The Prefrontal Cortex

Question 1

Difference in PFC in different animals

Answer 1

It looks quite similar to the chimpanzee but there’s a lot more complexity in the folds – in the sulci and gyri in the human brain.
Thus, it seems that the PFC must be doing something important that distinguishes us from other animals

Question 2

When you look at the picture of the connections between PFC and other brain regions, you can see that it is connected to …

Answer 2

When you look at the picture of the connections between PFC and other brain regions, you can see that it is connected to virtually every other region of the brain, which also suggests it plays a key role in human behaviour.

Question 3

Who was Phineas Gage

Answer 3

One of the first indications of the function of the PFC was provided by the unfortunate case of Phineas Gage, a railway worker who was carrying out some explosives work on the lines when an explosion went off unexpectedly and a metal bar shot up into his face, entered his brain and exited the top of his head.
This is a reconstruction of the approximate entry and exit points based on the damage he sustained.

Question 4

what is the Wisconsin Card Sorting Task

Answer 4

Patients are given a single card (here the one at the bottom) and must choose which of 4 decks to place the card on. They have to learn the rule governing which deck it should be placed on and continue to place different cards on the same deck according to the correct rule. The rule could be based on colour, shape, number of shapes etc. The patient must use trial and error to find the correct rule. Then, after 10 consecutive correct responses, the rule is changed and the patient must discover the new rule.

Question 5

what is task-set switching

Answer 5

The idea is that the patient must acquire a ‘set’ for task performance, which basically means a rule or set of rules. And this set can switch repeatedly during the task.

Question 6

Errors on Wisconsin Card Sorting Task pre and post surgery

Answer 6

Dorso-lateral prefrontal cortex seems to be involved.

Milner (1963)

Question 7

what did Milner 1963 do

Answer 7

Patients with frontal lobe injuries showed disproportionate impairment on the task.
What these patients did was perseverate – after the rule changed, they tended to carry on responding according to the same rule.

Normal healthy people and patients with brain damage in other regions would discover the new rule with a couple of tries. However, patients with PFC lesions would simply carry on placing the card on the deck with the same colour as the single card. They would perseverate.
The conclusion was that the PFC is doing something related to inhibition - enabling flexible behaviour by inhibiting (suppressing) previously relevant, but no longer relevant, responses.

Question 8

what did Shallice and Burgess do

Answer 8

They found that patients with frontal lobe lesions were actually surprisingly not very impaired on tasks that measure supposedly frontal executive functions.
e.g. Stroop task (attentional interference), Tower of London task (planning), even a modified WCST.

Question 9

who came up w six elements task

Answer 9

Shallice and Burgess

Question 10

what is the six elements task

Answer 10

• Limited total time (e.g., 10 minutes)
• Six different tasks to work on, such as picture naming, arithmetic, visual cancellation task.
• The goal is to work on all six of the tasks, and hopefully complete all six of the tasks, in the 10 minutes.
• The score is based on the number of tasks attempted, and score penalties are given for rule infractions or not spending an equal amount of time on each task.
To investigate this, Shallice and Burgess designed the 6 elements task, in which subjects had to do multiple different sub-tasks in a limited time. It wasn’t possible to complete all of the tasks so the optimal strategy was to spend a bit of time on each task and switch between the tasks.

Question 11

what was foudn from the six elements task

Answer 11

What they found was that patients failed quite dramatically on the 6 elements task – most patients only attempted 2 or 3 of the subtasks relative to the 6 of healthy controls, and spent a dispropotionately large amount of time on each of the sub-tasks.
They theorised that the deficits on this task were due to the breakdown in a unitary supervisory system, located in the PFC

Question 12

what is the The SAS model (Norman & Shallice, 1980)

Answer 12

The model is basically designed to explain how behaviour can operate in non-routine situations, in other words where well learned behavioural sequences are not sufficient. The WCST could be seen to be a model of this type of behaviour – the patient learns a particular rule governing responding but then has to re-learn this rule repeatedly

Question 13

problems w SAS model

Answer 13

Homunculus’ criticism
• There is no controller at the front of the brain. Otherwise someone has to control that control. This causes infinite regression
• Who controls the controller?
• Explains what is controlled but not how control is exercised

Question 14

what is cold cognition

Answer 14

Lateral parts of the prefrontal cortex. These are cold cognition. Doesn’t involve much emotional input

Question 15

what is hot cogniton

Answer 15

The orbitofrontal cortex is hot- this is driven by values, emotions

Question 16

what did Miyake do

Answer 16

In an influential study, Miyake and colleagues attempted to fractionate executive function into component variables using behavioural tasks and factor analysis.
They gave healthy subjects a variety of tasks:

Question 17

what tasks did Miyake give

Answer 17

• Number/Letter Task (Rogers & Monsell, 1995)
• Letter Memory Task (Morris & Jones, 1990)
• Stop Signal Reaction Time Task (Logan, 1994)

Question 18

what is Number/Letter Task (Rogers & Monsell, 1995)

Answer 18

For example, task switching – subjects have to perform two tasks – odd/even or vowel/consonant – depending on the location of the letters/numbers
Subjects are slower in switch trials than they are in repeat trials on this task.

