Getting started with ParaView
ParaView is both an application framework and a turn-key application for advanced data visualization that can be used to inspect anatomical MRI data and segmentations, in combination with triangular, tetrahedral and/or hexahedral meshes. It is useful to check the details of the meshes used in BEM and FEM forward modeling.
Example use
The following segments an anatomical MRI in 5 tissue types
cfg = [];
cfg.output = {'csf', 'gray', 'scalp', 'skull', 'white'};
% cfg.output = {'brain', 'skull', 'scalp'}; % this is what you need for BEM
segmentedmri = ft_volumesegment(cfg, mri);
From the segmentation, you can make meshes with
cfg = [];
cfg.method = 'isosurface';
cfg.numvertices = inf;
cfg.tissue = {'csf', 'gray', 'scalp', 'skull', 'white'};
surface = ft_prepare_mesh(cfg, segmentedmri);
cfg = [];
cfg.method = 'tetrahedral';
tetrahedral = ft_prepare_mesh(cfg, segmentedmri);
cfg = [];
cfg.method = 'hexahedral';
hexahedral = ft_prepare_mesh(cfg, segmentedmri);
Using the isosurface method of ft_prepare_mesh and the highest number of vertices makes a detailed surface description using the MATLAB isosurface function. The surface meshes created here are not suitable for a BEM model, as that requires nested surfaces that are closed and non-intersecting. To construct BEM meshes you can use the projectmesh method on the brain, skull and scalp segmentation.
The tetrahedral and hexahedral meshes can be used for FEM models.
We can write the anatomical MRI and segmentation to a .vtk file with ft_write_mri
% write the anatomical MRI, including the coordinate transformation matrix
ft_write_mri('mri.vtk', mri.anatomy, 'transform', mri.transform, 'dataformat', 'vtk')
% convert the different boolean/probabilistic volumes into one volume with indices
segmentedmri = ft_datatype_segmentation(segmentedmri, 'segmentationstyle', 'indexed');
% write the segmentation with the indices that represent the different issue types
ft_write_mri('segmentedmri.vtk', segmentedmri.tissue, 'transform', segmentedmri.transform, 'dataformat', 'vtk')
and the meshes with ft_write_headshape
% one surface per file
ft_write_headshape('surface_csf.vtk', surface(1), 'format', 'vtk')
ft_write_headshape('surface_gray.vtk', surface(2), 'format', 'vtk')
ft_write_headshape('surface_scalp.vtk', surface(3), 'format', 'vtk')
ft_write_headshape('surface_skull.vtk', surface(4), 'format', 'vtk')
ft_write_headshape('surface_white.vtk', surface(5), 'format', 'vtk')
% also write the tissue indices
ft_write_headshape('tetrahedral.vtk', tetrahedral, 'format', 'vtk', 'data', tetrahedral.tissue)
% also write the tissue indices
ft_write_headshape('hexahedral.vtk', hexahedral, 'format', 'vtk', 'data', hexahedral.tissue)
It is possible to split the tetrahedral mesh into the different tissue types and write those to separate files.
% the tissue and tissuelabel will become invalid
tetrahedral_csf = rmfield(tetrahedral, {'tissue', 'tissuelabel'});
tetrahedral_gray = rmfield(tetrahedral, {'tissue', 'tissuelabel'});
tetrahedral_scalp = rmfield(tetrahedral, {'tissue', 'tissuelabel'});
tetrahedral_skull = rmfield(tetrahedral, {'tissue', 'tissuelabel'});
tetrahedral_white = rmfield(tetrahedral, {'tissue', 'tissuelabel'});
% only select one tissue type
tetrahedral_csf.tet = tetrahedral.tet(tetrahedral.tissue==1,:);
tetrahedral_gray.tet = tetrahedral.tet(tetrahedral.tissue==2,:);
tetrahedral_scalp.tet = tetrahedral.tet(tetrahedral.tissue==3,:);
tetrahedral_skull.tet = tetrahedral.tet(tetrahedral.tissue==4,:);
tetrahedral_white.tet = tetrahedral.tet(tetrahedral.tissue==5,:);
ft_write_headshape('tetrahedral_csf.vtk', tetrahedral_csf, 'format', 'vtk')
ft_write_headshape('tetrahedral_gray.vtk', tetrahedral_gray, 'format', 'vtk')
ft_write_headshape('tetrahedral_scalp.vtk', tetrahedral_scalp, 'format', 'vtk')
ft_write_headshape('tetrahedral_skull.vtk', tetrahedral_skull, 'format', 'vtk')
ft_write_headshape('tetrahedral_white.vtk', tetrahedral_white, 'format', 'vtk')
and idem for the hexahedral meshes.
% the tissue and tissuelabel will become invalid
hexahedral_csf = rmfield(hexahedral, {'tissue', 'tissuelabel'});
hexahedral_gray = rmfield(hexahedral, {'tissue', 'tissuelabel'});
hexahedral_scalp = rmfield(hexahedral, {'tissue', 'tissuelabel'});
hexahedral_skull = rmfield(hexahedral, {'tissue', 'tissuelabel'});
hexahedral_white = rmfield(hexahedral, {'tissue', 'tissuelabel'});
% only select one tissue type
hexahedral_csf.hex = hexahedral.hex(hexahedral.tissue==1,:);
hexahedral_gray.hex = hexahedral.hex(hexahedral.tissue==2,:);
hexahedral_scalp.hex = hexahedral.hex(hexahedral.tissue==3,:);
hexahedral_skull.hex = hexahedral.hex(hexahedral.tissue==4,:);
hexahedral_white.hex = hexahedral.hex(hexahedral.tissue==5,:);
ft_write_headshape('hexahedral_csf.vtk', hexahedral_csf, 'format', 'vtk')
ft_write_headshape('hexahedral_gray.vtk', hexahedral_gray, 'format', 'vtk')
ft_write_headshape('hexahedral_scalp.vtk', hexahedral_scalp, 'format', 'vtk')
ft_write_headshape('hexahedral_skull.vtk', hexahedral_skull, 'format', 'vtk')
ft_write_headshape('hexahedral_white.vtk', hexahedral_white, 'format', 'vtk')