Master of Science, The Ohio State University, 2011, Computer Science and Engineering

Graphics processing units (GPUs) traditionally have been used to accelerate only parts of the graphics pipelines. The emergence of the new age GPUs as highly parallel, multi-threaded and many core processor systems with the ability to perform general purpose computations has opened doors for new form of heterogeneous computing where the GPU and CPU can be used together in accelerating the underlying computations. General-purpose computing on graphics processing unit (GPGPU, also referred to as GP2U) techniques can be used to perform highly data parallel computations and to accelerate some critical sections of an application. Accelerating the computation of fMRI analysis on a graphics processing unit is mainly attractive when used in a clinical environment. In this thesis, I discuss methods which try to exploit the capabilities provided by GPUs to accelerate the analysis of time varying data acquired during fMRI experiments for identifying regions of activity/inactivity. Static activation maps are obtained by inspecting voxels independently with the help of statistical methods in parallel using CUDA (Compute unified device architecture) threads. I provide an efficient strategy for mapping each individual time varying voxels to GPU kernel threads for data parallel analysis of fMRI data and present GPU version of methods used in the fMRI analysis pipeline based on voxel to thread mapping technique. Also, an efficient method for octree based hierarchical clustering of voxels on a GPU and using a combination of GPU and CPU for enhanced clustering speedup is discussed. A comparison between the data parallel methods implemented on GPU and the corresponding CPU implementations and overall speed up achieved using combined GPU and CPU implementations in octree based hierarchical clustering is discussed.

Committee: Raghu Machiraju PhD (Advisor); Gagan Agrawal PhD (Committee Member) Subjects: Medical Imaging

Doctor of Philosophy, The Ohio State University, 2024, Linguistics

Decades of psycholinguistics research have shown that human sentence processing is highly incremental and predictive. This has provided evidence for expectation-based theories of sentence processing, which posit that the processing difficulty of linguistic material is modulated by its probability in context. However, these theories do not make strong claims about the latent probability distribution of the human comprehender, which poses key research questions about its characteristics. Computational language models that define a conditional probability distribution are helpful for answering these questions, as they can be trained to embody different predictive processes and yield concrete surprisal predictors (i.e. negative log probabilities) that can be evaluated against measures of processing difficulty collected from human subjects. This dissertation evaluates Transformer-based large language models, which are artificial neural networks trained to predict the next word on massive amounts of text, as expectation-based models of sentence processing. Experiments reveal a robust and systematic divergence between the predictive processing of large language models and that of human subjects, the degree of which increases reliably with their number of parameters and amount of training data. This strong effect indicates that human sentence processing is not driven by the predictions made by these large-scale machine learning models, and highlights a fundamental shortcoming of large language models as models of human cognition. This dissertation additionally elucidates this discrepancy between humans and large language models. A series of analyses shows that large language models generally underpredict human-like processing difficulty by making 'superhumanly accurate' predictions of upcoming words, which may be a manifestation of the extensive real-world knowledge gleaned from large sets of training examples that are not available to humans. The learning trajectorie (open full item for complete abstract)

Committee: William Schuler (Committee Chair); Michael White (Committee Member); Micha Elsner (Committee Member); Tal Linzen (Committee Member) Subjects: Computer Science; Linguistics

Doctor of Philosophy, The Ohio State University, 2024, Psychology

While you are reading this sentence, have you ever thought about what is happening in your brain that allows you to perceive these abstract symbols as words and extract meaning from them? Have you ever thought about how learning to read, a uniquely human skill, plastically shapes your brain's functional organization? Strikingly, researchers found that within the left ventral temporal cortex—the high-level visual cortex that houses regions specialized for processing various abstract visual stimuli like faces and objects—there is also a small region that responds to written words and letters (thus called the visual word form area, VWFA) across many writing systems and languages. Critically, the VWFA serves as a unique example for us to investigate how the human brain develops its mosaic-like functional organization. In this dissertation, I aimed to: 1) comprehensively characterize the functional nature of the VWFA to settle any lingering debates, 2) explore factors that contribute to the emergence of the canonical VWFA, and 3) investigate the developmental trajectory of the VWFA and explore neural factors that might contribute to the developmental changes of the VWFA. These aims were accomplished through a series of studies in four corresponding chapters. In Chapter 2, I examined the VWFA's response to a wide range of visual and nonvisual stimuli to demonstrate its word-selective nature as well as to provide a more complete functional response profile of this region. In Chapter 3, by investigating functional connectivity in neonates, I asked to what extent the cortical tissue that will later become the VWFA already showed adult-like preferential connectivity with the putative frontotemporal language regions. In Chapter 4, I presented unique lesion data to explore the neural consequence of a missing typical language cortex—therefore disrupting the typical VWFA-language connection at birth—on the emergence of a canonical VWFA. In Chapter 5, I utilized longitudinal data (open full item for complete abstract)

