Stephen Wilson 0:06 Welcome to Episode Twelve of the Language Neuroscience Podcast. I'm Stephen Wilson. Today my guest is Roy Hamilton, Associate Professor of Neurology at U Penn, director of the Laboratory for Cognition and Neural Stimulation and director of the Penn Brain Science Translation, Innovation, and Modulation center. Roy is a behavioral neurologist and one of the leading researchers investigating the application of neural modulation, or brain stimulation to enhance recovery from aphasia. He uses transcranial magnetic stimulation and transcranial direct current stimulation in clinical trials with patients with post-stroke aphasia, and primary progressive aphasia. He also carries out basic science and cognitive neuroscience studies, not to mention writing numerous papers and serving in leadership roles to support the development of students, scientists and clinicians from underrepresented groups. I really admire his work, which is always grounded in thoughtful and carefully constructed theoretical models of neuroplasticity and language function. Okay, let's get to it. Hey, Roy, how are you? Roy Hamilton 1:06 Great, Stephen, how are you? Stephen Wilson 1:08 I'm good. It's very hot today. It's like really heating up and I had to turn off all of the AC and stuff so that I can record the podcast. I'm just hoping that I don't overheat in my front room here. Roy Hamilton 1:17 Well, I am actually recording from a different location. I'm not in my home. We we are doing this on the first day of my vacation, which is fine. And I am in a estate house in Milford Pennsylvania, the former home of a two-time governor of the state of Pennsylvania, apparently in years gone by. It's quite rustic, and probably pretty echoey for your podcast. I hope it works out okay. But it's also very cool and comfortable. Stephen Wilson 1:54 It looks really stately. Like it's got this very ornate fireplace in the background and everything. Roy Hamilton 2:01 It's lovely during the day and creepy at night. Stephen Wilson 2:07 So I'm sorry to be interrupting the start of your vacation. But thanks for taking the time to talk with me. Roy Hamilton 2:13 Of course. No, it's it's a great way to kick off my vacation. Stephen Wilson 2:15 Is it a long way from Philly? Roy Hamilton 2:17 It's about a three hour drive. Stephen Wilson 2:20 And what kind of activities are you going to be doing? You got your family there, I assume? Roy Hamilton 2:24 I do have my family there is a large grounds that can be explored. It's part of a National Historic Site. So you can get a home on the historic site. And so the historic site is called the Grey Towers National Historic Site. It has a large grounds, as I mentioned. There's a pool and other things that the kids will enjoy. And as far as I can tell in this eight bedroom home, only one television, which is great. Stephen Wilson 2:54 That is good. You know, it's going to be character building. Roy Hamilton 2:57 Yeah, exactly. I've already... Ask my kids, by the end of the week, they will have developed so much character. Stephen Wilson 3:06 So we met over 20 years ago, in 1999, when we were both taking an introductory graduate course on neuroscience. And I remember that we used to study together. And I particularly remember that you taught me basic concepts of neurobiology because I didn't have much background in that. Do you remember that? Roy Hamilton 3:23 I do. I do remember being in those courses with you. Certainly. However, I think that my recollection of events is entirely different than yours. Because from my recollection, you taught me much more than I taught you. Stephen Wilson 3:41 That's completely false. I distinctly remember, like, when you taught me about long term potentiation or LTP, because we were reading this textbook by I think it was Larry Squire. It was all confusing. And they're all these diagrams of like, biochemical processes that I didn't understand. And I remember, I just really remember like, we're at your apartment, and your just like listen, Stephen, this is what you need to understand about LTP. And you explained the entire thing to me in about three minutes. And it was the first and only time that I understood it. Roy Hamilton 4:17 Let's just say if I explained everything that I knew to you about LTP in three minutes that that should tell you something about my mastery of the topic. Stephen Wilson 4:26 No, no. It was a good time. And so I remember like we used to talk about the class and like we used to talk a bit about our lives and stuff and what was going on. But we never really talked much about our research back then. Because you were already doing important work in the lab of Alvaro Pascual-Leone, and our listeners may know that he's one of the pioneers of transcranial magnetic stimulation or TMS. You were working on Braille reading back then. And I don't remember that we talked about it as much as we should have. But can you tell me now like how you got into involved in that work in his lab and that early work that you did? Roy Hamilton 5:03 Sure. So just a little bit of personal background. So I think that unlike at least a few of the other persons who have been on your podcast, I'm a physician by background. I'm a neurologist. And so when I got to Harvard Medical School as a first year medical student it was right around the time that Alvaro Pascual-Leone was actually getting there as an assistant professor having just come from the NIH. And so he was just setting up his lab. And I bumped into him at a talk series, where he was introducing people who are interested in behavioral neurology, to this fantastic technology, which could be used to focally manipulate brain activity in order to explore structure-function relationships, as they pertain to cognition. I just thought that was the most amazing thing ever. I became one of the first people in his lab. And he was already engaged in research having to do with blind individuals, and their ability to engage the visual cortex for non-visual tasks and tactile tasks in general. And I was interested in the idea of inter-modal processing and the recruitment of brain areas for tasks that they may or may not have been natively engaged in. So that's how I kind of got involved. And the first thing that I did was actually some really simple behavioral work, in which we tested the tactile acuity of individuals who were blind, just to see whether or not they in fact had improved tactile acuity in their reading fingers compared to individuals who were sighted. Because at the time, it was quite debated, because it was felt by many that in fact, what was going on was just a greater ability to focus their attention on what was going on with their their reading fingers. And so the first demonstration was just superior tactile skill in those in the reading fingers. And then collaboratively, we did work demonstrating that that tactile processing tha if you were touching the fingers, those individuals are making read Braille, or doing other tactile tasks that it did indeed