Intra- and interobserver variation in CT-evaluated cerebral atrophy.

Four radiologists evaluated the internal and external atrophy of 94 patients into four grades. The CT scans were evaluated three times without measuring procedures in order to determine the inter- and intraobserver variation. In the third evaluation referential scans were used. A considerable inter- and intraobserver variation was experienced. When using referential scans, both inter- and intraobserver variation was reduced to two consecutive grades. In cases where the estimation of atrophy is of specific importance the visual procedure is not reliable.