YCSC Postdoctoral Fellowship in Childhood Neuropsychiatric Disorders (T32) Trainee Talks

June 07, 2023

Information

YCSC Grand Rounds June 6, 2023

Francesca Penner, PhD: Improving Our Understanding of Emotion Regulation During Pregnancy

Alan H. Gerber, PhD: Social Disruption and Loneliness in Autistic and Non-Autistic Youth during the COVID-19 Pandemic

Peter J. Castagna, PhD: Modeling Brain Dynamics and Gaze Behavior: Starting Point Bias and Drift Rate Relate to Frontal Midline Theta Oscillations

Moderated by Mike Crowley, PhD

ID10039

To CiteDCA Citation Guide

00:00Good afternoon, everyone.
00:04Thank you for coming today
00:05to our T32 Grand rounds.
00:07We're on a tight schedule.
00:08We have three talks,
00:10so I'll be brief and concise.
00:12I direct the T32 Codirect
00:15T32 with Michael Block.
00:17We're in, I think our 38th or 39th year.
00:21It's really one of the joys of
00:23my professional career to be a
00:25part of this following in the
00:27big shoes of Doctor Lechman.
00:29Not physically big, but big.
00:31Metaphorically,
00:35we are up for a newal next year,
00:37so we're going to be doing a mad
00:39rush and reaching out to all of
00:41you for materials to support us.
00:42And we really couldn't do it and
00:44succeed without you all and that the
00:46atmosphere that you bring to the center.
00:51Okay. They told me I need to hit
00:52click first here. There we go.
00:54So our T32 is growing.
00:56We only have six slots,
00:57Not only but that's what we have,
00:58but we have many others
01:00who participate with us.
01:01And here's a picture.
01:02I'm not a a Photoshop ace,
01:03so I was able to bring everyone in here.
01:06I'm going to ask for some help
01:08down the road from you all.
01:11And what we're going to hear about
01:14today are three talks from trainees
01:18Francesca Penner. Dr. Penner.
01:20Doctor Gerber and Doctor Kistagna.
01:22Dr. Penner will be telling us
01:24about her work on understanding
01:26emotional regulation and pregnancy.
01:28Doctor Gerber will be telling
01:30us about emotion disruption and
01:32loneliness and autistic and autistic
01:33youth during the COVID pandemic.
01:35And lastly, Dr.
01:36Kistagna will be talking about modeling
01:38gaze behavior and starting point bias,
01:41drift rate and frontal midline
01:43beta EEG oscillations.
01:45Before we get on to the three talks,
01:46I just want to say a few words
01:48about these three trainees.
01:50We would love to be able to have
01:51everyone speak and we did make a
01:53call to everyone and then these are
01:54three individuals who reached out,
01:56but we'll be catching other
01:58people next year to present again.
02:00Doctor Penner has really done an
02:03exceptional job as a T32 trainee.
02:06She got her own funding, F32.
02:09She published 38 papers up to this point.
02:12Not all in the T32,
02:13but you know that's what she's
02:15been doing across her career.
02:17And she landed a academic position
02:19at Baylor University of Department
02:21of Psychology and Neuroscience,
02:23so she'll be heading there.
02:24Doctor Gerber is in his first
02:26year in the T32 and he's rocked
02:28it already with two grants.
02:30He's got a autism grant and also
02:35from the Organization for Autism
02:37Research and also a Child Study
02:39Center Pilot Research Award.
02:40So Congrats to Doctor Gerber.
02:42And lastly Dr.
02:43Stagna.
02:46Also was quite prolific with 33
02:50papers and and F32 also that's an
02:54independent training grant that he
02:56received and he's landed a tenure track
02:59position at University of Alabama.
03:02So before I hand over the
03:06the mic to Doctor Penner,
03:09I just want to make a plug
03:11for these these F32 grants.
03:13We really only have a small number of.
03:16Spots on the T32 compared to the need.
03:19And so I encourage everyone here to
03:22try to to pursue these F32 grants.
03:24We have lots of support for you.
03:26We read them.
03:27Michael and I both sat on the study second
03:29committee for the review of the grants.
03:31We have examples so we can scaffold you to
03:33pursue these grants if you're interested.
03:35Anyway, thank you.
03:36Here's a treat for you.
03:39So Doctor Penner.
03:48Hi everyone.
03:49Thank you so much, Doctor Crowley.
03:50I'm really thrilled to be
03:52presenting at Greyhounds.
03:53It's very exciting.
03:54I like Doctor Crowley said.
03:56My name is Francesca Penner.
03:58I'm a postdoc working with Helena
04:00Rutherford in the before and after baby lab,
04:02and today I'm presenting some work focused
04:05on emotion regulation during pregnancy.
04:07I was excited to present this
04:09work in particular because it's
04:11something that I started on right
04:13at the beginning of postdoc,
04:14so I thought it would be interesting
04:16to kind of share the progression
04:18of work over the past two years.
04:20So to start,
04:21I wanted to begin with talking about
04:23why it's interesting and important to
04:26study emotion regulation in pregnancy,
04:28beginning more broadly with the importance
04:30of emotion regulation in general.
04:31So we know emotion regulation is
04:34a transdiagnostic factor relevant
04:36to many mental health disorders
04:38and symptoms of psychopathology.
04:40We also know it's a it's targeted in
04:42multiple evidence based treatment.
04:43So we have evidence that it can it's
04:46modifiable that we can improve emotion
04:48regulation and that by improving it.
04:50Or by decreasing emotion to circulation,
04:53we can prevent and reduce
04:55symptoms of psychopathology.
04:57And we also know that emotion
04:58regulation is important in caregiving.
05:00So it helps us be more sensitive caregivers.
05:03And it's also important in terms of
05:05modeling for children as they develop
05:08and learn emotion regulation as well.
05:10When you think about emotion
05:12regulation during pregnancy,
05:13you can think about some of the unique
05:15factors during this period that
05:16might affect our emotion regulation.
05:18So.
05:18Certainly there are lots of physical
05:20changes for the pregnant persons
05:22and physiological and brain changes
05:23that might affect the physiological
05:26experience of emotions during pregnancy.
05:28There are also psychosocial
05:29stressors that might come up,
05:31whether it's financial relationship
05:33or medical stressors that might
05:36challenge or require emotion
05:38regulation strategies during this time.
05:40And then we also know that pregnancy
05:41is a time of increased vulnerability
05:43for mental health disorders,
05:45especially depression and anxiety,
05:47which also makes emotion regulation
05:50really relevant during this time.
05:52And then finally,
05:53when we think about for new parents
05:56of the transition to parenthood,
05:57whether there might be changes
05:59in emotion regulation as sort of
06:01as new skills come online as we
06:03become parents for the first time.
06:05So thinking about all those ways
06:07that emotion regulation might
06:08change during pregnancy,
06:10but also its relevance for stress and mental
06:13health and caregiving during this time,
06:16we were interested in kind of
06:18looking at what's already known
06:19about emotion regulation during
06:20pregnancy in terms of the correlates,
06:23both during pregnancy and
06:25in the postpartum period.
06:27So early in 2022,
06:28Helena and I posted a paper and
06:30Archives of Women's Mental Health.