activate V1 for these individuals. Subsequently, I wasn't the driving force, I'll keep saying collaboratively because it was in Alvaro's lab. But one of the things that was especially fun in that lab was around the time that I was nearing my graduation from medical school there, we were engaged in a study in which we were bringing in persons who were sighted, and then blindfolding them for a prolonged period of time. Days on end, and making do all sorts of tactile tasks and tactile training and putting them in the scanner and demonstrating that, in fact, it is possible for individuals who are sighted to start activating the visual cortex in the context of performing tactile tasks, if you de- afferent that area of the brain so to speak, in terms of not giving them visual input. Stephen Wilson 8:31 Well, so that plasticity is way quicker than you might have expected. Roy Hamilton 8:35 It is quicker than I expect. The way we framed it is that if you're engaging the visual cortex, or if you're engaging some part of the brain that one doesn't typically think of as being involved in some activity. The possibilities, broadly speaking, are that you could be forging new connections and new networks, new pathways, or that you could be unmasking things that already exist. And so the basic inference or the basic intuition from work like that, is that if if it can happen in sighted people in the course of days, well, then chances are you're unmasking pathways that already could be recruited, rather than sort of forging new ones. Stephen Wilson 9:25 All right, yeah. That's a really good point. Roy Hamilton 9:27 Yeah. Which by the way I think it sounds on the surface, as though that body of work would have very little to do with anything I've done subsequently. But I do think a lot of the intuitions kind of remain in place for me the idea that brains can have some degree of reorganization with respect to how they how they operate, including in the face of injury. Stephen Wilson 9:55 Yeah, no, I think actually, the seeds are like so clear and that early work. Not only the year modulation piece but also the neuroplasticity? I mean, both of these have been like major themes of your work ever since. Roy Hamilton 10:05 Yeah, no, that's exactly right. And so I think that when people tend to view the work as disease based, then they don't see that connection. They say, oh, well, you, you worked with blindness. And all of a sudden, you changed over and you started working with individuals with stroke and aphasia. But to me, it seems like it's all a piece of one body of work, and certainly one avenue of interest. Stephen Wilson 10:31 Yeah. So was that when you sort of became like a researcher, in that lab? Or was that your first research experience? Roy Hamilton 10:39 It wasn't my very first research experience. So I had done my undergraduate work at Harvard as well. And I ended up doing my thesis work in a vision science lab there with Ken Nakayama, and ended up after that thesis work being hired on as a research assistant by Nancy Kanwisher for for a year. And so, you know, I did have the opportunity. And that was at a time when I was pondering my future. And the possible pathways that I was entertaining were to go into cognitive neuroscience as a PhD or to go to medical school. And during that year, I think I got a good first taste of, and obviously, during my undergraduate courses, as well, of what that research world would be like. I ultimately decided that I wanted to go into behavioral neurology and sort of research-based behavioral neurology and took the med school path. But you know, I already by that time would say that I had the seeds planted that what I want to do with my career would be more research based than than clinically focused, Stephen Wilson 11:59 Right. And what's your research-clinical division these days? Roy Hamilton 12:03 Well, I spend, I would say, about 60 to 70% of my time running some combination of my laboratory, the Laboratory for Cognition and Neural Stimulation, and more recently, in the last couple of years, I created a center at Penn for neuromodulation called rather and creatively, the Penn brainSTIM center. Although I have to say that in this case, brainSTIM stands for something. So it's the Penn Brain Science, Translation, Innovation and Modulation center. Stephen Wilson 12:47 Okay, how long did it take you to think that up? Roy Hamilton 12:49 Well, it took me a while sitting around with a post-it pad. And, you know, I think I had my eureka moment in the shower or something. Stephen Wilson 12:58 Yeah. That's where they usually take place. It would be embarrassing to kind of report how much time I spent, like thinking of possible lab names and center names and stuff like that. Roy Hamilton 13:11 Yeah, right. Yeah. No, it's a significant portion of my effort, apparently goes to thinking of acronyms. And then about 15% of my time, I would say, goes to my clinical practice, which is largely now patients with neurodegenerative disease. Although I do see some patients with primary progressive aphasias who I've gotten more interested in over time. I still do see some patients with post-stroke aphasia clinically. And then the remainder of my time, I haven't been keeping track of the math, but whatever remains from whatever percentages I just told you, is a devoted to my administrative roles. Stephen Wilson 13:58 I think it probably adds up to like maybe 150 160%, something like that. Roy Hamilton 14:06 As long as it doesn't add up to that on paper, right? Stephen Wilson 14:09 Yeah, I remember being sat down by like some HR guy who was who was explaining, like how we had to like count effort. And it was all about, like, how it had to add to 100. The actual hours was irrelevant. Roy Hamilton 14:20 Right. Stephen Wilson 14:23 Okay, cool. So, like we kind of alluded to, you've basically spent a lot of your career using neuromodulation TMS, transcranial magnetic stimulation, as well as more recently transcranial direct current stimulation or tDCS. I'm not sure to what extent our listeners are going to be familiar with these techniques. So I was wondering if we could sort of chat about those techniques, how they work, and then we'll talk about some other ways that you and your colleagues have applied them to aphasia Can you tell us about the biophysical action of those two stimulation methodologies? Roy Hamilton 14:58 Okay. I think that people mentioned them in the same sentence often, but they're actually fairly different in how they work. So I will take them in turn. So first, let's talk a little bit about transcranial magnetic stimulation. So when you're applying transcranial magnetic stimulation, this technology has been around longer. So my suspicion is that some of your listeners may be more familiar with it. But the way it basically it works is that you have an electromagnet. Alright, and the form factor of that is that for most researchers, it's a sort of figure of eight coil that is attached to, via a cable to something about the