MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel

Development and Customization of the RayStation Automatic Breast Planning Module: A Program to Adapt the Provided Module to Suit Clinical Needs

C Collins*, M Kaluarachchi, E Yu, J Hepel, M Schwer, Brown University Warren Alpert Medical School/Rhode Island Hospital, Providence, RI

Presentations

(Sunday, 7/12/2020)   [Eastern Time (GMT-4)]

Room: AAPM ePoster Library

Purpose: This work presents a new program, adapted from the RayStation automated breast planning (ABP) module, developed to perform institution-specific customized automated treatment planning. The default RayStation ABP module is not compatible with requirements in every clinic, this project rectifies this. Additionally, the study provides a comparison between the newly developed automated breast planning scripted workflow, the raw RayStation module, and conventional plans.

Methods: A customized RayStation automated breast planning scripted workflow was developed to include a physician-trained contour system, an institution-specific normalization and scaling standard, and customized settings and plan specifications. Fourteen breast tangent patients were planned with the new adapted system, the raw RayStation module, and by conventional forward planning. Plans were evaluated based on plan quality, contour quality, and workflow efficiency. For comparison, automated plans were normalized to conventional plan coverage achieved and all other metrics were evaluated.

Results: The new ABP program plans reduced the maximum heart dose by an average of 4.62Gy, and the ipsilateral lung D20 by 1.43Gy. Maximum plan dose, mean heart, heart D20, and heart D10 were comparable in respective plan comparisons. With the current clinical workflow, the conventional plans took 2.46 days total time to flow from CT-simulation to physician approval and 87.6 minutes active time. The new automated program took 18.5 minutes average total time with 8.3 minutes active time. Using the physician hand-drawn contours as the gold standard, the new automated program improved the dice coefficient in the heart contours by 0.73%, the left lung by 3.64%, and the right lung by 1.51% compared with the raw RayStation module’s contouring system.

Conclusion: The newly developed institution-customized RayStation ABP module provided organ at risk dose reduction and dramatic workflow improvements compared with the current clinical workflow. The system developed for this study also improved upon the raw RayStation contouring module.

Keywords

Treatment Planning, Breast

Taxonomy

TH- External Beam- Photons: Treatment planning using machine learning/Knowledge Based Planning/automation

Contact Email