Fig 5. Effect of electrical stimulation in AIP on the choices of monkey M2.

3D-shape categorization task: (A) shows the average proportion of choices (±SEM) that matched the preferred shape of the 3D-shape selective site (preferred choices) as a function of stereo-coherence for trials with (red) and without (blue) microstimulation. Positive and negative stereo-coherences relate to preferred and non-preferred shapes, respectively. Solid lines show the fitted psychometric functions. Microstimulation shifted the psychometric function towards a higher number of nonpreferred choices. (B) Histogram of microstimulation effects (N = 28 3D-shape selective sites) expressed as a shift of the psychometric function in terms of percent stereo-coherence. Positive values express psychometric shifts towards an increase in the number of preferred choices. (C) Histogram of microstimulation effects expressed as a shift of the psychometric function in terms of percent stereo-coherence. Positive and negative values are used for psychometric shifts towards more convex or more concave choices respectively. Microstimulation increased the number of concave choices. (D) Same as (C) but for nonselective sites (N = 19). TOAS task: (E) shows the average proportion of choices (±SEM) during the TOAS task in the direction of the choice target used for concave choices during the 3D-shape-categorization task for trials with (red) and without (blue) microstimulation. Positive TOAs indicate trials where the (right) concave choice target appeared first. Negative TOAs indicate trials where the (left) convex choice target appeared first. Solid lines show the fitted psychometric functions. Microstimulation during the TOAS task strongly increased the number of saccades towards the concave choice target. (F) Histogram of microstimulation effects (N = 10) during the TOAS task. Stimulation effects are expressed as the additional delay (TOA in ms) that would need to be added to the delay between the two targets on trials without microstimulation in order to match the behavioral performance that was actually observed on trials with microstimulation. Microstimulation had an effect equivalent to presenting the concave choice target 43ms before the convex choice target. Black bars indicate sites displaying significant shifts of their psychometric functions due to microstimulation (p<0.05).