Click here to


Are you sure ?

Yes, do it No, cancel

Minimizing Plan Quality Variation and Reducing Planning Time Via Automatic Platform for Breast Radiotherapy

K Yang*, L Romak , B Streifeld , Y Do , J Losito , E Geating , M Hofmann , S Day , F Mourtada , Christiana Care Hospital, Newark, DE


(Sunday, 7/29/2018) 5:05 PM - 6:00 PM

Room: Karl Dean Ballroom A2

Purpose: To minimize plan-quality variation and to reduce the planning-time of tangential-breast external-beam radiotherapy with the implementation of a novel CT-simulation workflow and the automatic breast platform in RayStation-V6.1.

Methods: Routine breast CT-simulation involves shift to the isocenter from 3-torso-tattoo-marker (3PT). The implemented automated breast planning-module relies on a detection algorithm of the breast-wire and border-marker. Anthropomorphic thorax phantom was scanned using a Siemens CT-Sim with/without 3PT and breast-wire to identify the minimal necessary modification of our routine CT-simulation process to enable this new feature in Raystation. A clinical patient study was then performed with the novel CT-simulation workflow, and concurrent tangential-breast planning (manual vs automatic-breast-platform) to depict the plan dosimetric-quality and planning-time efficiency.

Results: It is critical to reconstruct the planning-CT with fewer slices inferiorly to exclude 3PT, since the auto-breast algorithm failed to identify the border-marker and/or its laterality when the 3PT was visible in the CT-image. 3PT coordinate defined on the original CT with the same frame-of-reference facilitates our routine isocenter shift from 3PT. Compared to clinical plan concurrently generated manually by dosimetrists, the auto-plan further reduced V20,Ipsilateral Lung, V30,Ipsilateral Lung, DMean,Ipsilateral Lung, and DMean,Heart by 3.3%, 3.6%,1.2-Gy, and 0.1-Gy, respectively. It also improves the coverage in V100,cavity and V95,CTV by 8.1% and 2.0%, respectively, while the hot -spot difference in D1cc,CTV is found to be negligible. Our standard 10-day turnaround-time could be significantly reduced as automatic breast planning takes less than 10-min from the typical 2-3 hr manual planning. This allows the physicians to review the plan when contouring the cavity for the boost-plan, and the auto-plan could then either be accepted or used as a benchmark for routine breast planning.

Conclusion: Automatic planning allows safely reducing planning-time and minimizing the plan-quality variation for breast radiotherapy. More clinical cases are being collected and will be presented.


Breast, Optimization, Quality Control


TH- External beam- photons: General (most aspects)

Contact Email