Abstract
Previous studies have shown that response properties of neurons can be changed as a result of disease or training. New advances in functional magnetic resonance imaging, specifically population receptive field (pRF) modelling method, can be used to estimate the changes in response properties of neurons induced by reorganization and/or optimization of local neural connections. In this study, we characterize the pRF model properties in healthy volunteers who were trained on visual crowding task at a specific retinal location over 3 consecutive days (totally 2022 trials). Pre- and post-training fMRI images were acquired in all participants using 3-Tesla MRI scanner (Prisma, Siemens). The pRF parameters were estimated by the BOLD signals during a moving bar presentation where an aperture (exposing a flickering checkerboard pattern) was drifting across the screen in 4 orientations and two drifting directions (i.e. 8 bar configurations). Participants were to perform a fixation task while passively viewing the moving bar. Two-dimensional spatial pRF profiles for each voxel were estimated by using SamSrf 7.07 toolbox. To define the region of interest (ROI) of the trained area in the visual cortex, a separate fMRI localizer scanning session was performed where the crowding stimuli used in the behavioral training sessions were presented. We found that after perceptual training, participants who achieved substantial improvement in crowding task (experiencing less crowding effect) exhibited decreased pRF size (defined by the sigma, the spatial spread of the pRF), while others who have not shown behavioral improvement demonstrated wider pRF inside the crowding ROI. These results demonstrate that perceptual learning can modify basic neuronal response properties at visual cortex and provide insights into the understanding of plasticity across adult visual cortex.