Thesis Abstracts, Class of 2021

Ashley Cooper (Neuroscience), Racism’s Health Harm on Black Youth Mental Health: From (Neuro)Scientific Orthodoxy to Neuroscience as a Vessel of Visibility (advisors Marisa Silveri, HMS & Byron Good, Anthropology/FAS) – Perceived Racial Discrimination has been characterized as a chronic stressor in the lives of Black youth, and heavily associated with the development of poor mental health outcomes. Early Life Stress has been epidemiologically linked as one of the largest risk factors for developing adolescent depression, catalyzing a disruption of neural pathways necessary for proper cognitive function. This research seeks to expand the lexicon of Early Life Stress to incorporate how Perceived Discrimination may too impact neurodevelopmental trajectories. To this end, this study, analyzes data from the landmark Adolescent Brain Cognitive Development (ABCD) study (n = 9387) in order to characterize: (1) racial disparities in perceived discrimination and mental health symptomatology outcomes and (2) how experiences of perceived discrimination impact resting-state functional connectivity of large-scale networks. This study is, to our knowledge, the first-ever focused on the association between functional network connectivity and perceived racial discrimination in Black adolescents. We found increased prevalence of mental health symptomatology and perceived discrimination amongst Black youth, as compared to white youth. Further, we found main effects of race on rs-FC in all three of our network pairs of interest (Default Mode Network-Default Mode Network (DMN), Default Mode Network-Dorsal Attention Network (DAN), and Default Mode Network-Salience Network (SAL)). Finally, we found DMN-DMN within-network hypoconnectivity in youth reporting perceived discrimination and depressive mood mental health symptomatology. Our results, while exploratory, point to the potential of the ABCD study to longitudinally examine how perceived discrimination impacts neural developmental trajectories throughout the duration of adolescence and critical periods. Our research allows serves as a necessary catalyst, evidencing that future studies should further characterize how perceived racial discrimination, over time, may alter resting-state connectivity—specifically within the DMN—or depression outcomes in Black youth.

Emmanuel Garrison-Hooks (Neuroscience), Mouse Foraging Behavior in a Dynamic Virtual Environment (advisor Naoshige Uchida, Molecular and Cellular Biology/FAS) – As animals navigate through their environments, they must determine when abandoning a cluster of resources to search for another is better than remaining. Charnov’s Marginal Value Theorem offers a way to characterize animal foraging behavior, predicting that animals should leave a depleted ‘patch’ of resources as soon as the expected value of remaining on that given patch is exceeded by the average expected reward value of the environment. Using a virtual reality patch-foraging paradigm, wherein mice are head-fixed as they run on a 1-D track, we assessed how mice behavior is affected by different patch values by randomly varying the size and frequency of rewards across patches. During the task, the mice are presented with visual cues that signal that a patch is available. If they stop, water rewards are delivered probabilistically, with that probability monotonically decreasing over time to mimic the depletion of resources that is typical of natural environments as the mouse forages. Furthermore, we were interested in whether the mice could adapt their behavior based on net value of the environment by varying whether the mice could either receive only Small and Medium rewards or Medium and Large rewards across trial blocks. We found that the mice behavior was modulated by the quality of their environment, with longer patch residency times on higher-value patches in the high-value environment than in the low-value environment. Ultimately, we found cases where the mice remained on Medium value patches longer in the high-value environment but left sooner in the low-value environment.

