Research Masterclass in Radiotherapy Physics 2017

Erasmus-iCycle and Pinnacle Auto-Planning for plan quality enhancement with automated planning, are we done? Not at all: - Sequential constrained optimization - Too slow for fast planning (1 min) - Configuration may be really time consuming/impossible - Include beam angle optimization - Convex vs. non-convex cost functions

- Hard constraints? - Pareto-optimality? - Systems used/evaluated by very few users

Alternatives to automated planning for plan improvement?

 Manual planning with super fast optimization algorithms  Pareto Navigation (Raystation)  Fast graphical planning  …….

…. also complimentary to automated planning (not yet tested)

11