Neuroimaging

Question 1

MRI scanners

Answer 1

1. originally developed for structural imaging
2. more recently used for fMRI
3. non invasive
4. high spatial resolution
5. reasonably affordable
6. uses very strong magnets

Question 2

blood oxygen level dependent signal

Answer 2

1. when neurons become active, blood flows to the part of the brain to provide oxygen to fuel the cells
2. haemoglobin differs in how it responds to magnetic fields, depending on whether it has a bound oxygen molecule
3. the mri scanner detects these small changes in the magnetic field
4. with fMRI we are not directly measuring brain activation

Question 3

raw data to functional brain activation maps

Answer 3

1. brains don’t just light up in the scanner
2. test designed, data collected, preprocess the data by correcting for nontaskrelated variability in experimental data

Question 4

fMRI - experimental design

Answer 4

1. BOLD signal is arbitrary, it has no stable baseline
2. therefore the most important aspect of any fMRI experiment is providing an experimental and baseline condition
3. baseline - differs from the experimental condition only by the process of interest

Question 5

block vs event related designs

Answer 5

1. BOLD signal is slow - peaks 4-5 seconds after stimulus onset and takes around 16 seconds to return to baseline
2. all fMRI experiments originally employed block designs - long periods of alternating task/baseline performance
3. limitations - highly predictable occurrence of stimuli, inflexible for more complex tasks, ecological validity, can’t separate trials by performance
4. event related design - trials of different conditions are randomly intermixed and occur close together in time
5. advantages of event related designs - in freeing us from the necessity of block designs, event related enables us to design more complex and novel experiments, eliminates the predictability of block designs, avoids practice effects

Question 6

standardised space

Answer 6

montreal neurological institute combined 352 MRI scans on controls to create a standardised brain

Question 7

analysis

Answer 7

1. multiple regression is used to determine the effect of a number of independent variables on a single dependent variable
2. for each voxel, we use multiple regression to estimate how closely the signal correlates with the time course of each condition - resulting beta values tells us how close this correlation is
3. perform a contrast - ttest of beta values in condition 1 and 2
4. raw t map indicates the magnitude of the t statistic using a colour scale
5. the threshold we apply defines what we accept as a significant difference
6. is arbitrary

Question 8

correlation for multiple comparisons

Answer 8

1. brain images divided into up to 130000 voxels in the brain = 130000 t tests
2. chance of a type 1 error (false pos) increases with every test performed

Question 9

whole brain analysis approach

Answer 9

1. examine effects on a voxel by voxel basis across the whole brain
2. advantages - requires no prior hypothesis about the areas involved, includes entire brain
3. disadvantages - can lose spatial resolution due to intersubject averaging, can produce meaningless lists of areas that are hard to interpret

Question 10

region of interest analysis approach

Answer 10

1. restricting analysis to a particular brain region
2. advantages - hypothesis driven (no meaningless lists), simpler as data can be exported and treated as normal. generalisable across studies
3. disadvantages - easy to miss any other associations somewhere else in the brain

Question 11

limitations of fMRI

Answer 11

1. correlative data, we can’t say that a region activated during a task/function is essential, need converging evidence
2. temporal resolution is low, need additional evidence from EEG