Correction of metal-induced susceptibility artifacts for functional MRI during deep brain stimulation

• Published in: NeuroImage (Orlando, Fla.) Fla.), 2017-09, Vol.158, p.26-36
• Main Authors: In, Myung-Ho, Cho, Shinho, Shu, Yunhong, Min, Hoon-Ki, Bernstein, Matt A., Speck, Oliver, Lee, Kendall H., Jo, Hang Joon
• Format: Article
• Language: English
• Subjects: Algorithms; Animals; Boundaries; Brain; Brain - physiology; Deep brain stimulation (DBS); Deep Brain Stimulation - methods; Electrodes; Electrodes, Implanted; fMRI; Geometric distortion; Humans; Image Processing, Computer-Assisted; Intensity distortion; Magnetic Resonance Imaging - methods; Metallic artifacts; Metals; Point spread function; Point spread functions; Reverse gradient approach; Stimulation; Susceptibility artifacts; Swine
• ISSN: 1053-8119; 1095-9572
• DOI: 10.1016/j.neuroimage.2017.06.069

Functional magnetic resonance imaging (fMRI) is an emerging tool for investigating brain activation associated with, or modulated by, deep brain stimulation (DBS). However, DBS-fMRI generally suffers from severe susceptibility to artifacts in regions near the metallic stimulation electrodes, as well as near tissue/air boundaries of the brain. These result in strong intensity and geometric distortions along the phase-encoding (PE) (i.e., blipped) direction in gradient-echo echo-planar imaging (GE-EPI). Distortion presents a major challenge to conducting reliable data analysis and in interpreting the findings. A recent study showed that the point spread function (PSF) mapping-based reverse gradient approach has a potential to correct for distortions not only in spin-echo EPI, but also in GE-EPI acquired in both the forward and reverse PE directions. In this study, we adapted that approach in order to minimize severe metal-induced susceptibility artifacts for DBS-fMRI, and to evaluate the performance of the approach in a phantom study and a large animal DBS-fMRI study. The method combines the distortion-corrected GE-EPI pair with geometrically different intensity distortions due to the opposing encoding directions. The results demonstrate that the approach can minimize susceptibility artifacts that appear around the metallic electrodes, as well as in the regions near the tissue/air boundaries in the brain. We also demonstrated that an accurate geometric correction is important in improving BOLD contrast in the group dataset, especially in regions where strong susceptibility artifacts appear. •PSF mapping-based reverse gradient approach can be adapted to minimize susceptibility artifacts in fMRI.•This approach can minimize susceptibility artifacts that appear around the metallic electrodes.•Susceptibility artifacts near the tissue/air boundaries in the brain can also be resolved.•With the distortions minimized, local BOLD contrast in DBS-fMRI can be improved.