Image artifacts

﻿
﻿
﻿
﻿
*Wraparound in the superior-inferior direction. Notice the inhomogeneity in image intensity (presumably due to surface coil hardware issues). Also, possibly weird Gibbs circular type artifacts.*Wraparound in the superior-inferior direction. Notice the inhomogeneity in image intensity (presumably due to surface coil hardware issues). Also, possibly weird Gibbs circular type artifacts.
﻿
﻿
﻿
(**OLD**) Major bug, unusable, garbage. The SNR is "fine", but some weird pulse sequence or reconstruction or phase-artifact issue. 1.5/4?.(**OLD**) Major bug, unusable, garbage. The SNR is "fine", but some weird pulse sequence or reconstruction or phase-artifact issue. 1.5/4?.
﻿
﻿
*Massive ghost, horrible image quality.*Massive ghost, horrible image quality.
﻿
﻿
*This is tSNR results for a sample fMRI run. The holes are parts of the brain that aren't being listened to. Head coil with limited anterior coil elements*This is tSNR results for a sample fMRI run. The holes are parts of the brain that aren't being listened to. Head coil with limited anterior coil elements
﻿
﻿
*Checking out the acquired data right after you acquire it at the scanner. Very large low spatial frequency fluctuations from volume to volume. Pretty drastic, very bad for fMRI data.  The overall thermal noise level is pretty low (good). Notice the distortion stuff in the frontal aspect.*Checking out the acquired data right after you acquire it at the scanner. Very large low spatial frequency fluctuations from volume to volume. Pretty drastic, very bad for fMRI data.  The overall thermal noise level is pretty low (good). Notice the distortion stuff in the frontal aspect.
﻿
﻿
*This is a very high-res for inplane dimensions T2-weighted (HRT2). Notice the weird line-like artifacts near the top ("ghosting?"). Also, notice the "stippling"/"rippling" horizontal band (flow artifact from big arteries).*This is a very high-res for inplane dimensions T2-weighted (HRT2). Notice the weird line-like artifacts near the top ("ghosting?"). Also, notice the "stippling"/"rippling" horizontal band (flow artifact from big arteries).
﻿
﻿
﻿
*T1-weighted volume. There is rippling at the top (subject motion?). Fairly bad. Will cause issues with automated segmentation.*T1-weighted volume. There is rippling at the top (subject motion?). Fairly bad. Will cause issues with automated segmentation.
﻿
﻿
﻿
*Diffusion-weighted imaging. The right hand side is horrible (alarm bells). Maybe an interaction of head motion and eddy currents and the interleaved slice acquisition. Even after full pre-processing and full analysis of all of the diffusion volumes, the result on the left still has signatures of the artifact.*Diffusion-weighted imaging. The right hand side is horrible (alarm bells). Maybe an interaction of head motion and eddy currents and the interleaved slice acquisition. Even after full pre-processing and full analysis of all of the diffusion volumes, the result on the left still has signatures of the artifact.
﻿
﻿
﻿
Logan - Axial montage of EPI data, probably 0.8mm at 7T (maybe 0.5...). The rear door in the MRI room was slightly open for the aquisition on the left and closed for the one on the right. The artifact is visible as a horizontal bit of static above the brain. This time varying artifact looks minimal and outside of the brain, however the next image shows the temporal standard deviation over time.Logan - Axial montage of EPI data, probably 0.8mm at 7T (maybe 0.5...). The rear door in the MRI room was slightly open for the aquisition on the left and closed for the one on the right. The artifact is visible as a horizontal bit of static above the brain. This time varying artifact looks minimal and outside of the brain, however the next image shows the temporal standard deviation over time.
﻿
﻿
*Logan - Images are scaled to match maximum values - the timeseries is generally more noisy (higher stdev value) even within the brain, 'distant' from the artifact location. *Logan - Images are scaled to match maximum values - the timeseries is generally more noisy (higher stdev value) even within the brain, 'distant' from the artifact location. 
﻿
﻿
Logan - 7T, 0.8mm data, perhaps an off resonance artifact or the effect of bad shimming. The slice-specific nature makes me (logan) think this is more a hardware issue. It occured a while into the scan, so only some runs were affected. Takeaway - always look at your data while scanning!Logan - 7T, 0.8mm data, perhaps an off resonance artifact or the effect of bad shimming. The slice-specific nature makes me (logan) think this is more a hardware issue. It occured a while into the scan, so only some runs were affected. Takeaway - always look at your data while scanning!
﻿
﻿
﻿
Logan- movie visible phase artifacts (and drop out, distortion).Logan- movie visible phase artifacts (and drop out, distortion).
﻿
*Logan - These are two statistical images showing activation in response to task requiring foot motion - participant moved foot forward and back with specific timing. In the lower image, the only difference is that the participant held their breath while performing the task. The task activation is, surprisingly, still visible, but the remainder of the brain had a strong negative deflection due to breath holding - which becomes visible as a negative (blue) fit to the model prediction.*Logan - These are two statistical images showing activation in response to task requiring foot motion - participant moved foot forward and back with specific timing. In the lower image, the only difference is that the participant held their breath while performing the task. The task activation is, surprisingly, still visible, but the remainder of the brain had a strong negative deflection due to breath holding - which becomes visible as a negative (blue) fit to the model prediction.
﻿
﻿
*Logan - too much acceleration (MultiBand + GRAPPA) was used - the image itself is degraded, only the largest, most prominent features (edge of brain, ventricles) are visible, no clear gray/white matter contrast. *Logan - too much acceleration (MultiBand + GRAPPA) was used - the image itself is degraded, only the largest, most prominent features (edge of brain, ventricles) are visible, no clear gray/white matter contrast. 
﻿
﻿
clear striping artifactclear striping artifact
﻿
*These vertical bands are not biological. The stripes are "too" vertical and straight. Biology is never nice and regular to that extent.*These vertical bands are not biological. The stripes are "too" vertical and straight. Biology is never nice and regular to that extent.
﻿
﻿
﻿
*Clearly visible k-space related artifact*Clearly visible k-space related artifact
﻿
﻿
﻿
weird stripe in fMRI data caused by eyetracker camera communication to host computer(?).  leads to very high std dev (the first part of the movie)weird stripe in fMRI data caused by eyetracker camera communication to host computer(?).  leads to very high std dev (the first part of the movie)
﻿