06:32That summarizes this research area
06:34and it's a pretty small research
06:36area so far in terms of studies
06:38that have actually measured emotion
06:41regulation during pregnancy and
06:43association with other variables
06:45either in pregnancy or postpartum.
06:47So this figure from our paper
06:50kind of summarizes this
06:52area so far. It's definitely
06:53a growing area of research.
06:55So I expect that things have may have
06:56changed in the last year and a half,
06:58but in terms of what this
07:01figure represents, so the.
07:03Boxes in solid lines with solid arrows
07:06are correlates that we have evidence
07:09for from at least one study where the
07:12boxes that are grayed out with dashed
07:14lines are hypothesized correlates of
07:16emotion regulation during pregnancy.
07:18So some of the things you have
07:20evidence for so far are that emotion
07:23regulation measured during pregnancy
07:25are related to physical and mental
07:27health of the pregnant person
07:28both in pregnancy and postpartum.
07:30It's shown association with caregiving
07:33behavior measured during pregnancy as well.
07:35And then it's Even so shown some
07:39associations between motion regulation
07:40during pregnancy in the pregnant
07:43person and then with some infant
07:46outcomes like feeding interactions
07:48and infant attention and arousal.
07:50So we have some emerging evidence for.
07:54These significant links showing that
07:56emotion regulation in pregnancy might
07:58have implications for not only a mental
08:01health in the pregnant person but also
08:04caregiving and infant development.
08:05Which suggests that this is an
08:07important factor to study and also
08:10the potentially important factor
08:11for intervention because it could
08:14have these multi prompt impacts
08:16even into the postpartum period.
08:20So wanting to kind of build on this
08:22evidence base and study emotion
08:24regulation during pregnancy more,
08:25we conducted 2 studies with archival data
08:28focused on emotion regulation and perceived
08:31stress during the perinatal period.
08:33And in both of these studies,
08:35we were conceptualizing perceived stress
08:37in terms of appraisals of one's life
08:40as stressful versus objective measures
08:42of life of stressful life events.
08:45And we were thinking about emotion regulation
08:46in terms of James Gross's definition.
08:48So attempts to influence one's
08:51emotions and how they're expressed.
08:53In particular,
08:54thinking about the emotion regulation
08:56strategies of reappraisal and suppression
08:58with reappraisal thoughts to be a more
09:01adaptive emotion regulation strategy
09:02and suppression thoughts to be a more
09:05maladaptive emotion regulation strategy.
09:07So in the first study,
09:08we were focused on emotion regulation
09:11strategies and perceived stress
09:12and expectant mothers and fathers.
09:14During the third trimester,
09:17and this was a sample collected here at Yale.
09:20Of 83 expectant parents,
09:21about 50 of them were pregnant
09:23mothers and then about half of the
09:25sample were first time parents.
09:27They completed the Perceived Stress Scale
09:29and the Emotion Regulation Questionnaire.
09:30During the third trimester
09:33of pregnancy and 1st,
09:34we looked at associations between each
09:36of those emotion regulation strategies,
09:38reappraisal and suppression
09:40with perceived stress.
09:42And we saw associations in expected
09:44directions based on prior work
09:45with the emotion regulation
09:47questionnaire in other samples.
09:48So we saw that as parents reported greater
09:50levels of or greater use of suppression,
09:52they also reported greater stress.
09:55And as they reported
09:56greater use of reappraisal,
09:57they reported lower perceived stress.
09:59So this is really underlining that
10:01in a sample of expected parents,
10:03we're seeing associations in expected
10:06directions with these two emotion
10:08regulation strategies and perceived stress.
10:11We then were interested in whether
10:13the mothers and fathers differed in
10:15their levels of perceived stress
10:17and emotion regulation strategies.
10:19And we saw we were actually surprised
10:21to see this finding was against our
10:23hypothesis that perceived stress in
10:25mothers and fathers was not different.
10:28So during the third trimester,
10:29they're reporting similar levels of
10:32stress and they're also reporting
10:34similar levels of each emotion
10:35regulation strategies.
10:36There were not any differences
10:38between mothers and fathers.
10:41So this study, one limitation of it
10:43was that it was crosssectional data.
10:45We were interested in also trying to
10:48understand better the direction of effects
10:50between emotion regulation and perceived
10:52stress during the perinatal period.
10:53So we had some archival data
10:55from a collaborator at Texas
10:57A and M Rebecca Brooker,
10:58and that's what we examined and study too.
11:00So this study also looks at perceived
11:03stress and emotion regulation,
11:04those same 2 variables and
11:07measures across three time points
11:09in in the perinatal period.
11:12So because we had three time points,
11:14here was the 2nd trimester,
11:15third trimester and four months postpartum.
11:18We were interested in testing a cross
11:20like panel model to be able to look at
11:23both the stability of each of these
11:25constructs but also the cross lag and
11:28cross-sectional relationships between them.
11:31And this was a sample of 92 pregnant women.
11:35This these data were collected
11:36at Montana State University.
11:38And as I said,
11:39they completed the same 2 measures,
11:41the Emotion Regulation Questionnaire
11:42and the Perceived Stress Scale
11:44in the second trimester,
11:45third trimester and four months postpartum.
11:49So we looked at associations between
11:51suppression and perceived stress.
11:53We really only saw evidence for
11:54stability of each of these constructs.
11:57So there were no cross,
11:58lagged or cross-sectional associations
12:00between suppression and perceived
12:02stress in this sample.
12:05We looked at associations between
12:07reappraisal and perceived stress.
12:08We again saw evidence for the
12:10stability of each of these
12:12across the three time points.
12:14And then we saw cross-sectional
12:15associations in the same
12:17direction as study one. So as
12:21reappraisal increased, perceived stress
12:23decreased in the same time point.
12:27We also saw evidence for one cross lag
12:29effect, so giving us some potential
12:31information about the direction of
12:33effects where greater stress in the
12:36second trimester predicted lower
12:37reappraisal in the third trimester.
12:39So potentially suggesting that
12:41more stress in pregnancy could
12:44sort of get in the way of adaptive
12:48emotion regulation strategies.
12:49To summarize these two studies,
12:51so both studies together showed.
12:53Significant links between reappraisal
12:55and perceived stress cross sectionally in
13:00both mothers and fathers during during
13:02pregnancy Study two gave some evidence
13:05for the stability of emotion regulation
13:07strategies over the perinatal period.
13:09Although it's really important to know that
13:11this measurement of emotion regulations,
13:12your questionnaire measure,
13:13is thought to be more traitlike.
13:15So that's one reason why we're
13:18we're likely seeing stability here.
13:21And study two,
13:22we saw that higher stress or appraisals
13:24of stress in the second trimester
13:26sort of might get in the way of
13:28adaptive emotion regulation later on.
13:30So giving us information about the
13:32direction of effects and then study
13:33one suggests that there is a potential
13:35need to include both expected parents,
13:37so not just the pregnant parent in
13:39prenatal mental health screening
13:40and interventions because we saw
13:42these similar levels of stress
13:44reported by both parents.
13:47So I wanted to briefly go through two
13:50current studies that focus on emotion
13:52regulation during pregnancy that I
13:54got funding for as a postdoc and
13:56these are collecting data right now.
