Task 5

Question 1

Positron emission tomography (PET

Answer 1

 Measured variations in cerebral blood flow (glucose, …)
 Radioactive substance is injected (tracer, usually radioactive oxygen H215O or the recently found PiB to diagnose Alzheimer’s), which emits radiation that is monitored. The greater the blood flow in a region, the greater the signal emitted by the tracer in that region.

The gamma rays that are produced when the a positron and an electron collide (substance decays and emits a positron from the atomic nuclei). The PET scanner is therefore rather a gamma ray detector.
 Results are reported as a change in regional cerebral blood flow (rCBF) between a control and an experimental condition
 Where there is more blood flow, there will be more radiation
 PET measures relative activity, not absolute metabolic activity
Metabolic activity is resolved in regions / voxels (5-10mm

PET imaging requires sufficient time to detect enough radiation to create images of adequate quality . The participant must be engaged continually in a single given experimental task for at least 40 s, and metabolic activity is averaged over this interval.

Question 2

Functional magnetic resonance imaging (fMRI)

Answer 2

 Imaging focuses on the properties of deoxygenated hemoglobin (deoxyhemoglobin), which is paramagnetic compared to oxygenated hemoglobin.
 fMRI measured the ratio of oxy- to deoxygenated hemoglobin (blood oxygen level-dependent = BOLD)
 Active areas show an increase in oxygenated hemoglobin (opposite as expected) because the amount of blood directed to that area increases, causing an over-availability and excess of oxygen.
 Indirect measure of activity via BOLD

Question 3

why the BOLD response increase only multiple seconds after stimulus onset.

Answer 3

While neuronal processes occur within milliseconds, changes in blood flow occur much slower

Question 4

block design

Answer 4

the recorded neural activity is integrated over a “block” of time during which the participant either is presented a stimulus or performs a task. The recorded activity pattern is then compared to other blocks that have been recorded while doing the same task or stimulus, a different task or stimulus, or nothing at all.

Question 5

Event-related design

Answer 5

across experimental trials, the BOLD response will be linked to specific events such as the presentation of a stimulus or the onset of a movement. That way, a clear signal can be obtained by averaging over repetitions of these events.

Question 6

Limitations of PET and fMRI

Answer 6

Poor temporal resolution
(PET= decay rate of tracer, FMRI= dependant on hemodynamic changes that underly bold signal)

Difficulty interpreting data
(Data sets are massive and many differences arise in comparisons between experimental and control conditions)

Correlation ≠ Causation
PET= signal is vague and active areas are not necessarily critical

FMRI=Bold signal is rather drive by inpur then output
causation can not directly be implied , but fmri is helpful in establishing functional connectivity

Question 7

(f)MRI vs. PET

Answer 7

+Scanners are less expensive & easier to maintain
+ No radioactive tracers are needed and the same individual can be tested repeatedly
+ Higher spatial resolution (high resolution anatomical images)
+ Testing can be completed fast (under 1h)
+ Better discrimination between grey and white matter
- PET is less susceptible to signal distortion

While PET measures the blood flow, fMRI measured the concentration of oxygen in the blood. => both Hemodynamic measures

Question 8

Structural imaging

Answer 8

Measures of the spatial configuration of different types of tissue in the brain (principally CT and MRI)
 Based on the fact that different types of tissue have different physical properties that can be used to construct a static map of the physical structure of the brain
 Most common methods are CT and MRI

Question 9

Functional imaging

Answer 9

Measures temporary changes in brain physiology associated with cognitive processing; the most common method is fMRI and is based on a hemodynamic measure

Based on the assumption that neural activity produces local physiological changes in that region of the brain which are used to create a dynamic map of neural activity

Question 10

PET

Answer 10

Based on blood volume 
Involves radioactivity 
Participant is scanned only once 
Temporal resolution = 30s
Spatial resolution = 10mm
Must use a blocked design 
Sensitive to the whole brain

Question 11

MRI

Answer 11

Based on blood oxygen concentration
Involves magneticity
Participant is scanned multiple times
Temporal resolution = 1-4s
Spatial resolution = 1mm (depends on size of voxel)
Can use a blocked and an event-related design
Some brain regions are hard to image

Question 12

CT

Answer 12

• Constructed with X-ray absorption of different tissue types (amount of absorption is related to tissue density)
• Used in clinical settings to identify tumours, …
• Cannot distinguish between white and grey matter as good as MRI can and cannot provide any functional images

Question 13

Magnetic resonance imaging (MRI)

