Forward planning or inverse planning? In medio stat virtus

Abstract

Introduction Implementation of inverse planning algorithms into modern treatment planning systems (TPS) has supposed an improvement of the planning process especially in cases where intensity modulated planning solutions are desirable. On the other hand this improvements drastically reduced control that user might have over planning parameters. This situation generated an interesting debate among those who support forward planning and those who support an inverse planning approach. Purpose In this work we explore a third possibility to which we refer as initialized inverse planning and that may be considered a step in the middle between forward and inverse planning strategies. This methods actually combines flexibility of the first and optimization capability of the latter. Materials and methods TPS used in this work is Pinnacle 3 by Philips and its Direct Machine Parameter Optimization (DMPO) inverse planning algorithm. Our initialized inverse planning approach (to which we refer as iDMPO) starts from the solution provided by a forward planning strategy which is then optimized by the DMPO algorithm. In this second step we impose tighter constraints than the one reached with forward planning. Results In this work we compare this three methods (forward, inverse with DMPO and iDMPO) in the specific case of prostate cancer treatments. Nevertheless conclusions are also applicable to other treatments. Conclusion Results show that with iDMPO technique it is possible to improve treatments dosimetric results, compared to the forward planning, imparting lower doses to organs at risk while delivering more homogeneous dose distributions to target volumes and thus reaching the pure inverse planning approach solution. Disclosure Authors disclose any relationship that may bias this work.