13:58So if we go back to our initial model,
14:00what is known and unknown in
14:02terms of correlates?
14:03I was interested in better understanding
14:06emotion regulation in during
14:08pregnancy at a qualitative level,
14:10so understanding women's subjective
14:12experience of emotions and emotion
14:14regulation during pregnancy.
14:15So I received funding from Yale Women
14:17Faculty Forum to conduct a qualitative study.
14:20This is data collections going on right now,
14:22so I've done about half of the interviews,
14:25and the goal is to hear directly
14:27from pregnant women about their
14:29experience of emotions, stressors,
14:30and emotion regulation during pregnancy.
14:33It's specifically first time pregnant women,
14:36and the plan is to code these
14:38interviews once they're all complete
14:40using thematic analysis.
14:42The second study that's ongoing right
14:43now is focused on this association
14:45between emotion regulation measured
14:47in pregnancy and whether it can tell
14:49us anything about future caregiving
14:51behavior after the baby is born.
14:54So this study is funded by the
14:56Colleen Dobbins Foundation and the
14:57American Psychological Foundation,
14:58and it is recruiting 61st time parents,
15:01both mothers and fathers,
15:03to evaluate whether emotion regulation
15:05measured during the third trimester.
15:07Can predict caregiving 2 to four
15:10months postpartum.
15:10And in terms of caregiving,
15:12we're specifically interested
15:14in responses to infant crying.
15:16We're measuring that in multiple ways
15:18that so through questionnaire measures,
15:20through behavior during the still
15:22face paradigm through behavior
15:23during a baby simulator task that
15:25Helena has used before in her studies
15:27that is programmed to cry for a
15:29certain amount of time.
15:31And then also through measuring EE,
15:33G and event related potentials during
15:35audio of infant crying as well.
15:40So as I transition to faculty, I'm excited
15:42to continue building this line of research.
15:44So some initial thoughts are to expand the
15:46UP study which is that last study I shared to
15:48have a group of parents of psychopathology.
15:51I'm also really interested in physiological
15:53measures of emotion regulation, including EE,
15:56G&amp;ERP at multiple time points to understand
16:00other methods with other methods,
16:02whether there's stability or change in
16:04the motion regulation over this period.
16:05And then I'm very interested in,
16:08in this in the context of intervention,
16:11to really thinking about interventions
16:13that improve emotion regulation,
16:14whether they can really have this,
16:17these multipronged impacts in terms of.
16:20Both parental and child health and
16:23caregiving after the child is born.
16:26So I just wanted to end by thanking Helena.
16:29She's been the best postdoctoral mentor
16:31I could have imagined and as well
16:34as Doctor Crowley and Doctor Block,
16:36the T32 directors.
16:37It's been really great to be in the
16:40T32 and especially in terms of grant
16:43writing training and also wanted to thank
16:45funders and everyone in the Babel lab,
16:47coauthors and other post docs
16:49in the T32 seminar.
16:50Thank you.
17:01So we do have time for questions.
17:03We have 5 minutes. Make sure
17:06this is done. Any questions?
17:13That was very impressive and interesting.
17:15Thank you. Since you're interested
17:17in qualitative studies, I was just
17:19wondering what your thoughts are
17:20about measures of stress in the
17:22pregnant moms, if it's self-reports
17:25are better or some objective measures.
17:27What's your experience now being in there
17:29I think really that.
17:33Neither one is better.
17:34Like I really think it's
17:35important to do both.
17:36I mean these these studies, you know,
17:38were perceived stress and I do think
17:40there is a place for that because, you know,
17:43our perceptions of stress are important.
17:45And there's some evidence that it might
17:47overlap more with like mental health,
17:48like depression and anxiety.
17:50But I do think that it,
17:52I think it's also interesting in
17:54doing these interviews and kind of
17:56talking with women about the stressors
17:58they are experienced how sometimes.
18:01Like objective stressors are minimized
18:04in terms of like our reporting of
18:05them and so then being able to
18:08measure those because they might,
18:09they might be having some kind of
18:12biological effect that we're not,
18:14you know like acknowledging or or reporting.
18:18I do think there's some minimizing and.
18:21And I think also like I've learned
18:23from these interviews as well,
18:24I think also like pregnant women
18:25really get the message that they
18:27shouldn't be stressed during pregnancy.
18:29And so then they're like trying
18:31to minimize the stressors that
18:33they are experiencing.
18:37You have 3 minutes. Another question,
18:41Doctor Mcpartland.
18:48Having never been pregnant,
18:49I am surprised that the message
18:51received that you shouldn't be stressed
18:52during pregnancy As the husband
18:54of a woman who's been pregnant,
18:55we had a different experience,
18:56but I'm related
18:58to that. I'm curious when you I was surprised
19:00to see similar levels of stress
19:02between the between both partners and
19:04do you have a sense,
19:05how do you interpret that?
19:06And do you have a sense of
19:08the quality and the nature
19:08of the stress and whether they're
19:10stressed about the same things?
19:12Yeah, I think that's such a good question
19:14and I was wondering about that myself,
19:16like as I was doing this.
19:18Presentation again,
19:19just kind of re wondering about that result.
19:22I yeah, I'm not sure and some of them
19:24were couples and some of them were not.
19:26So I think there's also you know when
19:29we looked at whether we need to control
19:32for the fact that some of them were in
19:34couples that perceived stress like did
19:36have an effect at the couple level.
19:39So there were.
19:40So I think you know some of that
19:42is like whatever stressors are
19:43affecting both of them do seem
19:45to be a factor and then I think.
19:47And I think I'm definitely interested
19:49in kind of looking at that more.
19:51I'd like the partner effects in
19:53terms of like objective stressors
19:55and how stress on the mom might be
19:58affecting stress on the dad and
20:00we're on the nonpregnant parent.
20:02But then also understanding,
20:03I think,
20:04how the relationship can also be protective,
20:08like help to reduce stress or not.
20:15Thank you Doctor Penner. Okay,
20:17we'll have Dr. Gerber come up.
20:21All
20:28right.
20:33So that was a great talk,
20:35hard to follow. I am really,
20:38really excited to be here today and perhaps
20:40a little bit nervous to be speaking to
20:42such really great minds and people.
20:45Here that you know have
20:47inspired my work over the years,
20:49so I'm really excited to talk to you about
20:51some work that came out of my dissertation,
20:53which is done at Stony Brook University.
20:56I'm currently finishing up my first year of
20:59postdoc in Doctor Mcpartland's lab right now,
21:03and so I'll be talking about
21:05social disruption and loneliness
21:07in autistic and non autistic youth
21:10during the COVID-19 pandemic.
21:12So first of all what what do we
21:13mean when we talk about loneliness?
21:15So something we all a concept
21:17we're all familiar with,
21:18but really we're defining it as
21:21this mismatch between your desired
21:23and your actual social activity.
21:27So it's a really important and
21:29a major public health concern.
21:31There's a lot of data that we have
21:33pre pandemic even that shows that
21:35loneliness is associated with worse
21:38mental as well as physical health.
