Differences between revisions 1 and 2
Deletions are marked like this. Additions are marked like this.
Line 10: Line 10:
[[TableOfContents]] <<TableOfContents>>
Line 38: Line 38:
["mri_seghead"] binarizes the input volume such that all the voxels in the [[mri_seghead]] binarizes the input volume such that all the voxels in the
Line 71: Line 71:
["othercommand1"], ["othercommand2"] [[othercommand1]], [[othercommand2]]
Line 82: Line 82:
["References/Lastname###"] [[References/Lastname###]]



mri_seghead - binarizes volume so that all voxels in head are set to 255


mri_seghead --invol <invol> --outvol <outvol> --thresh1 <thr1> --thresh2 <thr2> [options]


Positional Arguments


Required Flagged Arguments

--invol <invol>

input volume

--outvol <outvol>

output volume

|| --thresh1 <thr1> || ?? ||

--thresh2 <thr2>


Optional Flagged Arguments


fill value (default = 255)

|| --thresh || single threshold value for 1 and 2. sets thresh1 and thresh2 to threshold. typical value = 20 ||


min number of consecutive hits (2). typical value = 2



output volume


mri_seghead binarizes the input volume such that all the voxels in the head are set to 255 (or whatever is passed with --fill). The result is stored in the output volume passed by --outvol.

The program first creates a binarized mass just a few mm below the skin. This mass is then grown out using a connected components algorithm so that most of the skin details are retained.

The initial mass is created in the following way. First, for every row and slice, the column is searched from both ends for the 'skin'. The skin is defined as the first consecutive nhitsmin voxels over thresh1. Once the skin is found coming from both directions, everything in between is binarized to the fill value. This process is repeated for the rows and slices. The initial mass is created by ANDing all three voxels.

After the initial mass is defined, the original volume is modified so that all the voxels in the mass are set to 255. This has the effect of filling in all of the subepidermal features that would normally be below threshold. A seed point is chosen at the center of the volume. The final binarized volume is computed as all the voxels above thresh2 connected to the seed point.


Example 1


Example 2




See Also

othercommand1, othercommand2



Methods Description




Reporting Bugs

Report bugs to <analysis-bugs@nmr.mgh.harvard.edu>



mri_seghead (last edited 2008-04-29 11:45:43 by localhost)