Improved Activation Detection in fMRI Studies Using Acquisition and Reconstruction Methods Robust to Susceptibility and Motion Artifacts
Pandey K, Noll D
University of Michigan
Head motion degrades accuracy and sensitivity of fMRI studies by introducing variability and causing erroneous activation. Motion causes dynamic changes in R2*, amount of signal loss and image distortions from susceptibility. Effectiveness of motion correction algorithms is limited since they do not account for varying susceptibility artifacts during image registration. In fMRI studies, we found that acquisition methods robust to susceptibility induced off-resonance, namely, combined forward and reverse spiral images, significantly improve the quality of activation detection in motion-corrupted data sets. Furthermore, iterative reconstruction with dynamically updated fieldmaps was able to preserve the fidelity of activation maps even in presence of large motion.