Laura Ahumada¹, Judith Cediel¹, Andrew F. Farkas¹, Faith E. Gilbert¹, Hannah Engel¹, Arash Mirifar¹, Katherine McCain¹, Caitlin M. Traiser¹, Sarah Gardy¹, Christian Panitz², Andreas Keil¹; ¹University of Florida, ²University of Bremen

Visuocortical changes in orientation tuning consistent with experience-dependent plasticity have been studied in learning paradigms, such as in aversive conditioning. Aversive conditioning consists of pairing an initially neutral stimulus (CS+) with an unconditioned aversive stimulus (US). Here we used an aversive generalization learning paradigm to compare changes in visuocortical tuning in two laboratories—a standard EEG laboratory (standard) and a 3-Tesla fMRI scanner (EEG-fMRI). This paradigm employs a generalization gradient of stimuli to quantify changes in tuning from the CS+ to gradually different orientations (GSs) not paired with the US, which may take the shape of sharpening or generalization. Four Gabor patches were used, where one orientation represented the CS+ and the remaining three the GSs. All patches flickered at 15 Hz and were shown foveally. For the standard study, a white noise was used as the US; for the EEG-fMRI study, an electrical shock was used. The steady-state visual evoked potential (ssVEP) of 31 participants in the standard study and 24 participants in the EEG-fMRI study were separately analyzed using a mathematical model for quantifying tuning functions, the Ricker wavelet. This model identifies the tuning pattern that best characterizes the data. Results showed that in both laboratories, the visuocortical response to the CS+ increased during the acquisition and extinction phases. During acquisition, similar generalization tuning was observed in both laboratories. However, the exact tuning patterns differed during extinction, with sharpened tuning apparent in the standard but not in the EEG-fMRI study. Overall, visuocortical tuning was reliably measured in both settings, demonstrating that aversive learning in combination with Ricker wavelet fitting is suitable for characterizing tuning functions in human EEG/ssVEP recordings. Results also suggest that the laboratory setting, including the nature of the US, may influence the experience-dependent re-tuning of neural mass activity as measured by EEG signals.