dafni

Getting Data

What’s the plan?

Unit	Topic
2	Data acquisition (scanning on 3T)

Data acquisition in the scanner

fMRI data will be acquired in ~30min sessions (in small groups) on one of our 3T scanners. Have a look at the webpage for the 3T scanners on campus to learn a bit more about the machines. Up until last year we ran our experiments on the 3t Achieva, but the SPMIC has now decommisioned that machine to make space for the new, national 11.7T facility which is currently being planned in detail.

Two important sets of things to consider:

• what are the parameters / settings on the MRI scanner?
• what is the subject doing in the scanner (stimuli & task)?

Scanner - Protocol

The protocol will be pretty standard for a cognitive neuroscience scanning sessions. The plan for the time in the scanner is as follows

1. Quick survey scan to allow “planning” on console (< 10s)
2. A test EPI to make sure that the slice positioning for the fMRI experiment is ok.
3. fMRI experiment (block): gradient-echo EPI, TR 1.5s (~4min)
4. T1w-MPRAGE: to illustrate detailed (1mm isotropic) anatomy (~5min)
5. T2w-FLAIR: (1mm isotropic)

Stimulus code

• stimulus code: inspect FFAlocaliser for lots of details.
• timing: block design 12s images (faces, objects), 12s rest: a 24s cycle. 10 repeats per scan - 240s or 160 dynamics (@ 1.5s). Stimuli were images from face and object image database (details below).

The code is written in matlab/mgl using the task library that comes with mgl. Written by Alex Beckett and DS based on a version of a working code from Justin Gardner

There are a couple of short youtube videos explaining a version of the FFA localiser and the fixation dimming task to control attention, which would be used in a real experiment. This should give you a sense of what the subject is doing inside the scanner.

The experiment runs as a simple block design in the following order:

[faces, rest] , [objects, rest] - …

The length of each [stimulus, rest] cycle is determined by the cycleLength (in TRs).

To run, make sure the stimulusCode folder is on the path and then simply run the following command. the Escape key can be used to stop the experiment at any point:

FFAlocaliser % quick test to see what's going on

To run at the MR centre, we also want to specify TR, not to run in a small window, etc. So probably worth setting a few parameters in the call like this:

FFAlocaliser('TR=1.5', 'debug=0', 'numBlocks=10', 'cycleLength=12')

Scanner - actual numbers from the day (2023/24)

2024-02-07, Denis Schluppeck

4 volunteers (sub-01 .. sub-04), scanned on the 3T GE scanner at the SPMIC QMC site. Scanner operator: AC. Start time 1400h.

(Data available via moodle link to a zip file on OneDrive).

For each person we obtained several scans. See json sidecar copied along for some details.

• T1w MPRAGE (1mm isotropic)
• T2w FLAIR (1mm isotropic)
• one or two repeats of an fMRI experiment (FFAlocaliser.m); 2.2mm isotropic, TR/TE 1500ms/35ms

fMRI experiment and timing

• the original scan was 168 timepoints long, but the first 8TRs (12s) were cut to allow for steady state.
• after removing these initial dummies, the time series is 160 timepoints long
• timing of the experiment is (Faces-rest-Obj-rest)*5)

12s ON (faces)
12s OFF (gray)

12s ON (objects)
12s OFF (gray)

... then each repeated for a total of 10 stimulus-rest blocks 

File organisation

The zip file contains data from the 4 participants in the following layout:

.
├── Readme.md
├── sub-01
├── sub-02
│   ├── sub-02-FLAIR.png
│   ├── sub-02-MPRAGE.png
│   ├── sub-02-flair.json
│   ├── sub-02-flair.nii.gz
│   ├── sub-02-fmri01.nii.gz
│   ├── sub-02-fmri02.nii.gz
│   ├── sub-02-mprage.json
│   └── sub-02-mprage.nii.gz
├── sub-03
└── sub-04

Notes

For FSL design matrix rules, this means for EVs

setting	Faces	Objects
Skip	36	36
Off	36	36
On	12	12
Phase	36	12
Stop After	-1	-1
Convolution	Gamma	Gamma

Previous scanning sessions

• 2022-23 cohort

• 2019-20 cohort

• 2018-19 cohort

• 2017-18 cohort

Notes

• working code with face and object images
• parameters to set cycle length, TR, number of block
• youtube clip explaining the fixation task and experiment
• write out text file in correct format for fsl/feat analysis.
• test with actual scanning parameters

Materials for stimuli:

We will provide the stimulus code (written in Matlab / MGL) in line with what happened in - Learning Matlab / C84NIM - a pre-requisite for this course.

We’ll run a “Faces versus houses / scenes localiser, as this works well and is a very robust experiment.

• Faces download: https://wiki.cnbc.cmu.edu/images/multiracial.zip

Stimulus images courtesy of Michael J. Tarr, Center for the Neural Basis of Cognition and Department of Psychology, Carnegie Mellon University, https://www.tarrlab.org/. Funding provided by NSF award 0339122.

• Objects download: https://bradylab.ucsd.edu/stimuli/Exemplar.zip

Object stimuli from: Brady, T. F., Konkle, T., Alvarez, G. A. and Oliva, A. (2008). Visual long-term memory has a massive storage capacity for object details. Proceedings of the National Academy of Sciences, USA, 105 (38), 14325-14329.

• Scenes download: https://timbrady.org/stimuli/Scenes.zip

Scene stimuli from: Konkle, T., Brady, T. F., Alvarez, G.A. and Oliva, A. (2010). Scene memory is more detailed than you think: the role of categories in visual long-term memory. Psychological Science, 21(11), 1551-1556.