Abstract
When stimuli are presented closely in time or space, the total evoked neuronal response is often less than the sum of responses evoked by each of the stimuli alone. Similar suppression is observed in fMRI BOLD signal, which has been attributed to neuronal origins. However, the suppression effect can also be due to nonlinearity in the hemodynamics, devoid of any neural correlates. We conducted two experiments with a subject who was born without an optic chiasm. The achiasmic subject's left and right visual hemifield representations are overlaid in both cerebral hemispheres. At each point in V1-V3, there are two population receptive fields (pRFs) situated symmetrically across the vertical meridian. These pRFs do not interact. For example, a high-contrast static or counter-flickering mask presented contralateral to a contrast-detection target has no effect on detection threshold. In one experiment, we measured the net BOLD response to a high-contrast peripherally presented stimulus that flashed twice with a 1 sec ISI. Subtracting the single-flash response from the net response yielded the second-flash contribution, which in V1 was 31% less than the single-flash response. However, a similar amount of suppression (28%) was observed when the second flash was presented to the visual hemifield contralateral to the first flash. Hence, less than 1/10 of the observed 31% suppression (an insignificant amount) can be attributed to a neuronal origin. In a second experiment, we found that a high-contrast spatial surround led to 100% suppression of a low contrast center in V1 when both center and surround were presented to the same hemifield. When the surround and center were presented to different visual hemifields, we observed 68% suppression of the center. Thus, only 1/3 of the observed suppression in the conventional surround suppression setup can be attributed to a neuronal origin. In both experiments, non-neuronal hemodynamic suppression dominated.
Meeting abstract presented at VSS 2014