micapipe from the Multimodal imaging and connectome analysis lab (http://mica-mni.github.io) at the Montreal Neurological Institute. Read The Docs documentation below
Is you look at the connectome aspect and look-up tables from micapipe, we can infer that the connectome is composed of (in order) :
48 subcortical and cerebellar ROI (infered from lut_subcortical-cerebllum_mics.csv)
150 cortical regions (infered from lut_subcortical-cerebllum_mics.csv), 75 for the left hemisphere (starting from ctx_lh_G_and_S_frontomargin to ctx_lh_S_temporal_transverse) and 75 for the righ hemisphere (starting from ctx_rh_G_and_S_frontomargin to ctx_rh_S_temporal_transverse)
Indeed 48 + 150 = 198, so this is consistent.
# slice the matrix remove subcortical nodes and cerebellum
SC = mtx_sc[49:, 49:]
SC = np.delete(np.delete(SC, Ndim, axis=0), Ndim, axis=1)
The line SC = mtx_sc[49:, 49:] slices the 48 first items which are the subcortical and cerebellar nodes (index from 0 to 47), but please why do you slice the 49th (index 48, which should be ctx_lh_G_and_S_frontomargin) ? Identically, the line SC = np.delete(np.delete(SC, Ndim, axis=0), Ndim, axis=1) slices the 76th item (Ndim = 75 for the aparc-a2009s parcellation), which seems to be ctx_rh_G_and_S_frontomargin.
Is the 75 cortical areas well the one included in the lut_subcortical-cerebllum_mics.csv please ? More generally, I think look-up tables for the full connectome in all parcellations would be of great help for micapipe users.
Those issues could be related : #124 #106 #119
Hello
We are performing structural connectome (SC) analysis with micapipe. The full-SC shape with the aparc-a2009s parcellation has a shape 198 x 198 (this is also precised here : https://micapipe.readthedocs.io/en/latest/pages/04.fetch/index.html?highlight=load_annot), which includes cerebellar, subcortical and cortical nodes (precised here : https://micapipe.readthedocs.io/en/latest/pages/04.matrices/index.html?highlight=full%20connectome#structural-connectome).
Is you look at the connectome aspect and look-up tables from micapipe, we can infer that the connectome is composed of (in order) :
in this webpage (https://micapipe.readthedocs.io/en/latest/pages/04.fetch/index.html?highlight=load_annot), a function load_sc is defined to load the SC in order to plot it on a figure, without the medial wall. We are a bit confused by this part of the function :
The line
SC = mtx_sc[49:, 49:]slices the 48 first items which are the subcortical and cerebellar nodes (index from 0 to 47), but please why do you slice the 49th (index 48, which should be ctx_lh_G_and_S_frontomargin) ? Identically, the lineSC = np.delete(np.delete(SC, Ndim, axis=0), Ndim, axis=1)slices the 76th item (Ndim = 75 for the aparc-a2009s parcellation), which seems to be ctx_rh_G_and_S_frontomargin.Is the 75 cortical areas well the one included in the lut_subcortical-cerebllum_mics.csv please ? More generally, I think look-up tables for the full connectome in all parcellations would be of great help for micapipe users.
Thank you for your help, and for this great tool.