Room: Exhibit Hall | Forum 4
Purpose: Patients with multiple brain metastases (>6) can be treated with the GammaKnife Icon system and require distribution of treatment over several sessions. This has the advantage of patient comfort with treatment broken up into shorter intervals and the proposed benefit of minimizing the risk of necrosis from treating adjacent metastases simultaneously. We propose a novel spatial-temporal approach for optimal distributed radiosurgery for multiple-brain-metastasis treatment.
Methods: Unlike conventional fractionation, spatial-temporal distributed radiosurgery aims to maximize tumor dose while reducing toxicity to adjacent normal tissues and balances treatment time in each session. We modeled the spatial-temporal solution, where each metastasis was represented by electric charges proportional to its volume multiplied by dose, and intra-session toxicity was represented by electric field potential energy of all metastases within the same session. Then, optimal spatial-temporal distributed radiosurgery can be obtained by minimizing the total potential energy of all sessions. The optimization was formulated as quadratic programming of binary decision variables with linear constraints that each metastasis belongs to only one session and the treatment time of each session is bounded and solved by the OPTI toolbox. All shots for a metastasis will be delivered in the same session.
Results: Both the optimized and manual schedules for clinical plans were of similar quality according to physicianâ€™s blind reviews, but the optimized schedules had lower potential energy and sharper dose gradients on the exterior rings of the targets.
Conclusion: We applied a novel approach to solve the spatial-temporal distributed radiosurgery problem for multi-brain-metastasis treatment. The definition of potential energy may be varied in this framework to simulate different clinical emphases, such as maximizing between-shot distance, absolute dose gradients, or radio-biological effectiveness. This heuristic model is effective and efficient in finding solutions, allowing automated shot scheduling for improved clinical workflow and treatment quality.
Funding Support, Disclosures, and Conflict of Interest: This work is partially supported by NIH R01 CA235723 and the seed grant of Radiation Oncology Department at University of Texas Southwestern Medical Center.