A conversation with Dr. Emily Finn (OHBM 2024 keynote interview series Pt.2)

Dr. Finn is an assistant professor at Dartmouth College, leading the Functional Imaging & Naturalistic Neuroscience lab. She completed her PhD at Yale University with Todd Constable and moved to the National Institutes of Health (NIH) for her Postdoctoral research with Peter Bandettini. In this interview, she discusses how she became a neuroscientist, accidental discoveries that led her down fruitful research avenues, and what we can expect from her upcoming keynote talk at this year’s OHBM Annual Meeting in Seoul. Read on to learn more! If you’d like to read more about Dr. Finn’s recent work, check out our Brain Bites summary, here!

Nabin Koirala (NK): Good morning! Thank you for taking time for us today. I am very excited to find out more about your upcoming  keynote lecture for the annual OHBM meeting 2024 and your personal experience in research. To start with, could you introduce yourself to the readers and tell us briefly who you are and what you do?

Emily Finn (EF): Sure! I do consider myself a neuroscientist, and I am one by training. I wasn't always thinking that I would be on this career path—my undergraduate major was actually linguistics. Although, by the end of my time as an undergrad, I had become really interested in the brain and how the brain processes language. That's what got me into doing my first functional MRI study, and the rest is history. I did end up finding my way into a Ph.D. program in neuroscience, where I worked with Todd Constable, and then my postdoc was with Peter Bandettini. Now I find myself in the Department of Psychology and Brain Sciences, here at Dartmouth, where my heart lies in sort of the theoretical questions in cognitive neuroscience, cognitive psychology, social neuroscience, and social psychology, but also on how we can bring exciting new tools in functional imaging to help answer important questions in those spaces.

NK: I read your New York Times article titled “How I Learned to Stop Worrying and Love Linguistics”, and it was brilliant. Could you talk a bit more about it and take us through your journey in science? What motivated you to do this line of research and how did you end up doing fMRI as a linguist?

EF: Oh, thanks for digging that up from the archives! So, as I disclosed in that article, I was not much of a science person growing up. I worked hard in school and I did okay in science, but it was not one of my favorite subjects. I was very interested in languages. In high school, I studied French, German, and a little bit of Spanish. While working on my college application I saw that “linguistics” was a potential major listed on the Common Application, and I just chose that without really knowing exactly what it was. I arrived at Yale where I did my undergraduate and graduate training and started taking some linguistics classes, which I found fascinating. I also studied individual languages which really piqued my interest in neuro linguistics. My first formal exposure to the topic was a class in linguistics called 'Language and Mind' that focused on aphasia; it went through the cognitive neuroscience of language and what we knew about how the brain processes language, mostly from lesion studies, including a little bit of fMRI. 

I also took Yale's Intro to Neuroscience course, and I loved it. That class really hooked me, and I did consider at one point double majoring in neurobiology and linguistics, but I ended up just sticking with linguistics. Fortunately, I was able to get some really awesome research opportunities as an undergrad. My senior undergrad thesis was an fMRI study with Todd Constable. Fast forward to senior year,  I was not really sure what I wanted to do, and I was applying to anything and everything—random fellowships and whatever jobs I thought might be mildly interesting, and I was not getting many hits. I was getting increasingly stressed and one day my undergrad advisor at the time suggested, why don't I just go to grad school?  To be honest, embarrassingly, that hadn't even really occurred to me. I think I didn't fully understand what it means to be a professor and have that combination of research and teaching. It was clear that the thing I was most passionate about was the research, so I ended up staying for a semester after I graduated to wrap up some of my studies in that lab as an undergrad, and submit my first few applications for PhD programs. However, by the end of that extra semester, I was experiencing something many people are probably familiar with: the honeymoon period with science was over in some ways. While the study I had done for my thesis worked out beautifully, all of our follow ups were miserable failures. I was struggling seeing the side of science that can be really hard because you just are not seeing a lot of reward for the effort that you're putting in. I decided that I wasn't ready to commit for a PhD right then. 