21:41So it's really great concern,
21:42but of course, as we all know,
21:44we all live through,
21:45during the pandemic this became
21:47almost one of the, you know,
21:49key or probably the key psychosocial concern.
21:52And even into today we're still
21:55experiencing a rise in social
21:57isolation and loneliness,
21:59especially in our youth or for our youth.
22:05So how is this affecting autistic
22:07individuals and autistic youth?
22:09Well, there were already some of
22:12these preexisting disparities,
22:13and the pandemic really exacerbated those.
22:16So, for example, mental health concerns,
22:19increases in stress, anxiety,
22:21depression for autistic youth who are
22:24already kind of at risk and importantly
22:26as well for their caregivers.
22:32So one thing we know is somebody who
22:35studies social isolation and loneliness is
22:37pre pandemic autistic youth were already
22:40experiencing some challenges with this.
22:42They were already at elevated risk
22:45for loneliness and social isolation.
22:47And so one thing to consider is that
22:51the pandemic could really put them
22:53at even greater risk for you know,
22:56based on its impact on social life.
23:00So despite the fact that you know,
23:02there's a clear interest in this
23:04and there are qualitative reports
23:06on this that will tell you,
23:07you know, autistic people will
23:09report missing of social contact,
23:11but there's actually, to my knowledge,
23:14has not been any qualitative or sorry
23:16quantitative examination of loneliness
23:18and autistic use during the pandemic.
23:23So we set out to do is really
23:25understand what were the trajectories
23:27of social disruption and loneliness
23:29for autistic youth during this
23:31early period of the pandemic.
23:32What was it like for them?
23:36I want to take you through a little bit
23:38of what the study recruitment looked like.
23:40So we began the study early June,
23:44so June 1st, 2020, if you can think back
23:46to a couple years ago what that was like.
23:49And we follow families and and
23:51youth for about six months.
23:53So that went from basically June until mid.
23:57Should I talking to the mic more?
23:58Can you can you guys hear me?
24:00Is it better with the mic?
24:02OK, hold on,
24:05it was meant for somebody taller I think.
24:09So in so basically the study ran
24:12from June until early December,
24:14early to mid-december 2020.
24:16So that over the period of six
24:18months we had participants in there
24:21and one caregiver fill out some
24:23questionnaires every two weeks.
24:24And so that total 12 total
24:27questionnaires over that period of time.
24:33All the families that came in
24:35and participated had already come
24:37into the lab and when they did.
24:39They completed standardized a gold
24:42standard diagnostic evaluation for autism.
24:45They also completed a
24:46cognitive assessment as well.
24:48So during this study,
24:50we asked participants to complete the a
24:54Standardized Self Report of loneliness,
24:56and that's the UCLA Loneliness Scale.
24:57So they they did that every
24:59other week for six months.
25:01We asked our caregiver to tell
25:02us a little bit about how the
25:04pandemic was impacting the family.
25:06And in particular,
25:07we're really interested in understanding
25:09social disruption in the family.
25:11And when I say that,
25:12what I mean is we were focused on the
25:14items that were related to family,
25:16anything that's limited or restricted
25:19family and social activities.
25:25So 76 youth participated in this study,
25:2851 were autistic, 25 were not.
25:32They range in age from 8 to 17 and what
25:36you can see here is as we would expect,
25:39there were differences.
25:40The autistic youth were had
25:43higher autism symptoms, severity.
25:45They also were more males,
25:47but other than that they were pretty
25:49evenly matched across the board.
25:50So there were no differences in
25:52loneliness or social disruption
25:53at that first time point.
25:59So what do we think?
26:00What's going to happen?
26:01Well, we hypothesize that social
26:03disruption would decrease over
26:05time for non autistic youth,
26:08but remain about the same for autistic youth,
26:11perhaps due to some of the stress
26:13and mental health challenges going
26:15on with parents and in youth.
26:18We also hypothesize that loneliness would
26:21decrease over time for non autistic youth,
26:24but remain about the same for autistic youth.
26:27And perhaps due to challenges
26:29in the change in routine that we
26:31all experienced in the pandemic.
26:33And finally,
26:34we hypothesize that greater social
26:36disruption would be associated
26:38with greater loneliness.
26:42So what did we
26:43find? Well, I want to walk
26:45you through this chart.
26:46So on the X axis, what you see basically
26:49is time since the study starts,
26:51so time over six months.
26:53And on the Y axis you can
26:55see their social disruption.
26:57So higher scores here,
26:59higher numbers means greater
27:02social disruption.
27:03And what you can see is
27:05that over time both groups,
27:07both the non autistic and the
27:09autistic groups decreased in their
27:12experience of social disruption.
27:14However, there was an interaction effect,
27:15so we did find that non autistic
27:17youth had a greater decline
27:19in social disruption over time
27:21compared to autistic youth.
27:26So what about loneliness?
27:27What happened with loneliness?
27:29Well, and I'll get into this later,
27:31this was a bit of surprise,
27:32but what you can see here is on the
27:35X axis you can see time again on the
27:38why you're now seeing loneliness.
27:39So higher scores here means
27:42higher self reported loneliness.
27:44And actually what we found here was that
27:46loneliness did in fact decrease over time,
27:48but only for the autistic youth in the study.
27:51So you can see in the blue.
27:54That's a statistically
27:56significant decline in the red.
27:59You're seeing non autistic youth
28:00and there's no statistically
28:01different change over time.
28:06So finally I want to show you the results
28:08for loneliness and social disruption.
28:11So you can see on the X axis now
28:13you're seeing social disruption.
28:14So again, higher numbers means
28:17greater social disruption.
28:19On the why, you're now seeing loneliness.
28:21So higher numbers means greater
28:23self reported loneliness.
28:25The colors are the same and what you
28:28can see is this interesting interaction
28:31effect where for autistic youth we did
28:33find the relationship we expected so
28:35we did find greater social disruption
28:38was associated with greater loneliness.
28:40But for the non autistic youth
28:41we did not see that,
28:42we didn't see that relationship.
28:48So what do we make of all this?
28:50So let's start with the
28:52findings on social disruption.
28:54But what we found was that
28:56social disruption declined
28:57over time for both groups,
28:59but it was a greater decline
29:01in the non autistic youth.
29:06So perhaps one way to look at this
29:08is that non autistic youth made a
29:11quicker return to social activities.
29:16So in thinking again.
29:17Into what this period
29:19was like for caregivers.
29:20There's quite a bit of research
29:22that suggests that, you know,
29:24caregivers of autistic individuals and
29:25autistic youth were already stressed.
29:27And the pandemic with challenges and
29:29getting services and all sorts of changes
29:32in routine were really quite stressful.
29:34And if you think about it or if
29:35there any parents in the room,
29:37parents tend to be the gatekeepers,
29:39the facilitators of social activity.
29:42And so perhaps one way to think about
29:44this is that it might have been hard.
29:46For those parents to reengage in social
29:49activity and to bring their kids to
29:52activities and things of that nature.
29:54And so I think it's really important
29:56to think about and the implications
29:58here for parents health particular
30:00or parent mental health,
30:01thinking about caregivers of autistic youth,
30:05both during the pandemic but also
30:07now that it can be have a really sort
30:10of profound impact on their kids.
