Detection systems for range monitoring in proton therapy: Needs and challenges

Abstract

In-vivo range verification has been a hot topic in particle therapy for about two decades. In spite of vast efforts made by research groups all over the world, clinical devices and procedures for routinely monitoring the range of therapeutic particle beams in the patient’s body and to ensure their correspondence with the treatment plan are not yet available. The paper reviews recent approaches with focus on prompt-gamma based methods of proton range verification and points to challenges that have not been discussed with the necessary depth and rigor in many (even recent) publications: First, the macro time structure of treatment beams in common proton therapy facilities requires detection systems with extreme load tolerance, throughput capability, and stability against load leaps. Second, the time period available for verifying the range of a single pencil beam spot is of the order of milliseconds, which limits the number of prompt gamma events that can be detected and processed. In view of these constraints it might be favorable to waive tight event selection by collimation or coincidence conditions as applied in most prompt-gamma based range verification techniques considered so far, and to move on to straight detection with uncollimated detectors combined with a multi-feature analysis deploying all pieces of information comprised in a registered event. Energy deposition, timing, and energy sharing between the involved detector segments in case of Compton-scattering or pair production are parameters bearing information on the beam track that could be extracted in a comprehensive analysis. This would maximize the number of valid events on the expense of ‘information sharpness’, but could eventually increase the total yield of information exploitable for range verification. Some aspects of such a strategy have already been realized with the Prompt Gamma-Ray Timing (PGT) and the Prompt Gamma Peak Integration (PGPI) techniques proposed recently. Data analysis schemes for a more generalized approach have not yet been developed, but the hardware to be used can already be sketched: Prompt gamma rays should be detected with scintillation detector modules consisting of single pixels with individual light readouts and independent electronic channels, similar to those developed for PET-MR. Prompt gamma-ray detection in this context is, however, much more demanding with respect to dynamic range, energy resolution, load acceptance, and stability. The corresponding requirements represent a challenge for the detector physics community.