Thursday, November 9, 2017

not all MB4 fMRI scans are free of gradients and banding

I was encouraged by the images in the previous post: it looked like the gradient and banding ("washing out") artifacts were not visible in the MB4 people. I need to revise that a bit: I do have bands and gradients in at least some MB4 people, though to a much lesser degree than in our MB8 datasets. How much this affects our task analyses is still to be determined.

On the left is the same MB4 person and run from the previous post (27 September 2017), voxelwise standard deviation over time of a task fMRI run, after going through the HCP minimal preprocessing pipelines. I was glad to see that the vessels were brightest (as they should be), though concerned about the frontal dropout. The person on the right is another person scanned at MB4 (doing the same task; same encoding, scanner, etc.); same preprocessing and color scaling for both.

The vessels clearly don't stand out as much in the person on the right. It's hard to tell in the above image, but there's a gradient in the standard deviation and tSNR images, with better signal on the edges than in the center of the brain. Below is another run from the person on the right, tSNR calculated on the images ready to go into afni for GLM calculation (so through the HCP minimal preprocessing pipelines, plus smoothing and voxelwise normalizing). This tSNR image is shown at six different color scalings; it's easier to see interactively (download the NIfTI here), but hopefully it's clear that the darker colors (lower tSNR) spreads from the center of the brain to the edges, rather than uniformly.

Here is the same person and run again, with the standard deviation (first two) and tSNR (right pane) of the raw (no preprocessing, first and lower) and minimally preprocessed (right two) images. I marked a banding distortion with green lines, as well as the frontal dropout. The banding is perfectly horizontal in the raw image, at 1/4 of the way up the image (see larger image), which makes sense, since this is an MB4 acquisition. I included the larger view of this raw image since all three banding artifacts are somewhat visible; in our dataset the inferior band is generally the most prominent.

The banding and gradient artifacts are certainly less visually prominent in our MB4 than our MB8 images, but they are present in some MB4 people. I haven't systematically evaluated (and probably won't be able to soon) all of our participants, so don't have a sense of how often this occurs, or how much it impacts detection of task BOLD (which is of course the key question).

Below the jump: movement regressors for the two runs in the top image; the person with the banding and gradients had very low motion; less than the person from the September post.

Realignment parameters from the minimal preprocessing pipelines, plus FD and enorm, for the two runs shown in the top image. The left-hand person (from the September post) is first, marked with an "S"; the person on the right (and other images) is in the second plot. This new person had very little motion (or even apparent motion) in their runs.

  Vertical grey lines mark one-minute intervals (entire run about 13 minutes); horizontal green lines are task blocks.


  1. Hi Jo, My bet is you're looking at another effect of head motion. It's been 5+ years since I looked specifically at motion sensitivity in SMS, but here is a quick assessment of what happens to tSNR for (intentional) motion during the single band reference acquisition, then no intentional motion during the accelerated time series, versus no intentional motion during the SBRef, then movement only during the time series.

    (These are old slides from a 2013 talk. I can't find the original data on my laptop but I almost certainly have them on a DVD in my office. Remind me next week!)

    What's interesting is the different motion sensitivity for SMS compared to acceleration in-plane, e.g. with GRAPPA. (I don't have GRAPPA comparison data here.)

    With in-plane GRAPPA, movement during the ACS has a huge effect on the tSNR, whereas movement during SBRef for SMS has more subtle effects. Note the main loss of vessels and edge signal for motion during SBRef in slide #2. Movement during the time series reduces tSNR globally and to a greater extent than during the SBRef, suggesting that the mismatch between the SBRef and the accelerated data in the time series is the bigger motion sensitivity for SMS. Put another way, any mismatch between the SMS time series and the SBRef produces substantial loss of SNR.

    This is is as far as I've tested motion sensitivity in SMS because other than the obvious - eliminate motion! - I don't have any ideas for increasing robustness. The same thing goes for post-ACS movement with in-plane GRAPPA acceleration. Clever folks like Jon Polimeni have come up with methods like FLEET to reduce motion sensitivity in ACS (conventional GRAPPA), but I've not yet seen anyone offer improvements in the face of mismatch between ACS or SBRef and the accelerated time series, for in-plane GRAPPA or SMS, respectively.

    1. What caught my interest is that the MB4 person with the visible standard deviation banding and gradients has *less* motion in the realignment parameters than did the MB4 person with better looking standard deviation plots (27 September 2017 post person). I haven't looked at the SBRef images at all; sounds like they might be useful for understanding signal quality.

    2. Hi Jo, definitely be interesting to see if the SBRef offer any clues. I recall when we were looking at in-plane GRAPPA motion sensitivity we realized many subjects might move at the start of a run either because the operator just talked to them and they need to swallow or readjust, or they think "hey I know not to move but the noise just started so I'll quickly scratch my nose then be still." Our tactic was to add a few "smart dummy" scans - say, five volumes - of no interest that occur after the ACS but before the first action in a stimulus script. (For resting state the tactic would be to add the five volumes to the total and then disregard the first five TRs of the series.) On a Siemens scanner this produces images you can see in the Inline Display. You make a go/no go decision based on what you see in those first five volumes. If there is any suggestion of movement you abort the run before any stimuli are presented, and start over. If the operator is inexperienced it may take more than five TRs to evaluate and call the ball, but most people can do it in 4-5 sec with experience. I've not been instructing people on this tactic for SMS-EPI because it hasn't seemed necessary.