The modeling of blood diseases (e.g. polycythemia vera) and the mathematical optimization of treatment options are highly complex tasks, which are subject of a number of recent studies. Usually nobody can be sure of how good the solutions really are, because only local optimization techniques are used. The thesis aims at combining the medical application with methods of global optimization for dynamical systems and thus being able to guarantee the optimality of a treatment. As especially global optimization techniques are computationally very expensive, mathematical complexity reduction might become a keystone of the thesis.