short-axis PROPELLER

One of the more interesting abstracts from last year’s ISMRM in Seattle has now been published as a full manuscript:

Propeller EPI in the other direction

A new propeller EPI pulse sequence with reduced sensitivity to field inhomogeneities is proposed. Image artifacts such as blurring due to Nyquist ghosting and susceptibility gradients are investigated and compared with those obtained in previous propeller EPI studies. The proposed propeller EPI sequence uses a readout that is played out along the short axis of the propeller blade, orthogonal to the readout used in previous propeller methods. In contrast to long-axis readout propeller EPI, this causes the echo spacing between two consecutive phase-encoding (PE) lines to decrease, which in turn increases the k-space velocity in this direction and hence the pseudo-bandwidth. Long- and short-axis propeller EPI, and standard single-shot EPI sequences were compared on phantoms and a healthy volunteer. Diffusion-weighted imaging (DWI) was also performed on the volunteer. Short-axis propeller EPI produced considerably fewer image artifacts compared to the other two sequences. Further, the oblique blades for the long-axis propeller EPI were also prone to one order of magnitude higher residual ghosting than the proposed short-axis propeller EPI.

Skare S et al, Magn Reson Med. 2006 Jun;55(6):1298. PMID: 16676335

The basic idea here is brilliant in its simplicity: rather than acquire individual blades with the readout direction going the length of the blade, acquire them with the readout direction along the (usually shorter) width. And use EPI instead of FSE to crank up the speed. The higher bandwidth allows for substantially reduced geometric distortions relative to traditional EPI, as the authors indicate.

It’s worth noting that this paper was also referenced by Chuang et al in the previous Journal Club paper. SAP-EPI can’t quite get as good results on its own as the aforementioned technique, but the two can certainly be combined. Note that this paper didn’t try a DTI acquisition, though they did do some diffusion-weighted scanning. I don’t have much more to add as the concept is pretty straightforward; I imagine that we will see more of SAP-EPI in clinical applications fairly soon.