Luis D. Ramirez^1,2, Jasmine Pan¹, Janneke F. M. Jehee³, Joseph T. McGuire¹, Sam Ling¹; ¹Boston University, ²University of California San Diego, ³Donders Institute for Brain, Cognition, & Behavior, Radboud University

Emotions guide perception, influencing threat and opportunity recognition in daily life. Because spatial frequency (SF) processing defines the scale of detail encoded by the visual system, emotional modulation of SF sensitivity could have a substantial impact on daily vision. Indeed, psychophysical work has suggested that negative emotions may augment SF processing. However, the mechanisms by which emotion affects visuocortical SF tuning remains unknown. To address this, we used fMRI to measure voxel-wise population SF tuning (pSFT) in early visual cortex (V1–V3), while participants heard sounds that contained either Negative or Neutral valence. To measure pSFT, we presented bandpass-filtered noise stimuli (0.5–12 cpd; 40 log-spaced steps; 80% contrast; 40-s blocks). During each pSFT block, participants also heard either Negative or Neutral valence daisy-chained auditory clips (IADS database), followed by a 2-s arousal rating window. In a separate baseline condition meant to map SF preferences in the absence of sounds, participants performed a silent luminance change detection task at fixation during pSFT mapping. Importantly, visual statistics were identical across blocks, and the only difference was what participants heard and the task performed. Negative emotion was validated by behavioral ratings, pupil dilation, and amygdala activation. Interestingly, we discovered that Negative sounds induced a systematic increase in preferred SF, both as a function of eccentricity, as well as a function of baseline pSFT peak. Furthermore, this shift was accompanied by increased selectivity (narrower population-level bandwidths), particularly in extrastriate cortex. Our results suggest that negative emotion may alter how we see by shifting population tuning preferences towards a higher visual resolution, allowing us to better discern the finer details of potential threats.

Acknowledgements: This research was funded by National Institutes of Health Grant EY028163 to S. Ling and supported by F99NS124144 and K00EY036804 to L.D. Ramirez.