Gazzaniga • Ivry • Mangun Cognitive Neuroscienceraphe.kaist.ac.kr/lecture/2016fallbis451/ch12 Cognitive...Gazzaniga • Ivry • Mangun Cognitive Neuroscience FOURTH EDITION Cognitive

Cognitive Neuroscience, Fourth EditionCopyright © 2014 by W. W. Norton & Company, Inc.

Chapter 12Cognitive Control

Gazzaniga • Ivry • Mangun

Cognitive Neuroscience

FOURTH EDITION



Cognitive Control (a.k.a. Executive function)

⬧ “The process that allows information processing and behavior to vary adaptively from moment to moment depending on current goals, rather than remaining rigid and inflexible”

⬧ Executive functions (also known as cognitive control and supervisory attentional system) is an umbrella term for the management (regulation, control) of cognitive processes, including working memory, reasoning, task flexibility, and problem solving as well as planning and execution.

⬧ The executive system is a theorized cognitive system in psychology that controls and manages other cognitive processes, such as executive functions.


What Controls Cognitive Control?

⬧ Prefrontal Cortex (PFC) ⬧ Four main subdivisions of the PFC

⬧ Lateral prefrontal ⬧ Frontal pole ⬧ Medial frontal ⬧ Ventromedial prefrontal (Orbitofrontal cortex)


▪The most caudal part of the frontal lobe contains the primary motor region. Here the ventromedial zone is also known as the orbitofrontal cortex.

▪There are extensive projections to and from the prefrontal cortex to all other lobes of the brain.


The purple region indicates prefrontal cortex in six mammalian species.

Although the brains are not drawn to scale, the figure makes clear that the PFC spans a much larger percentage of the overall cortex in the chimpanzee and human.


Lateral Prefrontal Cortex

⬧ Cognitive Control ⬧ Working Memory ⬧ Inhibition of

Prepotent Responses ⬧ Selective Attention


Frontal Pole

⬧ Cognitive Control ⬧ Memory Retrieval ⬧ Hierarchical

Representation of Action Goals


Medial Frontal Cortex: Anterior Cingulate Involvement

⬧ Cognitive Control ⬧ Error Detection ⬧ Resolving Conflict


Working memory (prefrontal cortex) and cognitive control


In this example, the woman’s goal is to tell her friend about the highlights of her recent trip to San Francisco.

Her knowledge of the Golden Gate Bridge requires activation of a distributed network of cortical regions that underlie the representation of long-term memory.



In n-back tasks, responses are required only when a stimulus matches one shown n trials earlier.

The contents of working memory must be manipulated constantly in this task, because the target is updated on each trial.


Fluid and crystallized intelligence

⬧ Fluid and crystallized intelligence (respectively abbreviated gF and gC) are factors of general intelligence, originally identified by Raymond Cattell and John L. Horn.

⬧ Fluid intelligence or fluid reasoning is the capacity to think logically and solve problems in novel situations, independent of acquired knowledge. It is the ability to analyze novel problems, identify patterns and relationships that underpin these problems and the extrapolation of these using logic. It is necessary for all logical problem solving, e.g., in scientific, mathematical, and technical problem solving. Fluid reasoning includes inductive reasoning and deductive reasoning.

⬧ Crystallized intelligence is the ability to use skills, knowledge, and experience. It does not equate to memory, but it does rely on accessing information from long-term memory.


⬧ Participants performed a working memory task. Trials were divided into those with lures where a mismatch was a stimulus that had been previously seen (and thus had potential for a false alarm) and trials without lures where the stimulus had not been seen.



(a) Regions in prefrontal and parietal cortex that had increased BOLD response on lure trials compared to no-lure trials.

(b) Correlation between individual scores on the Lure activity factor (Lure–No Lure) and measures of fluid intelligence (left) or a measure of working memory span from a different task (right).



(a) In the working memory task, the monkey observes one well being baited with food. After a delay period, the animal retrieves the food. The location of the food is determined randomly.

(b) In the associative memory task, the food reward is always associated with one of the two visual cues. The location of the cues (and food) is determined randomly. Working memory is required in the first task because, at the time the animal responds, no external cues indicate the location of the food. Long-term memory is required in the second task because the animal must remember which visual cue is associated with the reward.

Prefrontal cortex lesioned monkey shows both impaired working memory and associative memory.



Decision-making and cognitive control


Decision Making


Decisions require the integration and evaluation of multiple factors.

⬧ In this example, the person is asked to choose between two objects, each of which has an inferred value (offer values). The values involve some weighted combination of multiple sources of information.

⬧ Some sources are external to the agent: What will I gain (commodity), how much reward will be obtained, will I get the reward right away, and how certain am I to obtain the reward? Other factors are internal to the agent: Am I feeling motivated, am I willing to wait for the reward, is the risk worth it?