Jonathan Garzon (Neuroscience), Utilizing a novel naturalistic paradigm to systematically assess social need in WT and ASD mice (advisor Catherine Dulac, Molecular and Cellular Biology/FAS) – Social need motivates animals to seek social interactions and maintain social bonds. By contrast, social isolation induces a negative emotional state (loneliness) that enhances the drive for social interactions, leading in turn to an observable rebound in social investigation. Social need is disrupted in autism spectrum disorders (ASD), and children with ASD display impaired orientation to social stimuli and reduced response to social reward. Therefore, a better understanding of neural mechanisms underlying social need should provide significant insights into the basic mechanisms underlying ASD pathology and suggest new avenues for treatment. By using a novel naturalistic behavioral paradigm, I systematically measured the frequencies and durations of bouts of various social interactions in two wild-type (WT) mouse strains FVB/NJ and C57BL/6J, and two autism spectrum disorder (ASD) model mice Shank3b-/- and CNTNAP2-/- following a short period of social isolation. Both WT strains exhibit significant social rebound after isolation when compared to their baselines. Specifically, FVB/NJ mice showed stronger social rebound than C57BL/6J mice. Using the elevated plus maze and open field tests, I found no difference in stress levels between isolated and socially housed WT mice, which indicates that social rebound is not a result of isolation stress. In comparison to the WT mice in the same strain background (C57BL/6J), ASD model mice showed abnormal social behaviors during social reunion: CNTNAP2 mutants showed little rebound while Shank3b mutants showed hyperactivity in specific modules of social behaviors. To further reveal the behavioral details in ASD mice, I have trained a deep-learning network via DeepLabCut and have analyzed the body pose features in two interacting mice across video frames. Overall, my study uncovered unique social behavior patterns in both WT and ASD model mice following social isolation. This study provides a fruitful behavior reservoir for future mechanistic studies of the nature and function of neuronal circuits associated with basic social drive.

Debolina Ghosh (Neuroscience), Investigating the Role of Klf9 in the Division Pattern and Differentiation of Neural Stem Cells in the Adult Hippocampus (advisor Amar Sahay, HMS) – Adult hippocampal neurogenesis is maintained by the limited activation and proliferation of neural stem cells. Upon activation, radial-glial like neural stem cells (RGLs) symmetrically divide into more RGLs or asymmetrically differentiate into astrocytes or neural progenitors. A balance between quiescence and activation as well as the two modes of RGL division is crucial for maintaining hippocampal homeostasis, maturation, and response to pathogenesis in adulthood. Kruppel-like factor 9 (Klf9) is a transcription factor hypothesized to play a critical role in maintaining RGL quiescence and determining cell fate. This thesis investigates the potential regulatory mechanisms of Klf9 on RGL division and differentiation by characterizing adult-born hippocampal cell populations in Klf9f/f versus Klf9+/+ mice. Population analysis in Gli1-CreERT2 and Ascl1-CreERT2 mice revealed significantly higher proportions of RGLs in postnatal and young adult Klf9f/f mice after 1-4 weeks of lineage tracing. Therefore, Klf9 may be important for maintaining RGL quiescence and preventing symmetric expansion at these time points. Genetic profiling of young adult Klf9f/f mice confirms the upregulation of genes associated with stem cell activation and expansion as well as downregulation of genes associated with stem cell quiescence. Interestingly, population analysis of mature adult mice failed to show differences between Klf9f/f and Klf9+/+ mice, suggesting that the regulatory role of Klf9 in adult neurogenesis may diminish in the aging niche. Further investigation of Klf9 expression as well as clonal analysis and 2-photon imaging is needed to confirm the role of Klf9 in maintaining postnatal and adult RGL homeostasis.

Joshua Glauser (Neuroscience), Neural Measures of Social Attention as an Indicator of Communication and Social Development in Infants at-risk for Autism Spectrum Disorder (advisor Carol Wilkinson, HMS) – Identifying neural markers of autism spectrum disorder (ASD) before they clinically emerge can improve outcomes through early treatment. This study aimed to characterize the neural correlates of face recognition in 12-months-old infants at familial risk of developing ASD. The study’s objectives were (1) to compare face-sensitive event-related potentials (ERP) (Nc, N290, P400) between high-familial-risk infants who develop ASD (HR-ASD), high-familial-risk infants without ASD (HR-NoASD), and low-familial-risk controls (LRC), and (2) to determine how face-sensitive ERP components correlate with the core ASD clinical symptoms of communication and social development. 12-month old infants participated in the mother/stranger paradigm (Nc n = 102, N290/P400 n = 64), with EEG data collected as infants observed pictures of their mother and a similarly-looking stranger. Parent-reported and laboratory-observed communication measures were recorded at 12 months, and laboratory-observed social measures were conducted at 18 months. Multiple linear regressions were conducted with maternal education and outcome groups as covariates. For each of the Nc, N290, and P400 analyses, the amplitude difference between mother and stranger (Mother-Stranger) trials was not statistically different between the three outcome groups (Nc p = 0.72, N290 p = 0.88, P400 p = 0.91). Controlling for maternal education and outcome group, a significant association was observed between expressive communication measures and the Nc Mother-Stranger (R2 = 0.11, p = 0.002) and between receptive communication measures and the P400 Mother-Stranger (R2 = 0.35, p = 0.005); Larger P400 Mother-Stranger differences were also associated with higher social development scores (R2 = 0.17, p = 0.04).

