Well, let’s see how they plan to handle oscillating cavitation bubble collapse. Unless that’s accounted for, they’re essentially just sending microbubbles into the brain and hoping for the best. If they’re lucky, the bubble adsorbs the drug, maybe if it’s protein-based or has a slow adsorption curve with an induction period and non-spherical oscillations.

But even then, the local velocity field around the bubble could easily damage soft tissue. I really hope someone runs micro-PIV first to map background flow and bubble wake velocities, because otherwise, it’s just high-resolution trauma delivery.