Thread by @LennartVerhagen

The next generation brain stimulation is here! Reversible, non-invasive, deep, focal, long-lasting neuromodulation in primates with ultrasound. Two papers out now in @eLife and @NeuroCellPress. Thread with details and data: 👇Links: doi.org/10.1016/j.neur… elifesciences.org/articles/40541

A forty second ultrasound protocol triggers changes to brain activity, even in deep areas such as the amygdala and ACC, for up to two hours. Work together with @davidefolloni, @jerome_sallet, Matthew Rushworth, Cecile Gallea, Pierre Pouget, Jean-Francois Aubry, and many friends.

In the @NeuroCellPress paper we are aiming deep: targeting amygdala and the anterior cingulate cortex. These regions are out of reach for conventional non-invasive brain stimulation techniques (TMS and TCS), but simulations showed we can reach the amygdala with ultrasound.

After forty seconds of repetitive ultrasound we measured brain activity in three monkeys using functional magnetic resonance imaging (fMRI), for up to two hours. We use connectivity analyses to assess the impact of stimulation. Check out its impact on amygdala connectivity!

These effects are specific. If we target amygdala, only the amygdala is affected. Here is a heat-map of the change to brain activity after amygdala ultrasound. Yup, that red spot is in the amygdala. Strike one!

And if we target anterior cingulate cortex (ACC), only this region is affected, and not much else. Here is a heat-map of the change to brain activity after ACC ultrasound. The modulation matches the simulated ultrasound trajectory perfectly. Strike two!

Only at this point did I really started believing in these data. I must admit, I have been worried 💩-less that we measuring noise, artefacts, or whatever else. But this is just so strong, and so specific. This is real.

In the @eLife paper we go a lot further, stimulating two new frontal areas and checking every possible confound. For example, here is a simulation of the temperature changes. Its mostly the skull that heats up, the temperature rise in the brain is limited: no thermomodulation.

In this study we stimulated the supplementary motor area (SMA) and the frontal pole (FPC). The connectivity profile of SMA is shown in a radial plot. After ultrasound SMA connectivity with many areas is reduced (line is closer to centre, areas listed on circumference).

The impact of ultrasound on SMA connectivity lasts long. Up to two hours! You can also see that after reducing weak input to SMA this region starts to become more coupled with its preferred sensorimotor network. The connectivity profile sharpens.

While the above analyses were hypothesis driven, we also allowed ourselves exploratory analyses. When inspecting the data I got the impression there might be something lurking hidden behind our preprocessed, cleaned data. Something in the murky corners of our confound regressors.

We noticed that in fact ultrasound targeted at the SMA and FPC not only modulated signal in grey matter in a site-specific manner, but also modulated signal in the meningeal compartment. It lead to widespread meningeal signal coupling! Huh, what?!

Again, I have to be honest with you, I have no clue what drives this massive widespread signal change. But it is very striking after SMA and FPC stimulation. It is also strikingly absent after amygdala or ACC stimulation? What is different between these?

When we target the superficial SMA and FPC, but not the deep amygdala or ACC, we also deposit a lot of acoustic energy in the meninges. In fact, our ultrasound beam hits the sagittal sinus and other vascular structures on the midline. Have induced vasodilation, enhancing BOLD?

Increased blood flow might lead to the stronger coupling of meningeal signal. Whatever it is, this effect is strong, and persistent. See here the changes over time.

Previous work revealed that in rodents modulation of brain activity can be observed when the ultrasound protocol is modulated at a frequency within the hearing range of the animals. This can possibly account for a lot of previously reported effects. Can it explain our effects?

Nah, probably not. We did not modulate the ultrasound within the hearing range of the monkeys and measured brain activity many many minutes after the stimulation. But better safe than sorry, so let's check.

Nope, the connectivity of primary auditory cortex is not modulated by ultrasound at al. Neither when targeting SMA, nor when targeting FPC.

Now, real important... Is this safe? We did use a somewhat high intensity to maximise our chances of seeing an effect. Did this cause tissue damage?

Well, no sign of oedema on the T1w scans... No damage here.

And no sign of any microstructural damage on the histological slices either!

And then now for the million dollar question.... Can we do this in humans?

Yup, you can. But please be cautious! Ultrasound is already being used in humans, safely. But those protocols are short single-burst protocols that are well tested. We used a novel repetitive protocol that deposits quite a bit of energy in those 40 seconds.

To date all reported ultrasound neuromodulation effects were direct, online, during the stimulation itself. To make a real difference in the clinic, to help people recover from brain disorders, we need brain stimulation protocols that have a longer-lasting effect.

That is what we aimed for. And following our repetitive protocol we observed effects lasting up to two hours, a strong drive of brain plasticity.

This repetitive protocol has not yet been tested in humans, and if I would do so, I would probably start by using slightly less energy, or start in a population where the potential benefits outweigh the risks.

There is still an enormous amount of work to be done, but I believe ultrasound has a fantastic future ahead as a non-invasive brain stimulation tool.