Radiation Physics

PV QA 3 - Poster Viewing Q&A 3

TU_9_3197 - Treating benign cranial lesions - treatment planning system comparison

Tuesday, October 23
1:00 PM - 2:30 PM
Location: Innovation Hub, Exhibit Hall 3

Treating benign cranial lesions - treatment planning system comparison
D. Epstein, E. Shekel, R. M. Pfeffer, Y. Lipsky, R. Spiegelmann, and D. Levin; Assuta Medical Centers, Tel Aviv, Israel

Purpose/Objective(s): To compare three treatment planning systems for treating benign cranial lesions using stereotactic radiosurgery (SRS).


Three treatment planning software systems were compared for treating benign cranial lesions. iPlan (Brainlab, Germany) is a dedicated software for SRS planning using non-coplanar dynamic conformal arcs. The dose prescription, arc setup and MLC positions are planner controlled. Cranial SRS (Brainlab, Germany) (CSRS) is a new software for treating benign cranial lesions using up to 5 intensity-modulated arcs. The user can select the degree of modulation, (from minimal modulation to full VMAT). The couch angles, number of arcs and their lengths are optimization parameters and the treatment planning is highly automated.

Eclipse RapidArc (Varian, Palo Alto) (RA) uses an inverse planning approach where the planner determines the number of arcs, their lengths and couch angles, as well as the optimization objectives for the PTV and organs at risk (OARs). The optimization parameters can be adjusted "on the fly" as the optimization progresses.

The comparison metrics were conformity index (CI), gradient index (GI), volume of brain receiving over 12 Gy (V12) and mean brain dose between the three TPSs. We compared 10 plans for acoustic neuromas and meningiomas to plans generated and treated by iPlan, our current clinical standard. We also compared 4 cases that were treated with RA due to unusual shape, for which iPlan did not yield acceptable plans.


All CSRS and RA plans for all indications were judged clinically acceptable. The average CI was 1.38 and 1.30 for CSRS and RA, respectively, and the average GI was 3.76 and 4.11 for CSRS and RA, respectively. No metrics were significantly different compared to iPlan. However, for those lesions for which iPlan could not generate acceptable plans, treatment plans with CSRS were more homogenous than those generated by RA.

Planning times were significantly shorter for CSRS compared to the other TPSs. This was especially noticeable for iPlan, where MLC leaf shaping is a manual, labor-intensive process. Eclipse took much longer to perform the actual calculation using a 1 mm grid than either Brainlab TPS.

iPlan used less MUs compared to either RA or CSRS: an average of 1910 MUs for iPlan, 3883 MUs for RA and 2033 MUs for CSRS.


Planning times for CSRS were much shorter than for either iPlan or RA and planning was more efficient due to the high level of automation. However, the need to pre-define settings in a template outside the CSRS workspace can be cumbersome. Higher modulation in CSRS led to lower CI and GI values and lower OAR doses without affecting planning times.

For complex lesions for which iPlan was not adequate, both VMAT TPSs yielded highly conformal treatment plans that satisfied OAR constraints, however CSRS was faster to plan, gave more homogeneous dose distributions and used less MUs.

Author Disclosure: D. Epstein: None. E. Shekel: None. R.M. Pfeffer: None. R. Spiegelmann: None.

Dan Epstein, M.Sc


Send Email for Dan Epstein


TU_9_3197 - Treating benign cranial lesions - treatment planning system comparison

Attendees who have favorited this

Please enter your access key

The asset you are trying to access is locked. Please enter your access key to unlock.

Send Email for Treating benign cranial lesions - treatment planning system comparison