Daniella Gomez-Ochoa (Neuroscience), The effect of parental monitoring on epigenetic age and CRP levels among young adults in the ALSPAC cohort (advisors Erin Dunn, HMS & Alexandre Lussier, HMS) – Parental monitoring may be a protective factor against depression, potentially becoming biologically embedded through epigenetic marks and inflammatory factors such as C-reactive protein (CRP). However, it remains unknown whether epigenetic age and CRP influence the relationship between parental support and depressive symptoms. This study aimed to determine the extent to which epigenetic age and CRP levels in peripheral blood during adolescence can moderate the impact of parental monitoring on adult depressive symptoms. The analyzed data came from a subset of the ALSPAC cohort, a longitudinal study that has collected multiple variables of parental monitoring, depressive symptoms, and epigenetic data across development. To capture parental monitoring, I created a cumulative parental monitoring score that included five parental monitoring qualities measured at age 12.5 years. Furthermore, epigenetic age was measured from peripheral blood DNA methylation at ages 15-17 years using the Horvath and Hannum epigenetic clocks (n=966). CRP was measured from blood at age 15.5 years (n=720). Depressive symptoms were measured using the Short Mood and Feelings Questionnaire (SMFQ) at age 21 (n=455). This project tested three hypotheses. First, I assessed the effects of the cumulative parental monitoring score on epigenetic age, CRP concentration, and depressive symptoms. Second, I evaluated the biological influences on depressive symptoms. Finally, I gauged the moderation of the effects of the parental monitoring score on SMFQ by molecular outcomes. Overall, this research suggests that neither epigenetic age nor CRP concentration levels moderate the effects of parental monitoring on depressive symptoms as measured by SMFQ score. Considering the limitations of sample size in genomics research, future studies should consider performing imputation on missing variables to prevent a loss in power.

Catherine Ho (Neuroscience MBB), Aging, Episodic Retrieval, and Metacognition during Everyday Problem Solving (advisor Daniel Schacter, Psychology/FAS) – When faced with a difficult problem, people often rely on past experiences. Memory impairments, whether due to aging or hippocampal lesions, impair problem-solving abilities. On the other hand, brief training in recollecting the details of a recent event helps people to solve social means-end and personally worrisome problems. While episodic retrieval clearly benefits performance, feelings of ease during recall may also leave people feeling more prepared for problems than they really are. My thesis tests whether this illusion occurs in young and older adults, since aging increases the effort associated with retrieval and may mitigate overconfidence. Young and older participants learned tips for “worst case scenarios” (e.g., shark attack). Later, they listed steps to solve these problems (retrieval), as well as new ones (generation), and indicated how prepared they felt for each scenario. With a retrieval focus, young adults not only provided higher-quality solutions, but also felt more prepared for serious problems. Better performance in the retrieval condition did not fully explain young adults’ increased preparedness ratings. Older adults, on the other hand, performed similarly in both conditions and felt equally prepared when remembering old solutions and generating their own. These results suggest that remembering only misleads us to overestimate our abilities during problem solving when retrieval feels easy, or fluent.

