Topic 2: Imaging Review

Question 1

What is the persuasive power of the dead salmon poster?

Answer 1

the persuasive power of the dead salmon poster largely comes from the fact that it provides a rare example in which we actually know with certainty that a statistically significant result is a false positive

this was especially problematic given the large number of datapoints that are typically collected in a fMRI study, related to the concept of a “voxel”

Question 2

What is a type 1 error?

Answer 2

the more comparisons you make between various data points, the more likely you are to find a statistically significant result purely by chance (i.e., a type 1 error) if you don’t take the appropriate statistical steps in order to control for multiple comparisons

Question 3

What are functional magnetic resonance imaging (fMRI) analyses?

Answer 3

fMRI analyses “draw” what we can think of as imaginary boundaries based on three-dimensional pixels (voxels) that are superimposed onto the brain

analyses typically compare estimates for activity levels across different (groups of) voxels

e.g., are the voxels that overlap the region in space that the amygdala occupies more active during the “scary picture” condition, as compared to a control condition?

Question 4

How were false positives found in the dead salmon poster?

Answer 4

many early fMRI datasets were analyzed by comparing the activity level of each voxel to every other voxel in the dataset

this means that the rate of false positives for studies that did not correct for multiple comparisons are assumed to be greatly inflated

Question 5

What are the confounds that are involved in the Christmas spirit network?

Answer 5

one possibility could be a confound related to memory

for example, many people who celebrate Christmas and have positive associations with it have more memories associated with the Christmas stimuli, which could lead to greater activation (or “recruitment”) of memory-relevant brain regions

we could speculate about lots of other potential reasons why the implicated regions may have difference in activation levels across groups

among other things, this also demonstrates the gap between the objective data on the one hand, and the (inherently, at least somewhat) subjective interpretation on the other

Question 6

What is localization of function?

Answer 6

another consideration that could be relevant for methodological and/or interpretation reasons relates to localization of function, or the idea that we have specific brain regions (or networks) that are specialized for certain kinds of stimuli and/or engaging in particular kinds of processing

localization of function is often assumed, and may even be evident in data, yet can nevertheless sometimes be an oversimplification

Question 7

Is the fusiform face area a good example of localization of function?

Answer 7

the fusiform face area (or FFA) has been heavily studied and tends to become more active when we process stimuli that is more “face-like”

this has been taken to mean that the FFA is causally responsible for our ability to perceive faces

as more people publish studies that fit with this idea, more people continue looking for supporting evidence (confirmation bias?)

while there is lots of evidence to back up these claim, are there some other possibilities?

Question 8

What are other explanations for the localization of function in the fusiform face area?

Answer 8

one kind of alternative interpretation of the FFA is that is simply one of many regions in the brain that contribute to the perception of not only faces, but also various other kinds of stimuli

while the FFA has been observed to be approximately twice as active when people are presented with face and face-like stimuli (as compared to non-face stimuli), it is also sometimes found to be more active for non-face stimuli (as compared to baseline)

so, although it responds to faces, it can also respond to other kinds of stimuli too… does that justify how it gets discussed (“the face center”)? how does that line up with the conventional understanding of what it does?

Question 9

What is phrenology?

Answer 9

a belief system that attempted to relate variation in the shape of different parts of the skull with behavior and cognition

although not done with scientific rigor, this general idea is similar to the modern concept of localization of function

Question 10

What is mass action?

Answer 10

proposed that specific functions were in fact not localized to particular parts of the brain

Question 11

What is equipotentiality?

Answer 11

assumes any chunk of cortical tissue has the potential to support any brain function

Question 12

What is plasticity?

Answer 12

tissue in different parts of the brain can “take over” for other damaged parts and effectively relocate functions, albeit there are limitations

Question 13

What kind of evidence would support a mass action view?

Answer 13

kidney analogy: damage is similar regardless of whether it is located in the lower, middle, or upper portion, simply reduces it’s capacity/efficiency

Karl Lashley example: location of lesion didn’t seem to matter, just the size

Question 14

What early work supports the localization of function of language?

Answer 14

Broca and Wernicke’s area

Question 15

What early work supports the localization of function of visual perception?

