Please enable JavaScript.

Coggle requires JavaScript to display documents.

Task 6: (Taylor et al. (Reinforcement learning theory (Errors = negative,…

- - - - larger ERN even on tasks not related to OCd symptoms
      - corrleation between ERN magnitude and OCD symptoms
      - increased ACC activity in resting state
    - - greater ERN and feedback negativity
      - negative affect might be associated with elevated ERNs
    - - smaller ERN and dminished MFC activity in response to errors
      - rACC and not dACC as the source of smaller activity in go/no-go task
      - reflection of motivational deficit- negative symptoms, interacting with performance monitoring
  - - - dACC monitors for the presence of cognitive conflict + passes a signal to the lateral PFC to increase cognitive control when conflict is high
      - errors generate high conflict when the stronger, but undesired, response tendency reaches a sufficient threshold to command an actual response
    - - elevated both for trials with high conflict and for trials in which errors occurred
      - predicted dlPFC activity in a subsequent trial + adjustments following an error (post- error slowing)
      - same dACC region will process conflict-related activity and error-related activity
      - criticism: Several lines of research have challenged the contention that error signals reflect the same processes and neural systems as response conflict
  - - - dlPFC - controll over task selection
      - dACC/pMFC - conflict monitoring/task set region
    - - rACC - reward: evaluate outcomes, positive and negative
      - dmPFC - self-relevance
      - aMFC: motivation and affect - monitor and influence motivational contexts such as "correct is good"
    - - medial PFC
        
        endogenous control over action patterns
        
        part of the cognition related multiple demand network
      - (pre)SMA
        selection of actions: activity rises whenever a choice has to be made between competing responses
      - ACC
        
        processing unexpected outcomes of our actions (positive + negative) to remediate + optimize our action repertoire
        
        direct control over the reward system - through an inhibitory control via the lateral habenula over the midbrain dopamine neurons in substantia nigra and ventral tegmental area (VTA)
        
        action outcome monitoring network
- - - - EXP 2: (when introducing noninformative stimuli) in contrasting errors with informative (negative) feedback vs. correct trials with informative (positive) feedback, the results from experiment 1 were replicated
      - EXP1:
        
        positive feedback - increase activity in VS (Nucleus accumbens)
        
        negative feedback - rCMA, inferior anterior insula + epitalamus (habenular complex)
        
        rCMA reacted only to errors with negative feedback but not to errors without feedback (ruled out an influence of response conflict / uncertainty on its role in error detection by external signals)
    - - ventral striatum – reward
        
        activated only when positive feedback occured - phasic dopamine release
        
        inconsistent with previous findings that nonrewarding salient events may result in dopamine release in the nucleus accumbens
        could it be that people expect to be wrong => for negative feedback there is no prediction error and no striatal activity
        
        whereas if correct, there is unexpected positive feedback, leading to DA response in VS
      - habenular complex – reward processing + influencing DA system
        
        negative feedback (when expected rewards fial to occur)
        
        activated most when an error was made and negative feedback was received
        
        similar activity was observed when correct responses were not followed by informative (positive) feedback (interaction of response type and the occurrence of informative feedback)
        
        habenula inhibits the midbrain nuclei (VTA, SN)
        higher engagement of the habenula reduces the probability of phasic dopamine release in the reward system
        
        a) trials without feedback information  lower habenular activity for errors than correct trials  VTA and SN are less inhibited (feedback resulting in a phasic dopamine signal would be highly informative during the uncertain error trials)
        
        b) positive feedback on correct trials  positive error in reward prediction
        reflected in the decrease in the habenular activity when correct trials were followed by positive feedback  disinhibiting the dopaminergic midbrain areas
        
        c) negative feedback on errors assures the participant that no reward (knowing that the response was correct) can be received in the current trial  negative error in reward prediction  accompanying increased habenular activity  increased inhibition of the dopaminergic midbrain nuclei resulting in decreased dopamine output
      - model:
        
        decreased dopamine release during negative feedback errors can result in higher activity in the rCMAit is not rCMA that drives habenula, but the other way around!! We've been assuming that medial error detection drives habenula to inhibit midbrain DA neurons
        
        for correct trials, reward expectancy was slightly higher than during errors because of higher certainty
        
        difference in the habenular activity between informative and noninformative activity correlates with the error in reward prediction
- - - - RESULTS: activity of ACC was dependent on
        
        25% intention to move or not
        
        c) combination of movement and reward expectation 11%
        
        25% - expectation of reward or no-reward
      - METHOD: go-no go in response to picture cues, where the associations of pictures with instructions were sometimes reversed. Also changing the rewards
    - - METHOD: - taught monkeys to select one of two joystic movements after free delivery of one of two rewards correct movements were rewarded with a second similar reward
      - RESULTS: - ACC lesions disrupted performance on the reward guided action selection task without affecting visual stimulus selection
        
        The ACC lesion did not affect the ability to discriminate stimuli associated with rewards
        
        a) ACC neuron that encodes both a specific motor intention (No-go) + reward expectancy increases in activity when a visual cue is presented, although activity time is locked ,to the cues it is not specific to either cue. - But encodes the expectation of reward delivery on a No-go trial as opposed to the association of a reward on a Go trial or Go or No-go trials in which no reward is expected
    - - Method: - Rats performed a T-maze task in which the choice of one arm was always followed by a small reward whereas choosing the other arm resulted in a large reward large reward could only be obtained if the rats selected a more effortful action
      - RESULTS: Normal rats went for high reward, where ACC lesioned rats rarely did.
        
        impaired ability to integrate both the expected costs and benefits of an action
         rather than a simple insensitivity to reward differentials, failure to remember the size or positions of rewards, or difficulty climbing the barrier
- - - - OFC damage = disrupted ability to sustain the correct choice of stimulus after positive feedback (not for action-value task) - more likely to shift the choice after win in stimulus-value task. no effect for action-value tasks
      - dACC damage= more likely to shift to the other option after a previously winning choice (win–shift) than were healthy controls in the action-value task
        were no more likely to shift after a win in the stimulus-value task than were healthy subjects