30:16So what happened with loneliness?
30:18Loneliness declined over time,
30:20so we did find that, but actually
30:22it was only for the autistic youth.
30:24And if you think about that graph
30:26what it what seems to be happening
30:29is sort of they're coming close
30:31to their non autistic peers.
30:34This is really striking to me,
30:36really surprising because it runs
30:39counter this widely accepted idea
30:41that autistic youth are sort of
30:43universally lonely or or isolated.
30:46And so one one thing we thought
30:48about maybe this is actually
30:49related to the change in routine,
30:51but it in a positive way.
30:52So perhaps there was some
30:55flexibility or choice in who,
30:57how, when they were interacting,
30:59how frequently that led to
31:01reductions in loneliness.
31:06Another thing we really thought
31:07about though is if you guys remember.
31:10When you were in the pandemic,
31:11remember this appeared of of
31:13June 2020 and and on right?
31:15There was a big increase in who
31:17you were spending time with.
31:18It was whether it was your roommate,
31:20your your family.
31:21And so there was a big
31:23increase in family time.
31:24And perhaps one possibility is that
31:25this was actually a big positive
31:27for autistic youth that they enjoyed
31:29spending time with their family.
31:33So lastly, we found that increases
31:35in social disruption did lead to
31:37greater loneliness, but actually
31:39it was only for autistic youth.
31:43So this suggests to us that when they were
31:46actually experiencing social disruption,
31:48autistic youth were more vulnerable to
31:51feelings of loneliness than their peers.
31:54And so one thing we thought about was,
31:56you know, if you're experiencing the
31:58social disruption and you're sort of.
32:00Forced into only this
32:02digital social communication,
32:03we all remember the zoom fatigue,
32:05that zoom burnout of of 2020.
32:08This is something that might be in
32:10particular a challenge for autistic
32:12youth as they are experiencing
32:13more and more social disruption
32:15and this is for their only option.
32:17Although another possibility is that they
32:20didn't have anyone else to reach out to.
32:22Perhaps other teens were Facetiming all day,
32:24but autistic youth who are experiencing
32:26a lot of social disruption didn't really
32:29have other options and deep connections.
32:32So I'm thinking about what my next steps are.
32:34I'm really interested in continuing
32:36to examine loneliness and autistic
32:38youth and thinking about its
32:40relationship with suicidality.
32:42So I'm really grateful for funding from
32:44the Yale Child Study Center pilot grant,
32:46as well as the Organization
32:48for Autism Research.
32:49And what we plan to do is we'll
32:51have participants come in the lab,
32:53complete a naturalistic
32:54social reward paradigm,
32:56and then we'll have them fill out
32:58questionnaires through an app on their
33:00smartphone telling us about loneliness as
33:02they experience it outside of the lab.
33:04And ultimately,
33:05we hope to understand the
33:07relationship between social reward,
33:09loneliness,
33:10and and suicidality in autistic youth.
33:15So I just want to close
33:16by acknowledging the mic.
33:18Doctoral advisor Doctor Lerner as
33:20well as Doctor Mcpartlin who's in
33:22the room who've been really key in
33:25in getting all of this work done
33:27and for the great support of the
33:29Stony Brook team that was essential
33:31in in conducting this research.
33:33I also want to thank everybody in my lab,
33:35many of which are here and in particular.
33:38I did want to thank Doctor Keifer and Dr.
33:40Naples,
33:41who I know is in the room for their
33:43really essential and amazing work
33:44on this naturalistic paradigm.
33:46And finally,
33:47I'll conclude by thanking the
33:49funders which you can see there,
33:51as well as really all of the
33:53participating families who we really
33:54could not do any of this work without.
33:57So thank you very much for
33:58listening and I can take questions,
34:00questions
34:08for Doctor Gerber.
34:12Everything was so clear.
34:20Hi. I'm curious if you're defining
34:24loneliness as the mismatch between the
34:27social motivation and the and what kids
34:31are actually getting when when you're
34:33looking at the loneliness scores.
34:35When we know that social
34:36motivation might not change,
34:38but what they're getting might change.
34:40If there was a difference in.
34:44Initial social motivation in
34:46autistic and non autistic use,
34:48if that makes sense.
34:49So the loneliness scores might
34:51not be changing because they
34:53might have been lower initially.
34:55And the and
34:57yeah, if that makes sense,
34:59yeah. So I think this actually
35:01brings up kind of two questions.
35:02One is. The relationship between
35:05social motivation and loneliness and
35:07autism and this kind of gets to the
35:09heart of what I'm interested in.
35:11This idea that autistic people
35:12may not be socially motivated,
35:14they may not be interested in interaction,
35:16so how could they feel lonely pre
35:19pandemic though there's quite a
35:20bit of data at this point that
35:22suggests that that's not quite true,
35:24that they actually do feel a
35:26lot of loneliness.
35:27Now the other thing they are bringing up,
35:28which is kind of a challenge is and
35:31everybody I imagine experienced this,
35:32who did pre pandemic work.
35:34Or during pandemic work,
35:36right is we didn't have that
35:38information before the pandemic.
35:40So we do have some data on these kids,
35:42but we don't have their social
35:45motivation and loneliness prepandemic.
35:46So it would be really interesting to see if,
35:49you know,
35:50kids who are not socially
35:52motivated were totally fine,
35:53but we just don't have that.
35:54But it's a great question.
35:55More
35:58questions for Doctor Gerber.
36:04Hopefully this is a softball,
36:06it's going to be,
36:08it's not a super softball,
36:09but if you probably can
36:10answer just with yes or no.
36:12I was wondering if you'd done
36:15anything looking at the date,
36:17the data over time in a nonlinear fashion.
36:19Because I guess when I'm
36:21thinking about COVID,
36:21I kind of think about it is
36:23it was there was a lot of
36:24abs and flows of things and
36:26I'm wondering if you there's
36:27any use to parsing out
36:28the data looking at time or?
36:31Chronologically in terms of months of
36:33the year rather than time and and then
36:36also looking at when the lockdowns
36:38were and how that affected
36:40autistic versus non autistic kids.
36:42Yeah, this is this is a great
36:45question and I'm grateful to.
36:47I practice this in my lab and
36:49this question came up so.
36:51Always get to practice.
36:53It's a great question.
36:54We've thought about it.
36:56We have looked at some of these
36:57things in a long linear fashion,
36:59and I figured I'd only had 13 minutes,
37:01so I didn't get into it too much.
37:03But there is a quadratic
37:06relationship with social disruption.
37:09Where kind of dips over the
37:10summer and comes back up,
37:11which is it was just interesting.
37:14Loneliness didn't appear to change that much,
37:17which I also thought was interesting
37:19but wasn't shocking because
37:20if you look at the general,
37:23if you look at the data that's
37:25coming out now on loneliness,
37:26there was sort of this initial period
37:28where people didn't know what to do and
37:30people were feeling trapped and lonely,
37:32but people adjusted pretty quickly.
37:34And in the end,
37:35loneliness remained relatively stable.
37:37So we have data from June on.
37:40I think it would tell a different
37:42story if we had data in April
37:45and May in terms of a break,
37:48a breaking point when school starts.