Helen Huang (Computer Science), The Net Work of Networks: The Effect of Immersion on Analyzing Network Diagrams (advisor Hanspeter Pfister, Computer Science/SEAS/FAS) – Networks are powerful visualizations that can represent complex systems of relationships in datasets. However, as datasets have grown in size and complexity across a wide range of academic and professional fields, so have the node-link diagrams used to model and visualize them. Analyzing more complicated networks poses an issue in traditional screen and mouse environments due to the limitations of these 2D visualization and interaction methods. Immersive Analytics (IA) has recently emerged to address these issue by leveraging engaging analysis tools to support data understanding. Leveraging key ideas from IA, we report the findings of the first systematic study on how immersion can impact a user’s ability to complete network diagram tasks. We determined three fundamental graph tasks according to widely-used task taxonomy: (1) Find the common nodes between two highlighted nodes, (2) Count the number of triangles in a graph, and (3) Find the missing nodes between side-by-side graphs. We asked users to complete each task in four different conditions of increasing immersion (Fig. 1). Given the importance of determining a device’s realistic usability for data analysis, we also studied user workload throughout each condition to add additional depth to our findings. Results showed that while 2D Mouse and 3D Mouse had the fastest task completion times, more immersive conditions like 3D Trackball and 3D VR had better accuracy. 3D VR required the lowest amount of workload of all conditions in Common and Count, and post-study rankings revealed that 3D VR and 3D Trackball were seen by users as the most effective for node-link diagram tasks.

Stacy Jo (Human Evolutionary Biology), Voluntary pitch control: An experimental study of pitch matching in domestic dog howling (advisor Erin Hecht, Human Evolutionary Biology/FAS) – Human singing involves an ability that is rare in the animal kingdom: individuals are able coordinate their pitch patterns in relation to those produced by other voices or instruments. Yet this ability may not be uniquely human; wolves and dogs may possess a similar skill when howling with conspecifics. Whether there is a convergently evolved capacity for voluntary control over vocal pitch in humans and domestic dogs is an issue that can be examined by observing how dogs howl to specific sound stimuli which trigger howling (such as music); if a dog howls with a certain mean fundamental frequency (F0) when hearing a musical piece, one can test whether the dog shifts the mean F0 when hearing a pitch-shifted version of that music. Such a shift would lead to the idea that a dog can regulate the pitch of its howls with respect to the pitches of the music. A quantitative study was conducted using video and audio clips of 9 dogs howling to musical stimuli which triggered howling. By isolating the F0 contours of each dog's howls while hearing its howl-triggering stimulus, it was investigated whether there were changes in mean howl F0 when dogs heard versions of the same music which had been shifted up or down in pitch (upshifted vs. downshifted). The results of the study show that 3 of 9 dogs show significant differences in the mean F0 of their howls when hearing upshifted vs. downshifted stimuli. Alfie and Checkers (2 out of the 3 dogs whose mean F0 differences were significant) howled, respectively, 1.4 and 3.5 semitones higher in their upshifted trials compared to their downshifted trials while Bug-Z (the last of the 3) howled 5.6 semitones lower in his upshifted trials compared to his downshifted trials (which might be explained by a habituation hypothesis). The results conform to the hypothesis that motivated this study, namely that dogs have some incipient pitch matching ability. Such a finding may represent an instance of convergent evolution for complex social communication skills in dogs and humans through pitch-matching.

Emily Johns (Neuroscience), An Analysis of the Neural Correlates of Psychosis: From Antiquity to Modern Research (advisors Martha Shenton, Psychology/FAS & Mark Schiefsky, Classics/FAS) – Understanding the relationship between neural aberration and severity of hallucinations and delusions using modern neuroimaging techniques, as well as analysis of ancient medical documents is an important area of study. In this study, we aimed to characterize variation in cortical thickness of language processing brain regions as a function of patient status to determine if hallucinations and delusions have a central neural basis. We also aimed to analyze the medical text, On The Differentiae of Symptoms, written by Galen (129-216 CE) to understand how hallucinations and delusions were understood in antiquity. From a close textual analysis, we observed that hallucinations and delusions were characterized by Galen as distinct presentations connected by a central region of mental (brain) malfunction. Investigating this premise in a modern neuroscientific context, we examined subjects who participated in the Human Connectome Project for Early Psychosis, which included 157 individuals suffering from psychosis and 72 comparable healthy individuals. Structural differences in cortical thickness in three brain regions of interest (ROI), the left frontal, dorsal, and temporal gyri, were evaluated using analysis of covariances and post-hoc Tukey’s contrasts. We also observed a significant difference in cortical thickness in all ROI when comparing controls, symptomatic patients (i.e., hallucinations and/or delusions present), and asymptomatic patients (i.e., no hallucinations or delusions present). Of note, cortical thickness was reduced in symptomatic patients compared to controls, with asymptomatic patients intermediate in all ROI. These findings suggest that these language regions play an important role in patients suffering from hallucinations and/or delusions.

