The role of the pre-supplementary motor area in the control of action

Abstract

Although regions within the medial frontal cortex are known to be active during voluntary movements their precise role remains unclear. Here we combine functional imaging localisation with psychophysics to demonstrate a strikingly selective contralesional impairment in the ability to inhibit ongoing movement plans in a patient with a rare lesion involving the right pre-supplementary motor area (pre-SMA), but sparing the supplementary motor area (SMA). We find no corresponding delay in simple reaction times, and show that the inhibitory deficit is sensitive to the presence of competition between responses. The findings demonstrate that the pre-SMA plays a critical role in exerting control over voluntary actions in situations of response conflict. We discuss these findings in the context of a unified framework of pre-SMA function, and explore the degree to which extant data on this region can be explained by this function alone.

Fig. 1

Lesion localisation. (A) Activation in the region of the SEF – a marker for the rostral extent of the SMA – superimposed on T1-weighted anatomical scan normalised in MNI space (orange). The purple area identifies the location of the rostral pre-SMA region which we have previously shown to be activated by changing oculomotor plans during conflict. (B–D) Sagittal, axial and coronal anatomical slices showing the extent of the lesion.

Fig. 2

Schematic of the change-of-plan paradigm. Stimuli are not drawn to scale.

Fig. 3

Behavioural data. (A) Psychometric inhibition functions for patient AG derived using the change-of-plan paradigm. The line plots show maximum a posteriori (MAP) estimates for the relation between the probability of inhibiting the first movement and the SOA adjusted by the median RT (effectively the time available to inhibit a response) for the left (solid line) and the right (dotted line) hands. Scatter plots of the raw data (left, white circles; right, black triangles) are also shown. The SOA values were determined by a staircase tracking algorithm that targeted the 50% performance level and therefore the sampling was slightly different between the two sides, and weighted towards the middle of the function. Only SOA values for which at least 5 data points were available were used. The SOA values were: left, 100 ms (n = 9), 150 ms (n = 24), 200 ms (n = 25), 250 ms (n = 10); right, 100 ms (n = 6), 150 ms (n = 15), 200 ms (n = 24), 250 ms (n = 20), 300 ms (n = 6). So as to generate functions describing the relation between the probability of changing response and the effective time available to do so, the SOAs were adjusted by subtraction from the median RT on Go trials: left = 458 ms, right = 440 ms. The horizontal box plots show 0.05, 0.25, 0.5, 0.75, and 0.95 confidence boundaries. The two functions are significantly different (Bayesian p = 0.002). (B) Scatterplot of the left and right hand inhibition function 50% thresholds for AG (filled markers) and the ten control subjects (unfilled markers), derived as described in A. AG differed from controls only on the left (two-sample t test, p = 0.011), and none of the controls showed a significant difference between left and right based on Bayesian MAP estimates.