Transcranial Ultrasound Neuromodulation: The Studies

What does it do in experiments?

Searching Google Scholar for results on “transcranial focused ultrasound”, so far I found 19 studies to include in this post, including experiments on humans, rats, mice, and sheep.

The link to the spreadsheet is here.

(I expect to update this post with more studies when I have time.)

I restricted attention to experiments that measured either some kind of functional outcome (behavior, performance on cognitive or motor tests, sensory perceptions, etc), a physical effect on brain cells/tissue, or a direct measure of the effect of ultrasound on neuronal activity.

I excluded studies that only measured things like “ultrasound altered functional connectivity in fMRI” because it’s hard to evaluate if that matters. From the perspective of evaluating ultrasound as a neuromodulation tool, I want to know if ultrasound can increase or decrease neural firing in the target region, and also if it has any other clinically relevant effects.

What can tFUS do?

Well, to start with, and most importantly, noninvasive ultrasound stimulation can definitely focus on deep brain regions (such as the thalamus) and create measurable effects on brain activity. It does do a thing at all.

It can affect neuronal activity at baseline or in response to sensory or motor stimulation, making neurons either more or less susceptible to firing.

It can affect things like sensory perception (increasing or decreasing sensitivity to pain or touch, inducing sensory perceptions like bright lights or tingling). It can improve reaction time. Stimulating the motor cortex with ultrasound can induce muscle twitches, just as you get from electrode stimulation there.

It can block epilepsy in a mouse model (as you’d expect — if it can block neuronal firing at all it ought to be able to halt epileptic seizures), it locally increases blood-brain-barrier permeability, and might do some interesting things to glial cells.

And it changes EEG patterns and connectivity patterns and such, but we don’t really know what that means.

And, finally, there are some “actually useful” but IMO less credible results, like that it improves mood or (marginally) improves cognitive function in Alzheimer’s patients.

When is tFUS Excitatory vs Inhibitory?

Let’s define “excitatory” tFUS as one that increases neural firing (in amplitude or frequency) in the targeted region, and “inhibitory” tFUS as one that decreases neural firing in that region.

One study (Yoon et al, 2019) in the sheep thalamus directly compared the effect of different ultrasound pulse patterns, and found that “excitatory” modes had high pulse repetition frequencies (0.6-1.4 kHz) and high duty cycles (30%-100%) while “inhibitory” modes had much lower duty cycles (3-5%) and lower pulse repetition frequencies (0.03-0.1 kHz).

This is consistent with a human study (Yu et al 2020) finding stronger excitatory effects on the motor cortex from higher-duty-cycle ultrasound modes, and another human study (Kim et al 2022) finding that 70% duty-cycle stimulation was excitatory while 5% duty-cycle stimulation was inhibitory

It’s also a consistent pattern that ultrasound directed to the visual cortex is excitatory (and creates an effect similar to exposure to a flash of light; in fact humans report seeing lights and patterns when their visual cortex is stimulated.)

On the other hand, there’s some inconsistent results in different experiments targeting the human somatosensory cortex — in (Liu et al, 2021) a 6% duty cycle is excitatory while in (Wonhye et al, 2015) a 50% duty cycle is inhibitory.

I think we don’t totally understand how this works though and we’ll need a more systematic search through brain regions and ultrasound parameters before we understand when and where it does what.

What Do We Do Next?

I’ll touch on this in later posts, but there’s a few heuristics for where to look for “good” tFUS applications.

In brain regions where we already know that lesions, implanted electrodes, or noninvasive techniques like rTMS have a “useful” effect (clinically effective against a disease or cognitive/psychological improvements), it’s plausibly a good idea to try tFUS experiments on those regions and see if they have those effects.

More broadly, one could imagine a systematic program of “functionally mapping” the brain, first in animals, by trying different ultrasound protocols on different regions and running a consistent battery of brain activity measurements and behavioral/sensory/motor tests.