Committee: Zeynep Saygin (Advisor); Dylan Wagner (Committee Member); Julie Golomb (Committee Member) Subjects: Cognitive Psychology; Neurosciences; Psychology

Doctor of Philosophy, The Ohio State University, 2023, Psychology

This dissertation investigates the neural representation of complex social information by employing multivariate methods such as functional connectivity classification analyses and intersubject representational similarity analyses (IS-RSA). The present research examines the functional network associated with social interactions as well as the association between neural similarity and individual differences in emotional reactivity and empathy. Study 1 utilizes multivariate pattern analyses (MVPA) to classify different categories of social interactions based on connectivity patterns between brain regions involved in social perception. Study 2 uses IS-RSA to investigate how individual differences in emotional reactivity modulates the neural representation of different social interactions. Study 3 explores how idiosyncrasies in behavioral measures of empathy are associated with neural synchrony during the observation of naturalistic social scenes depicting specific characters and various types of interactions. Study 1 demonstrates that contextual categorical information about social interactions is better classified by a network of regions rather than within a single region of interest. Studies 2 and 3 suggest that idiosyncrasies in trait-like attributes such as empathy or emotional reactivity reflects differences in neural representation of complex social information. This research contributes to our understanding of how social information is processed in the brain and sheds light on the impact of individual differences on social perception.

Committee: Dylan Wagner (Advisor); Baldwin Way (Committee Member); Steven Spencer (Committee Member) Subjects: Neurosciences; Psychology; Social Psychology

Doctor of Philosophy, The Ohio State University, 2023, Speech and Hearing Science

Two essential components of language are semantics, or the meaning of language, and phonology, or the sounds that make up our words. Researchers have long sought to investigate the neural correlates of semantics and phonology; however, questions remain related to the specific brain regions comprising each network as well as the degree to which these networks coincide. Moreover, patterns of reorganization following injury to these networks in populations such as those with post-stroke aphasia remain unclear. Across three manuscripts, this dissertation addresses these questions, emphasizing the influence of aging on the language networks in the brain as well as reorganization during the process of recovery from post-stroke aphasia. Recent work examining the semantic and phonological networks in the brain has focused on neurologically intact younger adults. Considering many people who experience acquired language impairments are older adults, the first manuscript in this dissertation presents the results of a scoping review addressing the regions comprising the semantic and phonological brain networks in this aging population. The review finds that these brain networks are consistent with the networks of younger adults but may have subtle differences that should be further explored in a full systematic review or meta-analysis. The second manuscript in this dissertation specifically examines the resting-state functional connectivity of the inferior frontal gyrus, a region that has been implicated in both semantic and phonological brain networks and is often damaged in cases of post-stroke aphasia. Compared with younger adults, we again found subtle differences that may be accounted for in part by theories of age-related de-lateralization of the dominant left hemisphere. We next correlated significant resting-state functional connectivity with behavioral tasks targeting semantics and phonology, which did not support theories of semantic specialization at the anterior inf (open full item for complete abstract)

Committee: Stacy M. Harnish (Advisor); Jennifer Brello (Committee Member); David E. Osher (Committee Member) Subjects: Neurosciences; Psychology; Rehabilitation; Speech Therapy

Doctor of Philosophy, The Ohio State University, 2023, Health and Rehabilitation Sciences

Anterior Cruciate Ligament (ACL) injury is a frequently occurring sports-related knee injury which is surgically remedied with an ACL reconstruction (ACLR). ACL injury and ACLR create a cascade of events which negatively impact an individual's musculoskeletal function, neuromuscular control, and central nervous system (CNS). Individuals undergoing ACLR have immediate deficits in knee related biomechanics which remain unresolved for years. Deficits are noted early during gait and as individual's regain more function are also noted in higher level activities such as jumping, running, cutting, and pivoting. These biomechanics deficits coincide with, but are not always explained by strength deficits, patient reported function, and range of motion limitations. These deficits also have long term consequences with the prevalence of cartilage degeneration and knee osteoarthritis being higher in individuals post-ACLR compared to age and activity matched individuals. The broad array of musculoskeletal deficits noted following ACLR have remained, even with updated clinical practice guidelines leading to investigation of neuromuscular and CNS function. This dissertation strives to illuminate how brain activity is altered after ACLR by investigating connections between the brain and musculoskeletal system. This will be done by systematically reviewing prior literature on ACLR populations who have had their blood oxygen level dependent (BOLD) response evaluated using functional magnetic resonance imaging (fMRI). All peak voxel coordinates from articles included in this review will be entered in activation estimation likelihood (ALE) meta-analysis to determine if any regions are identified as systematically active across all studies (Chapter 3). We will then investigate how BOLD signal is associated with biomechanics during a run to pivot task to determine the neural correlates of knee loading in individuals who were cleared for full activity following ACLR (Chapter 4). Finall (open full item for complete abstract)

Committee: Jimmy Onate (Advisor); Laura Schmitt (Advisor); Scott Hayes (Committee Member); Jaclyn Caccese (Committee Member); Dustin Grooms (Committee Member) Subjects: Health Sciences; Neurosciences; Physical Therapy

Master of Science, The Ohio State University, 2023, Psychology

Introduction. Sustained attention is a ubiquitous human experience, and much work have been done to investigate its neural mechanisms within distinct stratified populations, such as in young or old adults. Yet, much less is known about how brain connections associated with sustained attention evolve across the life span. Here, we leveraged connectome-based predictive modeling to identify unique functional edges related to sustained attention. Additionally, we altered the CPM to incorporate age as a moderating variable; a novel modification to a well-established technique that allows for the identification of age-variant edges relevant to sustained attention throughout the lifespan. Methods. Using fMRI data from the HCP-A (n = 698, age 35 – 100), we derived two distinct CPMs: an age-invariant CPM that used edges stable across the lifespan to predict attention, and an age-variant CPM that used a different set of edges that change their contributions over time to similarly predict attention. Results. The age-invariant model, using edges stable across the lifespan, successfully predicted sustained attention across the lifespan (High Attention: r = .23, p < .001; Low Attention: r = .15, p < .001; Combined: r = .20, p < .001). Furthermore, the age-variant model – derived from edges that changed across the lifespan – was similarly predictive of sustained attention (Positive Interaction: r = .33, p < .001; Negative Interaction: r = .38, p < .001). Specifically, edges within the control, somatomotor, dorsal and ventral attention, and visual networks were overrepresented in the positive interaction model, while edges between the control and somatomotor networks, as well as between the default mode and dorsal attention networks were overrepresented in the negative interaction model. We also found the proposed age-moderating effect via the systematic change in correlations between observed d' and network strengths when we subdivided the sample by age-groups defined by th (open full item for complete abstract)

Committee: Ruchika Prakash (Advisor) Subjects: Psychology

Master of Science, The Ohio State University, 2023, Psychology

Language and nonverbal skills like face/body perception and theory of mind (ToM; the ability to infer others' mental states) are vital for effective communication in a social world. The superior temporal lobe (STL) has been widely implicated in each of these skills, as well as social communicative skills like face and body perception, but its functional organization as well as the potential modularity of these functions is unclear, particularly early in development. What is the functional organization of these skills in children who are still developing complex language and ToM? Do more complex cognitive functions like language and ToM emerge from an initially common neural substrate for social communication? In this study, 29 children ages 4-9 years old completed three functional MRI tasks to localizer language, non-verbal ToM, and dynamic face and body recognition. We first examine the general landscape of cortical organization of the STL. We find the left, anterior STL is strongly dominated by speech and language responses, whereas the right STL shows clusters for face, body, and ToM, primarily in the inferior anterior and superior posterior regions. Using subject-specific regions of interest, we find that language and ToM are represented by distinct modules in the STL, that are specific for their category, not sensitive to other functions, and show minimal overlap in their foci of activation. This minimal overlap is not associated with age or selectivity of non-overlapping regions. This study is the first to examine the modularity of language and ToM in children, and we see early functional specialization even in children still developing these skills, similar to prior work in adults with fully developed skills. These results suggest that language and ToM develop separately in the STL, not from earlier developing general social cognitive skills, and provide a valuable addition to developmental cognitive neuroscience literature by providing insights to the develo (open full item for complete abstract)

Committee: Zeynep Saygin (Advisor); Dylan Wagner (Committee Member); David Osher (Committee Member) Subjects: Psychology

Master of Science, The Ohio State University, 2023, Psychology

Accurately categorizing items requires humans to selectively attend to stimulus dimensions that are relevant to a task. However, learning to direct attention toward the relevant dimensions is often achieved through trial and error. Therefore, category learning models should seek to describe a neurally plausible account of how humans adjust their attention over time. The Adaptive Attention Representation Model (AARM) attempts to describe this dynamic process by employing a between-trial attention updating function in the form of a feedback-based error gradient. To provide neural validation for AARM's attentional mechanisms we conducted a simulation study, fit AARM to behavioral data from Mack et al (2016), and conducted three model-based fMRI analyses. The simulation demonstrated a priori expectations of the model's behavior in the context of the Shepard VI paradigm. The behavioral fits showcased AARM's capacity to capture choice accuracy and attentional dynamics in a complex learning environment. The fMRI analyses revealed a brain-wide system that supports flexible attention updating. This neural system includes areas believed to support attention orienting (prefrontal and parietal cortices), visual perception (visual pathways), memory encoding and retrieval (hippocampus and MTL), prediction error (basal ganglia), and goal maintenance (PFC). These results support AARM's specification of attentional tuning as a dynamic property of distributed cognitive systems.

Committee: Brandon Turner (Advisor); Julie Golomb (Committee Member); Vladimir Sloutsky (Committee Member) Subjects: Cognitive Psychology; Neurosciences; Psychology; Quantitative Psychology

Doctor of Philosophy, Case Western Reserve University, 2023, EECS - Computer and Information Sciences

The COVID-19 pandemic has necessitated disease surveillance using group testing. Novel Bayesian methods using lattice models Ire proposed, which offer substantial improvements in group testing efficiency by precisely quantifying uncertainty in diagnoses, acknowledging varying individual risk and dilution effects, and guiding optimally convergent sequential pooled test selections using Bayesian Halving Algorithms. Computationally, however, Bayesian group testing poses considerable challenges as computational complexity grows exponentially with sample size. This can lead to shortcomings in reaching a desirable scale without practical limitations. To overcome these challenges, I propose a high-performance Bayesian group testing framework named HiBGT, which systematically explores the design space of Bayesian group testing and provides comprehensive heuristics on how to achieve high-performance Bayesian group testing. I show that HiBGT can perform large-scale test selections ($>2^{50}$ state iterations) and accelerate statistical analyzes up to 15.9x (up to 363x with little trade-offs) through a varied selection of sophisticated parallel computing techniques while achieving near linear scalability using up to 924 CPU cores. I further propose to scale HiBGT using a lightning fast and highly scalable framework, named SGBT. In particular, SBGT is up to 376x, 1733x, and 1523x faster than HiBGT in manipulating lattice models, performing test selections, and conducting statistical analyses, respectively, while achieving up to 97.9\% scaling efficiency up to 4096 CPU cores. I propose algorithms and workflows for next-generation real-time analysis of fMRI data and dynamically adjustment of experiment stimuli through early stopping. To overcome significant computational challenges raised in this setting, I design a \underline{S}calable, \underline{P}arallel, and \underline{R}eal-\underline{T}ime \underline{S}equential \underline{P}robability \underline{R}atio \underline{T}est (open full item for complete abstract)

Committee: Curtis Tatsuoka (Advisor); Vipin Chaudhary (Committee Chair); Xiaoyi Lu (Committee Member); Vincenzo Liberatore (Committee Member) Subjects: Computer Science

PhD, University of Cincinnati, 2022, Medicine: Neuroscience/Medical Science Scholars Interdisciplinary