⬧ Normative Theories (economics) of decision-making The normative theories describe how we ought to think (e.g., expected utility theory).

⬧ Descriptive Theories of decision-making The descriptive theories describe how we actually do think.


Representation of Value: motivation

⬧ Primary Reinforcers: food, water, sex (pleasure) etc.

⬧ Secondary Reinforcers: money, social position, self-esteem etc.



Reward is expectation: prediction error



Mice were placed in a box with two levers. Contact with one lever resulted in optogenetically-triggered activation of dopamine receptors (active), whereas contact with the other level resulted in no stimulation (inactive).

Over time, the mice were more likely to contact the active lever when the stimulation was of the D1 receptors in the direct pathway. In contrast, the mice learned to avoid the active lever when the stimulation was targeted at the D2 receptors of the indirect pathway.



[Human study] People were presented with one of four cues. Over time, they learned that each cue was associated with one of two possible outcomes (or for Cue A, the same neutral outcome).


Prediction errors reliably predicted the BOLD response in the ventral striatum (the center of the positive prediction error response, green), which is slightly anterior to the center of the negative prediction error response (red).


Excitatory and inhibitory influences of positive and negative rewards






: Probability of payoff: Value representation



Goal representation becomes more abstract as you move forward along the anterior-posterior gradient of the frontal lobe.

1. Planning: identify goals, develop subgoals 2. Receive information about goals and means (rules and rewards) 3. Select task-relevant information, including where to attend and how to respond 4. Determine what temporal order is required to achieve the subgoals 5. Switch tasks if it will result in the hoped-for outcome 6. Check progress


Goal-directed vs. habitual decision-making

⬧ Action-outcome decisions (Goal-directed decisions) vs. stimulus-response decisions (habitual decisions)

⬧ Model-based means that the agent has an internal representation of some aspect of the world and uses this model to evaluate different actions. Model-free means that you just have an input-output mapping, similar to stimulus-response decisions.

⬧ Model-based decision vs. Model-free decision

⬧ Goal-directed decisions are based on the assessment of expected reward, whereas habits are actions taken that are no longer under the control of reward.


Cognitive control is a multiple processing.


In a delayed-response task, a set of intact faces or scrambled faces is presented during an encoding period. After a delay period, a probe stimulus is presented and the participant indicates if that face was part of the memory set.


The BOLD response in the lateral prefrontal cortex (PFC) rises during the encoding phase and remains high during the delay period. The magnitude of this effect is related to the number of faces that must be maintained in working memory.


The BOLD response in the lateral prefrontal cortex and the fusiform face area (FFA) rises during the encoding and retrieval periods.

The black dotted and red dotted lines indicate the peak of activation in the FFA and PFC.

During encoding, the peak is earlier in the FFA; during retrieval, the peak is earlier in the PFC.


Subregions of the prefrontal cortex are sensitive to either contents or processing requirements of working memory.

⬧ An instruction cue indicates the task required for the forthcoming trial.

⬧ Following a delay period, a series of pictures containing letters and squares at various locations is presented.

⬧ The participant must remember the order of the instruction-relevant stimuli to respond after the memory probe is presented.



Ventrolateral PFC is activated in a consistent fashion for all four tasks. Dorsolateral prefrontal cortex is more active when the stimuli must be remembered in reverse order, independent of whether the set is composed of locations or letters.



(b) Firing profile of a neuron that shows a preference for one object over another during the “what” delay. The neural activity is low once the response location is cued. (c) Firing profile of a neuron that shows a preference for one location. This neuron was not activated during the “what” delay.



• Prefrontal activation in an fMRI study increased along a posterior– anterior gradient as the experimental task became more complex.

• Activation in the premotor cortex shown in green was related to the number of stimulus–response mappings that had to be maintained.

• Activation in caudal LPFC shown in yellow was related to the contextual demands of the task. For example, a response to a letter might be made if the color of the letter was green, but not if it was white.

• Activation in rostral LPFC shown in red was related to variation in the instructions from one scanning run to the next. For example, the rules in one run might be reversed in the next run.


Behaviors That Succeed: Goal Representation


Self-Control

⬧ Ventrolateral Prefrontal Cortex (VMPFC)

VS. ⬧ Dorsolateral Prefrontal

Cortex (DLPFC)


Data were sorted according to the amount of time the animal stayed in one patch (from shortest to longest: black, red, blue, purple). For each duration, the animal switched to a new patch when the firing rate of the ACC neurons was double the normal level of activity.


The prefrontal cortex as a Dynamic filter


Why Do So Many Things Occur in the PFC?

