Transcranial focused ultrasound is a promising neuromodulation tool, but skull-induced aberrations to the ultrasound beam require subject-specific correction. To predict and compensate for aberration, the standard in transcranial focused ultrasound is to perform acoustic simulations based on x-ray computed tomography images of the head. However, computed tomography-based methods face shortcomings due to imaging resolution and uncertainty in acoustic parameter estimates. Relativistic through-transmit (RTT) is a recently proposed imaging-independent method that corrects skull-induced aberrations by comparing through-skull acoustic transmission to free-field signals from two opposing transducer arrays. Although RTT has shown promise, its robustness across subjects and conditions has not been fully characterized. In this work, RTT performance was evaluated in simulation as a function of several parameters in four unique non-human primate subjects using commercial linear phased array transducers operating at 1 and 2 MHz. Simulated results were validated experimentally at 2 MHz. In simulations, RTT restored pressure amplitudes to within 5% of the intended value for three skulls but achieved only 60% of the intended pressure in one, with focal shift ranging 0.4 to 5.5 mm. RTT performance depended on signal-to-noise ratio and physical uncertainties and was markedly impacted by steering. Modifications to RTT are proposed and characterized.
Donovan et al. (Sun,) studied this question.