Juvenile fibromyalgia (JFM) is a poorly understood chronic pain condition that predominantly affects female adolescents. It is characterized by persistent widespread musculoskeletal pain and often accompanied by symptoms such as fatigue, mood and sleep disturbances, cognitive dysfunction and multisensory hypersensitivity, leading to significant functional disabilities. The past two decades have seen important advances in our understanding of the pathophysiology of adult-onset fibromyalgia (FM), which shows that FM is a nociplastic pain condition characterized by augmented pain and nonpainful sensory processing in the central nervous system. However, the underlying mechanisms of JFM remain unknown. Although sharing many similarities, JFM and FM also have some differences in clinical features. In addition, adolescence is a critical period for the development of the brain, where pain perception is formed and regulated. Yet, how specifically adolescent brain development affects pain processing remains unclear. The aim of this research project was to characterize the brain correlates of pain processing in both healthy adolescents and adolescents with JFM. To this end, we conducted two studies using a noxious pressure task to elicit brain responses to pain during functional magnetic resonance imaging (fMRI) scans. In the first study, we compared pain-evoked brain responses between healthy adolescent girls and healthy middle-aged adult women to determine whether pain perception and brain mechanisms of pain processing during adolescence might differ from adulthood. We found that adolescent girls reported greater pain than adult women in response to low-intensity noxious stimuli and exhibited greater brain activation in multiple pain-associated regions including medial and lateral prefrontal cortex, supramarginal gyrus and rostral anterior cingulate cortex. Further analyses revealed that adolescents had heightened pain-related responses in the Neurologic Pain Signatur (open full item for complete abstract)

Committee: Mark Baccei Ph.D. (Committee Member); Robert Coghill Ph.D. (Committee Member); Jing Xiang Ph.D. (Committee Member); Marina Lopez Sola Ph.D. (Committee Member); Christopher King Ph.D. (Committee Member) Subjects: Neurology

MS, University of Cincinnati, 2022, Engineering and Applied Science: Biomedical Engineering

The objective of the current study is to investigate brain alterations in patients with localization-related epilepsy (LRE) compared to healthy controls using different brain activity measures including regional homogeneity (ReHo); functional connectivity density mapping (FCDM) with global (gFCD) and lobal (lFCD); amplitude of low-frequency fluctuations (ALFF) and fractional amplitude of low-frequency fluctuations (fALFF), based on resting-state fMRI data. ReHo, gFCD and lFCD, ALFF and fALFF were used to examine the alterations in the brain from 19 LRE patients and 19 healthy controls, by using the whole brain resting-state fMRI (rs-fMRI) data. For each method, a two sampled t-test was conducted; and all results were corrected for family wise error for group analysis. The LRE patients were placed into subgroups based on whether the seizure onset falls into the limbic regions. The two-samples t-test was conducted to investigate the brain alterations in the limbic LRE patients versus the healthy controls, the non-limbic LRE patients versus the healthy controls, and between these two epileptic patient groups. The current findings demonstrated ReHo, gFCD, lFCD, ALFF, and fALFF alterations in the LRE patients compared to healthy subjects during resting state. For both group analysis and for subgroup limbic analysis, we obtained specific brain regions with significantly altered ReHo, lFCD, ALFF, and fALFF values. The targeted brain regions with significant alterations might be contributing to the overall lower synchronization and more impaired functional activity in the processing of motor pathways, visual information, and emotional functions. The targeted brain regions were also in the salience network and the default mode network (DMN), as well as part of the language area, sensorimotor regions, and those associated with the posterior attention system. We also (open full item for complete abstract)

Committee: Jing-Huei Lee Ph.D. (Committee Member); Marepalli Rao Ph.D. (Committee Member); Zackary Cleveland Ph.D. (Committee Member) Subjects: Biomedical Research

