Circuit intervention

welcome in this lecture we're going to talk about brain connectivity and how we can use non-invasive brain stimulation as a circuit intervention we're going to walk through this lecture uh using several examples some of them you might have seen from other lectures but we're bringing it all here together to give you an overview of the different ways and how we can use brain stimulation to intervene and modulate with circuits we're going to look at both transcranial magnetic stimulation transcranial electric stimulation and ultrasonic stimulation let's dive in and just one of the examples of where circuit modulation is perhaps the most impactful or most often used this isn't even in research this is in a clinical setting for example when we're treating depression the most common produ are aiming to induce plasticity in deep circuits but that is not available or Out Of Reach for transc Cranium magnetic stimulation the most common tool used here and therefore people actually or clinicians use repetitive stimulation over cortical targets the dorsal lateral prefrontal cortex this is already an example of circuit modulation you're hoping to have deep effects but you're using using a cortical entry point that is strongly connected now let's take a look um here we're uh investigating or looking at details of an accelerated protocol this is the same protocol where you have very accurate and individualized targeting based on brain connectivity so a patient under goes a resting state functional MRI scan you look at the connectivity patterns from your deep cortical Target and where this is most strongly expressed on the cortex and in fact if you would use this as targeting for your TMS you have a circuit based targeting and if you up your doze with accelerated and as many interventions as you can people have observed quite strong clinical effects that are also lasting longer importantly the better your targeting based on the circuit in fact if we can quantify that as is how close you are to the strongest connection as measured with resting state fmri the stronger the clinical effects so here on the y-axis is the distance or sorry on the x-axis is the distance and on the yis we're looking at the clinical effect and a reduction here means a reduction of depression symptoms and two different um questionnaires either on the left and on your right you can see that there's a relation ship with your circuit targeting the more precise the stronger the clinical effect this means that for many treatments for example ha of depression we're using a circuit targeting based approach and the circuit definition can be on a whole array of information either the resting state FM as we've discussed before but also your connectivity or specific individualized measures of distance and many of this is brought together in a complete model that could also include an electrophysiological and an electrophysics model of the stimulation this is an example of circuit modulation in practice as it's used for clinical treatment in depression but it shouldn't um surprise us I I know it sounds fancy and and state-of-the-art but the very first example of TMS is already circuit modulation you're stimulating on the brain the signal transfers multiple synapses until it finally ends up in your muscle that is a circuit the effects of brain stimulation do not stay local they always travel through the connected circuit and this is often used um in cognitive Neuroscience so we have focal stimulation but our aim can be to perturb a whole circuit that is connected with our Target and we in cognitive Neuroscience often use a behavioral readout uh we're looking at TMs in this case as a causal interference tool to measure Behavior behavioral perturbation one of those examples we're interested in how we um look and control our actions this is sensory motor transformation and there has been a longstanding view like a textbook view that we still see back that there are two major pathways through Visual processing one here in red which is mostly related to where things are and how we control our actions and one related in blue along the temporal lob which called the ventral stream and that's mostly related to what we see to perception and both these streams the dors on red and the vento and blue have been associated with different processing speeds the Assumption or even the proposal has always been that the fastest processing happens in the dorsal stream in Reds and much slower uh identification of perception here in blue that has let people to suggest even assume for true that the doors of parietal processing should precede the slower blue temporal processing however uh uh based on these assumptions we have proposed or tested actually the reverse even slower processes can serve as a prior there you identify a Target look at all the characteristics that help you deliver prior for your actions to probe this as a circuit effect we've been uh uh developing a task where people have to rotate a bar that where we can look at the accuracy and the orientation where people might grasp it or mgrasp it in different directions and we are using multiple targets with TMS either on the dorsal stream or on the vental stream in a control as you can see in the middle if we use TMs on the prial tet in the dorsal stream we see that people their behavior is perturbed especially when the TS is delivered late but critically if TS is delivered on the ventral stream we see a behavioral perturbation only when we delivered a TS early so here you can see TS being used to probe circuits to disentangle their temporal contributions and the strength of TMS as an intervention tool is here that we have a causal inference that allows us very strong interpretations what actually the causal rle between these two different circuits should be flipped now another example of these are not into cortical circuits but where we're having a cortical Target