Version 1
: Received: 19 April 2023 / Approved: 20 April 2023 / Online: 20 April 2023 (02:36:28 CEST)
Version 2
: Received: 2 June 2023 / Approved: 5 June 2023 / Online: 5 June 2023 (09:53:36 CEST)
Rossi, M., Belotti, G., Mainardi, L., Baroni, G., & Cerveri, P. (2024). Feasibility of proton dosimetry overriding planning CT with daily CBCT elaborated through generative artificial intelligence tools. Computer Assisted Surgery, 29(1), 2327981.
Rossi, M., Belotti, G., Mainardi, L., Baroni, G., & Cerveri, P. (2024). Feasibility of proton dosimetry overriding planning CT with daily CBCT elaborated through generative artificial intelligence tools. Computer Assisted Surgery, 29(1), 2327981.
Rossi, M., Belotti, G., Mainardi, L., Baroni, G., & Cerveri, P. (2024). Feasibility of proton dosimetry overriding planning CT with daily CBCT elaborated through generative artificial intelligence tools. Computer Assisted Surgery, 29(1), 2327981.
Rossi, M., Belotti, G., Mainardi, L., Baroni, G., & Cerveri, P. (2024). Feasibility of proton dosimetry overriding planning CT with daily CBCT elaborated through generative artificial intelligence tools. Computer Assisted Surgery, 29(1), 2327981.
Abstract
In radiotherapy, cone-beam computed tomography (CBCT) is mainly used for patient positioning and treatment monitoring. CBCT is deemed to be secure for patients, making it suitable for the delivery of fractional doses. However, because of a narrow field of view (FOV), beam hardening and scattered radiation artifacts, loss of anatomical information and variability in pixel intensity occur that prevent the use of raw CBCT images for dose recalculation during treatment. To address this issue, reliable correction techniques are mandatory to remove artifacts and remap pixel intensity into Hounsfield Unit (HU) values. The present study proposes a deep-learning framework to calibrate CBCT images acquired with narrow FOV systems and demonstrate the potential use in proton treatment planning updates. Cycle-consistent GAN processes raw CBCT to reduce scatter and remap HU. Monte Carlo simulation enables a fair comparison between planning CT and calibrated CBCT dosimetry. Tests were performed on a publicly available dataset of 40 patients who received ablative radiation therapy for locally advanced pancreatic cancer. The CBCT calibration led to a difference in proton dosimetry of less than 2%, compared to the planning CT. The potential toxicity effect on the organs at risk (bowel and stomach) decreased from about 50% (uncalibrated CBCT) up the 2% (calibrated CBCT). These results may confirm that generative artificial intelligence brings the use of CBCT images incrementally closer to clinical translation in proton therapy.
Keywords
Deep Learning; image-to-image translation; dosimetry; cycleGAN; CBCT; CT; limited FOV; artifact correction; Hounsfield unit recovery
Subject
Engineering, Bioengineering
Copyright:
This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
