Medial axis optimal locations for min-max sphere packing (Q1592843)

The sphere packing problem considered by the author is motivated by radiosurgical treatment planning using gamma knives: Given a region \(R\) in the 3-dimensional Euclidean space, find a smallest number of spheres with centers in \(R\) such that the area of \(R\) covered by these spheres exceeds a given bound \(V\) and the number of points on the boundary of \(R\) touched by spheres is maximized. In previous papers a greedy algorithm was developed by the author, in which iteratively maximal spheres \(S(X_i)\) are packed in \(R_1=R\), \(R_2:=R\setminus S(Y_1)\), etc. While this algorithm seems to work well in practice, it is shown in this paper that it produces, in general, non-optimal solutions to the sphere packing problem. The paper is concluded by a section on further refinements of the packing problem to better model the radiosurgical planning problem. The resulting mathematical models seem to be very challenging.