Prompt-gamma based treatment verification: Multi-institutional benchmark experiment of a prompt-gamma-imaging and a prompt-gamma-spectroscopy prototype system using an anthropomorphic head phantom

Introduction
Prompt-gamma-based range verification has been clinically implemented using either prompt-gamma-imaging (PGI) or prompt-gamma-spectroscopy (PGS). Here, two proton therapy centers, currently investigating the clinical benefit of PGI and PGS, collaborated to systematically compare the two techniques in a set of benchmark experiments under equalized conditions.

Materials&Methods
The same anthropomorphic head phantom (CIRS, USA) was used for treatment planning, beam delivery and PG monitoring in both centers. Two pencil-beam-scanning fields (1GyE) targeting a brain lesion were optimized on a Dual-energy-CT using the same spot positions and energies in both centers, enabling spot-wise comparison. The horizontal short-range field was clinically realistic. The long-range, oblique field served as stress test. Absolute range verification accuracy against a blinded ground truth (SPR map) for both fields as well as the capability to detect relative range shifts, introduced by plastic slabs (2 and 5mm) on half of the short-range field, was assessed.

Results (µ±σ)
The absolute accuracy of PGI and PGS were (-0.5±0.8)mm or (2.4±0.8)mm for the short-range and (2.4±1.9)mm or (1.3±1.5)mm for the long-range field, respectively. Relative range shifts were detected with (2.0±0.9)mm or (4.2±0.8)mm accuracy for PGI, and (1.8±0.5)mm or (4.8±0.4)mm for PGS. Both systems show a performance worthy of clinical application for the detection of range deviations. Future improvements of their sensitivity are ongoing.

Conclusion
For the first time, two independent PG range verification systems utilizing different PG information have been successfully benchmarked under equalized conditions in two proton therapy centers. This marks an important milestone for translational research on proton treatment verification.