Published normative volumetrics of the hippocampus (HC) vary substantially. While the protocol suggested by Watson et al. (Neurology 42 (1992) 1743-1750; Arch Neurology 54 (1997) 1521-1531) is the most frequently adhered to, this leaves the posterior section of the HC tail unmeasured, which has been estimated to be in the order of 2-4 mm, representing 5-10% of total HC volume. The objective of the current study was to compare HC volumes according to the method of Watson et al. (Neurology 42 (1992) 1743-1750; Arch Neurology 54 (1997) 1521-1531) against those measured to include the posterior tail section. From a random community sample of 60-64 yr old individuals, 478 subjects underwent magnetic resonance imaging brain scans. Of these, 452 scans (238 males and 214 females) were adequate for hippocampal measurement. The scans comprised whole brain T1- weighted and T2-weighted FLAIR images. One hundred and fifty scans were randomly selected for the measurement of HC volumes beyond the opening of the crus of the fornix by manual tracings on T1-weighted images by a trained operator. Intracranial volume (ICV) and total brain volume (TBV) were measured using an automated program. We found that the posterior HC tail extended for a mean of just over 5 mm and comprised 11% of total HC volume. Males had significantly larger raw HC volumes, and while normalization with ICV or TBV reversed this pattern, it was significant only when the posterior HC tail was included in the measurement. In conclusion, this study showed that including the posterior part of the tail can influence the results of HC measurement. An argument is presented that accurate HC volumes should include the entire HC and not exclude the tail.
Copyright 2005 Wiley-Liss, Inc.