Longitudinal sampling and aliasing in multi-slice helical computed tomography

Helical pitches 3 and 6 are sometimes called “preferred” in 4-slice helical computed tomography (CT), because it is believed that the effective longitudinal sampling intervals at these pitches are equivalent to those in single-slice helical CT operating at pitches 1 and 2, respectively. While these equivalences have been supported by comparative studies of slice-sensitivity profiles in single- and multi-slice helical CT, artifacts have been observed in pitch-3 and pitch-6 multi-slice images that were not evident in their purported single-slice counterparts. These differences have inspired us to look more closely at longitudinal sampling and aliasing in multi-slice helical CT. In doing so, we have found that the sampling patterns in question are not strictly equivalent, and specifically that under certain conditions longitudinal aliasing is evident in multi-slice reconstructions that is not present in their single-slice counterparts. The difference between the multi-slice, pitch-3 and single-slice, pitch-1 results is attributed to the small cone angle in multi-slice helical CT, which introduces inconsistencies among the measurements of different detector rows. The difference between multi-slice, pitch-6 and single-slice, pitch-2 results is attributed to a combination of the cone angle and genuine differences in sampling patterns. It is argued, however, that the lack of strict sampling equivalence with single-slice counterparts does not necessarily undermine the claim that pitches 3 and 6 are “preferred” relative to other pitches in multi-slice helical CT