I ended up taking another two years off. I moved to Peru and worked for a non-profit coffee company in Lima for about a year. That was an amazing experience, but by the end of it, I realized that I was not as passionate about that as I had been about research. I moved back to the US and got a job at MIT in their news office to write press releases and public-facing articles for their website. I was getting to interview scientists about all the amazing findings that were coming out of their labs and try to write them up for a lay audience. I really enjoyed that, and I still think science communication is really fun and important. But to be honest, by the end of about a year in that job, I found myself just really wanting to be the one doing the science again, as opposed to the one just writing about it. So, I reapplied to PhD programs and at that point, I only applied back to Yale, because I was pretty sure that I wanted to just go back there for my PhD and work with Todd. Luckily, he took me back and I guess the rest is history. It was a bit of a long road but I think it's important to tell the story, because it wasn't necessarily a straightforward path as many would assume. 

NK: Thank you—it was fascinating to hear about your journey, and I think it would be very interesting for younger researchers who are thinking about careers in research. Let's switch gears a bit and talk about your keynote lecture. Would you be able to give us an idea of what you will discuss?

EF: Yes, I'm very excited and honored to be giving this keynote, as I consider OHBM my home meeting. The first meeting I attended was back in 2012, when I was not even technically a grad student yet. So, it's a huge honor for me to be giving this talk. I'm really thinking deeply about what message I want to try to convey. I think one theme that will definitely factor prominently will be this idea of how we are actually acquiring data, what we are having participants do—or not do—while they're in the scanner. When I started my PhD back in 2012, resting state was really big and was growing. I was very excited about resting state connectivity, and Todd's lab was doing a lot of work in that space during the time. In recent years, I've been trying to advance this idea that it might be time to take some of the great tools and analytic approaches and new ways of looking at the brain that we developed in resting state and take those forward to acquisitions where we impose some known paradigm on our subjects. Whether that's watching a movie or listening to a story, or some other sort of more controlled task. We can often get more out of our data by having some kind of ground-truth experience that we know could be influencing the subjects’ patterns of brain activity at any given time. 

NK: I recently read your article about resting state, where the message was, as the topic said, “let's put rest to rest.” I know that it is crucial for various demographics and patient populations to acquire resting state data. So, I think it would be great if you could explain to the readers what you mean by resting state and how it differs from a naturalistic paradigm? Why is one better than the other? 

EF: Yes. When I think of rest in its purest form, it's no external sensory stimulation—or “minimal” if there's a fixation cross. Obviously, they're doing something, they're thinking about something and they're experiencing something, but you just don't know what that is, right? For naturalistic stimuli, you have that sensory input, like movies or stories, that you've imposed on the subject. That really gives you leverage on the analysis side. On the one hand, you can treat the data like rest in the sense that you can analyze functional connectivity in the same way that you can do at rest. But you can also link the fluctuations that you're observing to actual known elements of the stimulus, whether those are punctate events or some kind of continuous feature time course that you're pulling out from the movie. You can also use inter-subject approaches where you don’t even have to model anything about the stimulus itself, but you’re using one brain as a model for another brain and looking at synchrony across different brains, or perhaps of the same brain over multiple acquisitions, in a way that licenses you to assume that correlations you see are largely driven by something about the movie. 

I want to be careful here, because obviously, my work and many other people's work has shown that it's not like the brain completely functionally reorganizes between these different states. It’s not the case that the brain at rest looks completely different than the brain while watching a movie—there’s no theoretical reason to think that would be true, and we also know empirically that that's not true. So, to me, the really powerful thing about naturalistic stimuli is not that they completely change the fundamental organizational backbone, but rather that we have this foothold into what these fluctuations might mean, because we can link them to some kind of known time course of what the person was experiencing.

NK: Yes, and I guess one of the biggest advantages of using a naturalistic paradigm is in identifying those inter-individual differences and being able to build naturalistic task based paradigms. Right? We recently had a talk from Nick Turk-Browne here at the Wu Tsai Institute, and he also talked a bit about the naturalistic paradigm, but also how you integrate different social norms and cultural aspects if you are collecting data in a large range of demographics. One would assume that the kind of movie or video that is used, taps into hierarchical emotions and social context but that varies significantly across population, right? So, is it the case that there we are making it much more complicated than is necessary to understand a specific brain function? Do you think we would ever be able to take that in and be able to design a pattern where we could actually understand the basic neural mechanism? 

