Cognitive control

• goal orientated behaviour dependent on functioning working memory. 
• working memory entails the ability to use current information to achieve behaviourally relevant goals. it is an active process that goes beyond storage. priotization which subtask needs to be perfromed first. enables flexible behaviour that is not simply reflective and stimulus driven.
• WM is about strong current novel information, but also about retrieval of information from LTM. It needs the integration of novel information with long term goals (stored in LTM).
• our working memory abilities critically depend on the lateral prefrontal cortex. 
• monkeys with lesions still form associative memories in relation to food retrieval but do poorly in WM tasks where they have stored the location of a baited container in WM for later food retrieval. 
• equally do poorly on other food retreival tasks which tax working memory load, but have little or no deficits in tasks that assess recognition memory or familiarity. 
• children younger than 1 - poor in WM tasks as their preforntal cortex is still very immature. 
• recording in lateral prefrontal cortex have identified neuronal activity that has the signatures of working memory signal 
• enhanced spiking activity that carries on after an indicitiave stimulus is extinguished + continues till either response is perminited to obtain a reward or by engaging in specific behaviour or until stimulus gives further which will result in action to achieve reward. 
• Miller- prefrontal cortex memory sufficently specific to encode 
• lesions to prefrontal cortex memory deficits to WM not LTM. 

Goal orientated behaviour

• lateral prefrontal cortex-  cruicially involved in the planning and execution of goal oriented behaviour (or cognitive control) 
• lateral prefrontl cortex situated in the frontal cortex. this part of cortex undergone substantial expansion in mammalian, and primate evolution. 
• frontal cortex houses the motor cortex, and the premotor areas. 
• further anterior areas are located which make up the prefrontal cortex and frontal pole. (grown substatinally during primate evolution). involved in the planning of behaviour. 

• lesions of prefrontal cortex= deficits in recency memory (tested by probing the ability to recall the order of images seen) recognition memory is not affected so not due to general memory deficit 
• recency task cannot simply be recalled by WM (>5-8)
• source memory- our ability to remember circumstances in which we formed an episodic memory. important retrieval cue 
• source memory also depends on intact prefrontal cortex. 
• low frontal lobe perfromers perfrom comparitively poor in source memory tasks. 

content based approaches. 

• Left hemisphere- active in WM tasks that tax verbal memory. 
• Right hemisphere- active in WM tasks that tax visuospaitalory memory
• Dorsolateral frontal cortex subdivided- 'what' and 'where' visual pathways. 
• Dorsolateral frontal cortex of lateral frontal cortex= more active in spatial memory tasks (where) 
• Ventral parts= verbal tasks (what)

process based approaches

• different mental processes, rather than mental contents 
• Lateral frontal cortex= more active in backwards tasks than forward
• very few tasks seem to differentially activate most frontal aspects to prefrontal cortex 
• not differentially activated in forward or backward (Verbal or visuospatial) tasks but strong activation in all of these task conditions throughout 
• this part of frontal cortex is known as frontal pole 
• frontal pole neccessary to organise the different behavioural demands that exist in each task into a global coherent whole, more and more involved the more difficult the required behaviours are. 

• each behaviural goal requires planning of subtasks that need to be perfromed to achieve the goal
• steps need to be organised
• frontal cortex= essential in organization process
• prefrontal damage patients may be able to understand the requirements in goal directed behaviour but have difficulty following these up in an appropriate manner or in identifying appropriate steps to achieve goal
• these patients have difficulty identifying adequate goals, balancing pros and cons of alternatives
• goal oriented behaviour also requires the suppression of actions that might yield short term gains but jepardize long term success (e.g. looking for date at party could conflict with being rich in future if party night before exam)
• all that is required is encaptured by term cognitive control of goal oriented behaviour
• frontal cortex- vital in this operation
• widely accepted test of frontal dsyfunction + dsyfunction in cognitive control is Wisconsin card sorting test
• subjects confonted with cards that contain different domains (normally 3:number, shape, colour). subjects have to sort newly obtained cards by matching them to cards that are laid out on table. not told what the matching rules are. find out by try and error. once rule has been discovered, experimenter changes and have to find out again.

• patients with prefrontal damage do poorly in the task. once they have identified rule, they have trouble letting go of the last matching rule
• they perseverate, trying it over and over again, even thougt they have been told that their sorting is incorrect

• as cognitive control requires working memory= wm capacity is limitied. filtering neccessary to determine what info is selected and processed in wm.
• = dynamic filtering hypothesis of prefrotnal cortex.
• dynamic filtering is a form of attention, it provides attention (processing) resources to task relelvant items, and neglects or surpresses non-task relelvant information.
• idea of dynamic filtering might explain might why source memory is affected with frontal lesions, while content memory less so.
• source memory (info.) is not as salient as content memory (info.) so by means of salience finds its way into the storage systems.
• preservance in Wisconsin task can be viewed as failure to select i.e. to dynamically filter. patients can still filter but dynamics are lost. they stick to a dimension that occupies the filter.
• prefrontal patients also affected in stroop task but only when incongruent stimuli is shown.
• do poorly in tasks that require associations with high filtering demand- i.e. verb generation task. subjects have to report verb associated with given noun. some nouns may trigger associations that have to be filtered. others trigger fewwer immediate associations + filter demands are lower.
• prefrontal patients have difficulty in the verb generations when many verbs are possible match to noun. dont have semantic deficits though- just cant decide which one to take.

