A quantitative comparison of three methods for inflating cortical meshes

Polygon meshes representing the cerebral cortex have been increasingly used for visualisation and analysis in biomedical imaging in recent years. Due to the highly convoluted structure of the cerebral cortex, a number of methods have been proposed for inflating cortical meshes to expose information buried in sulci and simplify the surface for further computation. Such inflations necessarily introduce geometric metric distortions to the surface. In this paper we present a quantitative comparison of three methods for inflating cortical surfaces: CARET, FreeSurfer and Diffusion-based Deformation. The methods were evaluated by how well they preserved a set of surface metrics, as well as their computational efficiency. We show that CARET is best for area preservation, while FreeSurfer performs best with respect to angular distortion and computational efficiency.