Question 19

what is Letter Memory Task (Morris & Jones, 1990)

Answer 19

A letter memory task that requires subjects to remember letters but also to update the letters in memory.

Question 20

what is Stop Signal Reaction Time Task (Logan, 1994)

Answer 20

And a stop signal task requiring subjects to withhold prepotent responses. So the task is to respond as quickly as you can to the direction of the arrow but to withhold responding if you hear a loud beep after the arrow is presented.
The researchers gave subjects 9 tasks in total measuring a variety of different executive functions like these.

Question 21

what did Miyake find

Answer 21

They then performed a factor analysis of all the tasks.
The figure here shows the 9 tasks on the left. They found that there were three very clearly distinct, latent variables that accounted for performance differences on the 9 tasks. These variables (or ‘factors’) are shown in the central part of the figure.
– shifting, which means shifting between task sets.
- updating, which means updating the contents of WM.
- And inhibition, which means inhibiting prepotent responses.
- This has been quite an influential model of executive function, and it tends to be used as a kind of template for understanding how executive functions can be fractionated.
- The idea is that any complex executive task can be accomplished by drawing on (some mixture of) these three functions.

Question 22

what was Stuss & Alexander (2007) ‘s goal

Answer 22

Our goal was to determine whether all focal frontal lesions produced a similar impairment in cognitive supervisory control or whether lesions in different regions produced specific impairments that might or might not appear on a task depending upon the particular demands of the task

Question 23

what did Stuss & Alexander (2007) do

Answer 23

• Tested frontal lobe patients (n = ~40) on a range of neuropsychological tasks including classic frontal tasks (WCST, Stroop), language and memory tests requiring executive functions, and attentional tests.
• Brain lesions were mapped out and location of brain damage defined by registration to a standard anatomical template

Question 24

what did Stuss & Alexander (2007) find

Answer 24

Stuss performed a neuropsychological study where he ran a battery of tasks on patients with different focal prefrontal cortex lesions. He found some correspondence between the Miyake model and patients, as well as some differences.
Right lateral – monitoring issues – Miyake’s “updating” variable included monitoring tasks
Left lateral – task setting issues– necessary for shifting as in Miyake’s model
Convergence in medial PFC – energising. No space for inhibition. Stuss actually goes as far as to say that inhibition may not exist at the psychological level, certainly that it wasn’t a necessary concept to explain performance on the tasks they used.
No sign of inhibition in this study.

Question 25

what did Carter et al 1998 do

Answer 25

Carter and colleagues looked at the role of the anterior cingulate cortex (ACC), part of the medial PFC, in executive function.
They asked healthy subjects to perform an ‘AX-CPT’ task (continuous performance task) in the MRI scanner. In this task subjects see two letters on each trial, one after the other. The first can be A or B and the second can be X or Y. The subject must press a button when they see an X but only when it is preceded by an A. This is a task that leads to quite high error rates because on some trials they will see an A which primes them to make a response, but then they see a Y. Quite often subjects press on these trials accidentally.

Question 26

what did Carter et al 1998 find

Answer 26

They measured brain activation during error and correct trials and found increased activation in ACC for errors relative to correct trials – activation in this region was higher when subjects responded incorrectly than when they responded correctly.
The authors suggested that the ACC was an error detection module – a dedicated module in the brain for detecting when we make an inappropriate or incorrect response. 

• ACC activation higher in interference than neutral trials even when the subject responded correctly
• ACC involved in conflict detection
• Tasks that require resolution of conflicts between competing information streams by sensory and/or response selection

Question 27

what did Bush et al 1998 do

Answer 27

Bush and colleagues followed this (carter et al) study up with an fMRI study of the stroop task. In this task subjects had to say how many words were on the screen on each trial. The words could be either neutral (e.g. dog) or could interfere with the main task (e.g. the word ‘two’ when there were 4 words on the screen).
They found that even when subjects responded correctly, ACC activation was higher on interference trials than on neutral trials.
The results are slightly inconsistent with the previous study because they show that ACC activation is not simply associated with errors. Rather, this study suggests that error-related activation occurs because of the detection of a conflict between competing information streams

Question 28

• Neuroimaging inconsistent with lesion evidence

Answer 28

– Bush/Carter – ACC activation related to conflict monitoring
– Stuss – ‘energising’ behaviour

Question 29

• Possible role for ACC in evaluating effort associated with a choice

Answer 29

– Grinband et al. (2008) – Stroop task
– ACC activation linked to time-on-task – greater activation for slow RTs than fast RTs even on congruent trials
– Potential resolution to inconsistency
– ACC sensitive to the amount of effort involved in task performance

Question 30

Results of a meta-analysis of 36 task switching studies

Answer 30

• Bilateral pattern of activation, distributed across frontal and parietal regions
• Some convergence with neuropsychological evidence but suggests a more distributed network involved

Question 31

what did kim et al find

Answer 31

The meta-analysis found a distributed network of frontal and parietal regions in which activation was greater for switch trials than non-switch trials. You can see that the overall pattern of activation is bilateral, although dorsolateral PFC (DLPFC) activation appears to be preferentially left lateralised.
So there is some support from neuroimaging for the link between left lateral PFC and task setting/switching but the evidence is also inconsistent in that other regions also appear to be involved– e.g. parietal cortex and right PFC. Thus, task setting may not be an exclusively mediated by the left lateral PFC but also require other brain regions

Question 32

what did Bode et al 2009 do

Answer 32

This study provides further evidence that the parietal cortex plays a key role in task switching.
They analysed fMRI data with the MVPA method – decoding cognitive states from patterns of activation.
They gave subjects a task-switching task in which subjects saw a cue (the letter A or B), followed by one of two coloured patterns.

