Click here to


Are you sure ?

Yes, do it No, cancel

Joint AAPM-ESTRO Symposium: Automated Treatment Planning in Clinical Practice

C Hansen1*, B Heijmen2*, M Hussein3*, K Sheng4*, (1) Odense University Hospital, Odense, Denmark, (2) Erasmus MC Cancer Institute, Rotterdam, ,(3) National Physical Laboratory, Teddington, England, (4) UCLA School of Medicine, Los Angeles, CA


(Wednesday, 8/1/2018) 10:15 AM - 12:15 PM

Room: Karl Dean Ballroom A2

There is increasing evidence that conventional, interactive trial-and-error planning has a substantial risk to result in sub-optimal treatment. This is bad for patients and also for tax/insurance payers, as their money is not optimally used. In recent years, new planning tools to streamline and automate treatment planning have been developed. In particular fully automated treatment plan generation has turned out to be highly effective to enhance plan quality. Both a posteriori (MCO, Craft et al) and a priori (Erasmus-iCycle, Breedveld, Heijmen et al.) approaches for definition of a Pareto-optimal plan have been clinically introduced. Many groups have investigated the use of machine learning based on historical treatment plans to propose a dose distribution for a new patient (Yuan et al., Moore et al., Verbakel et al.), which has resulted in a commercial product. Also the work by Cotrutz et al on the automatic control on regional doses has been an inspiration for a commercial product for fully automatic plan generation. Scientific reports have demonstrated that these four clinical approaches are effective in enhancing quality. On top of this, a huge reduction in planning workload is feasible.

The four talks in this session will present achievements, limitations and future developments of four solutions for automated treatment planning (Philips Auto-planning, Varian RapidPlan, iCycle automated a priori multicriteria optimization, automated planning for 4pi treatment delivery). The presenters will discuss to what extent planning is indeed fully automated (i.e. no manual input) for each of the presented options.

Learning Objectives:
1. Understand for four novel, clinically applied, automatic planning approaches the algorithmic principles.
2. Discuss for each of the approaches observed clinical advantages compared to conventional planning.
3. Compare the four approaches regarding advantages and disadvantages.
4. Discuss further future improvements.



Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email