Spline snakes have been proven to be efficient segmentation models in bioimaging. They enable shape representation (curves and surfaces) in the continuous domain with only few control points using compactly supported basis functions. Advantages of these models include fast optimization schemes and intuitive user-interactivity, because moving a control point directly influences local deformation of a shape. This project aims at developing a novel energy term that prevents the snake from twisting and self-intersecting. The underlying idea is to use Coulomb’s law to place charged electrons (discretely or in the continuum) at specific locations on the contour. Internal repulsive charges will favor non-self-intersecting shapes of the snake. The project consists in A) a theoretical aspect where the student should work out the equations based on physical laws for 2D and 3D shapes and B) a practical aspect where the corresponding energy will be implemented into our existing framework. The result will be tested with artificial data and ideally be applicable to segment the aorta in 3D MRI and CT images with our existing cylindrical snake. The student should be familiar with vector analysis and objective-oriented programming such as Java or C++.