to modulate subcortical or deeper regions and the primary example we've used throughout these lectures could be in controlling your emotional Behavior where we targeting the frontal pole and we also see effects deep down in the brain on the Amala but uh let's take another example maybe one where we have a more gold standed or clear reduct and here we're going to go back to the two to cortical circuits and what we are interested in is the interactions between different brain regions within a circuit we would expect that these interactions are time dependent uh two different brain regions they need to talk to each other at exactly the right time to share information and the effect should be dependent on the state whether you're executing one task but not another if we're building models and this is just my bad drawing of a circuit in a primary motor cortex we see that most of the influence from another brain region is projecting here on an inhibitory inurance and this could be a connected regions like a vental premotor region these are can be two Targets for TMS and you can imagine that you can stimulate and interfere on both of them and the timing by which you stimulate them has an impact on how they interact first let's deliver two pulses over the primary motor cortex if we do this with only three milliseconds in between in other lectures we're explaining that this might have an inhibitory effect so single pulse TMS has a motorok potential of a certain amplitude and if this is conditioned by a TS pulse before it's we see a reduced motor potential but when we move to two different brain regions we can suddenly see that if we're for example doing this during grasping and that's what involves both the motor cortex and the vental premotor cortex that the two pulses now suddenly have an excitatory effect you can see this at the bottom line so delivering a pulse as a test pulse over the motor cortex and you're conditioning this from another site we have a stronger result because the input from your vental premotor region has an excitatory projection on the primary motor cortex this effect is highly dependent on the state and on the timing there is a lot to unpack in this figure but let's start at the Top If we have on the x-axis the timing between the two pulses and at one line we're looking at two pulses delivered over the motor cortex we can see an inhibitory effect for short timings like 3 milliseconds but if we stimulate two different brain regions the timing suddenly takes a little bit longer it's about six or 8 milliseconds we still see this inhibition but critically that's at rest the summary we've seen before is what it happens if you're activating this circuit namely when you're grasping it then the projection that used to be inhibitory now suddenly flips and becomes an excitatory projection still with the same timing so here you see that we can use TMs to with an excruciatingly impressive detail tease apart millisecond Precision on how different brain regions communicate what the timing is on sharing that information 8 milliseconds but not 10 for example and how the effect either an inhibitory projection or an excitatory projection depends on the state whether you're addressed or um in a grasping task this is still using TS to probe the circuit probe the timing and the state dependency can we also use it this knowledge to modulate circuits maybe um we're looking at circuit Dynamics oscillatory activation that we can measure from the brain and different brain regions are coupled in their oscillations this coupling is often dependent on the exact phase and the amplitude of these oscillations they're just not just random they're very deliberately timed together to really maximize their communicative strength we're looking at the same Paradigm so controlling your emotional action tendency right it evolves the prefrontal or the frontal pole um so far we've looked at FM or in this case blood flow measures of activity but of course in the brain blood flow isn't the main thing that is happening we can also record using Meg looking at oscillatory activity and during this task we see that the prefrontal pole has a stronger increase in these lower fluctuations about five Hertz unfortunately it's a slightly more difficult to record oscillations in healthy human participants from deep brain structures but there's one unanswered um aspect in this triangle right we've highlighted the frontal pole we've discussed the uh Amala for automatic action Tendencies but how about implementing your controlled actions this should involve the sensory motor cortex and indeed here we also see oscillatory patterns but now in a higher frequency band so more in a gamma range and these are patent specifically coupled in Phase to the lower frequencies from the frontal pole this is how two different brain regions communicate within a circuit one slower fluctuations and trains higher frequency bursts can we use this information to modulate the Circuit we can go back to protocol that we've discussed before right where we with transcranial electric stimulation are imposing such patterns of oscillation with higher frequencies captured in the peaks of lower frequency um uh waves but this example was at the same site right where the high frequency and the low frequencies are one train over sensor motor cortex but we were looking at two different brain regions that are still coupled in a similar fashion so how about we place two sets of electrodes one on the frontal pole and the other on the sensory motor cortex and at one we induce low frequencies the Theta frequency and at the other we induce higher frequencies in the gamma range but critically we couple them because we saw that during relevant behavior these were tightly coupled in Phase where the high frequency bursts were coupled in the peaks of the lower frequency waves this is something that we can entrain with transcranial electric stimulation we can