In this paper we present a graph-based volumetric data segmentation method based on a 3D hexagonal prismatic lattice. We evaluate the advantages and disadvantages of using this lattice in contrast with classic ones. One of the main advantages are high isoperimetric quotient, near equidistant neighbours (ability to represent curves better, resulting in a better segmentation) and high connectivity. Disadvantages are due to the main stream lack of interest in this area and thus data sets must be converted back and forth from rectangular to hexagonal latices both in acquisition and visualization processes.