fMRI Stop Signal Changes

[nqthqn]

Hi @abadi199 if we could chat about this tonight that would be great :)

reminder: Water should be 50% blue, 50% gray, Healthy should be blue border (always), Unhealthy should be gray border (always)

Fixes/changes ordered by priority

• [x] Current Incorrect: Is not using invalid images and Is assigning gray border to valid
• [ ] Move instructions blocks to page that is awaiting trigger — concerned with consistent onset times.
• [ ] Try to replicate 404 - don't reuse /stopsignal route (/fmri/:username? Restrict to admin and staff role users.)

Per our call: May need to rethink fMRI stop signal implementation.

[abadi199]

Issue 1 is fixed by this PR #55 Issue 2 is a lot harder than since I don't currently have the game info/instruction when we're on the "Start Game" state, so I will need to move the game info outside of the game implementation, or I can cut corner and just duplicate the game info, which I don't really want to do. Do you think we can push back on this change?

[nqthqn]

@abadi199 it's ok to take the shortcut - the fMRI game is such a specific use case I doubt we will ever start pulling game instructions from the backend.

I'm not sure how to push back as I don't understand the motivation fully. Say you have two participants data:

ins = 10
image 1 = 20
image 2 = 30
image 3 = 40

ins = 120
image 1 = 132
image 2 = 136
image 3 = 146

subtracting 10 and 120 respectively from everything zeros it out and converts it to relative offsets:

ins = 0
image 1 = 10
image 2 = 20
image 3 = 30

ins = 0
image 1 = 12
image 2 = 16
image 3 = 26

Notice how image 1 from both sets have different onsets, now imagine the onsets of the brain scans and trying to align this across multiple sets of onsets.

From my perspective you can zero out from image 1

ins = -10
image 1 = 0
image 2 = 10
image 3 = 20

ins = -12
image 1 = 0
image 2 = 4
image 3 = 14

But then you are still left with different relative onsets for the instructions which must be accounted for per each participant - since the scan still started at the instructions block.

If instead the true start is image 1, Sonja won't have to disregard a fluctuating amount of time (the instructions block) for each scan.

I'm trusting Sonja on this that she knows what she wants.