Intra- and inter-session reproducibility of artificial scotoma pRF mapping results at ultra-high fields

Abstract

Population receptive field (pRF) mapping allows for the assessment of visual field dysfunctions from retinal diseases through fMRI measurements of the visual cortex. Here we assess the intra- and intersession reliability of artificial scotomata pRF data. Twenty healthy subjects (ten female; age: 25.9±3.6) were measured on a 7T MRI scanner in two sessions with two runs each. Functional data (1mm isotropic resolution, 32 slices) was aligned perpendicular to the calcarine sulcus. T1-weighted anatomical data were acquired at 0.7mm resolution. Participants were presented with a flickering checkerboard bar aperture moving through the field of view in eight different directions. The central 2° radius were not stimulated to simulate a central scotoma. Five subjects had to be excluded from analysis due to high motion and/or limited attention in one or more runs yielding 15 subjects for the final analysis in mrVista. Analysis input did not contain any information on the artificial scotoma. Reproducibility was assessed by calculating Spearman’s correlation coefficient for all respective pRF-mapping parameters between single-run results. Coverage plots for all included subjects show the central artificial scotoma by lower pRF centre densities, compared to full-field stimulation. Correlation coefficients yield high values for the pRF position for within (eccentricity: 0.91; polar-angle: 0.99) and between sessions (0.89; 0.99). As expected (Lerma-Usabiaga, 2020; Lage-Castellanos, 2020), much lower correlation coefficients were found for the pRF size parameter (within: 0.39; between: 0.34). Results corroborate previous studies performed without artificial scotomata (van Dijk, 2016) and show that even a major disruptive factor, such as a scotoma, does not impair the robustness of pRF-mapping. In conclusion, pRF-mapping is an objective approach applicable to longitudinal studies for monitoring treatment effects or disease progression in patients suffering from partial visual field loss. However, the low reproducibility of the pRF size parameter indicates the need for further improvements.