Encrypted login | home

Program Information

A Software Application for the Generation and Delivery of Robust Multiple-Metastasis Stereotactic Radiosurgery Plans

A Yock

A Yock1*, G Kim2 , (1) Vanderbilt University Medical Center, Nashville, TN, (2) University of California, San Diego, La Jolla, CA


MO-RPM-GePD-JT-1 (Monday, July 31, 2017) 3:45 PM - 4:15 PM Room: Joint Imaging-Therapy ePoster Theater

Purpose: We created a software application for depicting and minimizing the deleterious effects of positional variations on multiple-metastasis SRS treatment plans. The application can be used during both treatment planning and delivery.

Methods: Created in Matlab, the application uses data directly from the treatment planning system. It evaluates the dosimetric effects (DVHs and 3D dose distributions) of positional variations by transforming the target using a 6 degree-of-freedom, user-defined motion model and then resampling from the planned dose distribution. For 61 mets from 13 previously treated patients, the application was used to model the change in V100% and D95% versus met radius and distance to isocenter. A second function helped select the number and position of isocenters using k-means clustering to maximize a geometric target coverage metric. We illustrate the improvement in plan robustness from using the application-derived isocenters by comparing DVH statistics with the as-treated plan for an example patient. At treatment delivery, the application can be used to provide real-time feedback regarding the dosimetric effects of observed motion or to establish prospective, plan-specific action limits. The latter is demonstrated by comparing plans treated with either 1mm & 1° or 2mm & 2° action limits for the same example patient.

Results: Models of V100% and D95% for the 61 mets showed a systematic dependence on met radius and distance from isocenter (R² of 0.72 and 0.68, respectively). For the five mets of the example patient, the minimum V100% increased, on average, by 8.3% when the number and positions of isocenters were determined with the application. The minimum V100% decreased, on average, by 11.2% when using 2mm and 2° action limits compared to 1mm and 1° action limits.

Conclusion: Our application incorporates population-, patient-, and plan-specific information for the planning and delivery of multiple-metastasis SRS treatment plans robust to positional variations.

Contact Email: