Info

aWhole brain includes gray and white matter; cortical gray is measured on coronal and axial slices with no subcortical structure.

aWhole brain includes gray and white matter; cortical gray is measured on coronal and axial slices with no subcortical structure.

(mostly ruts). For any point on the outer cortical surface, the absolute value of xVin at the point is simply the distance from the point to the inner cortical surface. Using this measure, we obtain a thickness map between the inner and outer cortical surfaces, which can be used to study the normal thickness variations in different regions as well as abnormalities. We have recorded the spatial variation in normal thickness for N = 30 normal subjects as seen in Fig. 9. These measurements are in good agreement with direct measurement of thickness in postmortem samples, which average about 3 mm across the whole brain [3]. Moreover, the cortical thickness of the frontal lobes is significantly greater (paired t(28) = 10.4,_p<.0001) than the thickness of the posterior cortex. This is consistent with other sources on cortical thickness, including postmortem data [3]. Also, two areas of posterior cortex expected to be thinner, the postcentral gyrus and the primary and secondary visual cortices in the occipital lobe, across the 30 control subjects tested, are found to be consistently relatively thinner as expected, as seen in Fig. 9, providing a degree of further validation.

Cortical surface area is also easily captured and measured from our segmentation approach. Our method yields area measurements from the inner and outer cortical surfaces for the entire cerebrum. Both inner (GM/WM) (1823 + 170 cm2) and outer (GM/CSF) surface area estimates (1527 + 271 cm2) are commensurate with postmortem neuroimaging estimates of about 1600 mm2 [15,18]), suggesting that our approach is accurate. However, the smaller outer compared to inner surface area points to some error in our method due to incomplete representation of the outer surface deep within some of the more tightly packed sulci.

5.6 Sulcal Surface Determination

Cortical surface determination facilitates further analysis by the determination of sulcal surfaces [46]. First, sulcal curves at the top and bottom of the sulcus can be automatically traced after the specification of start and end points using dynamic

FIGURE 8 The inner and outer cortical surfaces of a brain colored according to the corresponding shape index [21]. See also Plate 11.

vpb+rlG*! îr-wç'ri nt n*4dL# »ML* #«iJ4S# rl4g+ &m+ ipiMPrliciil C'iJ fOE ildg* ffet-

FIGURE 8 The inner and outer cortical surfaces of a brain colored according to the corresponding shape index [21]. See also Plate 11.

Region (Lobe)

Left Frontal Right Frontal Left Posterior* Right Posterior*

FIGURE 9 Measurement of cortical thickness. {Left) Table reporting mean thickness values for N = 30 normal control males. Note: ^Posterior region encompasses all parietal, temporal, and occipital cortical tissue. (Right) Thickness plots of two normal brains {one brain: a,b; second brain: c) showing marked thinning in the postcentral gyrus and primary and secondary visual cortices in the occipital lobe. See also Plate 12.

FIGURE 9 Measurement of cortical thickness. {Left) Table reporting mean thickness values for N = 30 normal control males. Note: ^Posterior region encompasses all parietal, temporal, and occipital cortical tissue. (Right) Thickness plots of two normal brains {one brain: a,b; second brain: c) showing marked thinning in the postcentral gyrus and primary and secondary visual cortices in the occipital lobe. See also Plate 12.

FIGURE 10 Sulcal surfaces shown with cutaway view of brain (b) and on outer cortical rendering (c). See also Plate 13.

Qeelpiti] C on tr

Fa !lci rural Gyrus programming based on a surface maximum principal curvature cost function. Sulcal ribbon surfaces (shown in Fig. 10) can then be determined between these curves, deforming the surface between these curves based on the distance function of the surface within the sulcus.

0 0

Post a comment