Answer 15

lesions in one hemisphere cause problems in the opposite visual field (because of contralateral wiring), posterior damage affects vision but anterior damages doesn’t, etc.

Question 16

What early work supports the localization of function of motor functions?

Answer 16

“Jacksonian March” seizures, a particular type that starts at the fingers and spreads up the arm

the advancement of sterile surgery practices opened the door to the systematic/experimental study of brain lesions using stereotaxic surgery

Question 17

Who was Herbert Jasper?

Answer 17

one of the pioneers in the use of electroencephalograms (EEG) to study brain function

Jasper worked with Wilder Penfield, who used electrodes to stimulate various parts of the brain to map out what parts of the brain did what in particular patients (to know what parts of tissue should/shouldn’t be removed to help reduce epileptic symptoms)

e.g., could map out the motor cortex and avoid removing chunks of tissue that are required to control movement

Question 18

What is high spatial resolution versus high temporal resolution?

Answer 18

as a generalization, we can contrast imaging methods that have high spatial resolution with those that have high temporal resolution

high spatial: fMRI, PET

high temporal: EEG/ERP, MEG

Question 19

What is an electroencephalogram (EEG)?

Answer 19

EEG provides an overall measure of electrical activity emanating from the brain on the basis of the signal that reaches electrodes placed on the scalp

this electrical activity fluctuates between different characteristic frequencies, which are each associated with particular states of consciousness

there are various clinically-relevant applications for EEG technology (e.g., screen for seizure activity)

Question 20

What are event-related potentials (ERPs)?

Answer 20

while the raw waveform produced by an EEG can be informative in and of itself, many cognitive psychologists are (depending on the research questions they’re asking) interested in using EEGs to measure event-related potentials (ERPs)

ERP’s measure average characteristic changes in electrical activity associated with particular psychological events (e.g., a stimulus presentation)

calculation of ERP requires signal averaging to form a grand average for multiple subjects across many trials that is time-locked to a specific event

Question 21

What are the ERP naming conventions?

Answer 21

have two parts, a letter or a number

the letter: N (for negative deflection, i.e. decrease in voltage) or P (for positive deflection, i.e. increase in voltage)

the number: 1, 2, or 3 (or 100, 200, or 300), depending on how many milliseconds after the time-locked event (e.g. stimulus presentation) before the component typically appears

Question 22

What are the standard components associated with particular kinds of processes?

Answer 22

e.g., the P1 component is a positive deflection that occurs approximately 100 ms after a visual stimulus is presented

on the basis of the specific of the experiment, you can then try to make inferences about what differences in signals related to those components can tell you

Question 23

What aspects of an ERP are the most informative?

Answer 23

the presence/absence of particular kind of ERP in response to an event can be informative, through the timing and/or magnitude of the ERP elicited is often more informative

smaller amplitude and/or delayed onset of an ERP could indicate diminished or impaired processing

Question 24

How has ERP data been used to examine cognitive function in clinical population?

Answer 24

ERP data associated with Alzheimer’s patients (AD) and those diagnosed with mild cognitive impairment (MCI) show ERP signals that:

are “smaller” (the vertical dimension, or Y axis, plotting the signal as a function of amplitude)

have a delayed onset (the horizontal dimension, or X axis, plotting the signal as a function of time)

Question 25

How was the effect of attention on early sensory processing examined experimentally?

Answer 25

experimental participants are engaged in a go/no go task: press the space bar every time a number appears on screen, expect you need to withhold your response for one specific number the P1 occurs approximately 100 ms after something visual is processed participants are periodically interrupted and asked to report their mental state (either "on-task" or "wind wandering")

Question 26

Does not paying attention affect the early stages of basic visual processing?

Answer 26

the amplitude of the P1 is reduced when experimental participants report having been "mind wandering" immediately after a visual stimulus is presented, relative to when they report having been "on-task" this suggests attentional mechanisms affect early stages of sensory processing

Question 27

Why would the amplitude of an ERP component differ across people, conditions, etc.?