The researchers used a pattern classifier to try and decode which task (A or B) the subject was performing from trial to trial, based on their patterns of brain activation.
As the top row of brains shows, they found task rules could be decoded from parietal and PFC.

This is a key piece of evidence demonstrating that the PFC is not always the source of task setting signals – the parietal cortex plays a key role as well.

Question 33

Left lateral PFC – the view from neuroimaging

Answer 33

• Partial support for the Stuss idea that left lateral PFC involved in task setting/switching
• Neuroimaging reveals a more distributed system including right lateral PFC and inferior parietal cortex
• Possibly a key role for parietal cortex

Question 34

However, a key neuropsychological study by Aron and colleagues in 2003 suggested that a particular part of the right lateral PFC (the right inferior frontal gyrus) plays an important role in response inhibition (the process of inhibiting (withholding) a prepotent or inappropriate response).

what did they do

Answer 34

They gave patients with different brain lesions a Stop Signal Reaction Time task in which they had to respond whether an arrow was pointing to the left or right but on occasional trials withhold their response when they heard a loud beep.
They found that performance on this task was strongly related to the size of the lesion in the right inferior frontal gyrus – there was a positive correlation between the Stop Signal Reaction Time (a measure of how good the subject is at inhibiting their response) and lesion size in this region – the bigger the lesion, the worse the patient was at inhibiting their response.

Question 35

what did Hampshire et al 2010 do

Answer 35

Hampshire and colleagues gave subjects a very simple task in which they were required to monitor for occasional targets and found increased activation in right inferior frontal gyrus (and to a lesser extent left) when subjects detected a target. This supports the idea that this region is performing a more basic attentional role.

Question 36

Right Lateral PFC – summary

Answer 36

• Right inferior frontal cortex may play an important role in response inhibition
• However, this may partly be due to a role in attention
• Some convergence with neuropsychological evidence – e.g. Aron’s study demonstrating inhibition impairments after damage to this region
• But seems to go beyond Stuss’s ‘monitoring’ account

Question 37

Methodological Considerations may partially account for lack of convergence

Answer 37

• Neuroimaging methods show regions that are generally involved in task performance, not necessarily those necessary for task performance
• Neuropsychology reveals regions that are necessary for task performance but cannot reveal information about networks of brain regions
• Is it even possible to fractionate the PFC?

Question 38

what did John Duncan and Adrian Owen do

Answer 38

John Duncan and Adrian Owen in 2000 performed a meta-analysis of neuroimaging studies of executive function, in which they plotted the activations associated with multiple different processes, e.g. response conflict, task novelty, on a single brain.

Question 39

what did duncan and owen find

Answer 39

What they found was that rather than there being separate regions of the PFC dedicated to different processes, the different processes all activated remarkably similar regions. There was no clear separation between the different processes – it was more like there was a network of regions, encompassing regions in the lateral PFC, the anterior insula, the medial PFC, and also the inferior parietal cortex, that all seemed to show increased activation when subjects did something cognitively difficult.

Duncan called this network the frontoparietal ‘multiple demand’ network, reflecting the idea that this is a ‘multi-purpose’ network of brain regions that underlies cognitive performance in multiple different types of demanding tasks.

Question 40

What does the ‘multiple demand’ network do?

Answer 40

multiple demand network is involved in the construction of ‘attentional episodes’, essentially that neurons in this network have flexible, dynamic response properties, and are able to encode, or represent, whatever information is within the current focus of attention

Question 41

fMRI evidence for coding of more abstract information about task rules in frontoparietal cortex

Answer 41

Woolgar et al 2011

Question 42

what did Woolgar et al 2011 do

Answer 42

Subjects learned different rules governing mappings between stimuli and responses

However, a study by Woolgar and colleagues suggested the PFC may perform a slightly more complex role.
Here, subjects had to perform a task in which they had to learn different rules mapping locations to responses. So when they saw a blue screen they had to remember a particular mapping between locations and button presses but when they saw a pink screen they had to remember a different mapping.

Question 43

what did woolger find

Answer 43

They found that although frontal and parietal regions did encode information about the position of the stimuli, the colours and the responses, these regions showed strongest coding of rules.
Thus, the results demonstrate that the primary role of the PFC in this kind of task is encoding the information about quite abstract aspects of task performance such as the rules governing stimulus-response mappings. It could be claimed therefore that these regions are not simply directing attention to specific stimuli in WM but are in fact performing a slightly more complex function.