一种简易的脑片图像的半自动区域划分及细胞计数方法

摘要

脑科学研究往往需要对实验动物不同脑区的分子表达、细胞数目及神经网络连接模式进行量化和比较分析,然而目前缺乏简便易行的规模化的脑区划分方法.本文基于 Paxinos 和 Franklin 的第二版小鼠图谱以及 Photoshop 和ImageJ等图像处理软件,发展了一种简易的脑片图像的半自动区域划分及细胞计数方法.首先,基于小鼠脑图谱可获得脑区划分所需的标准模版;然后,利用 Photoshop 软件对标准模版进行变形矫正,可实现小鼠脑切片图像的快速规模化半自动配准及区域分割;最后,利用ImageJ可实现对配准及分区后的脑片图像不同区域的半自动神经细胞计数和分析.此方法适用于脑片免疫组织化学分析、神经元分布模式检测和神经环路标记等研究,其操作简便、易于获取、并且不依赖于其他昂贵的商业化神经图像分析软件,能为条件有限的科研工作者提供极大便利.%Quantitative analyses of molecular expression, cell counts and neural network connections among different brain regions are essential in brain science research. Based on Paxinos and Franklin's the Mouse Brain in Stereotaxic Coordinates (Ⅱ) and public image processing software such as Photoshop and ImageJ, we developed a convenient method for semi-automatic segmentation and cell counting on brain sections. A standard template for brain region segmentation was first obtained from the Paxinos and Franklin's mouse brain atlas. Photoshop was then used to transform the standard template semi-automatically into the space of brain sections, yielding masks of segmented brain regions. Finally, ImageJ was used to analyze the data in different brain regions. This method is useful for immunohistochemical and neuron distribution pattern analyses, as well as neural network labelling studies. The method does not require expensive commercial image analysis software, and is also easy to implement and use.