Alfredo Sklar¹ (), Jenay Kocsis¹, Dylan Seebold¹, Brian Coffman¹, Dean Salisbury¹; ¹Clinical Neurophysiology Research Laboratory, UPMC Western Psychiatric Hospital, Department of Psychiatry, University of Pittsburgh School of Medicine

Introduction: The influence of attention on neural responses to gradations in visual contrast has been extensively studied in humans using electrophysiology and functional neuroimaging. However, these investigations have produced conflicting results regarding the mechanisms by which attention modulates the contrast response function (CRF), and both techniques possess limitations. The present study leveraged the spatiotemporal resolution of combined magnetoencephalography (MEG) and MRI to examine attentional modulation of the CRF within visual cortex. Methods: MEG was recorded from 12 participants during a 2-alternate forced-choice task. Participants identified the orientation of lateralized gabor patches during valid and neutral location cue conditions. Structural and resting-state fMRIs were acquired to obtain individual Human Connectome Project cortical parcellations. MEG sensor data was localized to 5 divisions of the visual cortex: V1, early (V2, V3, V4), dorsal (V3A, V3B, V6, V6A, V7), ventral (FFC, PIT, V8, VMV1/2/3, VVC), and MT+ (FST, LO1/2/3, MST, MT, PH, V3CD, V4t). Results: Participants responded more accurately (p<.001) and rapidly (p<.001) to higher contrast stimuli. Responses were faster on cued trials (p=.005) and to right visual field stimuli (p<.001). Across regions, increasing contrast (p<.001) and valid location cues (p=.01) enhanced cortical activity. The effect of contrast, however, differed across region with no significant effect observed within the ventral division (p=.34). Conclusion: Sensitivity to stimulus contrast and effects of covert attention were identified within V1 and early visual processing regions as well as some (dorsal division and MT+), but not all (ventral division), higher-order visual processing centers. The spatiotemporal resolution of combined MEG/MRI analysis may help replicate data obtained from single-unit recordings in animals and address discrepancies observed with fMRI analyses. Future work will model the CRF of each participant and derive its parameters to better characterize the response properties of the human visual cortex and assess the mechanisms of executive attentional modulation.

Acknowledgements: Supported by K23 MH127389 (Alfredo Sklar, MD, PhD PI).