37:50Also an interesting thing that
37:52we haven't quite looked at,
37:53but it's a great point.
37:54Sorry,
37:55I saved time for that last question.
37:57Do we have
37:58one more question?
38:04Hi, first of all great presentation.
38:06I wanted to ask if you saw any difference
38:09in habituation to the routine between
38:11a non autistic and autistic youth.
38:15Yeah so the question is
38:18about habituation between
38:20between groups to their routine.
38:23So the short answer here is we can only
38:26measure so much and we debated heavily
38:28what we should put in to this study.
38:30And so we didn't really ask about
38:33habituation to change in routine.
38:35So in a sense, I think what we're
38:37looking at when we look at loneliness
38:38and we have some data that I didn't
38:41present today on anxiety and
38:42depression is kind of a proxy for that.
38:45But it's a great question and that's a good
38:48lesson learned for designing studies if.
38:51The change in routine happened differently
38:53and was quicker and perhaps mediate
38:56some of of of these relationships,
38:58but I'm out of time.
39:00So thank you for that.
39:02Thank you Dr. Gerber. Nice job.
39:07Last but not least we have Doctor Kistagno.
39:15Wait, Mike, is it just the next?
39:27Perfect. All right, all set.
39:31So thank you for this opportunity
39:33to share my research Today I'll
39:35present research recently published
39:36in our image entitled Modeling
39:38Brain dynamics and gaze Behavior.
39:40Starting point bias and drift rate relate
39:43to frontal midline Theta EEG oscillations.
39:47In this study we applied.
39:49Computational modeling to participants
39:51performance on the anti saccade task
39:53with eye tracking while collecting
39:55high density EEG to investigate the
39:57effects of trial by trial Theta dynamics
40:00on contingent eye gaze behavior.
40:01So I know that was a lot of words
40:04and I promise that a lot of them
40:06will make sense by the end.
40:07Important to start is that a
40:09saccade is just an eye movement.
40:11So. If you're moving your eyes,
40:13what you're looking at that is a saccade.
40:15If I point to one side of the room,
40:17everyone that looked to that side of the
40:19room, that would have been a saccade.
40:21Whereas if you didn't look,
40:22that would have been an antisychade.
40:24You inhibited your natural inclination
40:26to look to where I pointed.
40:28So that's a task we're dealing with,
40:29which I'll get into more in depth,
40:32just figured need to get that
40:34out of the way early.
40:35So why visual, visual attention?
40:38I gaze plays a critical role
40:40in many human behaviors.
40:41What grabs our attention grabs our thoughts
40:44from moral judgments to purchasing decisions.
40:48Another is in regard to clinical
40:50implications.
40:51Tension bias is well known play
40:52a role in the development and
40:54maintenance of anxiety disorders
40:56and depressed depressive disorders.
40:58A a critical aspect of adaptive
41:01goal directed behaviors,
41:02appropriate response preparation.
41:04This led to our motivating research question.
41:08Can we model effortful eye gaze
41:11behavior to improve precision
41:13when studying intentional biases?
41:15Fortunately for the field,
41:17there's a decent grasp on a specific
41:19neural marker of effortful control.
41:22Frontal and central midline Theta
41:24oscillations are robust domain general
41:27neural marker of cognitive control
41:29processes and therefore promising candidate.
41:31So what are oscillations?
41:33Just really quickly there are
41:35two main types of eg analysis.
41:37Typically people are familiar
41:38with ERP event related potentials,
41:40which are an average of a bunch
41:43of different waves.
41:44One of those waves is Theta,
41:45which occurs between roughly 4 and 8 Hertz.
41:49There are other frequencies here we're
41:52interested in Theta oscillations,
41:54and really what this is indicative
41:56of is a population of neurons
41:57that are firing together.
41:59So this is a neural signature that
42:02is thought to play important role.
42:04It increases in the magnitude
42:06in response errors,
42:08negative feedback to unexpected
42:12events during inhibitory control
42:15when resolving different.
42:16Competition between different
42:18responses and adjusting response
42:20strategies to our task demands,
42:25as well as following events that are
42:28novel or ambiguous after performance.
42:30The signals thought to reflect
42:31activity in the anterior,
42:33at least partially in the anterior
42:35singlet cortex and plays a central
42:37role in detecting when our
42:39expectations are being violated.
42:40So what we thought was going to happen,
42:42did not happen, is one way to think about it.
42:46Depending on the circumstances
42:47when this occurs,
42:48it can work to recruit.
42:57Oh,
43:02excuse me.
43:30We
43:34got it worked that way.
43:34You were here. No, wait.
43:37Yeah, right there. Yeah, yeah,
43:42sure. Is that working for
43:46them though? On Zoom. I
43:51have a spy. On Zoom
43:57we see purpose enter for you.
44:01They can okay,
44:05so we went through that.
44:07So some of the limitations of past
44:09studies of visual attention behavior.
44:11A button presses one step removed from
44:14the true behavior of interest here,
44:15which is simple attention
44:17or eye gaze behavior.
44:19Therefore we apply the drift diffusion
44:21model to participants eye gaze behavior.
44:24And I will get into what
44:25drift diffusion model is.
44:26But first we need to cover what the task is.
44:28The anti saccade task,
44:30which I briefly touched on in
44:31the beginning in the sense of
44:33that is the behavior of interest
44:35during the anti saccade task.
44:36It's a fastpaced inhibitory control
44:38task strictly driven by participants
44:40eye gaze behavior and that's a really
44:42important thing to remember here.
44:43There are no button presses,
44:45it's strictly where the participant
44:47is looking on the screen is driving
44:50the task paradigm during pro saccade.
44:52Participants receive a queue on
44:54screen either a white or black
44:56fixation cross during the pro
44:58saccade is a white fixation cross
45:00and that tells them I'll need to
45:02look at the upcoming probe.
45:07Next they'll see the probe and they will
45:09look in that direction hopefully and
45:12they'll receive feedback of correct.
45:14Now during an anti saccade they
45:16will receive a probe that is a black
45:20fixation cross indicating to them.
45:21I'll need to look away when
45:24I see the upcoming queue.
45:26When the queue comes,
45:27if they are engaging in the task correctly,
45:30they should inhibit their response
45:31to look at the white box and look
45:34away in the opposite direction of
45:36the screen of the box and therefore
45:39providing a anti saccade response.
45:43Now the important thing also
45:45to remember here.
45:46Apart from it being strictly
45:47driven by participants,
45:48eye gaze behavior is that it
45:50is acute anti saccade cast,
45:52which some people would call
45:54proactive cognitive control.
45:55In this sense, they know what's coming.
45:57They know that they're going to
45:59have to either inhibit A prepotent
46:01response or they're going to have
46:03to just provide the response that
46:05is their natural inclination,
46:06which is which is to look at the
46:09white screen in this very dark
46:11room on this computer screen now.
46:14Briefly,
46:14Introduction to a Drift Diffusion model.
46:17It's a broadly defined any model
46:19as a dynamic system.
46:21When presented with a time series,
46:22inputs such as reaction time and
46:26performance can produce simulation outputs.