size of a VCR from the 80s. Okay, and when you discharge it, you're going to run an enormous amount of current through that coil in very short period of time, which is going to create a fluxing magnetic field, that fluxing magnetic field by virtue of electromagnetic induction, is able to create current in nearby conductive bodies, right. And so if you have it against your head, since neurons work via conductive electrochemical principles, if you place that magnet there and you discharge it, you will generate current and underlying neurons. That generated current is sufficient to depolarize those neurons. Now, that's the basic mechanism of transcranial magnetic stimulation. Most systems operate using a coil, you put a single coil, you put on the surface of the scalp, there are sort of deep stimulation systems. But that's sort of beyond the scope of what I want to talk about right now. But for most systems, the strength of the magnetic field falls off as a square of the distance away from the coil. So what that means is that conveniently enough for most of us in cognitive neuroscience, that since the cortex is on the surface, you're able to focally target cortical neurons. And you know, by the time you get to most deep structures, the strength of magnetic fields falling away. And so that allows for focal manipulation of the brain, if it's done in single episodes, or, you know, single pulses, brief sessions, you can use it fairly readily for basic cognitive neuroscience purposes. If you if you apply it repeatedly to the same site, then you can actually start to induce neuroplastic changes in how particular areas of the of the brain work. And, you know, this may harken back to our old graduate school days of talking about LTP. Because, you know, a lot of what people think about how repeated sessions of of TMS might operate in the brain is that they might harness LTP and LTD long term potentiation and depression like mechanisms in order to facilitate neuroplastic changes in regions of the brain and in brain connections. And that a lot depends on the specifics of the pattern of stimulation, what frequency you deliver it at, over what period of time and how many pulses. Those kinds of details. Those are the kinds of changes that you're hoping to instantiate with multiple sessions of stimulation. Stephen Wilson 18:35 So TMS can ultimately be excitatory, inhibitory, and even drive neuroplastic changes, all depending on the pulse frequency and factors like that. Roy Hamilton 18:46 Yeah, that is exactly the thinking. And that's useful, it's useful to think of it as having these different capacities. Because it can be used briefly, or it can be used as a tool to probe the brain by briefly manipulating the activity of specific areas of the brain, and then seeing what effect those manipulations have on performance on tasks. And it gives you a sort of an approach that nears causality in terms of your inferential strength. You say, okay, I can manipulate this part of the brain. And when that happens, I change that aspect of behavior. And I know that I caused that. So that's useful as an inferential tool in cognitive neuroscience. But then also with repeated use, you can stimulate the brain, you can inhibit the brain to the point where it causes long term changes in brain function. So it has both of those utilities. tDCS, by contrast, in terms of its mechanics and in terms of how it works. It is a sub-threshold stimulation technique. So remember what I told you about TMS. For TMS, you're delivering stimulation at an intensity that has the capacity to immediately at the point of the delivery of the pulse depolarize neurons and cause them to fire right then and there. tDCS involves the application of very small currents, on the order of one to four milliamps. So it's a really small amount of current applied to the brain through the scalp using at least two electrodes. At least two because one of them has to be an anode, one of them has to be a cathode. Although you can have more electrodes as long as some are anodes and some are cathode. And so that current that you're running through the brain is going to cause a very small change in the resting membrane potential of many, many neurons. So just to go through that deserve the I don't know how grounded your listeners will be in the sort of the basics of neuronal function, Stephen Wilson 20:56 There'll be variability. Roy Hamilton 20:57 Okay, all right. So what we're talking about is the amount to which a neuron has to be additionally stimulated in order to fire, right, its action potential threshold. So the thinking is, you're making tiny, tiny changes one way or another, either towards that threshold or away from that threshold in many, many, many neurons. And so that increases the potential. Inthe right circumstances, for example, if you're engaged in some task. And at the same time that you're engaged in that task, you're engaging large portions of the brain, with, with this low level current that makes it more likely for many of the neurons to fire, you can imagine that facilitating the brain activity associated with that task. Stephen Wilson 21:49 Right. Roy Hamilton 21:50 So it's a different kind of approach than supra-threshold stimulation, like TMS. But you can also apply it as a cognitive neuroscience tool, where you apply it in in a single or a few sessions. Either in a facilitative way, or an inhibitory way to try and manipulate performance to give you some insight. Or potentially as a therapeutic tool, again, with repetitive episodes, multiple sessions of stimulation associated with some training typically. Stephen Wilson 22:24 Cool. Yeah, well, that was super clear. And I think that now our listeners will believe me, when I tell my story that you taught me how LTP works in three minutes, because I think you're very good at explaining things and breaking them down to first principles. So how did these techniques differ in terms of their spatial resolution? Roy Hamilton 22:46 Oh, quite a bit. So conventional TMS, the way that TMS is typically done, or by most people in the field... a lot depends on things like coil shape, and whatnot. But I think that most people still use what we refer to as a figure of eight coil. So it's about 70 millimeters in diameter. But what it gives you is an area of focality. That is, you know, let's call it a centimeter, centimeter and a half square on the cortical surface, so imagine that the area of your thumbnail, you look at that, and you sort of think of that as your active area across the surface of the cortex. Compare that to giving tDCS, which involves, if you're doing conventional tDCS, large pad electrodes placed on the scalp, and you're just running current between the electrodes. And the current does not care about your cognitive neuroscience study, it doesn't care about the discrete boundary is of whatever you're trying to study. Stephen Wilson 23:56 It doesn't know the difference between BA44 and BA45, even> Roy Hamilton 24:00 It apparently did not read the