Hi Donna, Thank you so much for your advice! Sorry for the multiple copies of the original message that were posted to the list. I was having some problems with my e-mail.
-Mellanie Quoting Donna Hanlon <[EMAIL PROTECTED]>: > Mellanie, > > It turns out these last two handles aren't that hard to patch after all. > > In fact, they are a result of the ventricle filling gone awry (although > some manual patching did worsen the problems in this area. > > After making the patches specified below, do FloodFill Vol2 and save the > resulting file as nov26_2002.R.full.segment_vent_corr18.mnc. Then, load > SEGMENTATION/nov26_2002.R.full.segment.mnc as volume 1 (with > nov26_2002.R.full.segment_vent_corr18.mnc still loaded as volume 2). > > Then, set your mask size to 20 and set mask center at 8, 97, 115. Then > do vol1->vol2. FloodFill Vol 2 again, and update handle count. > > Save your edits, and I think you're ready to prepare for flatteining. > > Occasionally, ventricle filling and/or automatic error correction > misbehaves. While trying to be helpful, it fills something it > shouldn't. If you notice the segmentation is white where the structural > MRI is dark, then compare the segment.mnc with the > segment_vent_corr.mnc. If the latter is white where the former is dark, > then either ventricle filling or error correction is the culprit. In > such cases, vol1->vol2 can come in quite handy. > > Donna > > On 08/05/2004 01:28 PM, Donna Hanlon wrote: > > > Hi Mellanie, > > > > There are four very obvious handles in your segmentation that will > > cause crossovers in your flatmap. > > > > First, do FloodFill Vol2 to get an accurate handle count. There are > > five remaining handles. I can only see four, but the visible ones > > will cause flattening trouble. > > > > One on calcarine sulcus can be fixed in a coronal view by toggling > > these voxels off: > > Toggled voxel at 9 32 69 off > > Toggled voxel at 8 31 69 off > > Toggled voxel at 9 31 69 off > > Toggled voxel at 10 31 69 off > > Toggled voxel at 10 30 70 off > > Toggled voxel at 9 30 70 off > > Toggled voxel at 9 29 71 off > > > > One near the ventral end of the parahippocampal gyrus can be fixed in > > a coronal view: > > > > Set Mask Center 28 92 54 > > Dilate once > > > > The two remaining handles are along the dorso-medial cortex, above the > > ventricles, in this neighborhood: > > > > 5,85-105,120 > > Coronal slice 85-105 > > > > Here, it looks like some previous patching may have gone awry, zapping > > some voxels that should be on, and filling some that should be off. > > These handles are much trickier to fix, and I think I might start by > > setting a fairly large mask and loading the original > > "*segment_vent_corr.mnc" as Vol1, and using the vol1->vol2 feature. > > This has the effect of starting over with the original segmentation > > within a limited bounding box (i.e., you don't lose all your patches > > to the medial wall, subcortical structures, and occipital lobe, but > > you go back to the drawing board in the region of these two handles). > > I think it will be easier to patch the original handles than this > > post-patch region. > > > > I'm afraid that's all the advice I can give right now. > > > > Donna > > > > On 08/05/2004 12:46 PM, [EMAIL PROTECTED] wrote: > > > >> Hi Donna, > >> Is it acceptable to proceed with flattening even though there are a > >> few handles that are apparent on the inflated surface? For example, > >> in the file that I have uploaded, there is an obvious handle on the > >> surface that I am having trouble localizing in the segmentation. > >> Will it cause distortions in the flatmap? > >> The file is: > >> nov26_2002.R.full.segment_vent_corr.patch10_vent_corr.patch5.mnc > >> Thank You, > >> Mellanie > >> _______________________________________________ > >> caret-users mailing list > >> caret-users@brainvis.wustl.edu > >> http://pulvinar.wustl.edu/mailman/listinfo/caret-users > >> > >> > >> > > > > > > _______________________________________________ > > caret-users mailing list > > caret-users@brainvis.wustl.edu > > http://pulvinar.wustl.edu/mailman/listinfo/caret-users > > > > > _______________________________________________ > caret-users mailing list > caret-users@brainvis.wustl.edu > http://pulvinar.wustl.edu/mailman/listinfo/caret-users >
