Where are the fMRI correlates of phosphene perception?

Pulses of transcranial magnetic stimulation (TMS) over occipital cortex can induce transient visual percepts called phosphenes. Phosphenes are an interesting stimulus for the study of the human visual system, constituting conscious percepts without visual inputs, elicited by neural activation beyond retinal and subcortical processing stages in the visual hierarchy. The same TMS pulses, applied at threshold intensity phosphene threshold (PT), will prompt phosphene reports on half of all trials ("P-yes") but not on the other half ("P-no"). Contrasting brain activity (P-yes > P-no) can provide unique information on neural mechanisms underlying conscious percepts, as has been demonstrated by published EEG studies. Yet to our knowledge no articles reporting analogous contrasts with functional magnetic resonance imaging (fMRI) have been published. Since it seems unlikely that such studies have never been performed, this straightforward and technically feasible idea may have been explored in multiple failed, and unpublished, attempts. Here, we argue why such unsuccessful attempts, even small-scale, best be shared. We also report our own failed attempt to find phosphene-related activity in fMRI. Threshold phosphenes are weak percepts, and their detection subjective and difficult. If fMRI correlates of phosphenes are obtainable with this contrast, small-scale ('pilot') measurements may not be sufficiently powerful to detect them. At the same time, due to the challenges and costs involved in TMS-fMRI, attempts might not often get beyond the piloting stage. We propose that the only way out of this quandary is the communication and sharing of such unsuccessful attempts and associated data