Answer 27

the general assumption is that this happens when there is more neuronal firing (leading to more changes in voltage), which could happen if more neurons are firing, the same neurons are firing at a higher rate, or some combination of the two in many ways this is similar to what the BOLD signal with fMRI tells you (neuronal "activation") we often find that larger amplitude ERP components are associated with "better"/deeper, etc. processing, which makes some sense in the context of that being indicative of greater neuronal involvement however, keep in mind that "better" performance (depending on how that is operationalized) is sometimes associated with smaller amplitude ERPs, providing a warning against oversimplifying interpretations

Question 28

What were the experimental conditions in the Habibi (2016) study on the effect of training on early sensory processing?

Answer 28

Habibi (2016) randomly assigned children to one of three groups: a musical training conditions a sports training condition a passive control condition

Question 29

What were the results of the Habibi (2016) study on the effect of training on early sensory processing?

Answer 29

the children in the musical training group showed enhanced abilities to detect subtle auditory changes (behaviorally) this was also reflected in changes to their ERPs (reduced P1 amplitude and latency while passively listening to piano tones) note that it isn't necessarily clear whether this change is related simply to enhanced auditory processing, attentional control, both, something else, etc.

Question 30

In what ways do ERP studies sometimes require making connections with prior work?

Answer 30

signals recorded at certain electrodes (i.e. positions on the scalp) are typically selected for analyses on the basis of the presumed origins of the signal being investigated for example, activity related to early visual processing should be concentered at the back of the brain, near the visual cortex this also provides an example of how understanding ERPs may sometimes require making connections with prior work another example could be looking for comments that have been associated with particular cognitive processes in the past (e.g., attention) and using them as your DV to test something about your topic of investigation

Question 31

What is magnetoencephalography (MEG)?

Answer 31

MEG maps subtle changes in the magnetic fields around the scalp caused by fluctuations in electrical activity in the brain very similar conceptually to an EEG, and can also measure ERPs although MEG is more sensitive than EEG, there are lots of additional limitations (expensive, requires a completely magnetically shielded room, etc.)

Question 32

What are the trade-offs between spatial and temporal resolution?

Answer 32

both approaches can be important for more fully understanding cognitive and neurological phenomena (e.g., they offer different perspectives) some methods allow for source localization using EEG (providing limited spatial resolution), and ERP's can also be combined with fMRI for more balanced spatial and temporal resolution

Question 33

What is functional magnetic resonance imaging (fMRI)?

Answer 33

magnetic resonance imaging (MRI) uses magnetic fields and radio waves to generate anatomical images of body tissues in addition to using functional magnetic resonance imaging (fMRI) to measure changes in the BOLD signal (blood-oxygen-level-dependent signal), which is correlated with cognitive activity, there are various other MRI-based applications that can be useful for cognitive neuroscience

Question 34

How can researchers use neuroimaging techniques to develop more complete theories?

Answer 34

can assess functional connectivity can assess structural connectivity can manipulate neural function

Question 35

How can fMRI be used to assess functional connectivity?

Answer 35

can use task-related fMRI to determine brain location associated with a specific task (these data are referred to as the seed location) measure the resting-state fMRI at the seed location measure the resting-state fMRI at another location (referred to as the test location) calculate the correlation between the seed and test location responses to assess the degree of functional connectivity

Question 36

What is diffusion tensor imaging (DTI)?

Answer 36

diffusion tensor imaging (DTI) uses MRI technology to map out major pathways (tracts: bundles of axons) in the brain can also be referred to as track-weighted imaging (TWI) this is useful for understanding how differences in the structural connectivity between different regions of the brain are related to various factors

Question 37

What is magnetic resonance spectroscopy (MRS)?

Answer 37

also sometimes called nuclear magnetic resonance (NMR), is a technique that uses an MRI machine to produce estimates related to metabolic changes in the brain this allows estimates related to the chemical composition i particular brain regions, and has applications for both basic research and clinically relevant purposes

Question 38

What is a major advantage of MRS?

Answer 38

it can be used to produce in vivo (i.e., within a living organism) estimates related to some simple neurotransmitters (e.g., GABA and glutamate)

Question 39

What is a major disadvantage of MRS?

Answer 39

at present, it is not capable of measuring more (chemically) complex neurotransmitters that are of interest of cognitive neuroscientists (e.g., serotonin, dopamine)

Question 40

What is transcranial magnetic stimulation (TMS)?

