- Welcome, everybody. My name is Adina Roskies. I am a professor of
Philosophy at Dartmouth and chair of the
Cognitive Science Program, and we are really delighted to
have Caroline Robertson here to give a talk on her work. She's a new assistant professor in the Department of
Psychological and Brain Sciences at Dartmouth, and her research focuses on cognitive neuroscience approaches to memory, perception, neurodiversity, and, in particular, on autism. Caroline received her BA
from Columbia University where she studied neuroscience and, to my delight, philosophy. It's always nice to find a kindred spirit who does both things. She received her PhD from
the University of Cambridge in 2013 as a Gates Cambridge scholar and an NIH Cambridge fellow. At Cambridge, Caroline worked in the labs of Dr. Chris Baker at the
National Institutes of Health and in the labs of Dr. Simon
Baron-Cohen at Cambridge. Her postdoctoral research was in the McGovern Institute
for Brain Research at MIT, where she worked with Nancy Kanwisher. She held a coveted junior fellowship in the Harvard Society of Fellows, and, among her accolades,
Caroline was named a fellow of The American Academy
of Achievement 2014, a NARSAD Young Investigator of the Brain and Behavior
Foundation in 2015, and a Kavli fellow of the
National Academy of Sciences in 2016. She's been awarded several grants from the Simons Foundation,
and her work appears in top neuroscience journals. Caroline uses a variety of
neuroscientific techniques to study autism, and her
groundbreaking findings have provided the scientific community with a novel way to think about
the neural basis of autism. She is the Director of the Dartmouth Autism
Research Initiative, or DARI, which involves collaborations
with Dartmouth-Hitchcock, with Geisel School of Medicine, and the McGovern Institute
for Brain Sciences at MIT. I think DARI was recently written about in The Valley News, if I'm correct. And so we're all here
to hear about Caroline's fantastic work, and I will turn it over to her with no further
ado, thanks for being here. (audience clapping) - Thank you, Adina. And thank you all very much for coming. Being in a lecture at 5:30 in the evening, I really appreciate your time
that you've taken to be here. And I really think the
turnout here tonight is a real testament to the appetite and hunger for a community
around autism awareness and understanding in the Upper Valley, which I'm also very
grateful for, and beyond. I know that many people
here today have come from a great distance beyond
just the Upper Valley, so thank you for being
here with us tonight. Now, before this talk
begins, I'd just like to mention one little thing,
which you may have noticed on your invitation that
we listed this lecture as a sensory friendly event. So what does that mean tonight? First, simply, the lighting and sound during this talk will be
moderate and predictable. The second is that I'd like to invite you to, as you need, feel free to stand up during this lecture and
move around the auditorium. And third, we've reserved a space upstairs on the third floor, which
will be a break room tonight. So if you need to take a break at any point during this lecture, right outside of that door
there are a set of elevators. You'll take them to the third
floor and go to room 335, which is actually our
newly renovated lab space, which has been left open tonight
as a place to take a break. So I'd like to start by
giving you an outline of what I'm going to
talk to you about today. The first question is simply
why do we need autism research? And I'd just like to begin with this topic to get a discussion going
about starting on the same page about what autism looks
like today in our society. And second, I'd like
to give you an overview about what autism looks
like from the perspective of human neuroscience. Specifically, I'm going
to give you a brief taste of three current theories
from the perspective of human neuroscience of what might be different in the autistic brain. And I'm gonna do that
from three perspectives of different kinds of
analyses that we bring to this question using
brain imaging techniques. And I think you'll notice
from this section of the talk that we are still a far way
away from having a great grasp on the human neuroscience of autism. And finally, I'd like to introduce to you the Dartmouth Autism Research Initiative, which is our opportunity
here and in Hanover to build a world-class
autism research center based here in the Upper Valley. I believe that this initiative will really have long, large-range impacts in terms of scientific progress
for understanding autism and also impacts on our community in terms of awareness and
support for individuals on the spectrum in the Upper Valley. So I'm happy that you're all here to support this initiative. So starting with this first question. Why do we need autism research? The basic reason why
we need autism research is that we know very little about it. We know very little about what alterations in neurobiology might be
causing this condition, but we also know very little
about what we ourselves mean when we give someone
a diagnosis of autism. We don't really know what
autism is a disorder of. Now this is very unusual
in clinical medicine. For practically every
other psychiatric condition that we know of, we're able to answer the very simple question, what's this disorder a disorder of? Alzheimer's, for example,
is a disorder of memory. Depression, a disorder of sadness. ADHD, a disorder of attention. But autism is not that simple. There's no heart, no core
