Help Privacy Policy Disclaimer
  Advanced SearchBrowse





Image Distortion in 7 Tesla and its significance for high-field Amygdala neurofeedback

There are no MPG-Authors in the publication available
Fulltext (restricted access)
There are currently no full texts shared for your IP range.
Fulltext (public)
There are no public fulltexts stored in PuRe
Supplementary Material (public)
There is no public supplementary material available

Van der Meer, J., Hellrung, L., In, M.-H., Gotting, F., Borchardt, V., & Walter, M. (2018). Image Distortion in 7 Tesla and its significance for high-field Amygdala neurofeedback. Poster presented at 24th Annual Meeting of the Organization for Human Brain. Mapping (OHBM 2018), Singapore.

Cite as: https://hdl.handle.net/21.11116/0000-0003-C6FF-F
Even though the higher spatial resolution in 7T imaging is a boon for real-time fMRI BOLD
imaging, a primary concern at 7 Tesla for obtaining a neurofeedback signal is image distortion in difficult-to-image regions near boundaries of tissue and air. Signal variability is -more so than at 3T- altered by image distortions, due to longer readout times of the EPI sequence (to obtain higher special resolution), and shorter T2*. Image distortions can be as large as 20-30 voxels with 160x160 readout matrix, and distributed along the entire brain. Therefore, we assess effects of distortion correction for fMRI Neurofeedback outcome measures.

Image distortion in EPI imaging can affect fMRI timeseries and as a consequence, could impair quality of a RT-fMRI NF experiment. In this work we present the effect of using a real-time distortion correction on the quality of the NF signal in an Amygdala neurofeedback experiment, by prospective off-line comparison of corrected and uncorrected timecourses [1]
The data were obtained from a 15 subjects at 7T using an echo-planar imaging scanning sequence covering a part of the brain including the amygdala and cingulate cortex, with a resolution of 1.4x1.4x2.0 mm3. A practice run was followed by training 1,2 and 3 runs concluded with a transfer run. The experiment consisted five 8-minute runs consisting of rest (r), up-regulation (u) and counting (cnt) blocks; see Fig. 1 (C). Interleaved with the blocks, feedback(f) was given, but not during Transfer Run. During the experiment, data was real-time motion corrected, real-time distortion corrected [2] and real-time transferred to an external computer using TCP/IP which drove the experiment with RtExplorer [1]. We retrospectively examine the effect of applying distortion correction on learning effects. Learning effects is defined by the increase-over-runs of % image intensity of (up - count)/count in the Amygdala Mask.
Our results show that the distortion correction restored the distortions in the frontal lobe (Fig 1A) by pushing back signal up to 20 voxels. We also found that 1) the average distortion-corrected timeseries have a different shape in Amygdala ROI than the uncorrected timeseries and 2) that the average image intensity within the Amygdala is higher for the uncorrected EPI images: 478 instead of 402. Within the off-line analysis of the self-regulation capabilities across runs we found similar learning effects for both distorted and undistorted data, albeit with differences in the rate of change across runs.
The most pronounced effect of image distortion can be spotted in frontal regions but expands also in more basal regions such as the Amygdala. Exploratory analysis of the effects of distortions in EPI images show that employment or omission of a real-time distortion correction has implications regarding neurofeedback training signal and effect sizes at 7 Tesla. These differences are likely due to the distortion correction 'pushing' the signal in the right anatomical location. The impact of distortion effects even increases with higher field strengths. At this point it is difficult to state whether the distortion-corrected or uncorrected images represent the 'true' BOLD changes throughout time. However, for anatomical correctness, an on-line EPI distortion correction is a critical step in the fMRI Neurofeedback processing pipeline at 7T.