老师，您好！
我目前自己探索了一个方法，很多步骤是参考了老师你的帖子《使用SPM12进行VBM分析》
1、先将小鼠MRI图像进行手动颅脑分割，然后将小鼠MRI图像和模板图像体素放大10倍。
2、将小鼠MRI图像进行重新定原点，使其和原点尽可能在同一原点
3、使用spm12中的old segment对图像进行分割，将其中的默认TPM换成小鼠的TPM，分割选项为native
4、然后将得到的“seg_sn.mat”作为Dartel Tools中的Initial Import，这样就会得到rc开头的文件（rc1 rc2 rc3）
5、得到的rc开头文件（灰质 白质 CSF），分别进行run dartel（creat templates）,生成各自的template6模板
6、然后使用normalize（estimate&write）工具，将上面生成的template 6 分别与小鼠标准脑模板的三个进行对应及写入
7、最后再进行old segment，然后将默认TPM换成上面normalize后的三个template6,并且分割选项这里选为modulation，这样子是否可以？
（老师您的帖子第6/7两步使用的是normalize to MNI space，我觉得小鼠的模板不是基于MNI空间的，所以不能用，不知道这个想法是不是对的）
（部分步骤也参考这篇文献的步骤，老师您可以参考一下，但是他的3-5步我不知道他怎么完成的？
each image was segmented into three tissue priors (Gray Matter, White Matter, Cerebrospinal Fluid maps) using the old segment tool in SPM12. In this step, the default tissue probability maps were replaced by the SIGMA template’s GM, WM, and CSF tissue maps.
To perform a more accurate analysis, all the segment priors were used to generate a subject-specific template by DARTEL (Diffeomorphic Anatomical Registration Through Exponentiated Lie algebra) algorithm in SPM12 as follows:
1 the “seg_sn.mat” documents generated during the old segment were used for initial import step;
2 GM, WM and CSF templates were created by DARTEL using imported three tissue priors respectively;
3 the Jacobian images acquired during DARTEL template creation of GM priors were applied to warp individual T2-weighted images;
4 the normalized T2-weighted images were averaged to generate final subject-specific template;
5 the GM priors were normalized to a subject-specific template and modulated.
Finally, the normalized GM maps were smoothed using an 8-mm FWHM Gaussian kernel.
For the group-level statistical image analysis, the voxel-by-voxel one-way ANONA was used across the whole brain by employing the statistical tools in DPABI. We excluded all voxels with a value of < 0.2 by creating an exclusive mask based on the subject-specific template to avoid possible edge effects around the borders between the tissue classes and to include only voxels with sufficient tissue class proportions. In the group analysis, significant voxels were identified using a threshold of p < 0.01 with false discovery rate (FDR) correction for multiple comparisons.）