Neuroimaging

Question 1

MRI scanner

Answer 1

• Incredibly strong magnet: usually 1.5 - 7 tesla
• 1.5 tesla = 15,000 gauss
• Earth’s magnetic field strength = 0.5 gauss
• Exeter MRI scanner = 30,000 X earth’s magnetic field strength
• NO METAL IN THE SCANNER!!!

Question 2

what does BOLD stand for?

Answer 2

Blood Oxygen Level Dependent

Question 3

BOLD signal

Answer 3

• When neurons become active, blood flows to the part of the brain to provide oxygen to fuel the cells.
• Haemoglobin (the iron-containing oxygen transporting protein present in blood) differs in how it responds to magnetic fields, depending on whether it has a bound oxygen molecule
• The MRI scanner, which is basically a giant magnet, detects these small changes in the magnetic field.
• With fMRI we are not directly measuring brain activation.
• We are measuring blood flow (more precisely, the magnetic properties of oxygenated vs deoxygenated blood)

Question 4

from raw data to functional ‘activation’ maps

Answer 4

1. Design a task to be used in the scanner
2. Collect some data
3. Preprocess the data
4. Analyse the data
5. Interpret your results

Question 5

fMRI experimental design

Answer 5

• BOLD signal is arbitrary: It has no stable baseline - means that the baseline BOLD signal may be something in one session for one subject, and completely different for the same subject on the next day
• Therefore, the most important aspect of any fMRI experiment is that we have to provide both a) an experimental condition and b) a baseline condition.
• Resulting functional brain map reflects the difference between these two conditions.

Question 6

what is a good baseline?

Answer 6

• One that differs from the experimental condition only by the process of interest
• In general, having the subject rest, i.e. do nothing, is not a good baseline, because there will be so many things differing between that and your experimental condition that you never know what process your brain activation reflects.
• Ideally you want a baseline that differs from the experimental condition only according to the cognitive process you’re interested in.

Question 7

block designs

Answer 7

• BOLD signal is slow - peaks 4-5 seconds after stimulus onset and takes around 16 seconds to return to baseline
• All fMRI experiments originally employed block designs - long periods of alternating task/baseline performance

Question 8

disadvantages of block designs

Answer 8

• Block designs often group together lots of trials
• Highly predictable occurrence of stimuli: subjects know what is coming and may alter strategies accordingly (not always a pro)
• Inflexible for more complex tasks: impact of oddball stimuli? or stimuli or events that occur uncontrollably?
• Ecological validity. Does blocking trials change the psychological process you are interested in?
• Can’t separate trials by performance - e.g. to look at activation associated with correct vs incorrect response

Question 9

event related design

Answer 9

Trials of different conditions are randomly intermixed and occur close together in time

Question 10

advantages of event-related design

Answer 10

• Flexibility and randomisation - eliminate predictability of block designs - avoid practice effects
• Post hoc sorting - (e.g., correct vs. incorrect, aware vs. unaware, remembered vs. forgotten items, fast vs. slow RTs)
• Can look at novelty and priming - rare or unpredictable events can be measured - e.g., P300
• Can look at temporal dynamics of response - Dissociation of motion artefacts from activation - Dissociate components of delay tasks - Mental chronometry

Question 11

collect some data

Answer 11

• 2 to 3 seconds to collect a single ‘volume’
• To reference each point in the brain in 3D space we divide image into cubes or ‘voxels’ - Typical voxel size = 3x3x3 mm - Refer to each voxel with coordinates (x, y, z)

Question 12

preprocess the data

Answer 12

• Correcting for non-task-related variability in experimental data - Getting rid of the ‘noise’
  1. High pass filtering
  2. Motion correction
  3. Slice time correction
  4. Co-registration
  5. Normalisation
  6. Spatial smoothing

Question 13

high pass filtering

Answer 13

• Remove low frequency oscillations such as scanner drift that introduce noise into your data
• Standard low pass filter is approx 120 secs
• The consequence of this is that we can’t contrast events more than a couple of minutes apart because they will be wiped out by our filter.

