Background Suppression with Simultaneous Chemical Shift Artifact Reduction In Arterial Spin Labeling
Abstract
Purpose
Chemical shift artifacts from fat in echo planar sequences are difficult to suppress in artificial spin labeling (ASL) imaging without negatively impacting the label signal. The goal of this work is to suppress these artifacts by creating an angular dependency on the complex-value fat, brain, and ASL signals that allows for suppression of fat signals. To accomplish this, we created a custom pseudo-continuous ASL (pCASL) sequence using two different background suppression (BS) timings in one human subject.
Methods
Images were collected from one healthy volunteer on a 3T MR scanner (Cima.X, Siemens, Erlangen, Germany) using a 64-channel receive-only head-coil. A pCASL sequence was created using Pulseq in MATLAB. Two conditions, A (0.369s, 1.047s, 2.204s, and 2.767s) and B (0.510s, 1.577s, 2.203s, and 2.845s) were created through different BS timings and ten label-control pairs were collected for each condition. Complex-value label and control images were averaged independently for each condition, with one condition scaled prior to subtraction across conditions. Resulting complex-value label-control pairs were then pair-wise subtracted to yield perfusion images.
Results
The combination of the conditions, A and B, after scaling allowed for fat tissue suppression while retaining the ASL signal when performing complex-value processing prior to magnitude reconstruction. Standard magnitude subtraction or complex-value processing without image scaling retained chemical-shift artifacts in the label and control images.
Conclusion
We demonstrated that chemical shift artifacts from fat-shifted signals can be suppressed without substantially impacting the ASL signal by varying the BS timings between acquisitions and performing complex-value processing and scaling prior to pair-wise subtraction.