Master of Science, The Ohio State University, 2022, Psychology

Older adulthood is typically marked by enhanced emotional well-being potentially resulting from greater reliance on adaptive emotion regulation strategies. However, not all older adults demonstrate this overall trend in enhanced emotional well-being and instead show greater reliance on maladaptive emotion regulation strategies. An important moderator of age-related shifts in emotion regulation strategy use is working memory. Evidence from neuroimaging studies find that working memory and emotion regulation processes often share underlying neural circuity. As such, individual differences in the integrity of neural networks underlying working memory may predict older adults' emotion regulation strategy preferences. The current study leveraged a connectome-based predictive model of working memory (wmCPM), derived from young adults, to predict working memory performance and acceptance strategy use in an independent sample of older adults. Ninety-one community older adults were recruited as part of an ongoing randomized controlled trial examining the impact of mind-body interventions on healthy aging. Participants' baseline behavioral and neuroimaging assessment data was used for the current analyses. Results revealed that a combined wmCPM network strength successfully predicted working memory accuracy scores but did not predict overall acceptance use in older adults. Linear mixed models examining moderations of image intensity on the relationship between network strength and acceptance use were not significant. These findings highlight that a robust neural marker of working memory successfully generalizes to an independent sample of healthy older adults. However, the specific neural circuitry predictive of working memory performance may not generalize beyond cognitive domains to predict measures of emotional functioning.

Committee: Ruchika Prakash Dr. (Advisor); Jennifer Cheavens Dr. (Committee Member); Andrew Leber Dr. (Committee Member) Subjects: Psychology

Master of Science, The Ohio State University, 2021, Psychology

Humans regularly make decisions based on the perceptual properties of the stimulus (perceptual decisions) or the subjective value of the stimulus (value-based decisions). The characteristics of evidence used in perceptual and value-based decisions are different, but the process of perceptual and value-based decisions has been explained by a common computational model, sequential sampling models (SSMs). Studies on the neurobiological mechanism of SSMs have proposed that the frontoparietal network is responsible for accumulating evidence both in perceptual and value-based decisions. However, little is known about whether the frontoparietal network is indeed involved in tracking the state of accumulated evidence over time and whether the overlapping brain regions are responsible for tracking the state of accumulated evidence in perceptual and value-based decisions. Also, it has not been much studied whether a temporal bias is consistent within an individual across different decision domains. This study addressed these questions by using a modified interrogation paradigm along with computational modeling and functional magnetic resonance imaging. Subjects were presented with 30 pairs of square grids or snack foods in a series and reported their estimates of average evidence multiple time points in a trial. The behavior results showed that subjects had a recency bias in the evidence averaging process and such temporal bias was consistent within an individual. Neuroimaging analyses showed domain-specific brain regions for tracking instantaneous evidence (IE). Parietal lobe was tracking IE in the perceptual task and ventromedial prefrontal cortex, ventral striatum, and posterior cingulate cortex were tracking IE in the value-based task. Domain-general and domain-specific brain regions for tracking average evidence (AE) were found in the neuroimaging analyses. Visual cortex was involved in tracking AE in both tasks, whereas dorsolateral prefrontal cortex was selectively eng (open full item for complete abstract)

Committee: Ian Krajbich (Advisor); Andrew Leber (Committee Member); Brandon Turner (Committee Member) Subjects: Psychology

PhD, University of Cincinnati, 2021, Engineering and Applied Science: Computer Science and Engineering

Deep learning methods, especially convolutional neural networks (CNNs), have revolutionized many domains of artificial intelligence (AI) including natural image classification/segmentation, speech recognition, and natural language processing. It is slowly being acknowledged that deep learning methods have a huge potential to advance medical image analysis, medical diagnostics, and general healthcare. Traditionally, medical imaging interpretations have benefited from machine learning methods. In machine learning, feature extraction is significant for model design and experiment implementation. However, the exact etiologies behind many medical diseases are unknown. It is challenging to develop a feature extraction system when there is a lack of domain understanding. Compared to traditional machine learning methods, feature extraction is part of the learning process and discriminative features can be automatically extracted in deep learning. It has been demonstrated that deep learning methods can produce physiologically meaningful features and reveal new associations from high dimensional medical imaging data. Different from conventional methods, deep learning is an end-to-end solution for medical imaging problems without the feature extraction process. Data can be fed to deep learning models in their raw form, and high-level representations are automatically extracted.

