Acoustic time-of-flight for proton range verification in water.

Purpose: Measurement of the arrival times of thermoacoustic waves induced by pulsed proton dose depositions (protoacoustics) may provide a proton range verification method. The goal of this study is to characterize the required dose and protoacoustic proton range (distance) verification accuracy in a homogeneous water medium at a hospital-based clinical cyclotron. Methods: Gaussian-like proton pulses with 17 µs widths and instantaneous currents of 480 nA (5.6×107 protons/pulse, 3.4 cGy/pulse at the Bragg peak) were generated by modulating the cyclotron proton source with a function generator. After energy degradation, the 190 MeV proton pulses irradiated a water phantom, and the generated protoacoustic emissions were measured by a hydrophone. The detector position and proton pulse characteristics were varied. The experimental results were compared to simulations. Different arrival time metrics derived from acoustic waveforms were compared, and the accuracy of protoacoustic time-of-flight distance calculations was assessed. Results: A 27 mPa noise level was observed in the treatment room during irradiation. At 5 cm from the proton beam, an average maximum pressure of 5.2 mPa/1×107 protons (6.1 mGy at the Bragg peak) was measured after irradiation with a proton pulse with 10%-90% rise time of 11 µs. Simulation and experiment arrival times agreed well, and the observed 2.4 µs delay between simulation and experiment is attributed to the difference between the hydrophone's acoustic and geometric centers. Based on protoacoustic arrival times, the beam axis position was measured to within (x,y) = (-2.0, 0.5)±1 mm. After deconvolution of the exciting proton pulse, the protoacoustic compression peak provided the most consistent measure of the distance to the Bragg peak, with an error distribution with mean = -4.5 mm and standard deviation = 2.0 mm. Conclusions: Based on water tank measurements at a clinical hospital-based cyclotron, protoacoustics is a potential method for measuring the beam's position (x and y within 2.0 mm) and Bragg peak range (2.0 mm standard deviation), although range verification will require simulation or experimental calibration to remove systematic error. Based on extrapolation, a protoacoustic arrival time reproducibility of 1.5 µs (2.2 mm) is achievable with 2 Gy of total deposited dose. Of the compared methods, deconvolution of the excitation proton pulse is the best technique for extracting protoacoustic arrival times, particularly if there is variation in the proton pulse shape. [ABSTRACT FROM AUTHOR]