Background/Aims: In liver anatomy and surgery, is portal and hepatic vein segmentation (French segmentation) to be preferred over arteriobiliary segmentation (Healey and Schroy, North American segmentation)? Methods: Several embryological arguments and an analysis of anatomical data from a personal collection of 110 vasculobiliary casts were made. Results: Embryological arguments: Portal vein branching appears first, arteriobiliary branching secondly follows the portal vein distribution. Segment II (the left lateral sector) is the development of the right lateral embryological lobe. The umbilical vein enters the left portion of the middle embryological lobe, forming segment IV on the right and segment III on the left: this is the left paramedian sector. So the left portal fissure (between left and middle lobes) transversally crosses the classical left lobe, which is not a portal unit. Segment VI is a late secondary prominence of segment VII, reaching the anterior margin of the liver only in man, Anatomical arguments: hepatic vein seg mentation must be added to portal segmentation; the academic left lobe is the left hepatic vein sector, and the left hepatic fissure separates the classical right and left robes. Portal vein segmentation must be preferred: portal vein duplication of branches of first order occurs only in 23.5% of the cases, while arteriobiliary duplication of first-order branches is noted in 50% of the livers, portal segmentation being much simpler. Conclusions: Portal and hepatic vein segmentation seems to be much more accurate. Copyright (C) 1999 S. Karger AG, Basel.
