In automatic subcortical segmentation, each voxel in the normalized brain volume is assigned one of about 40 labels, including:
- Cerebral White Matter, Cerebral Cortex, Lateral Ventricle, Inferior Lateral Ventricle, Cerebellum White Matter, Cerebellum Cortex, Thalamus, Caudate, Putamen, Pallidum, Hippocampus, Amygdala, Lesion, Accumbens area, Vessel, Third Ventricle, Fourth Ventricle, Brain Stem, Cerebrospinal Fluid
The automatic subcortical segmentation can take many (11+) hours to complete.
To view just the segmentation, use this command:
tkmedit <subject name> norm.mgz -aseg
Float your cursor over any voxel and the label assigned to it will be displayed in the TkMeditTools window.
If the voxels are incorrectly labeled, then you can re-label them yourself although we suggest contacting the freesurfer mailing list first to see if there is an automatic procedure that can be used to re-label them.. Refer to the TkMeditGuide/TkMeditWorkingWithData/TkMeditSelectionsLabels page for detailed information on manually editing the aseg.
Automatic subcortical segmentation of a brain volume is based upon the existence of an atlas containing probablistic information on the location of structures. This is decribed here:
Whole Brain Segmentation: Automated Labeling of Neuroanatomical Structures in the Human Brain, Fischl et al., (2002). Neuron, 33:341-355.
The atlas included with the Freesurfer distribution is found in the 'average' directory, and is called 'RB_all_2008-03-26.gca'. It is possible to construct your own atlas. This is described next.
Constructing an Aseg Atlas
- Using tkmedit, label each volume in the brain. Repeat for all subjects to be included in the atlas.
Run mri_ca_train to create the atlas.
See also rebuild_gca_atlas.csh script in $FREESURFER_HOME/bin.
See also AtlasSubjects
For accuracy evaluations see also SubcorticalSegmentationAccuracy
The original CMA segmentation scheme used for subcortical segmentation training is defined by (Filipek, et al, Cerebral Cortex, 1994) (paper). However, with better imaging resolution, anatomists Verne Caviness and Nikos Mkris developed the separation of the "thalamus proper" from a "ventral diencephalon" region that subtends many of the smaller nuclei and structures in the area inferior to the thalamus, such as hypothalamus, red nuclei, later and medial geniculate, ect. The first definition of the ventral diencephalon method is provided in Am J Med Genet. 1997 Sep 19;74(5):507-14. (paper). As defined by Makris in Makris, et al. Biol Psychiatry. 2008 Aug 1;64(3):192-202. (paper "...ventral diencephalon (49), which according to our morphometric definition contains the hypothalamus, basal forebrain, and sublenticular extended amygdala (SLEA), as well as a large portion of ventral tegmentum (which is included in our ventral diencephalon region by convention although part of midbrain)".
Also, from BIOL PSYCHIATRY L.J. Seidman et al 1999;46:941–954 (paper) "We used the same segmentation procedure of the thalamus developed in our preliminary MRI report (Seidman et al 1997). The central gray nuclei were divided into superior and inferior segments, the thalamus proper and the “ventral diencephalon,” respectively. The thalamus proper (including all thalamic nuclei except the lateral and medial geniculate bodies) was defined on all coronal slices on which it was present, from the first slice containing the anterior nucleus through the last slice containing the pulvinar nucleus. The structure was bounded medially by the third ventricle, and laterally by the internal capsule. The superior border was the body of the lateral ventricle, and the inferior border, the hypothalamic fissure. This structure definition differed from the central gray nuclei of Filipek and co-workers (1994) only in its inferior boundary."
The 5th ventricle, referred to as the "cavum septum pallucidum," is a bit of an enigma that tends to only occur on occasion. It's not reported in any manuscripts, but it was confirmed by neuroradiologist, P. Ellen Grant.