textbook. It's just trying to get from from one electrode to the other, or one category of electrode to the other category, right? So through the scalp to the skin, right, like, any which way. And so it's spatial resolution is terrible. It's terrible. So if you do finite element modeling, what you get with the most standard arrangement of electrodes, which is sort of one in the front and one over the side where you want to stimulate, you get this just broad swath, this smear of current running through. Now, there are so called high-definition tDCS systems which use smaller electrodes and will target. They might have electrodes of one polarity at the center and then smaller electrodes, you know, other electrodes surrounding it. And you get better spatial resolution with those techniques. But generally speaking, even then the spatial resolution of those techniques is not as high as TMS, which is can be quite focal. Stephen Wilson 25:10 Yeah. And like this paper that we talked about before we started recording where you guys use high definition tDCS over the angular gyrus. Maybe that's a good example of the kind of spatial resolution you can achieve. Roy Hamilton 25:26 Right. So that was an experiment where we were using this high definition tDCS, because it was important. This is from a few years ago. At that time, Amy Price was a graduate student working with Murray Grossman in the FTD center, but had an interest in using tDCS to elucidate a number of specific principles and concepts. So it was a it was a good collaboration. And she was interested in conceptual combination and the role the angular gyrus in conceptual combination in language. And so there, it was pretty important to specifically stimulate the angular gyrus and not areas outside of angular gyrus for the purposes of making appropriate inferences about what the effects of stimulation were. So we used high density stimulation, and I think achieve pretty pretty focal coverage. Stephen Wilson 26:26 So you had like one electrode place like right over some fMRI-derived hotspot in the angular gyrus. And then you put the opposite polarity electrodes around that and kind of a ring, right? Roy Hamilton 26:37 That's correct. Yes. Stephen Wilson 26:39 So the current like flows right in and then right out, and it never really gets too far out of the angular gyrus. Roy Hamilton 26:44 Yeah, that's the basic idea. We try to we try to trap it in the right spot. Stephen Wilson 26:49 And so you have the enough resolution with HD-tDCS that you can target a structure like that, something the size of a gyrus? Roy Hamilton 26:58 Yes. Although, I still think that it can be difficult. So, for instance, we're currently engaged in a HD-tDCS study in persons with PPA (primary progressive aphasia). And there, we aim to stimulate the inferior frontal gyrus. But when we did the finite element models, it's hard for us to not also hit some of the superior temporal gyrus. And we certainly can't achieve the kind of resolution at least with our montage that would allow us, for example, to disambiguate the pars triangularis from the pars opercularis in our stimulation. So it has limits, that would easily be circumnavigable with TMS. Stephen Wilson 27:49 Right. But it seemed to work in that study by Amy Price, and you guys. You had a pretty compelling effect, as hypothesized. Yeah? Roy Hamilton 27:59 Yeah, we definitely did. It's been a few years. So, you know, you'll forgive for not being able to dust off all the details. But what we were looking at was basically whether or not we could manipulate performance based on the ability of words to combine conceptually. And so she employed pairs of words, some of which were sort of sensible pairings. You know, a "red hat". Stephen Wilson 28:31 "Plaid jacket" was in the abstract. Roy Hamilton 28:34 Exactly. Right. And then other ones, which were not sensible pairing. So, my recollection from her many presentations on the topic; the paradigmatic one she used to like to use was "fast blueberry". Stephen Wilson 28:47 Yeah, that's the one that I took a note of to. So the PLAID JACKET condition and the FAST BLUEBERRY condition. Roy Hamilton 28:54 Right. And so you can demonstrate that you can manipulate performance on this task as a function of where these word pairings lay in terms of their ability to combine sensibly. Alright, and that there's sort of preferential effects in the word pairings that were able to combine. Stephen Wilson 29:17 Yeah. And you guys showed that the tDCS didn't make a difference on accuracy as usually doesn't. But it did make a difference in reaction time and sped up responses to the PLAID JACKET condition. But not to the FAST BLUEBERRY condition. Roy Hamilton 29:33 Yeah. And so that highlights a couple of things. But one of the things that it highlights is that it is often the case in tDCS studies that you're especially, in tDCS studies of healthy individuals in performance, that you're often dealing with reaction times more frequently than you're dealing with very large changes in accuracy. Stephen Wilson 29:57 Right. I think that's even true of TMS to a large extent, right. More often, TMS effects are in terms of reaction time than accuracy. Although you might get a little bit more. Roy Hamilton 30:08 I certainly think that that's true. I mean, it's it's easier with these neuromodulation techniques to elicit changes in reaction time. I think with TMS, you have a range, or you have a range of effects that you can produce. I mean, as you can only imagine, when you have a suprathreshold stimulation technique. I mean, it is possible, for instance, and, if you've never seen the videos of this I think they're still out there. But it is possible to induce speech arrest with with TMS, right. Which is something that you wouldn't imagine possible with tDCS a technology that can't even surpass the threshold of firing for neurons, right? You have a range where you can certainly elicit a variety of different effects with TMS just by nature of its mechanism. Stephen Wilson 31:02 Right. Okay. Yeah, that's true. I mean, speech arrest is kind of the ultimate error, really, isn't it? The error of not being able to produce speech. Cool. So yeah, so that kind of study by Amy Price, and you guys, is an example of like a cognitive neuroscience application of these techniques. Can we talk now about your use of these techniques in facilitating aphasia recovery? So thinking back to how you got into that line of work, like, what was the rationale that you were starting from? Did you kind of have a model of recovery in mind and an idea of how neuromodulation could promote it? Roy Hamilton 31:38 Yes. So first of all, let me start by giving credit where credit is due. Which is to say that I'm originally the way that I got into this work was with the next mentor in my career, who was Branch Coslett, who's a behavioral neurologist at the University of Pennsylvania, who I still work with fairly closely. And a collaboration between that mentor and my first mentor, Alvaro