Answer 13

To obtain MRI scans, the following sequence is applied:

• Strong magnetic field is applied across the brain (strength is measured in Tesla – the more Tesla (T) the stronger the magnetic field)
• Single protons (found in water molecules) have their own (weak) magnetic fields and will align with the magnetic field that is produced by the scanner
• The aligned protons are hit with a radio frequency pulse that knocks their original orientation by 90 degrees and makes them spin / precess. Before you apply this pulse, the protons are in the same direction but do not spin in the same precessing phase. By applying the pulse, we set all the protons in the
• The spinning protons produce a detectable change in the magnetic field and this is what forms the basis of the MR signal
• The scanner will pull the protons bac into their original position and let them relax and then repeats the whole process

Question 14

Different types of images can be obtained with this method: (MRI)

Answer 14

• Images that distinguish tissues / structural images: variations of the rate at which the protons return to the aligned state (T1 relaxation time: measures how quickly the protons give back the energy to the tissue. This happens faster in fatty tissues which causes them to appear whiter in the scan then water tissues)
• Functional images (fMRI): in the misaligned (90°) state, the MR signal decays because of local interactions with nearby molecules (T2 component: measures how quickly the spins diphase again, T2* is sensitive to the oxygenation in the blood). Because the deoxyhaemoglobin produces more distortions because it is paramagnetic and is attracted to the magnetic field. The oxygenized haemoglobin is diamagnetic.

Question 15

• Voxel-based morphometry (VBM

Answer 15

= A technique for segregating and measuring differences in white matter and gray matter concentration
 Divides the brain into many regions (=voxels) and estimates the concentration of white/gray matter in each voxel
 Here we can see the structure if the brain

Question 16

• Diffusion tensor imaging (DTI)

Answer 16

Uses MRI to measure white matter connectivity between brain regions

 We use diffusion weighted imaging and can see the connectivity of the brain
 water molecules trapped in axons tend to diffuse in some directions but not others. With functional anisotropy, we can find out to which extend diffusion takes place in some directions more than others.

Question 17

Hemodynamic response function (HRF)

Answer 17

Hemodynamic response function (HRF) = how the BOLD signal evolves / changes over time in response to an increase in neural activity. It has 3 phases:

1) initial dip: because of consumption, when neurons eat oxygen deoxyhaemoglobin increases at first and therewith reduces the BOLD signal (0-2s)
2) overcompensation: In response to the increased consumption of oxygen, the blood flow to the region increases. This increase is greater than the consumption, leading to an excess of oxygen at that region. This is the component of interest in fMRI (6-10s)
3) Undershoot: blood and oxygen dip before returning to their original levels (relaxation), causing a temporary increase in deoxyhaemoglobin. We do not have to wait for the BOLD response to return to baseline in order to present another trial, because different HRFs can be superimposed. Many experiments do wait until it returns to baseline (16-20s)

Question 18

Safety and ethical issues in functional imaging research

risks

Answer 18

Risks are small but a bit higher in PET (due to the radioactivity used).
In fMRI, large magnets (>3T) can cause dizziness and nausea (enter the field gradually to avoid this). Also, fMRI is very loud (120dB), so all participants must wear earplugs

Question 19

exclusion criteria

Answer 19

PET: pregnant women and children
fMRI: people with metal body parts, implants, pacemakers, women wearing contraceptive coils, people suffering from claustrophobia, etc. cannot participate, no eye make-up, no metal spectacles or metal objects in general

Question 20

To minimize individual differences, the date of many participants is averaged in the following way:

Answer 20

• Stereotactic normalization = mapping of individual differences in brain anatomy onto a standard template
• Smoothing = Redistributing brain activity from neighbouring voxels to enhance the signal-to-noise ratio
• Flow diagram: summarizes the sequence from initial hypothesis to data interpretation

furthermore participants are instructed to keep still as possible , they’re heads are restrained -becasue moving the head could lead to regions being harder to detect or false-positve results

Question 21

voxel

Answer 21

a volume-based unit (cf. pixels, which are 2D); in imaging research the brain is divided into many thousands of these

Question 22

The BOLD signal is more sensitive to

Answer 22

neuronal input than to output (sensitive to post-synaptic potential vs. action potential – similar to EEG). It could therefore be that areas that simply receive input (“listen”) appear to be active.

Question 23

activation

Answer 23

An increase in physiological processing in one condition relative to some other condition(s)
 areas that are more active in Task A

Question 24

deactivation

Answer 24

A decrease in physiological processing in one condition relative to some other condition(s)
 areas that are more active in Task B

Question 25

Functional imaging has its limitations:

Answer 25

• Some regions appear active but are not critical for the task
• The tasks that are compared could differ in a general way (i.e. difficulty that asks for more attention)
   Attention might be favourable but not necessary
   Functional imaging gives us a better idea of which regions may be sufficient for performing a particular task but not always which regions are crucial and necessary for performing a task

Functional imaging data can be enhanced when used together with other methods. It can however not replace other methods such as lesion-based neuropsychology (see figure).

Question 26

The radiofrequency signal emitted by the spins depends on

Answer 26

the number of spins involved

Question 27

the decay of the signal depends on

Answer 27

relaxation times that describe the interaction between spins and their surroundings.

Question 28

The name MRI reflects this process:

Answer 28

magnetic (nuclear magnetic spins) resonance (the matching of frequency between the RF pulse and the precession of the spins) imaging (the process by which the signal measured by the MR scanner is spatially encoded and the computer algorithm that produces the images).

Question 29

Haemoglobin has different magnetic properties depending on its concentration:

Answer 29

• Fully saturated with O2 (oxyhaemoglobin): diamagnetic
• Not fully saturated with 02 (deoxyhaemoglobin): paramagnetic

The proportion of deoxyhemoglobin relative to oxyhemoglobin dictates how the MR signal will behave in a BOLD image: areas with high concentration of oxyhemoglobin give a higher signal (a brighter image) than areas with low concentration

Question 30

BOLD

Answer 30

• Pros: practical and easy implementation, flexible
• Limitations: not quantifiable or related to a specific physiological parameter, but is a complex interplay between cerebral blood flow, volume, and cerebral metabolic rate of oxygen (CMRO2)

Question 31

Temporal resolution and experiment duration

Time to repetition (TR)

Answer 31

the parameter that regulates temporal resolution = the time it takes to collect one brain volume (composed of many slices) = the time between two excitation pulses

 The speed comes with the expense of space: the shorter the TR, the lesser the number of slices collected by TR  thus brain coverage is limited
 You can also reduce spatial resolution to get a high temporal resolution (you can be quicker and have a lower spatial resolution or we “decrease our field of view”)
- So, TR lower than 1.5s provides ~12% more statistical power, but other acquisition parameters (i.e., flip angle) have to be changed in order to avoid contribution from blood inflow in the signal

• In practice:
  o 2min expended per condition
  o Not more than 12min per run
  o Total section time less than 40 min

Question 32

The higher the voxel size,

Answer 32

the lower the spatial resolution

Question 33

There are 2 ways to receive both a high temporal and spatial resolution:

Answer 33

1) Jittering = Use of different delays between the start of the sampling of brain volume images relative to the start of stimulus presentation to the subject (different time points of sampling)
- If all images are collected with the same delay from stimulus presentation (the ‘time-lock’ strategy) all brain regions would be sampled at the same time points at every ISI, with periodicity rate exactly equal to the TR
- But, if one jitters (offsets) the stimulus presentation time to image acquisition, then different time points would be sampled at each stimulus presentation
- This can be achieved by using an ISI that is not a multiple of TR (fixed jittering scheme) or varying the ISI (variable jittering)
- The acquisition of one brain volume with high spatial resolution can take up to 6s, and will still be able to sample time points closely spaced in time
- = increased total acquisition time
- Cons: requires more trials
- Recommended if need full brain coverage + temporal resolution, but behaviour analysis and long scanning sections are not critical

2) Parallel acquisition schemes: technique based on the spatial coding of signals from coil sensitivity profiles
 We do not measure one slice after the other but multiple slices at the same time
- Reduces the acquisition time by a factor set by the experimenter (usually between 2 and 3)
- Reduces the amount of susceptibility artefacts, improving the detection of signal from basal frontal and mesial temporal regions
- Positive impact on studies involving memory, emotion, and executive function tasks
- How? uses the differences in the MR signal measured by the coils—which depends on the proximity of the part of the body
- Cons: SNR ratio is reduced, but this seems not to impact negatively with the same magnitude on the BOLD effect sensitivity

Question 34

What determines the boundaries of temporal and spatial resolution of fMRI?

Answer 34

• Trade-off between them
• Temporal resolution is limited by the sluggishness of the hemodynamic response (it takes 2 seconds for the BOLD response to starts, 4-10 seconds until it peaks, …)
• Spatial resolution is limited by the strength of the magnetic field (the higher the field strength the more we can “zoom in”) but we can never look at things that are smaller than our blood vessels

Question 35

FMRI PROS

Answer 35

1. Flexible
2. High spatial resolution
3. Relatively high temporal resolution
4. Lack of ionizing radiation or need for external contrast agent

Question 36

CONS FMRI

Answer 36

Need to spend time resolving questions related to
1. Spatial and temporal resolution
2. Limits to brain coverage
3. Image artifacts in sophisticated experimental designs
Or need to image artifacts in sophisticated experimental designs

Question 37

Studies of commonalities in brain activation

Answer 37

If two tasks lead to activation of common brain areas, then these two tasks or behaviours are likely to share some process or processes. Thus, fMRI can be used to infer the cognitive processes involved in one task by showing similarities in brain activation to a better understood task.
- Eisenberger et al.: brain activity during the experience of social exclusion is similar to the activity found during the experience of physical pain

Question 38

firm can help with

Answer 38

Studies of commonalities in brain activation
(If two tasks lead to activation of common brain areas, then these two tasks or behaviours are likely to share some process or processes. Thus, fMRI can be used to infer the cognitive processes involved in one task by showing similarities in brain activation to a better understood task.)

Studies of distinctiveness in brain activation
{(Researchers want to discover distinctive activations between two tasks. This infers that two tasks have different cognitive processes mediating them.
Most findings yield results of partial overlap in activations (distinctiveness that is found may be quantitative rather than a qualitative one).

Documenting individual differences
(- Canli et al.: found consistent activation in the amygdala among participants when viewing fearful facial expressions, but inconsistent activation when participants viewed happy facial expressions
 Individual variability could be predicted by measuring the person’s extraversion score (the more extraverted, the higher the activation in amygdala)

Testing psychological models

Question 39

firm can use both

Answer 39

fMRI can use both a block design or an event-related design, which improve the design because experimental and control trials can be presented randomly. Here, the experimenter can also choose to combine the data in many different ways after scanning

Question 40

physiological underpinning of fmri

Answer 40

The brain consumes 20% of our body’s oxygen, which it gets via blood. As metabolic activity of neurons increases, the blood supply to those neurons increases as well, in order to meet their additional oxygen demands. Since all neurons in the brain constantly need oxygen (in order to simply survive), the whole brain is “active” all the time. To see activity that is unique to a certain task, we must therefore look at the brain region that presents with a greater oxygen amount relative to other conditions and areas.

While PET measures the blood flow, fMRI measured the concentration of oxygen in the blood.
 Hemodynamic measures (measure brain activity indirectly by measuring a consequence of neural activity like blood flow or oxygen)

Question 41

PET Ablauf

Answer 41

Positron emission tomography (PET)
PET uses a radioactive tracer injected into the bloodstream. The greater the blood flow in a region, the greater the signal emitted by the tracer in that region. The most commonly used tracers are oxygen-15, administered in the form of water, and fluorine-18, administered in the form of a glucose sugar. We can also use radiolabelled neurotransmitters (i.e. to study the effects of drugs on the brain).
Once in the bloodstream, the tracer converts back from its radioactive form into a normal one and while doing so, emits a particle (called a positron) that then collides with an electron, releasing two photons (emit gamma rays) that can be detected by detectors positioned around the head.
It takes 30 sec for the tracer to enter the brain and a further 30 sec for the radiation to peak to its maximum. This is the critical window for obtaining changes in blood flow related to cognitive activity, which explains the mediocre temporal resolution of around 30sec.
Temporal resolution = accuracy with which one can measure when a cognitive event is occurring

Question 42

fmri Ablauf

Answer 42

Over the last 10 years functional magnetic resonance imaging (fMRI) has overtaken PET scans in functional imaging experiments because of its multiple advantages.
The component of the MR signal that is used in fMRI is sensitive to the amount of deoxyhemoglobin in the blood. When neurons consume oxygen they convert oxyhemoglobin to deoxyhemoglobin. Deoxyhemoglobin has strong paramagnetic properties and this introduces distortions in the local magnetic field. This distortion can itself be measured to give an indication of the concentration of deoxyhemoglobin present in the blood. This technique has therefore been termed BOLD.

Question 43

analysing data fro. functional imaging

Answer 43

Analysing data from functional imaging 
The pictures that we receive from these methods are not literal pictures of the workings of the brain but rather pictures of the brain regions that are computed to be statistically significant given the type of design used. Because statistics are involved, these pictures are susceptible to errors. 
Individual differences (like the size or location of the sulci) are dealt with by assuming that these differences do not exist or at least that they do not negatively impact the claims made about the brain function.