to the diagnosis of autism. Rather, is autism is a
constellation of symptoms. People with autism often have difficulties in social interactions,
sometimes have repetitive and physical behaviors, highly organized and detail-oriented patterns of thought, and differences in how they sense and perceive the world around them. As you can see, these
symptoms really touch every single domain of human behavior, from sensation to thought to action. And yet, it's not clear what might bind this constellation of symptoms together. This is really a central mystery
of autism research today. What neural differences might cause both low-level sensitivities
in sensation and perception as well as high-level differences in empathy or theory of mind. But working off of that
constellation of symptoms as diagnostic criteria, we
are able to say a few things about how autism appears in society. The first is just a statistic
that I'm sure everyone in this room is aware of, which
is that the prevalence rates of autism in the United
States today are one in 59 people, according to
the most recent CDC reports. And this prevalence has more than doubled in the past 10 years. But perhaps something that you may not be as well aware of is what this
looks like across the world. One important thing to know about autism is that the prevalence rates are not at all consistent across the globe, suggesting that either
our diagnostic criteria for autism are not
universal or our ability to recognize autism in different
societies varies widely. So notably prevalence rates are on a comparable order of
magnitude in the United States, the UK, and parts of east Asia. But in Iran, for example, the most recent estimate is one in 1,500. Now some of these numbers
are outdated today. Israel, for example, has recently had a new prevalence estimate, which puts it in the same range as the US. But still, there's a lot of heterogeneity as you look across the globe. But of disparity acknowledged, there's one consistent
feature of autism diagnoses across cultures or, in this
case, states in the US, which is that the frequency of autism is consistently higher
among males compared to females at a ratio of four to one, which means there are four times more men with a diagnosis of
autism compared to women. The consistency of this
observation suggests that it might hold a clue
about the biological origins of the condition, but another possibility is that autism manifests quite differently in women compared to men and that we are less equipped to diagnose it. I'll just add a few more things to this list about how
autism appears in society, including that it's thought to
be a developmental condition. It's first diagnosable by
the 18th month of life, it recurs in families, which suggests a strong genetic component,
and finally, it's known as a spectrum condition, which
simply means that we all have some degree of autistic traits. Everybody in this room has some level of sensory sensitivities,
social understanding abilities, et cetera, and those of us
above a certain threshold are given a diagnosis. So why do we need autism research? Well, despite all these
things that we know about how autism appears in society, there are three sobering facts that we do not know about autism today. The first is that, even
though that I've told you that given our current
diagnostic tools we're able to recognize autism as
early as 18th month of life, currently in the United
States, the average age at which we diagnose
autism is the age of four. It is quite difficult, very difficult, to recognize it early in life,
which leaves very little time for early developmental
behavioral interventions. The second is that, unlike
conditions like depression or anxiety, there are
currently no approved drugs to target the core features of autism, which for some people might
dramatically improve quality of life if we understood them. But currently, we don't understand what neuropathways we'd even want to target in autism, let
alone how to target them. And third, without understanding autism, we simply don't understand
human neurodiversity. As you've seen, autism
is incredibly common in our community, and all of us have some degree of autistic traits. So it seems inarguable that a rich piece of understanding who we are as people is tied to understanding
people with autism. So all in all, I think we can agree that there is a great
need for autism research. So the next question is what
role can neuroscience play? What does neuroscience have to offer here? Of course, there are many lenses into understanding autism, from the genes that might carry autistic traits
to environmental influences that might elicit these
traits in development. And our expectation is simply
that each of these things at some point influences the brain. And so it's particularly important to understand whether there
are any key differences in neural development in brain function, which are associated with autism. This isn't to say that these other areas of research aren't important. They certainly are, but
I'm a neuroscientist, and this is where my expertise lies, so this is what I'll be talking
to you about mostly today. So the most prevalent technique that neuroscientists use to
investigate the autistic brain is called magnetic
resonance imaging, or MRI. And this is a picture of an
MRI machine that's quite like, in fact, the machine that
we have just one door down in this basement of this building here. This is where we do our brain
imaging, brain scanning. And I just wanted to show you a picture to give you a sense of
what we're looking at when we look at MRI data. Using MRI, we're able to
