Methods of Cognitive Neuroscience

Question 1

What is the utility of brain imaging over lesion studies?

Answer 1

1. Brain damage is rarely selective
2. Selective brain damage to subcortical structures does not occur
3. The brain can reorganise itself following damage
4. Baseline/pre-injury data is often unavailable

Question 2

What does it mean that brain damage is rarely selective? And doesnt damage subcortical structures?

Answer 2

Head injury/impact - injury is twofold = coup injury (adjacent to injury site) + contracoup injury (opposite side of brain)

Stationary object hits head = coup injury
Moving head + stationary object = coup + contracoup injuries

The nature of both types of injury means damage is widespread and rarely restricted to deep structures

Question 3

What does it mean that the brain can reorganise following damage?

Answer 3

Neural plasticity is immense and other areas of cortex can accommodate for lost areas of function in some instances

Phantom limb is an example - following loss of the limb, the motor and sensory homunculi can redistribute the roles of the cells

Conversely, following a stroke that initially may prevent speech, after training and slow neural regrowth, other brain centres can replace areas of lost function

Question 4

What does it mean that we lack baseline data?

Answer 4

We have nothing to compare the current abilities with - original cognitive functioning could have been poor so the change isnt drastic following injury

Can be avoided (slightly) by comparing to similar people (ie IQ, EF, SES matched) and demonstrating no deficits

Question 5

Why else is neuroimaging data important?

Answer 5

Some questions are unanswerable by neuropsychology (ie lesions)

Even those that are require converging evidence for support

Question 6

What are the necessary assumptions for brain imaging?

Answer 6

1. There is consistent correspondence between brain activity and mental activity
2. When our thoughts or emotions (etc) change, as does our brain activity, and that this change is consistent
3. Thoughts, emotions, actions (etc) arise from changes in brain activity = a materialist/physicalist view

Question 7

How can neural activity be measured?

Answer 7

Ideally = electricochemical activity at the direct site of action = animal or human electrode implantation

ERP = electrical activity but at a distance

Imaging = PET/fMRI = monitor regional cerebral blood flow (rCBF) as a proxy of neural activity

Question 8

What is ERP and how is it used?

Answer 8

Brain tissue/skull/skin = conductive to electrical activity of neurons

Scalp electrodes conduct brains activity = EEG

ERP = a discrete segment of EEG that is ‘time-locked’ to a particular stimulus ie object presented - Xms later = ERP

Question 9

What are all the waves detectable on an EEG and what are their significance?

Answer 9

Alpha - 8-15Hz, relaxed/reflecting, eyes closed, inhibition control

Beta - 16-31, active thinking, focus, alert, anxious

Theta - 4-7Hz, inhibition of elicited responses, drowsiness/idling

Delta - <4Hz, slow wave sleep

Gamma - >32Hz, active perception combining two senses, STM matching of the same

Question 10

How is ERP data analysed?

Answer 10

Single trials contain lots of noise - repetition of trials with the same stimulus allows for the signal-averaging = INCREASED signal:noise ratio

Multiple electrode sites allows detection of electrical electrical activity change with time and electrode location

Question 11

How are ERP components named?

Answer 11

(C1 - visual processing)
P1, N1, P2, N2, P3 = one naming convention, giving ordinal values to the peaks; P+ve Vs N-ve
N100, P200 etc = second convention, gives m/s values to the peaks

Question 12

How are ERPs used to measure cognitive processes?

Answer 12

Cognitive processes are time-dependent and ERP affords a high temporal resolution eg recognition memory task:

Presentation of probe/stimulus - 0ms

Recognition of probe (familiarity) - c.300-500ms - FN400 (frontal negative)

Retrieving the information related to the probe/context (recollection) - C.500-800MS - left parietal old/new effect (different site to FN400 and can differentiate between old/new stimuli)

Question 13

What are some limitations of ERP?

Answer 13

Forward solution: we know how many sources of activity, their magnitude and where they are = we can model predicted electrical potentials some distance from sources

HOWEVER

Inverse problem:
although we know what data we can get from the skull, we dont know/cant infer the properties of the source - the regions below the electrodes might not be responsible for the activation

Question 14

What is magnetoencephalography?

Answer 14

Similar to ERP as are reliant on inherent electromagnetic fields generated by neurons action potentials = event-related fields

All principles applying to ERP also apply to ERF

Question 15

How do haemodynamic measures of neural activity work?

Answer 15

Reliant on neurovascular coupling - neural activity increases blood flow to areas of neurons that are expending more energy (greater need for oxygen + energy replacements and waste removal)

Neural activity increases - blood volume increases (c.5s)

Known as BOLD measures = blood oxygen level dependent

Question 16

How do PET scanners work?

Answer 16

Use radioisotopes that decay due to too many protons - emission of positively charged electrons/positrons - collision with nearby electrons = positroniums - positronium + positronium = annihilation event = gamma ray production - detection in the scanner

Question 17

How does fMRI work?

Answer 17

A strong electromagnetic field is applied to the body - polarisation of protons (some up, some down) - application of another magnetic field (radiofrequency pulse) - alignment of protons in the same direction as the pulse - when pulse stops, slow realignment/precession of protons to their original polarised states - emission of radiofrequency wave - detection using ‘antenna’

Question 18

What data does fMRI provide?

Answer 18

2D position information that can be build up with high sampling rapid scanning to generate 3D images of changes in blood flow

Question 19

What are the pros and cons of haemodynamic measures?

Answer 19

+ Spatially accurate, within mm, unlike ERP

• Not time-locked to stimulus presentation
• Haemodynamic response also takes between 12-20s to reset to baseline following activation
• Only measuring a proxy for neural activity; we still dont really know how neurovascular coupling works

Question 20

What are blocked designs?

Answer 20

Division of participants into subgroups = blocks, where intra-block variability is lower than inter-block variability; then subjects (randomly) assigned to experimental/control conditions

Reduces variability + confounding within conditions = better demonstration of effects

Question 21

What are the pros and cons of blocked designs for haemodynamic measures?

Answer 21

+ Fewer issues with temporal resolution if trials are grouped together
+ More powerful than event related designs - good to determine which voxels are activated
+ Fewer issues switching between trial types

• Potential confound due to different strategy use
• Less similar to behavioural studies
• We might be interested in certain types of trial which we cannot predict in advance ie when participant cannot retrieve answer they know

Question 22

What does fMRI on lying tell us? (Langlenblen, Schroeder etc etc 2002)

Answer 22

Certain regions of the brain have been implicated in truth and lying

if we assume that the brain is responsible for mental experiences (3) and the different patterns of brain activity support lying and truth telling (2) once you’ve identified this patterns you should be able to tell whether someone is lying or not (1)

Problems: can you differentiate between spontaneous and prepared lies? White lies? Lies of omission versus commission? Lying to protect another? Etc (realistically probably yes at some point with advanced computer modelling but the current technology does not allow)

Question 23

Why is it advantageous for participants to perform at near ceiling levels on fMRI tasks?

Answer 23

Reduces noise - we want to see what happens when participants make the choice they are required to make; the greater number of trials participants perform at our expectations, the greater number that can be used in our analysis