Weird layers in LN2_LAYERS when CSF and GM touch

[layerfMRI]

This is to keep this documented somewhere: The RIM file that is the input to some of the LayNii programs as consisting of segmentation files consisting of CSF border (value 1), WM border (value 2), and GM (value 3) should be provided in a way that WM and CSF are NOT touching. If they touch, it can result in some form of singularity that results in weird layer results.

This issue was described by Gabi Lohmann. She also described the solution. Thanks Gabi

example_data.zip