Question 14

slice time correction

Answer 14

• Because it takes a few seconds to collect a whole brain volume, different slices may be acquired up to a couple of seconds apart
• This means that our estimate of the haemodynamic response will be incorrect for slices acquired later
• Slice timing simply corrects this by moving them earlier in time

Question 15

coregistration

Answer 15

Each functional image is aligned with the subject’s structural image so that functional maps can subsequently be overlaid and anatomical landmarks identified

Question 16

normalisation

Answer 16

• Ultimately we want to do statistics on group activation maps, so we have to get all the brains into a standard space
• To do this we use complex algorithms to warp each subject’s brain into the shape of a template brain

Question 17

normalisation - standardises space

Answer 17

• Montreal Neurological Institute (MNI) space - Combination of 352 MRI scans on normal controls
• Can compare activation across subjects and do group analyses.

Question 18

smoothing

Answer 18

Application of Gaussian kernel - allows closer neurons to influence the value more strongly than distal ones.

Question 19

why smooth?

Answer 19

• Because neurons do not fire in isolation.
• If a neuron fires, neurons close to it will tend to fire as well.
• Smoothing attempts to model the data according to this property of neurons.

Question 20

analysis

Answer 20

• Multiple regression is used to determine the effect of a number of independent variables (our conditions) on a single dependent variable (brain activation)
• For each voxel, we use multiple regression to estimate how closely the BOLD signal correlates with the timecourse of each condition.
• Finally perform a contrast - simply a t-test comparing beta condition 1 beta values to condition 2 beta values
• Calculating the t-statistic for each of the 130,000 voxels in the brain is equivalent to asking “in which regions is activation greater in condition 2 than in condition 1?”
• Raw t-map indicates the magnitude of the t-statistic using a colour scale - Yellow = bigger t values
• The threshold we apply defines what we accept as a significant difference in activation between our 2 conditions
• Essentially arbitrary (as it is in any statistical procedure)
• Standard alpha level in psychological research is p < 0.05, i.e. we accept a 5% rate of false positives

Question 21

multiple regression

Answer 21

Simply a way to find out the explanatory effect of a number of independent variables on a dependent variable.

Question 22

correction for multiple comparisons

Answer 22

• Brain images divided into up to 130,000 voxels = 130,000 individual t-tests
• Chance of a type 1 error (false positive) increases with every test performed
• With an alpha level of p < 0.05 (standard used in psychological research) with one t-test we have a 5% chance of type 1 error.
• With 100 independent t-tests, we have a 99.4 % chance of type 1 error
• In other words, with 130,000 t-tests and an alpha level of 0.05 we are guaranteed to find some significant voxels simply by chance.
• Very important to correct for multiple comparisons - adjust alpha level (p-value)

Question 23

different approaches to data analysis

Answer 23

1. Whole brain analysis

2. Region of Interest (ROI) analysis

Question 24

advantages of WBA

Answer 24

• Requires no prior hypotheses about areas involved

- Includes entire brain

Question 25

disadvantages of WBA

Answer 25

• Can lose spatial resolution with inter-subject averaging
• Can produce meaningless “laundry lists of areas” that are difficult to interpret
• Depends highly on statistics and threshold selected
• Multiple comparisons problem

Question 26

WBA

Answer 26

examine effects on a voxel by voxel basis across the whole brain

Question 27

ROI analysis

Answer 27

Restrict our analysis to a particular brain region

Question 28

advantages of ROI analysis

Answer 28

• Hypothesis driven - avoids meaningless ‘laundry lists’ of activated regions.
• Avoids multiple comparisons problem - data summarised in a single number per subject reflecting mean activation across voxels in ROI.
• Simple - data can be exported and treated as any other type of data, requiring no special software for further analysis
• Generalisable - data can easily be compared across studies, e.g. for meta analysis.

Question 29

disadvantages of ROI

Answer 29

• Easy to miss things going on elsewhere in the brain.

- Not always simple how to define ROIs

Question 30

limitations of fMRI

Answer 30

• fMRI data is correlative, therefore we can’t say that a region activated during a task/function is essential for that task/function.
• Need converging evidence from TMS/neuropsychology to make stronger inferences about causation.
• Temporal resolution of fMRI is low
• Need converging evidence from EEG/TMS to provide finer grained temporal information.