Pascual-Leone. So this is part of a multi site R01 study that actually another person who was importantly influential on that site was Margaret or Marnie Naeser. And the idea was to pilot the use of TMS, in the right hemisphere, in individuals with chronic post stroke aphasia. The motivating rationale at the time, one of which I still think is invoked by a number of people who do neuromodulation in aphasia, was the idea that, when you have a lesion of the left hemisphere, what happens is you release the right hemisphere from inhibition. That basically the two hemispheres, project inhibitory inputs to each other, normally. They keep each other in check. And so if you, if you have one side down, the other side's activity is elevated, but that activity is not productive in nature, according to that theory. And that activity is either spurious or deleterious. Because those areas then could potentially send projections since they're intact, and more engaged, could send projections back to the left hemisphere, to include two areas that are perilesional, or areas that were untouched in the language circuit by the lesion that are trying their best to recover from aphasia. And so, right, according to that theory, you would apply inhibitory stimulation to the intact hemisphere of the brain in order to try and quell that over activity. Stephen Wilson 34:05 And this idea came out of the motor literature originally, right? Roy Hamilton 34:08 Yeah, a lot of it exists in the motor literature. You can demonstrate this using motor physiology, that you can have the two hemispheres have inhibitory connections that you can manipulate the activity one side by by manipulating the activity of the other. So that was the the rationale at the time. And so what we did was we brought individuals in and searched around and tried to find a region that, if we applied inhibitory stimulation for a period of time, that there would be at least a transient improvement in their naming ability. And so once we found that site, we would then stimulate those individuals five sessions a week for two weeks, and then follow them up to see if they had a longstanding changes in their language performance. And so that was the rationale. And so each person, by the way, I think I said this, but I'll emphasize this, we were hunting around for the best site. So we had sort of a site finding phase. And so what we found was a couple of things. The first was that we were for most of these individuals able to induce some improvement over time, particularly in naming. There are other measures as well that seemed to last. A number of these individuals seem to actually improve further over time. So significant improvement at the two month time point, and even further improve in the six month time point. The other interesting thing was that for most of these individuals, almost every individual, the site, that was the best site to stimulate temporarily, which ended up being the site that we stimulated it for two weeks was the pars triangularis on the right. And so the homolog of Broca's area on the right. So that was interesting. So, later on a few years later, Peter Turkeltaub, who's now at Georgetown, came to the lab and and he did an ALE (Activation Likelihood Estimate) meta analysis. Stephen Wilson 36:31 Activation likelihood estimate. Roy Hamilton 36:33 Yes, yes. A technique he created. And what he discovered using all the literature that had previously existed, using functional imaging in patients with chronic aphasia, was that there's a bilateral network of areas. It's something which was already known sort of by that both hemispheres tend to see activation. It's something we saw in, for example, Dorothy Saur's work from the mid aughts and whatnot. But basically that there is this bilateral constellation of areas that get engaged. And then also looked at which areas seemed to be not just homotopes, but functional homologues of areas in the left hemisphere that appear to be engaged in healthy individuals for the same patterns of tasks. And so the reason I mentioned all this is that one interesting finding out of that work was that an area, the area, that seemed to be a homotome, but the least homologous, in its patterns of activation, was pars triangularis on the right. Many of the other homotopic areas, if you looked at when they were active versus when they weren't versus when they were active for tasks, they seem to track quite well with left hemisphere areas. But this one area seemed to be somewhat orthogonal in its patterns of activation relative to to its left hemisphere homotope. So we kind of conceptualize that as well, not necessarily the right hemisphere is bad, but rather that maybe there are areas in this newly engaged network that are less efficient than other areas. And if we can manipulate them, or tweak the efficiency of those areas, that we would improve the overall efficiency of the network, even if even the right hemisphere areas may be beneficial. Stephen Wilson 38:34 So you still think it's important that you inhibited it, but not to stop it from inhibiting the left. But to stop it playing some role that's less language-like and less homotopic. Roy Hamilton 38:44 I will remain at least somewhat agnostic to whether or not it has no inhibitory effect on the left. I don't know. But it does seem to be the case, that modifying it, that site may have something different about it. Or another way to say what I'm trying to say is that I do not believe that the right hemisphere is monolithic in its inhibitory role, just by virtue of the fact that you have a left hemisphere lesion. It's not as if what we've done is found a particularly bad actor in a framework where all this increased activation is bad. So just to give you another example, that kind of solidifies this point. So we actually had a subject who, unfortunately, for her, had a second stroke in the course of her treatment. The second stroke was actually in the right hemisphere, it was subcortical but it was in the right hemisphere. So if you were the sort of hardliner, who believes that the right hemisphere, the entire purpose of it, with respect to aphasia recovery is its activation is deleterious and that any good neuromodulator would just tamp down the right hemisphere. And that's what you need to do. You would argue that well, okay, if you've now severed some of the connections, some of those deleterious connections from the right hemisphere to the left a person's language function... They might lose other functions, because brain lesions are rarely good, but maybe this inhibitory influence to the left hemisphere will be sort of quelled by virtue of this new lesion. But it turns out that doesn't improve language at all. In fact, that patient got selectively worse in their language ability compared to all their other abilities. We were actively testing them. And so yeah, that suggests to me that for the most part, a lot of these areas seem to be doing something useful. It's just that even in this person, for whom most of the right