All right, our next speaker will be Dr. Ellen Ayer, and she'll be talking about modeling brain anatomy and function to simulate and guide surgery. Great. Thank you so much for having me. These are my disclosures. So how do we model brain function? You know, that's kind of a big topic. So I started to think about this as we look at the brain functioning on several different levels. We can think about it at the level of individual nuclei. We can think about it as fiber tracts and what various pathways do. And also at the level, more macro level of networks and the connectome. And as our understanding of the structure-function relationships grow, we can use this information to develop additional tools to guide surgery. These concepts do apply widely, I think, in terms of applying to things like epilepsy surgery and tumor surgery as well. But I am going to focus here on movement disorders just because they have been a subset of diseases where we have thought about them in a very circuited structure and that we are interacting with them as such. I usually have kind of viewed us as being the little guys inside the computer circuit trying to figure out which chip is wrong and how do we fix that chip, right? But in that model, and particularly for many years now in movement disorder surgery, we've been, it has been built on this older paradigm of looking at it from the direct and indirect pathways where we have an abnormal node in the system and now we need to block another node in order to kind of counterbalance that. And in that view, it's been so important that the key to success has been location, location, location, right? It's been all about getting to that particular node in the system where we can affect a change. And we've really bought into this in a lot of ways in the sense that, you know, we've gotten a lot of confidence in our ability to locate a structure to the point where many of us have moved entirely away from functional or microelectrode mapping to guide our location for surgery and use systems such as interoperative CT or interoperative MRI, basing our targets purely on imaging. However, you know, what are we really targeting in this paradigm? I mean, we may be specifically trying to get an electrode into the STN or into the GPI or the thalamus, but there are a lot of passing fibers in this area that we are also likely impacting. And we have to keep in mind that the cortex and the cerebellum both play a significant role in this as do the connections between them. So maybe as we think about it, we need to come up with a new roadmap and move past things like indirect targeting, microelectrode recording, and our most recent trend of direct targeting and think about things in terms of using functional imaging and computer modeling to help guide the surgery. So what are some options for that? Well, one is diffusion-weighted imaging and tractography. And the concept behind that is that water diffuses and can diffuse in completely Brownian motion, completely random motion. But in the brain, it tends not to because it's much easier to travel along inside a myelinated structure than to try and cross it. And so using that and computer algorithm, we can identify what's called the diffusion tensor and the directionality of flow. And we can even do additional tractography in order to basically determine which fiber tracts are likely moving between structures. So how can we use that in our approach to surgery? Well, it can be just simply by giving us a better idea of what objects and tracts are around us as we're trying to target and making sure that in a 3D sense, we're not just looking at the nucleus, but we're looking at adjacent fibers and structures to get into an optimal position. But what about actually potentially targeting a fiber tract? So in this example, in essential tremor, we can look at the dentorubral thalamic tracts and do a, actually do tractography to identify that tract by picking out regions of interest up in the brain and thalamus and the red nucleus in this case and in the cerebellum so that we can identify the tract that as it moves through there. And we can actually, it's actually been shown people have been able to use this in treating essential tremor. So in this example from Conan and all, they had a patient where they implanted bilateral thalamic nuclei. And on one side, they got pretty good control. But on the contralateral side, the patient would only get good tremor control if it was turned up enough that they ended up with gait instability. So looked back at the tractography and DWI for that patient from prior to surgery and saw in this kind of yellow and gold area, pretty significant variability between the two sides in this patient's brain. And in that situation, they added a second electrode on the side where the patient was not getting such good effect. And then were able to further model this and show with the medial lemniscus being in green here and the dentorubral cerebellar tract here in the sort of gold color, how they could use both and end up using both contacts or excuse me, contacts from both leads in order to focus the stimulation in the needed tract and avoid stimulation in the medial lemniscus. So I think tractography and the extensions of that are going to continue to be useful as we look at now, how are we actually stimulating neuronal populations? How does that electrical stimulation spread and what are we actually activating? But