
Among the many people who have measured curvature in volumes, our work is partly inspired by Monga et al., who analyzed curvature with 3rd derivatives to find feature lines of extremal curvature,Others have used principal curvature directions to indicate surface shape with textures and kinetic visualization, but our paper is most directly inspired by Hladuvka, who first showed that a 2-D space of curvature magnitudes provides an intuitive and effective way to highlight volume features.
As far as non-photorealistic volume rendering, Rheingans and Ebert mapped many NPR methods to volume rendering in their Vis 2000 paper. Oher people have worked on pen-and-ink [Trevett and Chen], or contour-based [Csébfalvi], or stippling-based [Lu et al.] methods.
More recently, NPR volume rendering has become interactive with graphics hardware. This paper [Nagy et al.], for example, is interesting for its interactive use of curvature-directed strokes and hatches.