46:29And drift diffusion models were
46:31specifically created in order to
46:33relate response times to underlying
46:35latent cognitive processes,
46:36which is the really important
46:38part to understand here is that we
46:40feed in the behavior of interest,
46:43in this case their sequential
46:44behavior on the anti saccade task,
46:46the reaction time, their performance.
46:49And what is generated is individual
46:51estimates of certain parameters.
46:53These parameters are latent constructs.
46:54They don't actually exist,
46:56but they're thought to relate to
46:59real world underlying cognitive
47:01processes that are a closer step
47:04towards what is going on in the
47:06brain than simple reaction time,
47:08which is an amalgamation of many,
47:09many, many cognitive processes.
47:12For the drift diffusion model,
47:14it parses it between drift rate,
47:16which is thought of as information
47:19processing.
47:19You can think of a drift rate as
47:21being an individual's subjective
47:23experience of task difficulty.
47:25So every individual in this task,
47:28once we feed in their behavior,
47:30response time and performance,
47:32we get an estimate of their
47:34specific drift rate during the task.
47:36And their drift rate estimate for
47:38an individual would be how difficult
47:40say they thought the pro saccade
47:43or the anti saccade trials were.
47:46How efficient they were at processing
47:48that and engaging in that task.
47:50There's also a threshold separation,
47:53which is the boundaries shown on the
47:56right there where the red lines are
47:59going and meeting in the star forms.
48:01That is the decision boundary.
48:03So once that boundary is reached,
48:04whatever boundary that is,
48:06that boundary is a decision that is made.
48:09And here the boundary,
48:11the top boundary is indicative of.
48:14Providing a pro saccade response,
48:16where is the bottom boundary
48:17is the anti saccade response,
48:19so they also have a bias or a starting point.
48:23So where in the middle of that?
48:26The decision threshold or
48:28the threshold separation?
48:29Where are they starting?
48:30Are they starting in the middle
48:32or do they have a bias where
48:33they need more information to
48:34gather to make one decision,
48:36much less to make the alternative decision?
48:40And finally, there is also
48:41a non decision time.
48:43I'm not going to get too much of the
48:45non decision time because of the
48:46amalgamation of a lot of cognitive
48:48processes that aren't related
48:49to the decision making process
48:51like early orientate orienting,
48:54early perceptual encoding and
48:55later processes that are non
48:58decision related such as the
49:00execution of a motor response.
49:03But let's walk through what this actually
49:05is so you have a better understanding
49:08cuz me giving you definitions
49:09is probably not going to do it.
49:12You have the decision threshold
49:14here for the anti saccade task that
49:16if the drift rate reaches this top
49:19boundary then they are going to
49:20produce a pro saccade or decide
49:22to produce a pro saccade response.
49:24And then you have a bottom
49:26decision threshold.
49:27If the drift rate reaches this threshold,
49:29they provide an anti saccade response.
49:32And you have a bias parameter or the
49:35starting point is what it's also known
49:37as and you can have a drift rate.
49:39So here's a blue drift rate
49:40indicating a pro saccade response.
49:42It's viewed as a noisy process
49:44which is beyond the scope of this,
49:46but that is why that is a jagged line.
49:49You'll often see jagged lines.
49:51They might also have a.
49:53This is a hypothetical anti saccade
49:56decision deciding to provide an
49:58anti saccade response so you can
50:00have a decision threshold.
50:02Like I said, top is a pro psychotic response,
50:04bottom is an anti psychotic response.
50:07You can also,
50:08so you can think about it as someone
50:10who has large decision thresholds.
50:13This would be an individual
50:14where the parameter estimates is
50:16larger than average for a group.
50:18You could think of them as having a
50:21conservative style of decision making,
50:23at least on this task.
50:24So they need much more evidence
50:26to come to any decision.
50:27They need a lot of information they
50:31are favoring. Accuracy over speed.
50:33There could be also people with
50:35more of an impulsive style where
50:38they favor speed over accuracy.
50:40You can imagine now they need
50:42much less evidence regardless of
50:44what decision they're going to
50:45come to to come to a decision.
50:49And now the bias parameter as well.
50:52It can do a little dance on the where
50:55determining where that starting point is.
50:57It can be high, it can be low.
51:00And altogether, this is hypothetical,
51:05several trials of the pro saccade or
51:09the anti saccade task for both pro and
51:12anti saccade conditions and just for
51:14to show what a drift rate where bias
51:18is shifted downward would look like.
51:20And this might be something to remember
51:23for when I talk about the results
51:25very shortly you see there's much
51:27more information that needs to be.
51:29Garnered to come to a prosychod
51:33response and alternatively much less
51:35information needs to be acquired
51:36to come to a antisychod response.
51:39This would be an individual with a strong
51:41bias towards the antisychod boundary,
51:44and you can see how that's different
51:46from the threshold separation where
51:48they generally for either decision
51:50are either conservative or impulsive
51:52in their decision making style.
51:58Now jumping into the results here,
52:01interestingly we found larger
52:02drift rate drift rates for
52:04the anti psychotic condition,
52:06which indicates that there was actually
52:08more efficient processing occurring
52:10during these high conflict trials,
52:11potentially reflecting a burst
52:13in frontal midline Theta that's
52:15not as strong in the Prosecco
52:19condition which I'll get into very shortly.
52:23There's also meaningful differences in the.
52:27Highest parameter as well.
52:30So specifically when cued of an upcoming
52:33challenge anti saccade condition,
52:35there tended to be a shift downward
52:38towards the anti saccade boundary.
52:40Therefore less evidence was required
52:44to provide that inhibitory response,
52:46but much more evidence was needed to
52:49incorrectly provide a pro saccade response.
52:52I think of this potentially as
52:55indicating A compensatory strategy to
52:58facilitate fast performance but accurate
53:00performance the more during the more
53:03difficult anti saccade condition.
53:05During the pro saccade condition,
53:07on the other hand,
53:08there was no there was a more
53:09neutral approach shown with the
53:11bias parameter estimate where equal
53:14amounts of evidence was needed.
53:16For either decision.
53:17So when they were cued of
53:19this upcoming challenge,
53:20they tended to have a shift
53:22downward in their bias,
53:23which gave them a buffer such that
53:25they could still respond accurately
53:28and quickly is what we are thinking
53:31might be underlying these group
53:32differences during the task from
53:34a drift diffusion framework.
53:35Now what about those neural
53:38oscillations we're talking about?
53:40Here are the head plots.
53:41I'm going to Orient you to the
53:43grand average in the bottom here.
53:45On the left in red is the anti
53:47saccade and on the in blue on
53:49the right is the pro saccade.
53:51You can see there's a pretty routine
53:54and reliable neural response to
53:56both pro and anti saccade response,
53:59but the difference can be shown much.
54:03It becomes much more salient in the
54:05time series output here where I'll
54:07Orient you to the delay period.
54:08So this is the period after they're
54:10told they're going to need to either
54:12provide a pro or anti saccade response
54:14to Remember that white or black
54:16fixation cross so they know what's
54:18coming during that short delay period.
54:20Before they see the white probe,
54:22there is a stronger burst
54:23of frontal midline Theta,
54:25remember that is indicating that
54:27expectations might be violated.
54:29You might need to get the right,
54:32get the cavalry to.