what about, you know, cortical networks? There have been a lot of work and this is from my mentor Phil Starr's work looking at what happens at the cortical level with stimulation in the basal ganglia. But also there's been work out there by the Stanford group looking at optogenetics in rodents where actual activation of the STN, those neurons themselves was not the key to ameliorating the Parkinsonian features of the animal. It was looking at activation of passing fibers or fibers in the STN and cortical fibers. So what about looking at this from a broader perspective from being able to say, okay, it's not just this one node or this one fiber, but we need to think about it from the perspective of networks as a whole. And so I just want to share a little bit of some preliminary exploratory work that has been done by my colleagues and Dr. Schwab's colleagues at Henry Ford using magnetoencephalography. So the concept of magnetoencephalography is that it identifies brain networks by being able to pick up the oscillatory activity which when synchronous in a population of neurons is able to be picked up by a magnetic field or creates a magnetic field that is able to be picked up. And coherence identifies and quantifies the strength and frequency of that oscillating neuronal activity. And so we're able to basically pick out very strongly related areas in the brain and certainly looking at the cortex specifically. So there's some advantage to MEG. It is a direct measure of neuronal function. It does have excellent temporal resolution measuring that connectivity at the time of the neuronal activity and has pretty good spatial resolution as well. Unfortunately, it's limited in how available this technology is, we're lucky enough to have it at Henry Ford. But it is susceptible to interference from external sources. So in some work, again, by our colleagues, they looked at patients with cranioservical dystonia compared to patients who were normal and areas of coherence in the various cortical structures. The patients that are in these first three rows here are all the dystonia patients. So what we actually see is that the relative to controls, dystonia patients, this is like no coherence, right, all the coherence relatively speaking is in the patients with, who are normal. So what we see then is an increase in the frontal striatal, occipital striatal, parietal striatal, and temporal striatal areas in control patients as, or in controls as compared to patients with cranioservical dystonia. Now, subset of these patients were then examined after peak dosing of Botox, so after a few weeks after their last dose, and we actually see that with the injection of Botox, and this is the pre-post comparison, there is an increase in coherence in various areas of the brain. And interestingly, in one patient, they looked at just what happens with the sensory trick. So this particular patient's sensory trick was a yawn, and there was noted increased coherence in the occipital and temporal structures when the patient performed the sensory trick as compared to baseline. So certainly, this is an interesting methodology that, you know, may provide us some insights and potentially point us to some of the directions or where the aberrant mechanism is in disease states such as dystonia, and, you know, with the data from the botulinum, post-botulinum injections, you know, is botulinum having some central effect here? We don't know. But certainly, I think that we need to step back and look at advanced imaging techniques as being new tools and ever-growing tools to help us both understand the brain networks and also as ways to lead to potentially new targets as well as new approaches, including things like closed-loop stimulation. So with that, I just want to thank my colleagues at Henry Ford, in particular, Abhi Mahajan, who provided, was kind enough to share with me a number of his slides, as well as Dr. Jennifer Sweet, who I'm sure is going to fly through her boards this weekend, who also shared with me a bunch of her slides. So thank you very much. Thank you. Thanks very much, Ellen. Any questions for Dr. Ayer? I mean, so something that I guess we all, you know, the incorporation of these advanced imaging techniques like the tractography, et cetera, are things that we probably, you know, all DBS practitioners are thinking about or doing to some degree. So I wonder, you know, what your experience is with, you know, how much of an expert do you have to become, right, in understanding how this is done? Because, you know, you can almost get whatever you want, depending upon how it's done, right? Or how much do you use the built-in packages that our navigation systems have? You know, like what level of proficiency do we need to have as, I don't know, you know, imagers to do this properly? You're absolutely right on that. Most of these techniques are very algorithm-dependent and driven. And so I do think that any of the information that we get from those needs to be placed in the context of what we do know from anatomical studies and which circuits are involved. Because you are going to potentially pick up a whole lot of tracts that have absolutely nothing to do with what you're trying to accomplish in a surgery. And also recognizing we need to, that we need more sort of validated methods for some of these algorithms.

modeling brain anatomy

surgical procedures

brain function

functional imaging

fiber tracts