Answer 40

TMS uses a magnetic field and an electrical current to modulate brain activity at specific sites (generally decreasing or increasing firing) can create what is referred to as a "virtual lesion", producing a temporary impairment and thus providing a means for ethical manipulation of neural function in humans (allowing causal inferences to be drawn) repetitive TMD (rTMS) can produce longer lasting changes in neuronal activity, though there is still some debate about the safety and efficacy of this method

Question 41

What is functional near-infrared spectroscopy (fNIRS)?

Answer 41

is a relatively cheap, non-invasive method for estimating changes in neural activation hemoglobin absorbs a significant amount of infrared light, more so than most of the other substances/tissue in our brain changes in the relative proportion of light that is reflected back can therefore be used to estimate changes in hemoglobin concentration (which can, in turn, be used to infer changes in neural activation)

Question 42

What is the "traditional" approach to fMRI studies?

Answer 42

involved what is sometimes referred to as region of interest (ROI) or "univariate", analyses, involve looking for changes in activation in specific regions i.e., looking for "hot spots", for example checking whether the amygdala becomes more active during one condition of another, whether there is more activity on one side or the other, etc.

Question 43

What is multi-voxel pattern analysis (MVPA)?

Answer 43

looks for reliable patterns of distributed activity across the brain this could be said to (at least potentially?) better take into account distributed representation and network activity in models/theories, etc. that are informed by such analyses

Question 44

What is the hemodynamic response function?

Answer 44

the increase in blood flow that accompanies neural activation is not instantaneous and, at least compared to how fast electrical signals can travel, is rather slow this lag is reflected in the HRF, which describes the time course of increased blood flow in response to the presentation of discreet stimuli (e.g., a picture shown during an experiment) this needs to be accounted for when analyzing fMRI data and can be (at least in some ways) a limiting factor

Question 45

What is positron emission tomography (PET)?

Answer 45

PET was an early functional brain imaging method in use by the 1970's, which involves the injection of radioactive tracers/dyes that are rapidly "taken up" (i.e. metabolized) by particular cells until gradually breaking down this allows a map to be produced showing where the radioactive dye goes, which correlates with certain kind of physioglical processe (e.g. 2-DG, structurally similar to glucose, can provide estimates of general neural activity) the use of other kinds of radioactive tracers can allow for the imaging of more specific metabolic processes, e.g., locating/quantifying amyloid plaques hypothesized to play a role in Alzheimer's disease

Question 46

What is optogenetics?

Answer 46

the use of optogenetics in cognitive neuroscience involves genetically engineering neurons that are light-sensitive (e.g., ion channels that open when exposed to a particular wavelength) this allows neurons to be effective controlled, or turned on/off, by exposing them to light

Question 47

What are case studies?

Answer 47

undertaken in cognitive neuroscience involve in-depth investigations into individuals presenting with atypical neurology (e.g. what effects does not having a hippocampus have on our various memory systems?)

Question 48

What is a major advantage of case studies?

Answer 48

they provide unique research opportunities that would not otherwise be possible with conventional experimental methods

Question 49

What are major disadvantages of case studies?

Answer 49

relate to generalizing the results (which can be particularly complicated given the heterogeneity of differences associated with case studies), as well as disentangling cause and effect

Question 50

What are lesion studies?

Answer 50

typically involve training a non-human animal to perform a task, damaging a part of the brain, then observing how that damage affects the organism's ability to perform the previously learned task

Question 51

What is a major advantage of lesion studies?

Answer 51

they allow relatively straightforward causal inferences to be made about the contribution of various brain regions

Question 52

What is a major disadvantage of lesion studies?

Answer 52

relates to the difficulty in generalizing observations from non-human animals to humans (since this method cannot be used with humans)

Question 53

What is the interpretation of ERP components?

Answer 53

over time, ERP components become associated with particular processes (and new ones are sometimes discovered) these associations can be more or less general and it isn't always clear why components are affected (e.g., changes in latency or amplitude) by particular factors (e.g. manipulations, group level differences, etc.) accordingly, like other kinds of data interpretations, ERP interpretations are inherently somewhat subjective, and certainly fallible (i.e. can be wrong)