⬧ People must resolve the conflict of what to pay attention to: ignore irrelevant tasks and focus on the relevant tasks.

⬧ Shimamura (2000) argued that the PFC acts as a dynamic filter which is processing ‘relevancy’ (Shimamura, The role of the prefrontal cortex in dynamic filtering. Psychobiology, 2000).

⬧ PFC is involved in performing many tasks for cognitive control and working memory.


The verb generation task can be performed with nouns that are associated with many actions (high filtering) or few actions (low filtering).


In these scans, prefrontal cotext areas showing higher activity in the high-filtering condition are shown in red.


Lesion overlap in patients who had difficulty in the high-filtering condition. Colors indicate the percentage of patients with damage in the highlighted regions.


Task-switch, multitasking, and flexibility


This task is cued by either a color (a) or a word (b). (c) Switching cost, the time required to switch from one task to the other (e.g., from naming the digit to naming the letter), is measured as the difference in response time (RT) on switch trials and no-switch trials. Patients with prefrontal lesions showed impairment only on the color cue condition.



Functional connectivity between prefrontal cortex and perceptual-motor areas for a visual-manual task (green) or auditory-vocal task (red).



(b) Connectivity strength before and after two weeks of multitask training. Connectivity was strong in a task-specific manner and showed little change, even though the participants became extremely proficient in performing the two tasks simultaneously.

(c) Activity in prefrontal cortex remained high, and shifted to an earlier latency after training in the dual-task condition, suggesting persistent cognitive control.



(a) TMS was targeted to disrupt activity in the frontal eye fields (red) or a control site (blue).

(b, c) A series of five TMS pulses were applied during a 570-ms interval that separated phases during which fMRI data were collected while participants viewed either visual stimuli (b) or a blank screen (c).

By comparing these conditions, the experimenters showed that the retinotopic maps in visual cortex were altered when top-down signals from the frontal cortex were disrupted.


Inhibitory control in the prefrontal cortex


Inhibitory Control

a planned response that should be aborted is erroneously executed 헛수윙

치려다 멈춤


(a) Successful actions sometimes require the ability to abort a planned response.

(b) BOLD response in motor cortex (M1) and right inferior frontal cortex (IFC) on trials in which a required response is performed (Go), a planned response is successfully aborted (Stop Inhibit), or a planned response that should be aborted is erroneously executed (Stop Respond).

The IFC responds on all stop trials, regardless of whether the person is able to abort the response. In M1, activation is high at the start of the trial on failed stop trials, likely reflecting a high state of anticipation in the motor system.

(c) DTI imaging reveals an anatomical network linking IFC with presupplementary motor area (preSMA) and the subthalamic nucleus (STN) of the basal ganglia.


▪Patients with prefrontal injuries may fail to engage successfully in goal-oriented behavior.

▪Knight and Grabowecky (1995) instructed subjects to monitor tones in either the left or the right ear.

▪Analysis: Attended tones minus unattended tones. ▪Healthy population: the effects of attention are seen at approximately 100 msec, marked by a larger negativity (N100).


Inhibitory Control: The left and right prefrontal cortex

⬧ Patients suffered from frontal lobe dysfunction.

⬧ Patients were presented with tones.

⬧ No response was required.



▪Analysis: Attended tones minus unattended tones. ▪Healthy population: the effects of attention are seen at approximately 100 msec, marked by a larger negativity (N100).

▪Major finding

▪Right prefrontal lesion patients show normal attention for ipsilesional tones but not for contralesional tones. ▪Left prefrontal lesion patients show less attentional effects for both contralesional and ipsilesional tones.



Medial prefrontal cortex is also involved in inhibitory control.

(a) Participants view a stream of letters and press a key whenever the letters alternate between X and Y (go trials). In a small percentage of the trials, the same letter appears in successive displays, and here the participants must withhold their responses (no go).

(b) Activation in the medial frontal cortex is lower in the cocaine users on all no-go trials—whether inhibition of the response succeeds or fails.



Actions are linked to schema control units. The perceptual system produces input to these control units. Selection of these units can be biased, however, by the contention scheduling units and the supervisory attentional system (SAS). The SAS provides flexibility in the response selection system.


Highlighted regions indicate cingulate voxels (potentially supervised attention system) that showed significant connectivity with eleven different brain regions. (conflict monitoring)



Perseveration

⬧ Perseveration is the tendency to continue giving a particular response even if the context has changed and the response is no longer appropriate.

⬧ Loss of Inhibitory Control

Documents

Gazzaniga • Ivry • Mangun Cognitive Neuroscienceraphe.kaist.ac.kr/lecture/2016fallbis451/ch12 Cognitive...Gazzaniga • Ivry • Mangun Cognitive Neuroscience FOURTH EDITION Cognitive