safely measure activity inside of the brain as
it processes information. So we're able to see, say, the red spots in the brain as regions which are firing as a person is watching a
movie like this compared to the blue spots on the
brain, which are regions of the brain that are deactivating as a person's watching a movie like that. But most notably, MRI allows us to investigate the
structure and the function of the autistic brain at
three different scales, which are very important
to look at in autism. Using one kind of MRI,
we can ask questions about which regions of the brain elicit different activity in autism. Using a second kind of MR imaging, we can ask questions about
how different regions of the brain talk to each other or are connected to each other in autism. And finally, using a
third kind of MR imaging, we can measure the molecular
composition of the brain and ask how different molecules, which contribute to signaling
and electrical activity in the brain, are acting
in the autistic brain. So today, my plan is to show
you three different hypotheses about autism which we've been able to test using these three
different scales of neuroimaging. And I'm gonna start with the first. Can we pinpoint a particular region of the brain which might
be functioning differently in people with autism? So one region of the brain which routinely has shown different activity
in people with autism is this swath of cortex here on the side of the brain behind your
left and right ears. In this region of the
brain, cells are sensitive to various types of dynamic motion. You might wonder what is dynamic motion have to do, potentially, with autism? Well, one key component of our ability to understand the world
around us with agility is our ability to integrate dynamic and moving information over time. In order to understand the movements of this ballerina or the facial
expressions of our friends or the eye gaze of a
person telling us a secret, we must be able to process
fast-moving information and integrate into a
complex representation. People with autism often show differences in processing dynamic information, and this has been shown to correspond to reductions in activity
in this region of the brain, which has led researchers to believe that this region of the brain might be a key component of
autistic neurobiology, perhaps a very simple processing step which could underlie a whole host of higher order symptoms in the condition. So that was one example of a
finding using brain imaging to implicate a specific
region of the brain in autism. But as we discussed a few minutes ago, autistic symptomatology is not simple. It's very unlikely, it's
very hard to imagine that a single region of the brain would ever be able to explain all of it. So recently, neuroscientists have turned to setting patterns of
connectivity across the brain. Perhaps there's no particular region which can single-handedly
explain autistic traits, but instead an atypical pattern of wiring across the brain, which changes the way that different regions of
the brain talk to each other. To study the connectivity
between brain regions, what we do is take a brain like this and divide it into
thousands of little nodes. And then we look to see how
many structural connections exist between two nearby nodes like this one and a node
that's nearby to it, as compared to two farther
away regions in the brain. And then if you were to line
up all these different regions of the brain next to each other, you might end up with a map
that looks something like this. Typically in the brain,
regions which are nearby to each other are highly interconnected, mostly because regions
nearby to each other in the brain tend to
process similar stimuli. So sharing information is very important and highly efficient. These are called short
range, or local, connections. But the brain also has many long-range, or global, connections which allow regions on opposite ends of the brain to share information with each other. So this way the part of your brain which processes information
about visual information, visual things, can speak to the part of your brain which
helps you make decisions about what to do with that information. And one prominent theory
in autism research is that the autistic brain might be marked by a relative overabundance
of local connections but a reduced number
of global connections. This architectural model of
what might be different here is appealing because it resembles
the broader symptomatology in autism that would potentially result in a highly accurate local representations but difficulty integrating
these representations together across the brain, reminiscent
of the classic phrase of seeing the trees but not the forest. Unfortunately, a close examination of this body of research reveals some pretty conflicting findings. Some studies have found evidence in support of more local
and less global connections in autism, but others have
found an overabundance of both local and global
connections in autism, and still others have found fewer local and fewer global connections in autism. However, one intriguing observation, which might in part explain these results, is that individuals
with autism tend to have a more unique connectivity
profile compared to controls. So, in other words, when you take a bunch of individuals without autism, here I'm showing you
three different people labeled control one, two, and three, and just ask which regions of their brain are spontaneously connected to each other when they're just at rest,
you see three patches of brain that seem to