impose it on the brain this we refer to as inphase oscillatory coupling there's a very crisp and clean control Condition it's the outer phase where you experience exactly the same thing but we tease it apart in time and we completely oppose the relationship let's look at these effects if we measure with fmri and we measure with behavior we see that there is a very specific State dependent and intervention dependent effect let's first look at the behavior when we deliver inphase oscillatory stimulation over these two brain regions we are expecting to enhance the communication between them and in fact we also observe an enhanced behavioral performance people now suddenly make less errors in controlling their automatic Tendencies this is a a boosting of behavior a behavioral Improvement the effect is high dependent on the brain state of the individuals we see the strongest effects also for the individuals that have the strongest modulation as measured with bold FM now let's wrap this up to a last example where we move from cortical circuit interventions to whole brain circuit interventions and now we will have to step away from electromagnetics and we're going to ultrasonic stimulation the example we've given repeatedly is is here in non-u primates we're focusing on a deep cortical unit the anterior singlet cortex for example or a deep sub corticol collection of nuclei the amydala and we're using a repetitive ultrasound protocol stimulating at 10 Herz for 40 seconds that is expected to have a longer lasting effect a delayed effect inducing early phase plasticity we can record the impact on neuronal um coupling using resting state FM and we saw the strongest modulation often at the site of the focus either at the singular cortex or when stimulating on the amydala right there the way that we've been measuring this um just to unpack it here for the amydala is often in a very simple way we're looking at the coupling patterns of natural spontaneous fluctuations in bold FM if we place a seed in the Amala we're looking at how the Amala signal goes up and down and we're looking at which brain regions co- fluctuate these are colored in hot colors um you can see the temporal pole the orbital frontal cortex all well-known connections of the Amala but it's this coupling pattern that is disrupted after ultrasound repetitive ultrasound is delivered on the Amala right it's actually these kind of analysis that we can repeat for every point in the brain and that allows us to create such heat maps to see where we observe the biggest modulation but let's do this again for another circuit now we're not focusing on the Amala we're interested in actions and decision on when to act and this involves again the anterior singular cortex but also the basil forbrain we now looking at the septo region let's look at it from the side it's the same uh bold uh blob here it becomes a target for ultrasound uh we're simulating where we deposit the acoustic energy we're overshoot shooting our Target so after 40 seconds repetitive stimulation here we move the transducer to the other side and repeat it you're now looking at the overlap that is more spatially constrained and we're going to record using resting state FM we again see a focal modulation right there where we were stimulating we see that we have modulated the coupling strength of this septo Bas of for brain region but there's an important aspect this brain region as any other brain region doesn't act alone it's part of a circuit but the circuit here is special namely the Bas of forbrain often has projections as to Coline projection throughout the rest of the brain this is critical if you're deciding when to go from rest into acting you need to suddenly activate your whole brain to get into an active State and that relies on these widespread projections from this single small brain region so let's look at those projections at the cortex of the brain um we're going to use the same color coding if there is no modulation in the circuit at all everything should be black but instead we observe that the perturbation in the basil forbrain leads to a massive and widespread uncoupling decoupling of nearly the rest of the brain with a few important exceptions that actually match the strongest projection in this task here you can see focal intervention with a circuit effect indeed just like any brain stimulation technique the intervention doesn't stay local but travels through the anatomical circuit and this is often used now in non-human primate models but also in human research where we're using many different tasks to investigate the function and the co computational contributions of different brain regions in cognition and behavior in this lecture we've uh investigated or we've looked deeper into using non-invasive brain stimulation as a tool for circuit intervention we've highlighted how any intervention never stays local but often has Network effects spreading through its circuits we've given examples of how we can use it either in the clinic for circuit targeting in depression treatments in cognitive Neuroscience to make inferences how different circuits are related to each other we've given examples of very de detailed timing and state dependency in the motor system where we can learn a lot about the neurophysiology and an incredible Precision but most commonly it's used in cognitive Neuroscience for task interference however if we know enough about our neurophysiology we've given you one example where we're using circuit interventions to enhance Behavior because we're imposing or even mimicking um endogenous brain oscillation using transcranial electric stimulation and lastly we've given you a glimpse of what might be coming next using ultrasound where we can reach any brain Target both cortical and deep cortical and where we're expanding from focal to Circuit interventions as a future in cognitive neuroscience and clinical applications thank you very much