Author: Niels Bassler, Jonathan Bortfeldt, Carlos Granja, Guyue Hu, Margarita Kozak, Julie Lascaud, Kirsten Lauber, Grigory Liubchenko, Jasper Nijkamp, Cristina Oancea, Prasannakumar Palaniappan, Katia Parodi, Marco Pinto, Per R. Poulsen, Marco Riboldi, Brita Singers Sørensen, Matthias Würl 👨🔬
Affiliation: Department of Nuclear Medicine, LMU University Hospital, LMU Munich, Department of Radiation Oncology, LMU University Hospital, LMU Munich, Department of Medical Physics, Ludwig-Maximilians-Universität München (LMU Munich), Danish Centre for Particle Therapy, Aarhus University Hospital, Department of Medical Physics, Ludwig-Maximilians-Universität (LMU) München, Department of Experimental Clinical Oncology, Aarhus University, Department of Oncology, Aarhus University Hospital, Advacam 🌍
Purpose: To apply in-vivo proton radiography (pRad) to aid image registration and treatment plan refinement for a novel precision image-guided small animal proton irradiation platform.
Methods: pRads were acquired on a novel small animal irradiation platform using a scattered low-intensity clinical proton beam. The proton imaging setup consists of two silicon TimePix3 detector units, which provide spatially resolved detection of individual energy depositions. Water-equivalent thickness (WET) calibration of the imaging system was performed using PMMA slabs with validated WET. This enabled converting the registered energy deposition into WET. A calibration phantom (SMART phantom) containing tissue-equivalent inserts of well characterized relative stopping power (RSP) values was imaged. Median WET values were calculated over a circular (∅ 2.5mm) region-of-interest within each insert to assess the WET accuracy. For nine anesthetized mice, twelve projections in 30-degree increment were acquired before treatment. WET images were reconstructed with 0.1×10.11mm2 pixel size to be used for registration between 2D pRads and 3D micro-CT, and to verify tissue WET values based on CT-RSP calibration.
Results: For each mouse, pRad acquisition and WET reconstruction was done in ≤15 minutes, and used for subsequent 2D/3D co-registration [1] for target position correction and on-site verification of the planning CT-RSP calibration for pre-treatment decision on the plan-of-the-day. A preliminary analysis of the WET estimation accuracy for the SMART phantom yielded a mean absolute percentage error over the inserts of 2.08%, corresponding to 0.43mm. An observed systematic underestimation due to day-to-day variation in detector energy response is currently being addressed.
Conclusion: We performed the first in-vivo experiments with a silicon-based compact proton imaging setup for an image-guided small animal proton irradiation platform. This work has demonstrated the feasibility of the proton imaging approach for tissue WET validation and target position correction.