Moderator of 1 Session
Presenter of 1 Presentation
SINGLE-SHOT THERMOACOUSTIC MEASUREMENTS DURING FLASH DELIVERY WITH A CLINICAL SYNCHROCYCLOTRON
Abstract
Background and Aims
Often, FLASH therapy will be hypofractionated so developing a dose verification feedback loop is critical. We investigate feasibility of real-time thermoacoustic range verification during FLASH delivery by a Mevion Hyperscan proton therapy system.
Methods
A Mevion Hyperscan S250i in physics mode delivered 15-20pC/pulse to a 6”x6”x3” Lexan block into which two transducers were embedded at 90-degree angles relative to the proton beamline. During each experiment, one transducer was distal to the beam and the other was positioned lateral (Fig.1a-b).
Nuclear emissions were detected by a large (0.5m) plastic scintillator+PMT assembly and compact (3cm) radiation detector. The large assembly provided a measure of beam intensity as a function of time; the compact detector provided a trigger.
Thermoacoustic emissions were detected simultaneously using a 4-channel digital oscilloscope (Fig.1c). Results are not signal averaged.
The beam profile was characterized using Gafchromic film, which provided estimates of σx=29.0mm and σy=27.4mm. Therefore, the “lateral” transducer was in the beam whenever WE range exceeded 7.62cm=3”.
The beamlets in Lexan were modeled using TRIM software to estimate initial pressure. Thermoacoustic emissions were simulated in kWave assuming point receivers. The Grüneisen and soundspeed of Lexan were assumed to be Γ=2 and νs=2.035mm/μs.
Fig 1. left: dose maps, right: measured(color) vs simulated(black).
Results
Thermoacoustic signals were visible on the oscilloscope, despite broadband noise that caused shot-to-shot variations. After bandpass filtering, however, DC levels were consistent and shot-to-shot variations seemed to be primarily in amplitude. Moreover, signal detected by the distal transducer agreed with simulations (Fig. 1d). However, lateral placement within the beam induced strong EMI (Fig.1e).
Conclusions
Thermoacoustic signals can be detected at distal transducer locations on a pulse-by-pulse basis during FLASH radiotherapy. For 10.7 WE range, accuracy was 0.7+/-0.4mm. Experiments should be repeated in clinical mode with small diameter beamlets to determine utility of lateral transducer locations.