Committee: H. Howard Fan Ph.D. (Committee Chair); Yizong Cheng Ph.D. (Committee Member); Carla Purdy (Committee Member); Ali Minai Ph.D. (Committee Member); Lili He (Committee Member) Subjects: Computer Science

PhD, University of Cincinnati, 2021, Medicine: Neuroscience/Medical Science Scholars Interdisciplinary

Reading is an evolutionarily new human invention that is an indicator of academic success for children in school. The Simple View of Reading (SVR) model suggests that reading comprehension depends on decoding and linguistic comprehension abilities. Further exploration of this model shows that cognitive abilities or executive functions (EFs) including working memory, planning behavior, and inhibition to name a few are also relevant for reading comprehension. Ten to 15% of school-age children are diagnosed with reading difficulties (RD) or dyslexia based on deficits in components of the SVR model by 3rd grade. Functional magnetic resonance imaging (fMRI) data has shown that children with RD have decreased activation in temporoparietal circuitry important for phonological processing, occipitotemporal circuitry involved in visual word recognition, and increased activation in frontal regions due to the cognitive demand of reading. Although RD commonly occurs due to endogenous-genetic reasons, children can also suffer from reading challenges due to exogenous/environmental reasons such as inadequate stimulation in the home reading environment or lack of resources. Environmental variables such as maternal reading fluency and educational attainment are associated with behavioral and neurobiological outcomes for reading acquisition. Therefore, the over-reaching goal of the current study was to examine the relations between maternal variables (i.e., reading fluency and education) and the behavioral and neurobiological correlates of reading in pre-readers and school-age children using task-based and resting-state fMRI data. Using this approach, we hope to expand the SVR model by adding an exogenous component to it, focusing on maternal education and reading fluency. We conducted three studies aiming to 1) determine the relationship between maternal reading fluency and the functional connectivity between the language network and regions related to cognitive control and visual pr (open full item for complete abstract)

Committee: Jennifer Vannest Ph.D. (Committee Chair); Mark Difrancesco Ph.D. (Committee Member); Tzipi Horowitz-kraus Ph.D. (Committee Member); John Hutton (Committee Member); Jeffrey Tenney (Committee Member) Subjects: Neurology

Doctor of Philosophy, University of Toledo, 2022, Psychology - Clinical

Many individuals develop posttraumatic stress disorder (PTSD) and/or major depressive disorder (MDD) following exposure to a traumatic event. There is a need for research on underlying mechanisms that can predict the development of PTSD and MDD following traumatic exposure, given that these mechanisms could be targeted in evidence-based interventions. The purpose of this study was to multimodally examine the predictive influence of reward responsiveness (RR), a biobehavioral mechanism that may be related to symptoms of anhedonia and low positive emotionality in MDD and PTSD, on the 3-month clinical trajectories of individuals with recent exposure to a traumatic event. A sample of 39 adult participants completed a behavioral fMRI task assessing RR, with a subset of participants (n=22) completing the task within two weeks of traumatic exposure. Participants completing the task within two weeks of traumatic exposure also completed self-report measures of RR and clinical symptoms at the time of the visit, and again three months later. The behavioral fMRI task elicited activation in regions relevant to RR, as well as regions involved in top-down regulation of attention and emotion. High self-reported RR at Time 0 predicted a better prognosis at 3-months, while high neural RR predicted a worse prognosis. Greater Time 0 neural activation in top-down regulatory regions also predicted worse 3-month clinical outcomes. These results speak to the complex neural sequalae of traumatic event exposure, and highlight the need for more longitudinal research to elucidate the role of neural processes in MDD and PTSD symptom trajectories. Although preliminary, these results may highlight the potential utility of brief interventions to target underlying neural mechanisms among individuals with recent traumatic exposure.

Committee: Matthew Tull Ph.D. (Committee Chair); Xin Wang MD, Ph.D. (Committee Member); Kim Gratz Ph.D. (Committee Member); Jon Elhai Ph.D. (Committee Member); Andrew Geers Ph.D. (Committee Member) Subjects: Clinical Psychology

Master of Science (MS), Wright State University, 2021, Anatomy

Because magnetic resonance (MR) and positron emission tomography (PET) scanning sessions last long durations, motion blur during scanning constitutes a problem for clinical interpretation. To counteract this, motion-correction algorithms have been developed to reduce smearing between scan slices of MRI, but these algorithms are not commonplace for PET. This feasibility study determined if applying MRI motion-correction algorithms to simultaneously acquired PET data improved PET signal clarity in specific brain regions. Seven subjects received increasing levels of PET tracers while undergoing two separate simultaneous PET/MRI scans. We modified existing fMRI algorithms to apply them to the accompanying PET data. We hypothesized gray matter activity was low due to motion-blurring, and correction would result in increased signal intensity. We evaluated this and other internal brain regions using a Wilcoxon Signed-Rank statistical test. We failed to reject the null hypothesis as no regions showed significant differences between the motion-corrected and raw data.

