Wing Kwan Hannah Chu¹ (), Bas Rokers^1,2,3, Kartik K Sreenivasan^1,2; ¹Psychology, New York University Abu Dhabi, ²Center for Brain and Health, New York University Abu Dhabi, ³Aspire Precision Medicine Institute Abu Dhabi, New York University Abu Dhabi

Items held in working memory (WM) can be decoded from fMRI activity in visual cortex, motivating the hypothesis that visual regions store information to support WM. Causal evidence for this hypothesis, however, is lacking; transcranial magnetic stimulation (TMS) studies have yielded decidedly mixed results, and cannot rule out unwanted downstream perturbation of higher order regions. We used artificial scotomas as a novel means of modulating visual cortical activity to test its role in WM. An artificial scotoma is a simulated blind spot, which can be induced by presenting texture or motion around the scotoma location. The scotoma location is filled in by information interpolated from the surround presumably via feedback from higher-order regions. Forty participants completed a WM task that required them to memorize the motion directions of two drifting gratings in opposite quadrants, hold these in WM over a memory delay, and report the motion direction of the grating cued at the end of the trial. During the delay, we induced an artificial scotoma by presenting a gray patch superimposed on a drifting plaid background. The scotoma could occur in one of three locations on each trial: the location of the to-be-probed (i.e., cued) memory item, the location of the uncued memory item, or a non-memory item location. Induction of the scotoma resulted in increased memory error, decreased memory precision, and a greater propensity to misreport memorized motion direction by 180°. Critically, performance was disrupted only when the scotoma was induced in the location of the to-be-probed memory item and only on trials when subjects confirmed that a scotoma was successfully induced. By using spatially selective artificial scotomas that target early visual processing, we overcome the limitations of previous fMRI and TMS studies to provide causal evidence that visual areas play an active role in WM storage.

Acknowledgements: This work was supported by the NYUAD Center for Brain and Health, funded by Tamkeen under NYUAD Research Institute grant CG012