Table 3.

Parameters used for the ANIMAL algorithm for mouse brain group‐wise nonlinear transformation estimation

Generation	Step size (mm)	Iterations	Gaussian blur (mm)
1	0.7	20	0.3
2	0.6	8	0.2
3	0.5	8	0.2
4	0.24	8	0.2
5	0.12	8	0.1
6	0.06	8	0.06

All values for lattice diameter were set to 1.5 x Step_size. Regularization parameters of stiffness, weight, and similarity were set to 0.98, 0.8, and 0.8, respectively. In all cases both intensity and gradient features extracted from the blur were used.