be appearing the most, two in yellow and one in blue. And in those three
people, those patches fall in approximately the same place. Everyone has a yellow spot
there at the back of the brain and a middle band of the,
yellow band down the middle, and a blue thing towards
the front of the brain, approximately the same
anatomical location. But here are three
individuals with autism, three different individuals. And what you might notice here is that, again, everyone
has those three patches of brain showing up during this time when we're just measuring
spontaneous connectivity. But the locations that are appearing are slightly different,
slightly idiosyncratic between for each individual, suggests that in autism perhaps the best way to understand connectivity
might be thinking of uniqueness among each
individual that you're studying rather than a distinct
connectivity profile that categorizes the entirety
of the autistic spectrum. So having explored a
technique which allows us to map connections and
connectivity patterns across the autistic
brain, I'm gonna turn now to a final brain imaging
technique which allows us to look at the architectural
building blocks of the brain. So this brain imaging technique is called magnetic resonance spectroscopy, and it lets us look inside of each region of the brain and study the molecules which regulate signaling
inside of these regions. To do this, we have to /dive
onto a much smaller scale and look inside of what might be happening inside a single region inside the brain. So I'm gonna play a
little movie here for you which is something to illustrate that, in every region of the brain, neurons are organized into units or columns. And electrical activity travels up and down these columns
like cars on a highway. In order to regulate this
kind of electrical signaling, there are two molecules with
are particularly responsible for this kind of information
flow in the brain. One molecule, glutamate, is a go signal. It facilitates this cascade
of electrical activity that you just saw. Another molecule, GABA, is a stop signal. It holds back, or dampens, activity. And we recently discovered
that this second molecule, GABA, is acting differently in parts of the autistic brain and contributing to sensory sensitivities
in people with autism. To show this, we measured the
amount of GABA and glutamate in the brains of people
without and also with autism. And what we found is
that GABA, that molecule which dampens neuronal activity, supported visual filtering
in people without autism. So the more of that inhibitory
stop signal you have on the X axis there,
the better you are able to filter visual information and sensation if you don't have a diagnosis of autism. Instead, with people with
a diagnosis of autism shown here in red, GABA seems
to exert no effect at all in this particular visual
filtering paradigm. So this new neuroimaging
technique allowed us, for the first time, to
link a specific molecule in the brains of people with autism to real behavioral differences that they are describing in their daily life in terms of visual sensitivity. All in all, so far at least, I've told you that motion-sensitive areas of the brain show differences in people with autism. And we expect that this might impact an individual's understanding
of the social world, which is very dynamic, as
well as the sensory world. I've also shown you that
there are differences in connectivity patterns
across regions of the brain in autism, and I've shown you a bit of discrepancy in that literature, mostly for the point of showing you that, in terms of the
neuroscience of autism today, there are lots of conflicting findings, and we're still doing a lot
of work to try to figure out what in this noisy world is true, is real. And finally, I've shown you
that we've recently discovered a specific molecule on the brain which seems to impact
an individual's ability to filter sensory
information and show that to be different in people with autism. At this point, I think a
question we should return to a little bit is where does this get us? Where does the little bit that we learn from the perspective
of neuroscience take us in terms of those three sobering facts that I brought up earlier in this talk? And I'd like to say that, first of all, our hope is that neuroscience
studies like these will be able to help us
find signatures in the brain which can be used for early diagnosis. The trick with diagnosing a kid who's less than 18 months of life is really children don't
speak at that point. It's very hard to standardize
a type of assessment, and yet you can imagine
that using techniques like brain imaging need to be able, potentially, to find signatures
in early, early days of life that could help us move
our diagnostic time points earlier and earlier. This is something that
we're really actively working on in our field. Second, these studies help us to identify neural pathways which we might target down the road in terms of drug therapies. As you saw with our recent findings like in GABA to sensory
sensitivities with autism, it's possible to identify
specific neural pathways which might link to specific
symptoms that people have. And finally and most importantly, we're starting to understand
the unique neural profiles of people with autism, which is something that is as interesting
to a person with autism as to just a general scientist who cares about understanding the brain
and who we are as people. So all in all, I think we
are still a long way off from understanding the
neurobiology of autism. But we're making exciting progress. And I think that we've
seen that neuroscience can play an exciting
