PMID-17237780[0] Switching from automatic to controlled action by monkey medial frontal cortex.
- SCLIN's blog entry
- task: two monkeys were trained to saccade to one of two targets, left/right pink/yellow. the choice was cued by the color of the central fixation target; when it changed, they should saccade to the same-colored target.
- usually, the saccade direction remained the same; sometimes, it switched.
- the switch could either occur to the same side as the SUA recording (ipsilateral) or to the opposite (contralateral).
- found cells in the pre-SMA that would fire when the monkey had to change his adapted behavior
- both cells that increased firing upon an ipsi-switch and contra-switch
- microstimulated in SMA, and increased the number of correct trials!
- 60ua, 0.2ms, cathodal only,
- design: stimulation simulated adaptive-response related activity in a slightly advanced manner
- don't actually have that many trials of this. humm?
- they also did some go-nogo (no saccade) work, in which there were neurons responsive to inhibiting as well as facilitating saccades on both sides.
- not a hell of a lot of neurons here nor trials, either - but i guess proper statistical design obviates the need for this.
- I think if you recast this in tems of reward expectation it will make more sense and be less magical.
- would like to do shadlen-similar type stuff in the STN
questions
- how long did it take to train the monkeys to do this?
- what part of the nervous system looked at the planned action with visual context, and realized that the normal habitual basal-ganglia output would be wrong?
- probably the whole brain is involved in this.
- hypothetical path of error trials: visual system -> cortico-cortico projections + context activation -> preparatory motor activity -> basal ganglia + visual context (is there anatomical basis for this?) -> activation of some region that detects the motor plan is unlikely to result in reward -> SMA?
____References____
|