Computational Neuroscience Commons^™

Elucidating Neuroinflammation In Multiple Sclerosis By Network Analysis, Nora C. Welsh

Dartmouth College Ph.D Dissertations

Multiple sclerosis (MS) is a heterogeneous disease, differing on many variables, including disease course, sex, and overall activity. Key characteristics of the disease encompass demyelination, axonal damage, neuronal loss, glial cell activation, and the infiltration of peripheral immune cells. Molecular proxies of these functions are secreted proteins, including cytokines and immunoglobulins, which, in the central nervous system (CNS), can be secreted into the cerebrospinal fluid (CSF). A detailed analysis of these secreted proteins can offer insights into the evolving immunological and neurodegenerative features as the disease progresses. To understand the dynamic biological processes involved in MS, I used network analysis …

A Causal Inference Approach For Spike Train Interactions, Zach Saccomano

Dissertations, Theses, and Capstone Projects

Since the 1960s, neuroscientists have worked on the problem of estimating synaptic properties, such as connectivity and strength, from simultaneously recorded spike trains. Recent years have seen renewed interest in the problem coinciding with rapid advances in experimental technologies, including an approximate exponential increase in the number of neurons that can be recorded in parallel and perturbation techniques such as optogenetics that can be used to calibrate and validate causal hypotheses about functional connectivity. This thesis presents a mathematical examination of synaptic inference from two perspectives: (1) using in vivo data and biophysical models, we ask in what cases the …

Utilizing Ai Integrated Neuroimaging Technology To Expand Upon Machine Learning In Positron Emission Tomography Technology With The Aim Of Detecting Amyloid Beta Biomarkers Early In The Onset Of Alzheimer's., Ethan S. Terman

Undergraduate Research Posters

Early intervention in Alzheimer's is vital for treatment. The earlier a professional can detect symptoms and make a diagnosis the earlier a prognosis can be implemented. With the prevalence of data in our day-to-day world combined with Artificial intelligence (AI), utilizing both for machine learning can pave the way for more accurate and efficient detection of Alzheimer's and other neurodegenerative diseases. AI combined with Machine learning (ML) increases diagnostic efficiency and reduces human errors, making it a valuable resource for physicians and clinicians alike. With the increasing amount of data processing and image interpretation required, the ability to use AI …

Dna Methylation-Based Epigenetic Biomarkers In Cell-Type Deconvolution And Tumor Tissue Of Origin Identification, Ze Zhang

Dartmouth College Ph.D Dissertations

DNA methylation is an epigenetic modification that regulates gene expression and is essential to establishing and preserving cellular identity. Genome-wide DNA methylation arrays provide a standardized and cost-effective approach to measuring DNA methylation. When combined with a cell-type reference library, DNA methylation measures allow the assessment of underlying cell-type proportions in heterogeneous mixtures. This approach, known as DNA methylation deconvolution or methylation cytometry, offers a standardized and cost-effective method for evaluating cell-type proportions. While this approach has succeeded in discerning cell types in various human tissues like blood, brain, tumors, skin, breast, and buccal swabs, the existing methods have major …

The Effects Of Brain Control: A 3-D Agent-Based Model For Studying Pain, Kayla Kraeuter, Carley Reith, Benedict Kolber, Rachael Miller Neilan

Annual Symposium on Biomathematics and Ecology Education and Research

No abstract provided.

Invariant Object Recognition In Deep Neural Networks And Humans, Haider Al-Tahan

Electronic Thesis and Dissertation Repository

Invariant object recognition, a cornerstone of human vision, enables recognizing objects despite variations in rotations, positions, and scales. To emulate human-like generalization across object transformations, computational models must perform well in this aspect. Deep neural networks (DNNs) are popular models for human ventral visual stream processing, though their alignment with human performance remains inconsistent. We examine object recognition across transformations in human adults and pretrained feedforward DNNs. DNNs are grouped in model families by architecture, visual diet, and learning goal. We focus on object rotation in depth, and observe that object recognition performance is better preserved in humans than in …

Selective Recruitment Of Cerebellum In Cognition, Ladan Shahshahani

Electronic Thesis and Dissertation Repository

Previous studies of cerebellar function in humans have shown that it is activated by a myriad of tasks ranging from motor learning and language to working memory and more. These studies have prompted a deviation from the traditional view of the cerebellum as a purely motor structure. However, the precise contribution of the cerebellum to these tasks remains ambiguous.