role in autism research. So next I'd like to
tell you about the role that we hope to play here at Dartmouth doing neuroscience research. So starting with who we are. I have described to you the Dartmouth Autism Research
Initiative as our name, and we're calling ourselves DARI, and we're more than just me. We're really a group of neuroscience and psychology researchers across campus and the medical school who are focused on understanding autism at Dartmouth. We have expertise in
neuroimaging and eye-tracking and also virtual reality, and
a lot of our funding comes from national autism
foundations which are invested in helping us to build a world-class autism research center here at Dartmouth. I describe our mission like this. I say that DARI is
dedicated to understanding the biocmedical causes of
autism spectrum conditions, and we lead programs
in both basic research, trying to understand
autism from the perspective of biology, as well as
translational research, trying to bring some of these insights into tools that could potentially
impact people's lives. Our research areas are things
like studying brain activity and sensory perception
and language processing, and these are just a few. So I thought that I would
give you a little bit of a deeper preview into
a few concrete types of studies that we are
running today at Dartmouth. So one major focus of DARI is on understanding the
neurobiology of autism. So along these lines, as
I've mentioned to you one of these already that we've
recently discovered a number of different markers
of autism in the brain, especially in sensory perception. And now we're trying to understand whether these differences might be able to serve as early diagnostic
markers of the condition by moving them into kids
and also translating them into tools for translational research. This is a very molecular level question, and the next kind of
question that we ask at DARI are really broad-range questions that relate how people with
autism represent and engage with real-world sensory
and social environments. To ask these kinds of questions, we often techniques like
virtual reality headsets and eye-tracking, where we can measure how people look around and engage with a 360 environment from
their own personal space and try to understand
differences in how people allocate attention in the
world and filter information. Finally, we're interested
in understanding the role that genetics play in autism. So we have some projects where we work with individual who have very, very rare single genetic known
causes of their autism. These individuals tend to
live around the country, and we fly to them, and
we try to understand whether differences in sensory
behavior look different in different kinds of
genetic causes of autism. We also are interested in generally how general variation in genes which we all have co-vary and predict autistic traits, which we all have. So we do genetic assays here at Dartmouth as well as in these national studies. A final thing I'd like to say is just that we are very motivated
by the burgeoning motto in the autism community of
"nothing about me without me." And what this means to us is we're trying in every way possible to
involve people with autism in our research, in the
design, and in the execution and in the communication of our studies. It seems like the best way for us to make scientific discoveries that might end up being able to translate into people's lives. So this takes the form mostly
right now involving students who are on the spectrum,
whether at Dartmouth or also we're working
closely with Landmark College in Putney, Vermont in our research, in getting our research
initiative off the ground. Secondly, science communication. You are witness to that today. We are interested in
trying to bridge the gap that currently exists
between neuroscientific work on autism and people who are invested in understanding autism by giving talks. And also, if you haven't yet
visited our Facebook page, we're trying to regularly
cover on social media new scientific papers that
come out every other day that have something to
do with a breakthrough related to autism. It's often hard to filter these things. I even find it hard to
filter these things, and I have accessed the full PDF, and (laughs) I know in
the general community you usually just have access
to the abstract, so that's part of the work that we're
trying to help with here. And finally, we're working
with community partners to increase autism awareness
in the Upper Valley. So we're collaborating in particular with the Special Needs Support
Center and Keene Perspectives to begin a sensory-friendly event series in the Upper Valley. So stay tuned to learn a
little bit more about that. And the last slide is really just to say if you have a diagnosis of autism or know someone who does, we would love for you to join our contact list. All of the research we do depends
on people getting involved and helping us to understand autism. Plus, our research is usually fun. In addition to having the gratification of helping with science, you
also get paid for your time, pictures of your brain, to
experience virtual reality, and it can be a great experience with us. So if you don't visit our website or go to our Facebook group or email me, you can also just go
see AJ, my grad student who's sitting outside to
the left of that door. So thank you very much, and
also wanna thank the many faces on this slide who are our clinicians and also scientists in my
building who have been helping to begin our first
research studies in DARI. (audience applauds)
