2311 - Novel SRS Program Using Universal Inflatable Headrest and Automatic Planning: Retrospective Validation of Patient Immobilization Using Mid-SRS CBCT

S. N. Huh^1,2, J. Y. Park^1,2, and R. Dagan^1,2; ¹Department of Radiation Oncology, University of Florida College of Medicine, Gainesville, FL, ²University of Florida Health Proton Therapy Institute, Jacksonville, FL

Purpose/Objective(s): We established a linear accelerator-based stereotactic radiosurgery (SRS) program using the universal inflatable headrest developed for patient immobilization and automatic planning. The patients motion during treatment was evaluated using mid-SRS cone-beam computed tomography (CBCT) to demonstrate the inflatable headrests clinical utility and accuracy. We implemented automated planning to improve plan integrity and the effectiveness of dose optimization. Materials/

Methods: The universal inflatable headrest consists of a blood pressure cuff, an air pump, and a pressure meter. The cuff with a small adult size (12 cm × 22 cm) was customized to fit an indexed base-of-skull headframe. The headrest was designed to immobilize patients head motions and chin movements using adjustable air pressures (80–100 mmHg) within a thermoplastic mask. The effectiveness of the immobilization system was evaluated in 45 SRS cases using mid-SRS CBCT acquired after half of the treatment beams were delivered. Compared to the initial setup, the translational and rotational variations in the mid-SRS CBCT were used as surrogates of patient motions during treatment. In addition, the automatic SRS planning script was developed to create pseudo-planning structures and beams. The optimal dose objectives and beam weightings were determined using a heuristic algorithm to achieve dose conformity (CI) and gradient index (DGI).
Results: The inflatable headrest showed a patient immobilization error of 0.6 mm (95% confidence interval) and a maximum of 1 mm in Cartesian coordinates from 140 mid-SRS CBCT images. The headrest demonstrated an effective reduction in anterior-posterior movement with air pressure adjustments. The developed immobilization allowed universal application for single and fractionated SRS cases with photons and protons. The automatic planning enabled initial planning and optimization in 3.5 minutes, including 40 seconds for structure creation. The heuristic algorithm successfully created optimized doses to meet CI = 0.8–0.9, DGI = 0.3–0.4, and a maximum target dose of 125–135%.
Conclusion: The inflatable headrest showed statistically significant precision in immobilizing SRS patients with less than 1 mm error in retrospective data analysis. It provided a comfortable and reliable posterior neck and chin position with a noticeable reduction of anterior-posterior motions. It can be clinically utilized for brain cancer treatment that requires high precision for critical organ-sparing and particle therapy. The automatic SRS planning led to standardizing the SRS protocol and effectively finding the optimal solution to achieve dose conformity and sharp gradients to critical organs.