Surface rendering in electron tomography: A heuristic approach

The final result of an electron tomographic reconstruction, computed from several projections, is a 3‐D array of data which resides in the computer memory as a long string (from 323 to 2563) of numbers. It is vital to retrieve information relevant to biology from this vast amount of numbers, and to share it with the scientific community. This is done with the use of visualization processes that may represent the most time consuming task in a reconstruction. The present article describes a technique of surface rendering designed by the authors to represent structures characterized by the presence of a number of subunits and of features such as branching, holes and cavities. The software used produces a ‘structure’ (an array containing different types of variables, peculiar to the C programming language) that is able to exploit the resources of the graphic engines embedded in modern workstations and of graphic libraries. This structure can be used to produce visual presentations in the form of wireframe‐ and surface‐rendered models with shading determined by the Gouraud algorithm.