“…The first, Euclidean distance δ, is simply a measure of the squared error between the resulting tomo gram and the actual test profile. It is defined by [12] (Xj X)) 1 (7) where η is the number of cells, Xj is the reconstructed estimate, and X) is the assigned test profile. Another, in a sense more qualitative, method of com parison is that of boundary definition by minimum variance partitioning [13].…”