Hi, Lijaz,

Assume
source image     --- ${src}
template image  --- ${temp}
label map         --- ${label_src} (in the source image space). This map may contain different labels for different tissues/regions, eg,
                           ${lable_value_1},
                           ${label_value_2},
                           ${label_value_3},
                           ${label_value_4} ...
                           (these labels are usually labels for WM, GM, CSF, but you can also have labels for ventricle or other structures. e.g, you can label WM as 100 and GM as 50, the values do not matter, as long as different tissue/regions of interest have different labels.)


to create RAVENS maps, this is the command I would suggest
dramms -S ${src}   \
            -T ${temp} \
            -L ${label_src} \
            -l ${label_value_1},${label_value_2},${label_value_3},${label_value_4} \
            -R ${RAVENS_prefix}

The first 4 lines in this command are inputs (you can have at least 1 and at most 5 label values in line 4).
The last line is the prefix for a series of RAVENS output.

Upon finish, you will get a series of RAVENS maps. One RAVENS map for one label (or more precisely, for one labeled region). They are of names
${RAVENS_prefix}_${label_value_1}.nii.gz
${RAVENS_prefix}_${label_value_2}.nii.gz
${RAVENS_prefix}_${label_value_3}.nii.gz
${RAVENS_prefix}_${label_value_4}.nii.gz
...


Note that the RAVENS maps are in the template space. The value at a voxel in the RAVENS map is the ratio of the volume of the corresponding voxels/region in the source image to the volume of the voxel in the template image. For example, if the ventricle in the source image is twice as big as the ventricle in the template and is homogeneous, then the RAVENS maps for ventricle will have values 2 all over the ventricle regions in the template space.


Hope this helps,


Best regards,
Yangming