Author: Marios Myronakis, Kyriaki Theodorou 👨🔬
Affiliation: Medical Physics Department, Medical School, University of Thessaly 🌍
Purpose: The development of an integrated application framework for rapid, seamless, and automated calculation of absorbed organ dose for individual patients undergoing kV CBCT imaging over the course of radiation therapy.
Methods: The application consists of three parts: a) automatic acquisition of patient information from treatment planning system (TPS) data, namely the CT images, and the contoured organs; b) automatic GPU-based rapid Monte Carlo simulation of dose deposition in individual patient phantom, constructed automatically from the CT images; c) calculation of organ dose from simulated distribution and generation of dose report for each patient. The framework was preliminary verified on a test database of 31 anonymized patients treated for prostate cancer and imaged with Varian On-Board Imager (OBI). The average time required to perform all tasks per patient, and the average, and median organ dose values were calculated.
Results: The average time required for the GPU-based Monte Carlo simulation was approximately 55 seconds to achieve a statistical uncertainty <5%. The average time required for processing and storing data was 30 seconds. The average and median dose per fraction was 6.6 and 6.7 mGy for bladder, 3.6 and 3.4 mGy for small bowel, 5.5 and 5.6 mGy for rectum, 8.3 and 8.1 mGy for left femur, and 6.8 and 6.3 mGy for right femur. A report with the average organ dose per fraction was generated for each patient and each contoured organ in the TPS.
Conclusion: This work is an ongoing pre-release verification of an integrated framework for the calculation of organ dose from kV CBCT imaging during radiation therapy treatments. Patient organ doses were automatically calculated without user input. A user-friendly report was generated to assess individual doses from the imaging procedure. Preliminary verification demonstrates that the framework can be robustly integrated within the clinical workflow without constrictions.