hemisphere seem to be useful, this one site seem to be a kind of inefficient contributor, and that for that person to that downregulating activity, that site seemed to be helpful. Stephen Wilson 41:18 Right. That's interesting. Yeah, I mean, like, there's been quite a few observations that subsequent right hemisphere strokes are not good when you're recovering from aphasia, even dating back to the 1800s. Barlow has this early observation. And then Luria says the same thing. And Basso has observed this, too. So yeah, like you said, you don't wanna just cut off the right hemisphere? Because it's not that simple. Right? Roy Hamilton 41:47 Right. Right. So I think that a number of people who do research in neuromodulation for aphasia recovery, have bought into this idea that the right hemisphere is that the sole strategy with respect to the right hemisphere ought to be to inhibit it because of this model of interhemispheric inhibition. But I think our our lived experience suggests that it's much more complicated than that. And I think a lot of our attention these days, or more of our attention, is moving towards trying to understand what exactly are the properties that cause an area to get engaged in language tasks in the setting of aphasia, and what differentiates a site that's an efficient contributor from one that's not? Stephen Wilson 42:49 So after your first experience in this field, in this study that you've just been telling us about, did you follow that up with other studies that explored different models of recovery? Roy Hamilton 43:04 Well, we explored a few different avenues. So we've sort of gone down a few different paths, since. So a couple of those paths include using tDCS, initially in individuals with post-stroke aphasia. And here again, I think that the work that we did, does inform or did inform our strategic approach because we didn't start out invested in the idea that the right hemisphere needed to be inhibited. We basically chose a montage, we sort of move things around and chose a montage that seemed best for each individual rather than stay overly invested in the idea of the right hemisphere being bad. And actually found in some cases that the appropriate montage for some persons seemed to involve putting the anode, what is traditionally thought of as the facilitating electrode, over the right hemisphere in some of our subjects, right. So we did, we did find that the prior work informed some of the work we did. And then we've also sort of branched out now what we do in the tDCS work, more work in individuals with neurodegenerative aphasias. But the the TMS work, cycling back to that. We are picking that up. And so for the last couple of years, our group and Branch Coslett, who's still an investigator in the Laboratory for Cognition and Neural Stimulation now, is actually the PI of this work. We're doing a clinical trial, a larger clinical trial, using that approach, that TMS approach that we used, originally, in a larger cohort of patients, to reaffirm whether it has efficacy. Stephen Wilson 45:08 And you still using these individualized stimulation sites, or are you honed in on that pars triangularis? Roy Hamilton 45:14 For this particular clinical trial, because the vast majority of subjects in our prior data had pretty much the same site of stimulation that was optimal. We're targeting the right pars triangularis. Stephen Wilson 45:30 We keep coming back to that, I guess. It is an area that is involved in inhibition, even in healthy brains, right. It's like, sort of the canonical inhibitory force of the brain, the right IFG. Roy Hamilton 45:46 Right. And, and that may have something to do with it having a special role, or a role that isn't entirely contributory to language production, in individuals with left hemisphere lesion. Stephen Wilson 46:04 I'm just gonna say something that's really, really stupid, and I'm probably gonna regret later. Maybe it's like, sometimes when you're having trouble writing the essay, and you just need to stay up all night and drink a bottle of wine and write it. Maybe, you know, when you have aphasia your language is not going to come out as as it did before. And it's gonna be very frustrating and difficult. But if you could just let go and do something, it might actually be better than you expected, you know? So it's just a matter of like, you know, releasing inhibitions and the positive results that can sometimes come from releasing inhibitions. Roy Hamilton 46:43 I don't think it's crazy. So much as i think that it's impressive that you would be able to drink a whole bottle of wine and still be able to write that essay better than... I'd go with the glass myself. But anyway. If one were to make a sort of cognitively based argument as to how it could be the case, that the right IFG would have such a consistent role such that inhibiting it seemed to have some benefit. I think that is a reasonable argument to make. As I mentioned before, we started to get more interested in what some of the properties are that actually cause brain areas to get engaged. And if we can, ultimately, if we can make predictions about which brain areas ought to be inhibited versus excited. A lot of the approaches that we've started to take more recently, and this is work again, in the name of giving credit where credit's due. I work a lot with John Medaglia at Drexel, a network neuro-imager and cognitive neuroscientist. And so we started getting interested in how some network features of language areas, and in particular, we've gotten interested in sort of structural network features, how they dictate the roles that particular brain areas have in networks. And whether it's the case that and this sort of goes into the world realm of speculation, but whether or not there's something about the roles of brain areas in the right hemisphere or even perilesional areas that get engaged once brain areas get injured, in order to accomplish some of the tasks that those injured areas had been performing. Is there something about the network roles of those areas that is predictable, such that you'd say, oh, yes, they those have the same kinds of computational affordance, however, you want to frame it in your mind that would suggest that they would be good candidates to start getting engaged if this area ever went down. And conversely, this other area has these these properties, and in the network that would suggest that its role would be if it was acting inefficiently, would be harmful to the overall functionality of the network. That's sort of, if we can answer those questions in the next five years, or the next series of grants, that'd be super exciting for me and my group. Stephen Wilson 49:30 Just as you're impressed by my suggestion that a bottle of wine would be the appropriate amount to facilitate better essay writing, I'm impressed by the idea that you think that you can answer any question within five years. Whenever I think about questions I want to answer it's like, well, maybe I'll know that by the time that