EF: I think the point about cultural and social differences is a very good one. Lots of us, myself included, initially think of using is Hollywood movies, right? That's the easiest place to start. But it’s important to remember that those are very deliberately crafted products and designed to be very powerful in generating cognitive and emotional experience, and you can certainly see how a lot of the content of those films might be culture- and/or demographic-specific. Essentially the idea is that Hollywood movies typically have a very clear style— the way directors use scene cuts and montage, there's kind of stylistic elements that might be very familiar to certain audiences and less so to other audiences. And then obviously there’s the content of that movie, which, depending on the demographics of the people involved in the storyline, is going to resonate with some people and not others. So certainly, selecting a stimulus in the first place is an art that needs to be thought carefully about. One thing my lab has been doing in some of our work is using simpler, more stripped down stimuli —for example, Inscapes, geometric shapes moving around the screen, etc. I still consider this under the umbrella of “naturalistic” as long as it has a continuous timecourse and is something that's engaging enough to actually keep people's attention. But I also think there are ways you can use this rich experience with more complex stimuli and can go directly after the cultural differences, which could be really interesting—for example, taking the same movie and showing it to people from different backgrounds. 

However, I think it would be a fallacy to think that our more traditional tasks don't have some element of cultural specificity. How people engage with purely “cognitive” tasks is probably influenced by what educational system they came up through, for example. I think people's social and cultural backgrounds are probably going to influence how they do any task and how they process any type of information that's coming at them. 

NK: Yes, I agree. Let’s move now to some of your other works over the years. It's somehow very related but I think has a different emphasis. Would you mind explaining to us the connectome fingerprinting work you’ve done?

EF: The original fingerprinting paper from 2015, that discovery was made basically by accident. The whole project started shortly after I arrived in grad school when we were working with an early release of the Human Connectome Project Data. In the lab, we were interested in how functional connectivity reorganizes between rest and various task states. The Human Connectome Project is a nice testbed for this because each person did resting state as well as seven different cognitive tasks. We were basically generating sliding window connectivity matrices from each person in these various task states and were throwing them into clustering algorithms. Our hypothesis, or what we hoped we might see, was that connectivity matrices would cluster based on task, such as the gambling task, or the emotional faces task, etc. But when we ran the algorithm, we kept getting clusters of subjects, rather than task states. That was the story the data really wanted to tell, and we thought, well, why don't we just lean into this? That led to that first paper where we were able to show that there is a lot of consistency to functional connectivity within a person across brain states—so even though task states do obviously change the connectome a little bit, there’s some kind of intrinsic backbone or foundation of the connectivity pattern that is present in an individual no matter what they're doing, and it's stable enough within a person and unique enough across people such that we can actually identify people based on their functional connectivity matrix with a pretty high degree of accuracy. We were very excited to show that you can actually get meaningful signal from a single individual and it's stable within that person over multiple observations, and even multiple brain states but it's different enough from other people in a meaningful way, so we called it fingerprinting. For us, that was more of a methodological win to be able to trust fMRI signals coming from individual brains and gave us some optimism that we could someday use these as biomarkers that could lead to real-world tools with some kind of practical utility. 

The other piece of that paper was that we wanted to prove that these functional connectivity patterns within a person were meaningful beyond just being this sort of uniquely identifying barcode, so to say. In that sense, the fingerprint metaphor is not a great one, because yes, every individual human has a unique fingerprint, but there's nothing in the pattern of actual grooves and ridges that tells you anything meaningful about that person, right? It's just kind of a barcode that identifies people. So, we started looking to see if these brain connectivity fingerprints also contained some kind of meaningful information that could be linked to another behavior or trait or characteristic of that person. We found that we could predict, at least with some degree of statistical accuracy, people's levels of fluid intelligence as measured by an out of scanner test based on their functional connectivity patterns. That gave us the confidence that not only are these things unique, but they actually carry some meaning and they index something about the real-world output of the brain. 

NK: Yes, there is a certain level of statistical validity that you would predict a certain behavioral outcome based on the connectome itself, and I think if you would add other forms of data, like multimodal data, or genetic data, that would improve the predictive ability. Anyways, I think we have talked enough about science today, how about we move slightly out of it now. While focusing on young researchers, one question that I'm really interested to know is your perspective as a woman in science. I have heard some brilliant talks from Danielle Bassett about the gender biases and the difficulties a woman faces in science on so many different levels. What's your take on this as a very successful neuroscientist, as well as a woman? What’s your experience so far? 