Ahan Malhotra (Computer Science), Improving API Reference Documentation: A Cognitivist Approach to Design and Evaluation (advisor Elena Glassman, Computer Science/SEAS/FAS) – Using API reference documentation is a crucial task for developers, especially when working with less popular programming languages and libraries. Although such interfaces are ubiquitous, there has been limited research into their effectiveness. The role of working memory in the comprehension of complex information at scale is also not well understood. Here, we introduce an interactive interface for viewing and exploring reference documentation from a cognitive science perspective. Our interface features a tabular layout with type sorting and search features, allowing users to see different individual functions more efficiently. We implemented this documentation interface for the functional programming language Ocaml and performed a within-subjects study against existing OCaml documentation. We investigated the impact of working memory on the performance of these interfaces. While users prefer our novel interface qualitatively, there are only marginal improvements in function lookup time and understanding. Further, we show that working memory is a strong predictor of function lookup time across both interfaces, suggesting that lookup times may increase during high cognitive load situations. Our results highlight the importance of leveraging cognitive science insights in future human-computer interaction research.

Kira Medish (History and Science), What Lies Behind Our Eyes: Technological, Epistemic and Ethical Challenges of Lie Detection (advisor Vi Le, History of Science/FAS) – Mind-reading is a trope in science fiction and a topic of scientific inquiry. The notion of lie detection is predicated on the possibility of decoding an individual's cognitive states — their hidden minds — using tools of physiological measurement. The reliability of these techniques, such as the polygraph, is far from validated. In the past 15 years, new digital technologies, such as functional magnetic resonance imagining (fMRI) and high-precision eye scanning, have been applied to lie detection. The field is thus moving from low-tech methods to high-tech digital biometrics attempting to pierce the veil of the human mind. This study finds that, despite confident claims of technical progress, abiding epistemic and ethical concerns continue to challenge the field. Bio-medical technologies such as those used for lie detection are inherently multi-use and morally ambiguous. Lie detection demands close epistemic and ethical scrutiny of its uses and abuses from the standpoints of scientific validity and democratic norms.

Claire Millett (Neuroscience), Calcium Signaling in Schizophrenia: Exploring the Role of CACNA1C in Human Synaptogenesis with High-Throughput Imaging (advisor Kevin Eggan, Stem Cell and Regenerative Biology/FAS) – Schizophrenia is a crippling neuropsychiatric disorder whose cause—which is in large part genetic—has remained elusive for centuries. However, recent studies suggest alterations in neuronal connectivity, potentially caused by decreased numbers of synapses, could underlie key schizophrenic symptoms. CACNA1C, a gene which encodes for a subunit of a calcium channel found on neurons, is both commonly mutated in patients with schizophrenia and is responsible for regulating a variety of processes important in synaptic growth. Although CACNA1C is widely recognized to be an important locus for schizophrenia research, there are currently few studies of CACNA1C which utilize human neurons to study how this gene impacts synapses. Therefore, this research creates a new methodology for analyzing the development of iPSC-derived human neurons. This method combines immunohistochemistry with machine learning to produce accurate and objective measures of neuronal development and synaptogenesis. After editing neural progenitor cells in the CACNA1C gene using lentiviral CRISPR constructs, this new technique was used to determine the effect of CACNA1C mutations on human neuronal growth. We found that mutations in CACNA1Ccausehighly significant reductions in synaptic density(numbers of synapses per unit area) of cortical neurons .This not only further defines the role of CACNA1C in schizophrenia, but also supports the idea that severe alterations in synaptic connectivity could be the cause of schizophrenia.

