Abstract
Various functions of the cerebral cortex are systematically organized on a highly-folded 2D surface, and surface-based analysis has been widely used to project neuroimaging data onto this 2D surface for analysis and visualization, such as retinotopic and object category localizers. Compared to the 3D volumetric analysis, 2D surface-based analysis affords better inter-subject alignment, higher statistical power, and more accurate localization of functional areas. The most commonly used surface templates are fsaverage and fs_LR, which were created using the cortical surfaces of 40 individuals. Vertices of these templates were defined based on the inflated spherical surface. Due to geometric distortions of the inflation process, the distribution of vertices on the original anatomical surface is far from uniform: The central and lateral sulci are densely sampled, whereas many visual and frontal regions are sparsely sampled, resulting in inadequate use of information and reduced statistical power in these regions. Here we present the "onavg" surface template, which was created using high-quality MRI scans of 1,031 individuals from 30 OpenNeuro datasets. Vertices of the onavg surface template were defined based on the geometry of the anatomical surface rather than the spherical surface, affording uniform sampling of the cerebral cortex. In a series of analyses, we found that the onavg surface template consistently provided higher accuracy for multivariate pattern classification and higher noise ceiling for representational similarity analysis, especially in previously undersampled visual regions. The same classification accuracy and noise ceiling based on fsaverage or fs_LR can be achieved with only 3/4 of the original amount of data based on the onavg template. To sum up, the onavg surface template affords better sampling of the cortical surface and improved access to information encoded in spatial patterns, making it an ideal template for studying the neural mechanisms of cognitive processes (e.g., vision, attention).