EF: Yeah, good question. Because both my PhD and my postdoc labs were predominantly physicists and engineers, those skew male. I think at one point in one of the labs I trained in, I was the only woman in a lab of 15 or 16 people. We had some fabulous research assistants at that time who were women but not much at the senior level, so I guess I was aware of this early on. Ten or twelve years ago when I started my PhD, I think I would have said, you know it'd be nice if there were more women around, but I don't really feel like it impacts me, and I don't feel discriminated against. I have been very fortunate that both my PhD and postdoc mentors being male were fabulously supportive of me and I didn't feel that I was treated any differently either by them or by anyone because of my sex or gender. However, I will say that the further on I get in my career I do see a bias; of course, society has changed and there's a lot more awareness of these issues in the last few years and it's hard to decouple those things from me getting later on in my career trajectory. Not always against myself, but I see it in my surroundings. I teach undergrads and I wear a lot of different hats in this job, as many of us do, and one thing I have noticed is the expectations of emotional labor—there's this sort of expectation that because I'm a relatively young woman, I'm just going to be more open and more sympathetic and more willing to work with people and adjust and give people what they want or need. It's hard because, you know, you want to be accommodating and supportive, and you want to play that role in people's lives, but it can also really add up if you're trying to do that for all of your own trainees, other people’s trainees, your undergrads, your colleagues, and then also at home and for your family. I'm not trying to say that my male colleagues don't do this, and I think there's a lot of emotional labor in this job generally, but I do get the sense that people (often subconsciously) expect more of this from women, and particularly young women. 

NK: You raised some excellent points there. So, if you would have a Harry Potter wand to change anything, what would that be? I mean, what do you think the field needs to do to address some of these issues? 

EF: Yeah, it's heartening to see more awareness around this stuff in the last few years. My—I won't call it pessimistic, maybe realistic—view is that things are changing, but it's just going to be a slow and kind of halting change and there's not going to be a magic bullet that cures all these issues right away. That's not to excuse us from anything—I mean, we need to be actively thinking about them and talking about them and setting up structures to help with this shift. But I think, given how deep-seated this is in society, it's just realistically going to take a while for things to truly shift. If I had a magic wand the thing that I would be really interested in is quantifying some of this stuff—like, how much of my day do I spend engaging in these small emotional labors or taking on small organizational or administrative tasks or anything that I'm having to do more of on average than our male colleagues. A lot of this is not very visible even to ourselves—I have a vague sense of how I spend my own time, but it's difficult to know that very specifically. Maybe big-brother AI will get us there pretty soon but I would love to quantify where everyone’s time is going down to the minute and be able to say, look female professors spend x amount of their time on these sorts of emotional labor activities, and male professors, on average, spend less than that. We’re scientists, we want the data—if we could actually quantify that stuff, in terms of time, that would show a direct impact on research, which would make it easier to argue upon.

NK: I think that's a brilliant idea and probably something that’s needed in the field. and I agree it's important to bring out some of this stuff in quantified ways to start talking about the change. Anyways, I think we are almost at the end of an hour. So, lastly, since this is for a blog article particularly targeting young researchers, I would like to ask you something which I know bothers many of them. How is the life of a neuroscientist outside of science? For example: What do you like to do when you're not doing science?

EF: Yeah. I'm located at Dartmouth which is in a very rural area in northern New England. It's very beautiful but there's not a lot of city life. So, these days, a lot of my free time is spent doing outdoor stuff like hiking, skiing when there's snow, boating, swimming, kayaking in the summer, and I really enjoy that. I have an 18 month old now who keeps me busy, so a lot of the social life just revolves around him and being out in nature and enjoying the outdoors. I also like to read, especially a lot of fiction. I try to always have some book I’m reading, even if it's just five minutes before I fall asleep at night. I play guitar badly, but that's a nice outlet for me because there's no expectation from me or anyone else that I'm good at it, which is very liberating. In general, I think having hobbies is great. I've gone through phases of my career where I've been working harder or less hard but no matter what, devoting some number of hours to hobbies is surely needed to bring in the balance. These days, I don't work most weekends unless there's a big deadline coming up (haha!). I think it's really important to understand that this career is a marathon and not a sprint and it's very easy to burn out unless you build in time to do things that make you feel like a complete human. 

NK:Thank you Emily. It was a pleasure talking to you and knowing not only about science, but the journey and inspiring stories behind that. I really appreciate the time you took today for this and hope to see you in Seoul. Thank you so much.