(Sunday, 7/14/2019) 3:30 PM - 4:00 PM
Room: Exhibit Hall | Forum 7
Purpose: To characterize a real-time interactive planning(RTIP) tool which could help physicians to evaluate and optimize clinical goals for a pacific patient.
Methods: The RTIP tool uses a proprietary automatic clinical goal optimization(ACGO) algorithm which takes as input only a prioritized list of planning goals. The ACGO algorithm then uses fast GPU based, non-constrained fluence optimization to search for an optimized dose distribution which matches the goal hierarchy. The prototype system utilizes Halcyon linac beam model and support multiple OARs and target constraints with various priority settings. To prove the clinical feasibility of using RTIP, ten prostate cancer patients previously treated with 70Gy in 28 fractions are included in this study. For each patient, starting with initial clinical constraints, a set of optimized and practical goals were generated through RTIP by adjustment of priorities or direct manipulation of DVH curves. The optimized clinical goals were then used to guide a manual planning process in Eclipse with the Halcyon linac.
Results: It took ~3min to achieve a set of optimal clinical goals after manipulating priority and the DVH graphically in real-time. With the final goals derived from RTIP, plans were generated by dosimetrist within an average of 10.5±5.5min. 9/10 plans meet both coverage and max dose constraints of PTV, and there was no difference (p>0.05) for key OAR constraints in mean dose to constraint volumes of penile bulb, rectal V57Gy&V35Gy, bladder V57Gy&V35Gy and femur V45Gy between RTIP goals and values from manual plan, which indicates the RTIP suggested OAR constraints are achievable and practical.
Conclusion: It is feasible for physician to use RTIP tool early in the planning stage to review, evaluate and optimize the clinical goals. The derived practical goals can be used to guide and accelerate the treatment planning process and minimize the back-and-forth consultations between physicians and planners.
Funding Support, Disclosures, and Conflict of Interest: This project receives funding from Varian Medical System
Keywords
Optimization, Dose Volume Histograms
Taxonomy
Not Applicable / None Entered.
Contact Email