Question 44

Stereotactic normalization

Answer 44

Each brain is divided up into voxels, each with different spatial coordinates.
Voxel = a volume-based unit (cf. pixels, which are 2D); in imaging research the brain is divided into many thousands of these
This enables us to map the voxels of one brain onto another and we need to do this to be able to compare the brains to each other. We do this by stretching the brain out into a “standard space”. To interpret this standard space brain we use the brain atlas by Talairach.
Talairach coordinates = Locations in the brain defined relative to the atlas of Talairach and Tournoux, which is based on anatomical data from a single post-mortem brain
 Origin at anterior commissure
 X-coordinate = left and right
 Y-coordinate = front and back
 Z-coordinate = top and bottom

Question 45

Smoothing

Answer 45

The smoothing process also occurs on the standard space brain. It spreads some activation to neighbouring voxels (the closer the neighbour the more activation it gets).
Figure: consider voxel D4 and L8
 D4 gets switched on
 L8 gets switched off
Smoothing thus enhances the signal-to-noise ratio and allows for analysing groups of participants. It increases the spatial extent of active regions therefore, the probability to find common regions of activity in multiple participants is greater.

Question 46

Statistical comparison

Answer 46

Usually, statistical analyses on the (corrected) data are analysed according to the question whether the activity in one voxel in the experimental condition was greater than during the baseline. In functional imaging however, we have thousands of voxels to consider and if the significance level would be p<0.05 (1 in 20 chance level) as usual, many voxels would be active just by chance.
A more conservative criteria (lower significance level) could be implied but this could lead to a type 1 error (failure to detect regions of importance).
 If the sea level (significance) is too high, there are no islands to observe. If the sea level is too low, there are islands everywhere.
The application of the Bonferroni correction (dividing the p-value by the number of tests / voxels) brings the problem that activity at each voxel is not independent. To solve this, general methods of correction, the family wise error and the false discovery rate were developed.
Family wise error (FEW) = An approach for correcting for many statistical comparisons based on the number of tests being conducted
 Involves choosing a statistical threshold based on spatial smoothing
 Look at clusters of voxels rather than on single voxels
False discovery rate (FDR) = An approach for correcting for many statistical comparisons based on the number of positive results obtained
 a more conservative statistical threshold is used for datasets in which lots of voxels are active than in a dataset in which only few voxels are active
We use these corrected (more conservative) significance levels to prevent detecting lots of regions just by chance. However, if the interest is in one particular voxel, then it is possible to use an uncorrected significance level.

Question 47

Inhibition vs. excitation

Answer 47

The fact that a region is more active in one task than in another does not always imply that this region is essential for the task. The increased signal could simply reflect the use of general mechanisms (increased attention, …) or the fact that a region receives input but does not respond to it (inhibition).

Inhibition = A reduction/suppression of the activity of a brain region (or a cognitive process), triggered by activity in another region/ process

Question 48

Interpreting data from functional imaging

Answer 48

Activation vs. deactivation
Inhibition vs. excitation
Necessity vs. sufficiency

Question 49

Necessity vs. sufficiency

Answer 49

• Some regions appear active but are not critical for the task
• The tasks that are compared could differ in a general way (i.e. difficulty that asks for more attention)

Question 50

Activation vs. deactivation

Answer 50

Activation and deactivation refer to the sign of the difference in signal between 2 conditions (Task A – Task B = +/-), while inhibition and excitation refer to the nature of the signal.
Activation = An increase in physiological processing in one condition relative to some other condition(s)
 areas that are more active in Task A
Deactivation = A decrease in physiological processing in one condition relative to some other condition(s)
 areas that are more active in Task B

Question 51

fmri-how it works

Answer 51

Nuclei / protons in the blood and other body parts (like water) have a magnetic moment (due to their spin) and align with the magnetic field produced by the MRI. We will then have a certain number of protons (proportional to the water content of the tissue) aligned with the main magnetic field.
The effect of these aligned spins is to produce a bulk magnetization that precesses (the circular motion that displays as it spins under the influence of gravity) around the direction of the magnetic field with a specific frequency (known as the Larmor frequency), directly dependent on the magnitude of the magnetic field

Larmor frequency = the rate of precession of the magnetic moment of the proton around the external magnetic field.
 Determines the precessing rate and depends on the strength of the magnetic field
 We need to know this frequency in order to apply the correct radiofrequency pulse
By applying a radiofrequency (RF) pulse with a frequency exactly matching the precession frequency, the orientation of the spins can be changed until their magnetic moments are perpendicular to the main magnetic field. In this orientation, the precessing spins will induce a voltage in a surrounding electrical circuit (like spinning magnets within a generator produce electricity).
After the RF pulse ceases, the spins slowly return to their original orientation, but not before this radiofrequency voltage can be detected by a suitable antenna (or coil), placed around the area of the body / brain.
The radiofrequency signal emitted by the spins depends on the number of spins involved and the decay of the signal depends on relaxation times that describe the interaction between spins and their surroundings.

Question 52

fMRI additionally makes use of BOLD contrast imaging. Haemoglobin has different magnetic properties depending on its concentration0ß:

Answer 52

The proportion of deoxyhemoglobin relative to oxyhemoglobin dictates how the MR signal will behave in a BOLD image: areas with high concentration of oxyhemoglobin give a higher signal (a brighter image) than areas with low concentration. This is because the increase in blood flow related to neuronal function is also accompanied by an increase in oxyhemoglobin concentration in a particular ‘activated’ area of the brain (even though we would have expected the opposite).
The BOLD effect is also influenced by cerebral blood flow and volume, and as such is not a simple measurement parameter.
A set of images covering the whole brain (a brain volume) is typically acquired every 2–3 s, and (to increase sensitivity) hundreds of brain volumes are typically accumulated during the execution of a complete fMRI scan, lasting around 2-10 min. The acquired data is the compared to a model of the expected BOLD response and any signal changes detected are statistically tested for significance, allowing detection of small increases in the signal of the brain areas correlated with the behaviour.

Question 53

Spatial resolution and brain coverage

Answer 53

One would want data with the smallest voxel size possible, and acquisition of the whole encephalic tissue available in a subject. The spatial resolution unit is the voxel (volume element) that represents the minimum unit of brain tissue sampled in each image.
 The higher the voxel size, the lower the spatial resolution
If we want to reduce voxel size in order to produce more spatially specific information (less susceptibility to artefact) we must keep in mind that this has a negative impact on the signal-to-noise ratio of the images, reducing the sensitivity of the BOLD (smoothing).
If the spatial resolution in increased, while the temporal resolution remains fixed, the amount of brain tissue (number of image slices) has to be reduced. If, in another situation, the spatial resolution is increased and the brain coverage is maintained, then the temporal resolution has to be decreased.
 We need to find a compromise
Smoothing tends to reduce the effective spatial resolution but facilitates the conventional spatial resolution. This is used in group mapping techniques and signal modelling.

Question 54

What determines the boundaries of temporal and spatial resolution of fMRI?

Answer 54

• Trade-off between them
• Temporal resolution is limited by the sluggishness of the hemodynamic response (it takes 2 seconds for the BOLD response to starts, 4-10 seconds until it peaks, …)
• Spatial resolution is limited by the strength of the magnetic field (the higher the field strength the more we can “zoom in”) but we can never look at things that are smaller than our blood vessels

Question 55

Image analysis strategies

Answer 55

• Choice of image analysis depends on the hypothesis
• No consensus on best method
• Common steps used by most data packages:
  1. Data are pre-processed: movement correction (and realignment), spin history correction, and optional steps (spatial/temporal filtering, re-sampling and re-ordering of data)
  2. The HRF is modelled to represent the fMRI signal evolution in time in a hypothetically activated region: General rule: use mathematical functions representing the empirical ‘shape’ of the BOLD response, more often represented by gamma or Poisson functions
  3. Statistical inference is performed via parametric or nonparametric approaches: Gaussian random Weld theory, permutation testing, analysis at cluster level, correction for multiple comparisons, and methods to deal with signal noise
  First level analysis: This will result in an activation map of a single subject (the areas surviving a statistical threshold are presented in a colour)
   Single participant
  Second level analysis: interrogate questions based on group statistics  voxel-based transformations are used in algorithms that ‘normalize’ and ‘register’ each subject’s brain volume = all data are analysed in a common space
   Average data over whole group
   You can treat 10 subjects as 1 subject (you just put your trials together by random-effects analysis)

Question 56

how can scanner acoutic noise be minimised

Answer 56

a) Reducing the source of the noise
b) Using the hemodynamic delay of the BOLD response, and inserting ‘silent periods’ in the acquisition process  most common method
c) Parallel acquisitions
- present the stimulus and collect data insensitive to scanner acoustic noise, but to the cost of an increase in acquisition time