Committee: Robert M. Lober M.D., Ph.D. (Committee Chair); Matthew S. Sherwood Ph.D. (Committee Co-Chair); David R. Ladle Ph.D. (Committee Member) Subjects: Anatomy and Physiology; Biomedical Engineering; Biomedical Research; Medical Imaging; Neurobiology; Neurology; Neurosciences

Doctor of Philosophy, The Ohio State University, 2021, Linguistics

The predictions of theories of incremental human sentence processing are often cached out in word-by-word measures, but the mind is a dynamical system that responds to language in real time. As a result, there may be a complex alignment between the properties of words in language and the influence those properties exert on measures of human cognition. One possible aspect of this alignment is temporal diffusion, whereby sentence processing effects are realized in a delayed manner (Mitchell, 1984). For example, because of real-time bottlenecks in human information processing (Mollica & Piantadosi, 2017), encountering a surprising word may increase cognitive load not only at that word, but also on subsequent words as the rest of the experiment unfolds (Smith & Levy, 2013). In this thesis, I argue that effect timecourses are of direct or indirect importance to many central questions in psycholinguistics, that failure to account for these timecourses can have large impacts on the results of scientific hypothesis tests, and that existing discrete-time approaches to estimating and controlling for effect timecourses are not well adapted to many experimental designs in psycholinguistics, which involve non-uniform time series in which events (words) have variable duration. I define and implement an analysis technique that addresses these concerns: continuous-time deconvolutional regression (CDR). CDR estimates continuous-time functions that describe the shape and extent of a predictor's influence on the response over time, thus directly illuminating and controlling for temporally diffuse effects. I show empirically that CDR accurately recovers ground-truth models from synthetic data and provides plausible and detailed estimates of temporal structure in human data that generalize better than estimates obtained using existing techniques. I apply CDR to measures of naturalistic sentence processing in order test several theoretical questions in psycholinguistics. In one st (open full item for complete abstract)

Committee: William Schuler PhD (Advisor); Micha Elsner PhD (Advisor); Paul Subhadeep PhD (Committee Member) Subjects: Computer Science; Linguistics; Neurosciences

PhD, University of Cincinnati, 2021, Allied Health Sciences: Communication Sciences and Disorders

In typical child and adult speakers, speech generation requires coordinated activation of a network of inferior frontal, temporal, and subcortical brain regions to carry out multiple linguistic and speech motor processes. However, a portion of children who exhibit speech sound errors in development persist in these errors beyond age 9, which can lead to broader, long-term consequences in scholastic achievement, literacy, and social-emotional well-being. The goal of this project was to investigate the neural underpinnings of residual speech sound disorder (RSSD) and its remediation through a speech therapy program. In Study 1, I investigated the neural activity of children with RSSD in comparison to children with typically-developing speech (TD) at baseline (Time 1). I had anticipated to observe significant differences between RSSD and TD groups. However, in a whole-brain analysis (at p<0.05 and with Bonferroni corrections for multiple comparisons), I did not observe statistically significant differences in activation on either the SRT-Early Sounds or SRT-Late Sounds. In Study 2, I followed up with a region-of-interest approach of activation at Time 1 and Time 2. I did not detect any significant differences across task, group, or time comparisons. While this finding was not expected, it implies that, when task performance is similar, children with RSSD do not show differences in neural activity from their typical peers. I also explored the relationship between change in activation and progress in therapy. I found that children with RSSD who made more progress in therapy tended to show a decrease in activation in the left visual association cortex on the SRT-Late Sounds (R2=0.78). The left visual association cortex is not a core component of the speech production network but may indicate differences in the children's reliance on sensorimotor integration or internal speech visualization processes. Using a seed-to-voxel approach, I also explored function (open full item for complete abstract)

Committee: Suzanne Boyce Ph.D. (Committee Chair); Edwin Maas Ph.D. (Committee Member); Jonathan Preston Ph.D. (Committee Member); Erin Redle Ph.D. (Committee Member); Jennifer Vannest Ph.D. (Committee Member) Subjects: Speech Therapy