Hitoshi Fukasawa
Department of Pathology, Research Institute for Brain anb Blood Vessels, AKITA
Abstract.
A computer-assisted morphometric method is described for estimating the volume and surface area of the cerebral structures from their cut-surface images.
Procedure: After fixation in 15% formalin, the supratentorial brain is sectioned coronally at intervals of d (ca. 10mm); the thicknesses of the slices at both ends are measured (d', d"). All the cut surfaces are numbered (1 
i n), photographed and printed at a constant magnification of about 1.5. The boundary lines of the main structures seen on cut surfaces are classified into several categories: the cortical line, subcortical line, ventricular line, etc. Each of the classified line is reduced to a set of constituent simple closed curves. These are extracted in digital form, input into the computer, and the perimetric lengths as well as the encircled areas are calculated. The input data are also processed to be displayed on CRT as lines of different colors according to the classification.
The cerebral cortical volume (Vc) is estimated from the series of its cut-surface area (Ai) using the equation
Vc = 
A1 + d 
Ai + 
An
on the basis of the trapezoid rule. The cortical surface area (Sc) is obtained by the equation
Sc = 
{(n-1)d + d' + d"} 
Li/n
where Li is the length of the cortical lines on each cut surface. This is derived from a basic equation of stereology relating the surface density to the boundary density on test area. A transformation of a plane concave figure into the smallest circumscribed convex figure is introduced; segmental application of this transformation to the cortical line is an effectual means to obtain a standard measure for the cortical folding.