Thermoacoustic range verification in the presence of acoustic heterogeneity and soundspeed errors – Robustness relative to ultrasound image of underlying anatomy

Purpose To demonstrate robustness of thermooacoustic range verification to acoustic heterogeneity and discrepancies between assumed and true propagation speed, i.e., soundspeed errors. Methods A beam sweeper was used to deliver 250 ns pulses that deposited 0.26 Gy of 16 MeV protons and 2.3 Gy of 60 MeV helium ions into water and oil targets, respectively. Thermoacoustic signals were detected by a 96‐channel ultrasound array with a 1–4  MH z sensitivity band (−6  dB ), bandpass filtered and backprojected to create thermoacoustic images in the plane of the ultrasound array. The same soundspeed and transducer array were used to estimate range and generate the ultrasound images onto which Bragg peak locations were overlaid. An air‐gap phantom that displaced the Bragg peak by 6.5 mm demonstrated accuracy. Robustness to soundspeed errors was demonstrated in a waterbath as the assumed propagation speed scanner setting was altered by ± 5 % . Tissue‐mimicking gelatin and a bone sample were introduced to demonstrate robustness to acoustic heterogeneity relative to ultrasound images of the underlying morphology. Results Single ion pulse measurements sufficed during the helium run, but signal averaging was required for protons. Range and entry point into the target were estimated from data collected by transducers placed at least 6 cm distal to the Bragg peak. When ultrasound images depicted the air–target interface where the beam enters, estimates of the entry point agreed with ultrasound images and range estimates agreed with Monte Carlo simulations to within 300 μm, even when thermoacoustic emissions traveled through a strongly scattering bone sample. Estimated Bragg peak locations were translated 6.5 mm by the air‐gap phantom and correctly identified scenarios when the beam stopped inside the bone. Conclusions Soundspeed errors dilate and acoustic heterogeneities deform ultrasound images. When thermoacoustic receivers are co‐located with the ultrasound imaging array, the same transformations shift thermoacoustic range estimates. Therefore, thermoacoustic range verification is robust relative to ultrasound images of underlying anatomy. When the treatment target is visible in ultrasound, e.g., prostate, online thermoacoustic range estimates could verify that the treatment spot is inside the target.