Kevin Ogonuwe (Neuroscience), The Marginalized Mind in Crisis: Racism-Related Adversity Alters Neurobehavioral Response to Threat and Portends Risk of Internalizing Psychopathology (advisors Kate McLaughlin, Psychology/FAS & Alexander Rodman, Psychology/FAS) – Racism is a pervasive aspect of our society that actively harms people of color in a number of complex and diffuse ways. From a biopsychosocial perspective, chronic racism-related stress has been found to contribute to the development of negative physical and mental health outcomes in people of color, though the underlying mechanisms are not clearly understood. To that end, a nascent body of research has sought to ascertain how exposure to discrimination impacts children of color. Within this thesis, we sought to extend the literature in our consideration of exposure to racial discrimination as a form of early-life adversity. The objective of the present work was to investigate whether discrimination resulted in neurobehavioral threat responses that mediated associations between discrimination and psychopathology in children of color. In this sense, we aimed to characterize a neuropsychological process whereby racism-related adversity has developmental consequences that may contribute to the cumulative wear-and-tear effects of racism-related stress. We examined this neuropsychological model using behavioral and neurobiological measures drawn from two complementary datasets: one of which consisted of a sample of 158 youthaged8 –17, while the other was a nationally representative sample of 4,738 youth aged 10 –13from the Adolescent Brain Cognitive Development study. We confirmed that exposure to discrimination was associated with increased psychopathology severity. We also observed that reduced neural threat responses in the dorsal anterior cingulate cortex(dACC) and anterior insula were associated with increased depression severity. Moreover, we found that reduced neural threat response in the right dACC mediated depressive severity in non-White youth. These preliminary findings validate the neuropsychological implications of racism-related stress during child development, supporting the need for further research within this domain.

Ifedayo Omotunde (Neuroscience), Effect of SRGAP2 Knockout on Temporal Properties of Critical Periods in Visual, Auditory, and Anterior Cingulate Cortices (advisor Takao Hensch, Molecular and Cellular Biology/FAS) – Human brain development lasts relatively longer than other animals. However, the mechanisms governing the trajectory of brain development in humans remain unknown. An intriguing candidate is SRGAP2, a gene uniquely duplicated in humans implicated in regulating the maturation of excitatory/inhibitory balance. Deletion of SRGAP2, or overexpression of the human-specific SRGAP2C which inhibits SRGAP2, delays E/I maturation. Evidence indicates that E/I balance is linked to critical period timing suggesting that SRGAP2 may regulate critical period plasticity and circuit maturation timing. We can use animal deletion of SRGAP2 to model SRGAP2C’s inhibition. We predicted that SRGAP2 knockout animals would display characteristics of delayed critical periods. We performed immunohistochemistry tests on the visual, auditory, and anterior cingulate cortices to get a broad view of SRGAP2’s role in circuit maturation throughout the brain and across different critical period trajectories. We used immunohistology to study parvalbumin and perineuronal net expression in Het knockouts of SRGAP2. Our immunohistology found mice with a heterozygous knockout of SRGAP2 broadly displayed lower PV intensity and PV/WFA colocalization (representing perineuronal net formation), pointing to effective delay of critical periods. We then performed an optomotor task to test the visual behavioral response to SRGAP2 knockout. The optomotor task confirmed delayed visual development in Het mice. We concluded that SRGAP2 knockouts displayed delayed critical periods and had more juvenile visual behavior. These results help us answer essential questions about human evolution and development, and open avenues to new therapies such as life-long learning, treatments for neurodevelopmental disorders, and rescue of lost sensations.

Paulina Piwowarczyk (Neuroscience), Predictability and detectability affect processing of syntactic category violations: An EEG study (advisor Jesse Snedeker, Psychology/FAS) – In language comprehension, listeners can use information to make predictions about the words and sentence structures that they are about to hear. This thesis asks specifically what happens when predictions about sentence structure are violated (e.g. “It takes two the tango”). We used electroencephalography (EEG)to record participants’ neural responses to structural errors (specifically syntactic category violations)that were spliced into a recording of a children’s story. We reasoned that if listeners had stronger expectations about upcoming words, there would be faster and/or more robust neural responses to the errors. Thus, we manipulated the predictability of the environment in which the violations occurred (high or low). We also investigated how detectable our violation manipulations were. Given the prior literature, we expected to find an Early Left Anterior Negativity (ELAN) in highly predictable environments, reflecting a rapid error response and increased processing difficulties associated with the errors. We also predicted a late positive-going response (P600) regardless of predictability, reflecting the slower process of reanalyzing the violations. However, we did not find either response; we instead found a sustained negativity, which has been argued to reflect difficulties in integrating information during comprehension. Moreover, this sustained negativity only appeared when the errors were highly noticeable and occurred in place of highly predictable words. Taken together, these findings suggest that, in naturalistic contexts, only errors that are detectable and that violate strong expectations lead to increased comprehension difficulties; all other errors may be disregarded or simply not recognized when listening to an engaging story.

