Room: Track 2
Purpose: To experimentally achieve prescribed image-noise distributions using a pre-clinical proton CT (pCT) scanner by employing optimized, dynamically-modulated proton fluence fields. The resulting inhomogeneous image-noise allows obtaining relative stopping power (RSP) maps required for particle therapy treatment planning at a reduced dose while maintaining image quality inside a region-of-interest (ROI).
Methods: We obtained fluence-fields to achieve a given prescribed image-noise by employing an optimization algorithm for fluence-modulated proton CT (FMpCT). Because the algorithm relies on a Monte Carlo simulation of the pCT scanner, an experimental validation is indispensable. Fluence is modulated laterally for every projection angle to achieve low noise inside an ROI and low imaging dose elsewhere. Optimized fluence-fields were decomposed into contributions from a grid of pencil beams and eventually to machine instructions for the accelerator, which needed to be executed synchronously with the pCT scanner acquisition. Image-noise maps were reconstructed using a variant of filtered backprojection and were compared to their respective prescription and to the planned noise distribution from a simulation study.
Results: The noise distribution of experimental FMpCT scans of two phantoms using an ROI conforming to a typical treatment beam configuration agreed well with the prescribed noise. In particular, the image-noise target inside the ROI was achieved with mean value deviations of 19% or lower, consistent with the simulation study, where deviations of up to 20% were observed. These deviations are expected and do not impair noise prescription if predictable. A slight misalignment of the delivered fluence patterns and the pCT scanner caused minor fluctuations in the image-noise maps, which can be avoided in future FMpCT acquisitions.
Conclusion: We have, for the first time, experimentally demonstrated the feasibility of dynamic fluence modulation for pCT using an optimization algorithm. This opens an interesting perspective for low-dose image-guidance in particle therapy using proton CT.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by the German Research Foundation (DFG) project #388731804 and the DFG's Cluster of Excellence Munich-Centre for Advanced Photonics (MAP), by the Bavaria-California Technology Center (BaCaTeC) and by the Franco-Bavarian university cooperation center (Bay-France).