I die. Roy Hamilton 49:51 It's not that impressive, because usually the answer to a question in our field is more questions. Stephen Wilson 49:56 Yeah, I think I could do that. I think I can, you know, generate more questions. Roy Hamilton 50:04 So it sounds like you're pretty convinced at this point that neural modulation is effective in aphasia in facilitating aphasia recovery? Well, I certainly think that it is effective. I feel like you're asking the wrong person this question. It's like, is neural modulation effective in aphasia? Or is your entire adult life a waste of time? Well, even framed that way, I'm going to go with the former rather than the latter. But we've demonstrated it, and at this point, certainly, we're not the only people. So we've also in our group, run a number of meta analyses across TMS studies and tDCS studies. So I do think that there is convincing signal there. Now, whether that signal is sufficient to radically alter the sort of clinical course of disease remains to be seen. I think we need to work on stratifying, who is most responsive to what approaches. And, obviously, not all aphasias are the same, right. We need to also figure out how to optimize our methodologies. But I do think that there's signal there. Stephen Wilson 51:29 Right, cool. Yeah. I mean, I was talking to Julius, a couple of months back now, I guess. And, he really emphasized to like the importance of individualized treatments for specific patients, and how much stronger effects we might see in some of these clinical trials, if we were doing the appropriate interventions with the appropriate people? Because, yeah that's one of the things that one of my students, Sarah Schneck, and I recently did a meta-analysis of neuroplasticity after stroke in fMRI and PET studies. And among the many conclusions that we drew, one of them was that any patterns of reorganization are probably pretty individually distinct. Because there's not any super salient patterns of reorganization that happened so consistently, that you see them in group studies. And we probably do need to be thinking that, you know, the nature of reorganization is going to differ very much, depending on the nature of the lesion and the aphasia. Roy Hamilton 52:23 Well, that's exactly right. I mean, it's unreasonable. I'm sorry, I probably offended a bunch of people by saying it's unreasonable. But it strikes me that one shouldn't expect that a condition that is caused by numerous lesion locations, of different sizes, of varying mechanics, like mechanically, what causes them in terms of pathophysiology... that they should all cause the same changes in language system reorganization and activity. It's got to be something more individualized. In order to meet that challenge with neuromodulation, neuromodulation has to move towards being more like precision medicine. Where each person has a set of targets or a specific approach that is germane to their specific constellation of areas that are engaged by language when they attempt to employ it. Stephen Wilson 53:28 Yeah, I was really intrigued when you were talking about that a few minutes ago about how you would kind of identify the optimal site for each patient. I guess, you just kind of try different things while they name and just move the montage around. I guess it was TMS. So you move the coil around, and you just kind of experiment and see what works? Roy Hamilton 53:47 Yeah, I mean, it's kind of brute force. That that doesn't sound like an ideal strategy for the future. So I think that in order to really make progress in this field, individuals who do neuromodulation have to figure out some way to sort out which areas would be the best locations to stimulate ahead of time. What are the underlying principles that determine which areas are likely to be engaged by which individuals in which ways? Stephen Wilson 54:21 Yeah, and that kind of relates to your collaboration with John Medaglia, right? I mean, like, this idea that you could use network theory to make sensible predictions. Even if you don't know, in advance, like where each individual is gonna be best stimulated, that you can make predictions and have like a strategy that was informed. You have a strategy. Roy Hamilton 54:42 Right. Because, you know, hunting around with a coil. I mean, it's a big brain out there and hunting around with a coil for exactly the right spot for each individual. In the long run, that's not going to be the way to go. We need rules. Stephen Wilson 54:59 We lead rules, yeah. Okay, so I don't want to take you too long because, you know, you're the first day of your vacation. But I wanted to talk a bit about some of the leadership stuff that you've been doing lately. I read a paper that you are a co-author on the effect of COVID-19 on the scientific enterprise. It's in Science Translational Medicine. Can you kind of talk about that paper? Like, you know, what do you see the challenges being? Roy Hamilton 55:30 Well, I think that probably listeners to this podcast, who do human subjects research, are very familiar with the challenges that have been posed to researchers during the COVID-19 pandemic. For many of us, it at least temporarily resulted in a complete work stoppage. And in many cases, lack of use or waste of resources, both material and the work of our personnel. So it's very costly from that perspective. And so the paper itself, focused on the disproportional burden that these kinds of events have on groups of individuals who are already disadvantaged in academic settings and academic communities. And so there had been some work, not in this paper, but there had been some work in the literature, on the disproportionate impact of COVID-19 on researchers who were women. This piece focused on the disproportionate impact of COVID-19 on researchers who are from underrepresented racial and ethnic minorities. And, you know, some of the challenges that are faced by individuals in these groups are that, in general, they have a lower, or I should say smaller and less robust scientific colleague networks, and mentorship networks. It's harder for them to achieve solid mentorship and sponsorship in research. There's been evidence, work that's been put out that demonstrates that it is harder for individuals from these groups to achieve success in grant funding. So they're often coming from a less robust funding base, right. And so in that context, they face special challenges, when something comes along, to really dismantle, for an undetermined period of time, their research operations. And so the piece focused on that, on that disproportionate burden. And made some suggestions to institutions as to how they could try to mitigate those those burdens, not in ways that are preferential per se. But in ways that improve everyone's possibility to survive the COVID pandemic as a researcher. But in ways that would benefit individuals who are suffering from that burden the most. Stephen Wilson 58:25 Yeah, so what kind of suggestions did you guys make for how this could