Christian Windischberger¹ (), Dominik Zuschlag¹, Rüdiger Stirnberg², David Linhardt¹; ¹Highfield-MR, CMPBME, Medical University of Vienna, Austria, ²MR Physics, DZNE, Bonn, Germany

Population-receptive field (pRF) mapping is an approach based on functional magnetic resonance imaging (fMRI) where visual stimuli are presented in the MR scanner while BOLD-weighted images are acquired. In the analysis of pRF, activation time courses in visual areas are associated with the stimulus to obtain information about receptive visual field areas processed in a given part of the visual cortex. While 2D echo-planar imaging (EPI) is still the workhorse in current fMRI brain mapping studies, EPI with three-dimensional k-space encoding (3D-EPI) has the potential to provide BOLD-weighted images with increased spatial resolution and signal-to-noise ratio (SNR). Here we compare pRF results acquired with 2D-EPI (CMRR; Moeller et al., 2010) and 3D-EPI (Stirnberg et al., 2021) protocols to evaluate potential benefits of employing 3D acquisitions. We recorded 2D and 3D data in eight young, healthy participants on an ultra-high field 7 Tesla MR scanner (Siemens Magnetom 7T) with a spatial resolution of 1.2mm isotropic (TE/TR=25/1000ms). In order to test performance at even higher spatial resolutions as required for laminar imaging, we also acquired 0.7mm isotropic resolution data in the same participants (TE/TR=25/2000ms). Minimal preprocessing was performed using fMRIprep v.25.0.0 (https://fmriprep.org). pRF analyses were performed using prfprepare v7.0.0 (https://github.com/dlinhardt/prfprepare) and GEM-pRF (Mittal et al., 2025). 3D-EPI acquisition showed medium to strong effects of improved model-fitting quantified by variance-explained values than 2D-EPI over all early visual areas (V1: Cohen’s d=0.56 , V2: d=0.44 , V3: d=0.47). Wilcoxon signed-rank testing showed highly significant increases in explained variance values of 3D over 2D EPI results while no changes in pRF parameters were apparent. Ultrahigh-resolution runs (0.7mm) were not affected by obvious imaging artefacts and showed the expected retinotopic maps across the early visual cortex (V1-3). Our results clearly demonstrate the advantages of 3D-EPI acquisition protocols in pRF studies with millimeter or even sub-millimeter resolution.

Acknowledgements: This research was supported by the FWF (10.55776/P35583; 10.55776/PAT8722623; 10.55776/ESP1445525)