Program Information
Poor Man's Knowledge-Based Planning for Solitary SRS Brain and SBRT Lung Lesions
O Blasi*, K McCullough , K Lofton , Colorado Assn in Medical Phys (CAMP), Colorado Springs, CO
Presentations
SU-I-GPD-T-562 (Sunday, July 30, 2017) 3:00 PM - 6:00 PM Room: Exhibit Hall
Purpose: Knowledge-based planning can be useful for determining the extent and time to spend on optimization. Our institution has retrospectively analyzed dose gradient index as a function of tumor size for both solitary SRS brain and SBRT lung lesions in order to prospectively optimize dose falloff in our planning process for stereotactic lung and brain solitary lesions.
Methods: 72 Cyberknife patient dicom plans (38 solitary lung and 34 solitary brain lesions) were exported to our MIM Maestro workstation where a workflow was scripted to generate a table of dose and target volume parameters including dose gradient index. The first 40 of these patients were retrospectively analyzed to build up a knowledge database of expected dose gradients index as a function of tumor size. 32 patients were then prospectively planned using the historical data as a guideline. Plans with dose gradient indexes above the modeled function would be replanned attempting to increase the dose falloff while maintaining planning target volume dose constraints.
Results: A linear least squared fit was used to plot dose gradient index vs the diameter of PTV. For lung lesions our historical data yielded the following slope, y-intercept, and R²: -0.332, 6.41, 0.163 respectively. Similarly, for prospective lung plans, our data yielded -0.294, 5.54, 0.591. Likewise, for the brain lesions, historical data yielded -0.712, 7.03, 0.434. For the brain prospective plans, the data yielded -0.569, 5.57, 0.597. In both cases, improvements in dose gradient and consistency was shown when using the modeled function for planning.
Conclusion: Building a simple knowledge based planning database of dose gradient indexes vs planning target volumes in solitary brain and lung lesions improved our dose falloff rate and consistency in planning.
Contact Email: