氏染色鼠脑内细胞数量的方法初探-生物医学工程专业论文.docxVIP

华中科技大学硕士学位论文 华 中 科 技 大 学 硕 士 学 位 论 文 II万方数据 II 万方数据 为 1.286×105 mm-3，杏仁核区域细胞平均密度为 1.308×105 mm-3，运动皮层区域细胞 平均密度为 1.301×105 mm-3，在运动皮层深度方向，观察到细胞密度呈现明显的分层 现象。 本文发展了一套三维细胞自动检测方法，可以比较准确地检测显微光学切片断 层成像系统获取的尼氏染色图像数据集中的细胞，为定量研究小鼠脑内细胞数量和 密度分布奠定了基础。 关键词：定量分析 细胞检测 图像二值化 尼氏染色 显微光学切片断层成像 III万方数据 III 万方数据 Abstract* Brain is the most complex and important organ, and plays indispensable role on maintaining life activities progression. The basic brain function and mechanism are dependent on the cells’ synergy effect. And Brain diseases are often accompanied by cytoarchitecture changes and cell number variation. So, quantitative study circumstances of the cell distribution and cytoarchitecture is quite important on understanding brain function and its working mechanism. Rapid development of nerve cell marking technology and microscopic imaging technology have laid solid foundation for cell numbers’ quantitative study. In the field of nerve cell marking, the nissl body inside the cells can be colored by alkaline dye, Thionine, through which all of the cells in the mouse brain will be labelled. In the field of nerve imaging, Micro-optical sectioning tomography is able to acquire nissl-stained whole mouse brain image data set, with TB scale data amount. Manually counting all of the cells in the mouse brain is unrealistic, as the cell number is absolutely huge. It is necessary to develop completely automatic cell quantitative counting method, which will reduce time consumption and laborious manual intervention. To quatitatively count cell numbers in nissl-stained mouse brain image data sets acquired by Micro-optical sectioning tomography, the information of cell gray scale distribution, morphology and size has been considered in this study. Optimal threshold value is calculated on the basis of otsu algorithm after gray intensity linear transformation. Next, analyze typical feature parameters in each connected component and remove incorrect identified foreground regions in the generated bi