54:34Help with this upcoming challenge since
54:36they were cued that this upcoming challenge,
54:39the anti saccade shown in orange
54:41there tended to be a larger
54:43burst of frontal midline Theta.
54:44So what about all that talk of trial by
54:47trial changes in frontal midline Theta?
54:50So when taking the behavioral neural
54:53physiological findings together,
54:54the drift drift diffusion model input
54:57includes participants trial by trial,
54:59reaction time, response,
55:01empower or strength of their event,
55:03locked Theta neural response
55:05during each task queue.
55:07So within the model is an estimate
55:09of their the specific participants
55:12Theta during that response queue.
55:16They're in that queue where I
55:18showed you between after the queue
55:20and prior to receiving the probe.
55:23Put differently,
55:24we examined the within subject
55:26effects of this trial by trial frontal
55:28midline Theta on drift rate and bias
55:30those two parameters that were found
55:32to differ in their performance.
55:34And allowing for different levels
55:35of difficulty,
55:36so pro and anti saccade to
55:38exert influence via drift
55:40diffusion regression model.
55:41This allowed us to directly examine
55:44eye gaze behavior and trial by trial
55:47changes in frontal midline Theta
55:49within an individual model together
55:52within subject in a Bayesian space.
55:54And this allowed us to to
55:57directly examine where these
55:59changes in frontal midline Theta.
56:01Over the course of tasks has a significant
56:04influence on the drift rate and bias.
56:09And finally these were the
56:11results of the trial by trial
56:13effects of frontal midline Theta.
56:14Here these are posterior distribution.
56:16So I oriented you to zero
56:18there with that line.
56:20And the important part here is
56:21if a posterior distribution in
56:22this context passes through zero,
56:24then is not a meaningful.
56:27Effect here for both pro and anti
56:29saccade shown in the blue and the red.
56:31You can see there was a positive
56:34effect of frontal midline Theta on
56:36pro during pro and anti saccade
56:38conditions with an individual which
56:40shows that which is consistent with
56:42those head plots you saw before
56:44because there were first the frontal
56:46midline Theta during both conditions.
56:47Although the time series input did
56:49show that they were stronger during
56:51the anti saccade condition however.
56:54Being probed that there was an upcoming task,
56:58a challenge,
56:58something to do look at the probe
57:01or look away elicited frontal
57:03midline Theta and both of those
57:06increased individuals processing
57:07efficiency during the upcoming demand.
57:10Now interestingly the bias parameter here.
57:14You can see the prosychot directly
57:16passes through zero,
57:17so there's no effect of frontal
57:18midline Theta within an individual
57:20on their prosychot response.
57:22So during the prosychot trials,
57:23there was no effect of frontal midline Theta.
57:26Very interestingly though,
57:27there was an effect, a negative effect,
57:30on the antisychotic condition which
57:32relates to that shift downward
57:35in that bias parameter.
57:37That shift downward,
57:38which I showed in that schematic earlier,
57:40is what's going on here.
57:42Where?
57:43These results indicate that that
57:45burst of frontal midline Theta during
57:47that anti psychotic condition not
57:49only increased processing efficiency
57:51via the drift rate but also shifted
57:54that bias parameter downward on that.
58:00Allowing their starting point
58:02bias to be shifted downward.
58:03Therefore, they need much more
58:05evidence to accumulate to erroneously
58:07provide a pro saccade response,
58:09but much less information need to
58:11accumulate to provide correctly
58:13the anti saccade response,
58:14if you remember,
58:15is that bottom threshold.
58:20Finally, we're also interested in
58:23potentially showing the utility of using.
58:26Computational modeling to
58:27decompose task based behavior.
58:29So we included reaction time in the
58:31first block which was not significant.
58:33In the second block we introduced
58:35the drift diffusion parameters.
58:36Bias was a significant predictor.
58:40Drift rate was not in this case,
58:41but in subsequent regressions where we
58:44weren't interested in showing the utility,
58:46but just examining whether drift
58:48rate and bias predicted frontal
58:49midline Theta during the task.
58:51Both of those were predictors with
58:54significant predictors without reaction time.
58:57In the in the model and the
59:00overall variance explained was
59:04fairly robust. Finally the take
59:08home here increased Theta power was
59:11associated with increased processing
59:12efficiency and a shift in starting
59:15point bias which facilitated accurate
59:16and fat but fast responding and finally
59:20modeling proactive cognitive control.
59:24At the level of eye gaze from a
59:26drift eye gaze, behavior from a
59:28drift diffusion framework improved
59:29our measurement precision,
59:32as shown through our regression analyses.
59:40And oh, there it is.
59:42And for acknowledgments,
59:43I'd like to thank Courage Lab and
59:45our members and Doctor Crowley,
59:47my mentor, as well as my other cowork
59:51coauthors on the on the paper,
59:53Stefan and Purr, as well as my
59:57funding the F32 as well as the T32.
59:59And Doctor Block who Co
01:00:02runs the T32 with Mike.
01:00:03So thank you.
01:00:10Thank you. Nice job, Peter.
01:00:11Sorry for the technical snafu.
01:00:12No worries. We have time
01:00:14for one question for Peter.
01:00:17Come
01:00:23on, there's gotta be a computational model
01:00:25and person in the crowd. There's Taylor.
01:00:33I wanted to go back to
01:00:34this one to show this is.
01:00:36I made this slide to show kind of
01:00:38what that effect was hypothesized
01:00:40for that effect of frontal midline
01:00:42Theta on anti sacod conditions,
01:00:44what that look like and that is kind of
01:00:47what that shift downward would look like.
01:00:49If anyone's interested,
01:00:50I wanted to go back to it.
01:00:51But right now I have a question, Peter.
01:00:54So where can we take this research
01:00:56studying anxiety for instance?
01:00:58Yeah, so I think I've thought a lot about
01:01:02using attentional biases to threat.
01:01:04And oftentimes we'll use a dot pro task or
01:01:08pretty much any kind of task we use really.
01:01:11We're inferring where their
01:01:13attention is via button presses.
01:01:15And I think it'd be it shows
01:01:18that we can use the Drift Drift,
01:01:19diffusion modeling framework to
01:01:22decompose gaze behavior into these
01:01:24late and underlying constructs which
01:01:27may allow us to better relate to.
01:01:31Neural dynamics, whether it be frontal,
01:01:33midline, Theta, A joint model as
01:01:36seen here can also be applied to FM,
01:01:40RI through bold response.
01:01:42It doesn't need to be necessarily EE,
01:01:45G or Austory dynamics,
01:01:47but what's really important with
01:01:49this type of modeling is having
01:01:52that trial by trial changes and.
01:01:54Obviously,
01:01:55the temporal specificity of veg
01:01:56lends itself very nicely to a
01:01:59computational modeling approach
01:02:00to something like this because
01:02:02of that temporal specificity as
01:02:03opposed to a bold response.
01:02:04But there are ways to kind of lag
01:02:06that so that it matches up with the
01:02:09behavior which is kind of interesting.
01:02:11So I think using this to study
01:02:13attention biases with with eye
01:02:15tracking is is something that's
01:02:17really cool and in the future.
01:02:20Thank you very much.
01:02:21Thank you for coming everyone.