Nivedita Ravi, Neuroscience), The Modern Infant Listening Subject: Examining White Matter Organization, Music Environment, and Listening in Early Childhood (advisors Nadine Gaab, GSE & Suzannah Clark, Music/FAS) – Longitudinal studies offer evidence for musical training-induced white matter plasticity in school-age children. The literature also suggests that putative neural predispositions may engender neural substrates for musical perception prior to formal musical training. The present research is three-fold: Study I examines relationships between music exposure and white matter organization in infancy, Study II explores impacts of the COVID-19 pandemic on childhood interactions with music, and the interdisciplinary final chapter imparts a humanistic approach to characterize the ​modern infant listening subject​. Study I used diffusion tensor imaging and Automated Fiber Quantification to estimate fractional anisotropy of key white matter tracts in 22infants (ages 2-13 months). The amount of music exposure was quantified using a parental questionnaire assessing infants’ music environment. Controlling for age and socioeconomic status, findings suggest that relationships between white matter and music exposure are evident within the first year of life. In Study II, the COVID-HELP questionnaire was developed and administered to analyze the impact of the pandemic on music exposure and engagement in 1568children (ages 0-11 years). During the pandemic, children are exposed to a richer home music environment but engage with fewer musical learning experiences when compared to pre-pandemic conditions. Findings carry implications for the hypothesis that neural predispositions for musical training are established prior to training onset and emphasize the importance of future research that considers the impact of dynamic home environments during times of crisis. Ultimately, the present research advocates for the integration of scientific and humanistic approaches to interdisciplinary inquiry.

Aba Sam (Neuroscience), The Fever Effect: Behavioral Changes Induced by Fever in Autistic Mouse Model CNTNAP2-/- (advisor Catherine Dulac, Molecular and Cellular Biology/FAS) – Autism is a disorder characterized by deficits in social communication, restricted repetitive patterns of behavior, and language impairment; its etiology remains largely unknown. The fever effect is a phenomenon in which subsets of children with autism experience temporary abatement of aberrant behaviors while febrile. Patients improve in cognition, communication, repetitive behaviors and social interactions during fever. The fever effect offers an opportunity to investigate the neural circuits implicated in autism, provided that one can identify model organisms that exhibit analogous changes in behavior while febrile. In order to explore potential model mice, my project has attempted to characterize changes in behavior caused by fever in Contactin Associated Protein-like 2 knockout autism model mice (CNTNAP2-/-). Previous investigation from the Dulac lab found that CNTNAP2-/- mice showed a significant increase in time spent in close proximity with other mice during fever. However, a precise characterization of the observed behavior was not performed. Thus, we sought to identify the specific changes in social behavior during febrile episodes. Specifically, we examined sniffing, allogrooming, fighting and huddling behavior and hypothesized that fever would cause an increase in sniffing and huddling behavior. To test this, we analyzed behavioral changes in CNTNAP2-/- mice during social interactions with either sick or healthy conspecific mice. Our results indicate that fever causes an increase in huddling behavior in CNTNAP2-/- mice, specifically when interacting with other febrile mice. This thesis identifies the differences in behavior caused by fever in autism model mice, which may help better understand how fever may affect the behavior of children with autism spectrum disorder.