be addressed? Roy Hamilton 58:30 Well, some of the things are pretty intuitive, you know. They aren't the kinds of things that, I would say, require groundbreaking innovation. But institutions becoming more invested in bridging and pilot funds that can support individuals, if they're going through periods, especially if it seems like they're on the cusp of promising things, they just are having trouble getting through a particular period of time. We also suggested that there was a time when people were shut down in sort of patchy and heterogeneous ways, across departments, divisions, things like that. And so we said, okay, if you could set up kind of sort of labor pools almost within institutions to say, okay, well, we need help here and individuals who do work, you literally have persons in your lab who cannot do their work because of the work stoppage. Can we do something to sort of manage our resources less at the level of individuals and more at the level of institutions, sort of pool skills and personnel and resources that we think could be more fluid. That would also be helpfu. A number of other sort of ways to build opportunities, in particular for individuals who are laid low by things they couldn't get from sort of stoppages in their work or permitting, say, wet bench labs resources, like physical resources, that they lost access to. Or in some cases, actually, if you're raising a colony of animals, for example, that you missed critical windows? In order to be able to share those resources more broadly, or be more fluid with them, to people avoid those kinds of crises. Stephen Wilson 1:00:28 Right. Are there any challenges that are kind of specific to neurology? Or the kind of research that you do, in contrast to science in general? Roy Hamilton 1:00:43 I don't know if I would say that there are challenges that are unique to neurology. Let me say that one reason for that is that neurology as a field of science is so heterogeneous. It covers all the ground, right? I mean, think about what I do with human subjects, essentially studying behavior with a particular intervention. Generally, my bread and butter is measuring human behavior, compared to individuals who are running wet bench labs or doing animal work. And we all call it neurology, right. So it spans the whole gamut. In my particular work, I think some of the challenges, obviously, when there's a, a pandemic, that is especially lethal to individuals who are of an age when you're studying diseases, like neurodegenerative diseases that lead to aphasia, obviously, there are special concerns and special impediments when you're trying to conduct your research in an at-risk population that make it harder to get that research up and going again. Stephen Wilson 1:01:48 Right. So you're an Assistant Dean in the School of Medicine at Penn. How long have you been in that role? Roy Hamilton 1:01:56 I've been Assistant Dean for Cultural Affairs and Diversity at Penn. Well, the title changed recently, but the role really didn't. So I've been in the role since 2012. Stephen Wilson 1:02:09 All right, a long time. What are the kind of major goals that you have in that role? Roy Hamilton 1:02:16 It is my job to support all of our student cultural affinity groups in that role. So what that means is, the students who come to Penn, they often have groups to support aspects of their identity. So oftentimes, that that refers to students who come from historically marginalized racial and ethnic groups, so Black or African American students, students who are Hispanic or Latino, LGBT students, students who are women. But also students who come from low income or first generation college or medical school backgrounds. I also support students from different religious identities and approaches. And the idea is to make the school the kind of place where every student from every walk of life and background feels like they can thrive and feels like they belong there. And so that's a big part of my job is to foster that community. Stephen Wilson 1:03:27 It's kind of incredible that you can have that role and be an extremely productive researcher and clinician. How do you fit all those things into your eight hour workday? Roy Hamilton 1:03:43 I think you elucidated the central falsehood that allows me to engage in those various activities, which is I do not have an eight hour work day. But it is a constant juggling act. I should also mention that more recently, I've become the Vice Chair in the Department of Neurology, for the same general principles. I'm the Vice Chair for inclusion and diversity in the Department of Neurology. Now, my job is also to make Neurology the kind of place where everyone feels like they belong and, you know, it belongs to them. So it's a juggling act, for sure. But on I think that is especially worthwhile. I approach it as being, especially worth wild because, in part, there are relatively few individuals from historically marginalized backgrounds of all types who get to relatively senior levels of research accomplishments. Or becoming faculty members. As a person from one of those groups or a couple of those groups. The feeling of always building the plane while you're flying it is, you know, being the first person to go through all these experiences, is a while it's one that I'm I'm proud of, I think that it's one that I would like to spare other individuals who are coming from these backgrounds. And to make it to make it easier. And if there's a way I can do that and sort of build both the diversity of science and also, I think, strengthen the overall enterprise of science. Because I think that having people come from more backgrounds only strengthens the enterprise. Well, then, yeah, that's something I think it's important to pursue. Stephen Wilson 1:05:35 It must be really rewarding to be kind of helping others to follow in your footsteps. Roy Hamilton 1:05:42 I'll tell you, I have now gotten old enough that I can actually start to see some of the individuals who I worked with closely as you know, undergraduates, medical students, residents, come into their own as faculty members, as researchers, and it is it is richly rewarding. Stephen Wilson 1:06:06 That's great. Well, I won't take any more of your time. I know that you have a vacation to get to and you've gotta explore the grounds and do something that doesn't involve watching TV. So thank you so much. It's great to catch up with you. Roy Hamilton 1:06:21 It really was. Thank you for inviting me on. I had a great time. Stephen Wilson 1:06:25 Cool. I'll see you around. Roy Hamilton 1:06:27 All right. Take care. Stephen Wilson 1:06:28 All right, bye. Okay, well, that's it for Episode 12. I hope you enjoyed my chat with Roy as much as I did. I've linked his lab and center websites on the podcast website, as well as the papers we talked about. You can find those links at www.langneurosci.org/podcast. I'd like to thank Latane Bullock for editing the transcript of this episode. Thank you all for listening, and see you next time.