• .. if prefrontal cortex is involved in this operation, differential activity in described task conditions should be found in prefrontal areas. experiments have supported this notion.
• if dynamic filtering is a key component of prefrontal actions, then prefrontal patients should be poor at task switching (shown by Wisconsin task)
• however, task switching abilities should also depend on the specific cognitive demand involved in a task. direct instructional cues that dont require wm load, should not affect task switching abilities where those that do should.
• shown to be case. normal patients- direct cue does not will not result in slower reaction times immedietately after task switching + same with prefrontal patients.
• however, if colour of background indicates which items needs to be reported then normal subjects have a switch cost. as wm needs to identify what the current colour means, and dynamic filtering needs to be readjusted.
• switching cost much bigger in prefrontal patients suggesting wm and dynamic filtering are handicapped.
• modified computer version of winsconisn task- allowed to measure cortical activation following switching cues directly. after 10 successive trials were instructed rules now changed (but not what new rule was) prefrontal activation varied systemattically following switch cue + activation levels parametrically related to number of dimensions employed in task.

• dynamic filtering related to top down contol of attention in a way the concepts interchangeable
• Df aids attending and selecting relevant- done by enhancing task relelvant information or by surpressing irrelevant. depends partially on age
• combined FMRI and TMS (briefly disrupts cortical activity locally)- TMS pulses over FEF decreases activity in parts of visual areas that represent foveal visual space + increases activity over part of of visual cortex that represent peripheral visual space.
• idea is that large parts of FEF are involved in fixation a specific location which is attended to during fixation.
• fixated location corresponds to part of visual space represented by FOVEA. since TMS over FEF disrupts that, the suprresion of activity over peripheral parts is reduced + consequently stimuli presented there are processd more efficiently.
• stimuli at fovea processed less efficiently when FEF is briefly distributed.
• selective filtering hypothesis more directly tested in patients with diff forms of cortical damage, when compared to controls in auditory tasks.
• in a passive listening task= prefrontal cortex damage showed much larger evoked potential activity originating in auditory cortex compared to controls. no such difference in other patient groups.

• in modified version of task- subjects asked to selectively listen to information from either the right or left ear.
• patients with right prefrontal damage showed absence of filtering from left ear. left prefrontal damage= reduced filtering for info for either ear
• this lack of filtering might also explain the dissiciation seen with recency memory and recognition memory
• assume no filtering happens. information that has been presented may not 'decay' as quickly _ likewise strong memory trails present
• if recency memory is based on strength of an ongoing memory trace, then decay may actuallly help to tell recent traces apart
• older ones will show less activity
• in absence of decay- all traces will be similar and the order of presentation is harder to establish
• enhancement of task relevant information- tested where subjects had to remember faces or scenes engaged in a pasive viewing task.
• when subjects engaged in memorize faces, activity in fusiform face area was increased relative to passive viewing.

• when task memorize scenes, parahippocampal place area showed higher activity.
• both areas showed reduced activity relative to passive viewing when the task demands did not require their involvement
• supression was only seen in college aged students. older age did not show supression but showed enhancement
• cognitive decline that comes about with ageing, partially due to lost ability to suppress task irrelevant information?

• another form of cognitive control requires inhibition of action
• neural mechanisms involved are related to suppression of irrelvant info as discussed.
• located in prefrontal cortex.
• test used= stop-signal task
• respond to specific items by pressing button- on some trial a stop signal is presented after stimulus presentation which indicates to subject to inhibit motor response
• right inferior frontal cortex= strongly activated when stop signal presented. irrespective of whether stopped or not
• why would there be no difference when there was a behavioural difference
• resolved by analysisng activity over left motor cortex
• activity on failed strop trials already high before the stop signal(anticipatory response higher) and thus stop signal failed to cancel action
• patients with right inferior frontal damage perform poorly in the stop task

• model by Norman and Shallice tries to link what has been discussed to the outputs (the actions)
• model proposes perceptual inputs automatically active what they call schema control units. these are units which trigger specific action pattern, many of which will be learned
• certain situations require a set of almost steretypical actions
• these schemas linked by inhibitory connections, which ensure actions dont interfere with one another
• according to model- controlled by a unit called 'content schedulling'. this acts automatically, and cancells incompatible actions
• the schema control unit is under the control of supervisory attentional system (SAS)
• schema control units could be conceptualized as being part of basal ganglia where action pattern are learned and stored
• SAS ensures behaviours are flexible and can adapt to short as well as long term goals- psychological model
• fails to account for a description where these functions are located and how exactly they are implemented
• we need a moniotring system that allows altering stratergies

• medial frontal cortex (MFC) proposed to be involed in outcome monitoring
• MFC parts seem to increase their level of activity with the level of task difficulty and number of subtasks that are neccessary to succeed in performing tasks
• cingulate gyrus also involved in this process. - gets active when novel task performed where alot of monitoring required.
• once we become better at task- activity shifts to more posterior and higher sensory areas; reflecting the fact that less monitoring is required
• problems with disccused examples= all shift responsibility for cognitive control to higher ventres
• incorrect to assume ultimate supervisor
• some insight comes from electrical signals that occur when subject makes mistakes
• provided subjects realise, electrical signal recorded over anterior cingulate cortex shows a signature which is known as error related negativity. could be viewed as wake up call to entire system to refocus on task and stop drifting away.
• medial prefrontal cortex would then not be directly involved in selection of task relevant information but in the monitoring of behavioural outcomes, helping refocus when these areas not desired
• MPC monitors level of response conflict. in stroop task conflic occurs in incompatible

• when response conflict high, moniotring system alters attentional system to increase its efforts
• FMRI adaption of stroop shows this.
• here activity in lateral prefrontal cortex highly correlated with activity coming from ACC on previous trial
• if ACC is activity it alerts the other systems invollved in goal related behaviours