Taylor Shirtliff-Hinds (Neuroscience), Troublesome Teenagers: Developing an Experimental Paradigm to Quantify Socially Mediated Alcohol Consumption in Juvenile Mus musculus (advisor Catherine Dulac, Molecular and Cellular Biology/FAS) – The common laboratory mouse, Mus musculus, exhibits a myriad of social behaviors that provide unique insight into the complexities of the mammalian social brain. Interestingly,juvenile mice display increased alcohol consumption when interacting with peers, similar to what has been observed in human adolescents. However, the neural circuitry underlying the socially mediated increase in juvenile mice alcohol consumption has not yet been established. Based on previous studies on reward circuitry in the mouse brain, we hypothesize that interactions between the prefrontal cortex, which regulates impulsivity; the ventral tegmental area, which encodes reward prediction error and social reward; and the nucleus accumbens, which encodes alcohol and social rewards, modulate this risk-taking behavior. Probing this pathway first requires establishing an experimental paradigm to quantify alcohol consumption in juvenile and adult mice in the presence or absence of peers. As such, we tested a custom-built drinking setup where mice are presented with four bottles of liquid fitted with an electronic lick detector. We developed a data analysis pipeline to accurately characterize drinking behavior for individual mice and report behavioral trends over a continuously monitored period. To corroborate the lick sensor data, we set up a video camera illuminated by infrared light to manually count licks at night as a positive control. In sum, this thesis introduces a novel drinking setup and data analysis pipeline to measure mouse liquid consumption over multiple days, thus establishing a protocol for future experiments to examine social risk-taking in juvenile mice.

Zoi Urban (Neuroscience), Mapping the Emotional Mind: Investigating the Fornix Using Diffusion MRI-Based Tractography Techniques (advisor Anastasia Yendiki, HMS) – The fornix is the major white matter projection of the hippocampus and it may play an important role in the neuropathology of mental illness. Yet despite its establishment as a prominent limbic white matter pathway, very few neuroimaging studies using high quality data and large sample sizes have sought to understand how the fornix is implicated in depression and anxiety disorders, due to its unique curvature, small size, and location. Diffusion magnetic resonance imaging (dMRI) tractography allows for the 3D in vivo mapping of white matter pathways in the human brain, and can be used to quantify tissue properties in much greater detail than what is possible with other neuroimaging approaches. The contribution of this work is twofold: i) we provide accurate, manual tractography dissections of the fornix on high-quality dMRI data, that we used as a training dataset for an automated probabilistic tractography algorithm; ii) we use this algorithm to automatically reconstruct the fornix in a cohort of 206 adolescents (ages 14-17 years) on a spectrum of mental health and we quantified fornix-specific microstructural differences compared to 68 healthy adolescent controls. The manual dissection protocol presented here will be available for use by other studies, and the automated reconstruction of the fornix will be available to the community in the public software program, FreeSurfer. Our results i) show we could reconstruct the fornix manually in 16 subjects with high-anatomical accuracy, ii) indicate that the fornix microstructure, proxied by along-tract fornix-specific diffusion-derived measures, is significantly different in adolescents with anxiety, compared to healthy controls. Our findings suggest that the microstructure along the fornix may be used as a biomarker of anxiety disorders, as we observed a lateral disruption in fornix integrity in correlation with disordered mental health.

Karina Walter (History and Science), "Help Wanted:" AI Hiring Tools and the Future of Diversity in the Workplace in Corporate, Media, and Activist Discourse (advisor (advisor Meg Perret, History of Science & Studies of Women, Gender, and Sexuality/FAS) – This thesis analyzes representations of AI hiring tools in corporate, media, and activist discourses in the United States from 2010 to 2021. This thesis contributes to the emerging body of scholarship in critical algorithm studies by applying its methodologies to an unexamined history of technology: the rapid implementation of AI in talent acquisition in the early 21st century. Using discourse analysis, the thesis reveals how companies, activists, and media represent the implications of AI hiring tools for the future of diversity in the workplace. Applying feminist science studies frameworks, this thesis examines the co-emergence of technology, culture, and power. Ultimately, this thesis argues that while corporate discourses advance that AI hiring tools will secure more objective, diverse, and efficient workplace futures, political and activist discourses argue that the tools will instead amplify historical patterns of bias and oppression against marginalized candidates. Media discourses support or reject corporate discourses’ liberatory promises for the future of the workplace, depending on the political orientation of the publication. The thesis concludes with recommendations for reimagining the future of workplace diversity through technoscience.