A prevalent assumption in fMRI studies is interpreting BOLD activation as evidence of the cerebellum's involvement in specific tasks. However, this interpretation is potentially misleading, especially considering that the BOLD signal predominantly represents cerebellar input, with output activity largely absent. Consequently, …

Visual Cortical Traveling Waves: From Spontaneous Spiking Populations To Stimulus-Evoked Models Of Short-Term Prediction, Gabriel B. Benigno

Electronic Thesis and Dissertation Repository

Thanks to recent advances in neurotechnology, waves of activity sweeping across entire cortical regions are now routinely observed. Moreover, these waves have been found to impact neural responses as well as perception, and the responses themselves are found to be structured as traveling waves. How exactly do these waves arise? Do they confer any computational advantages? These traveling waves represent an opportunity for an expanded theory of neural computation, in which their dynamic local network activity may complement the moment-to-moment variability of our sensory experience.

This thesis aims to help uncover the origin and role of traveling waves in the …

Neural Dynamics Of Visual Processes In Challenging Visibility Conditions, Saba Charmi Motlagh

Electronic Thesis and Dissertation Repository

In our daily visual experience, our brain effortlessly categorizes countless objects, enabling us to perceive and interpret the world around us. This core object recognition process is vital for our survival and adaptive behavior, allowing us to recognize objects despite variations in appearance. The incredible speed at which we accomplish this task is a testament to the efficiency of our visual system and the significance of visual processing is evident in the allocation of nearly half of the neocortex in primates to this function. Unraveling the intricacies of how the human visual system tackles this complex challenge has long been …

Neural Dynamics Of Target Processing In Attentional Blink, Mansoure Jahanian

Electronic Thesis and Dissertation Repository

The attentional blink (AB) phenomenon refers to the failure to report the second target (T2) if it appears 200-500 ms after the first target (T1) in a stream of rapidly presented images. The present study aimed to investigate the neural representations of target processing under conditions where AB does or does not occur. We recorded EEG and behavioral data while participants viewed a rapid sequence of natural object images embedded with two face targets presented at two lag conditions: lag 3 (targets were 252 ms apart) and lag 7 (targets were 588 ms apart). Consistent with AB, our behavioral results …

Self-Supervised Pretraining And Transfer Learning On Fmri Data With Transformers, Sean Paulsen

Dartmouth College Ph.D Dissertations

Transfer learning is a machine learning technique founded on the idea that knowledge acquired by a model during “pretraining” on a source task can be transferred to the learning of a target task. Successful transfer learning can result in improved performance, faster convergence, and reduced demand for data. This technique is particularly desirable for the task of brain decoding in the domain of functional magnetic resonance imaging (fMRI), wherein even the most modern machine learning methods can struggle to decode labelled features of brain images. This challenge is due to the highly complex underlying signal, physical and neurological differences between …

Temporal Dynamics Of Natural Sound Categorization, Ali Tafakkor

Electronic Thesis and Dissertation Repository

While extensive research has elucidated the brain’s processing of semantics from speech sound waves and their mapping onto the auditory cortex, the temporal dynamics of how meaningful non-speech sounds are processed remain less examined. Understanding these dynamics is key to resolving the debate between cascaded and parallel hierarchical processing models, both plausible given the anatomical evidence. This study investigates how semantic category information from environmental sounds is processed in the temporal domain, using electroencephalography (EEG) collected from 25 participants and representational similarity analysis (RSA) along with models of acoustic and semantic information. We examined information extracted by the brain from …

Machine Learning Techniques For Improved Functional Brain Parcellation, Da Zhi

Electronic Thesis and Dissertation Repository

Brain parcellation studies are fundamental for neuroscience as they serve as a bridge between anatomy and function, helping researchers interpret how functions are distributed across different brain regions. However, two substantial challenges exist in current imaging-based brain parcellation studies: large variations in the functional organization across individuals and the intrinsic spatial dependence which causes nearby brain locations to have a similar function. This thesis presents a series of projects aimed to tackle these challenges from different perspectives by using advanced machine learning techniques.

To handle the challenge of individual variability in building precise individual parcellations, Chapter 3 introduces a novel …

The Consolidation Of Memory Associations, Kyle A. Kainec

Doctoral Dissertations

Creating memories is a fundamental challenge for the human brain. To create memories, defining features of experiences must be stored distinguishably without forgetting other memories. Memory associations represent co-occurring features and defining features across experiences. Memory associations are represented as networks of information that are stored in the brain. New memory associations are encoded during experiences and can be used to update existing memory associations during offline intervals. However, the mechanisms that underlie how encoded memory associations are stored within existing networks during offline intervals remains unclear. The experiments in this dissertation address a significant theoretical gap in understanding the …

Phenotyping Regression In A Female Mouse Model For Rett Syndrome Using Computational Neuroethology Tools, Michael J. Mykins

Doctoral Dissertations

Regression is defined as loss of acquired skills over time and is a key feature of many neurodevelopmental disorders such as Rett syndrome (RTT). RTT is caused by mutations in the X-linked gene Methyl CpG-Binding Protein 2 (MECP2) and is characterized by a period of typical development with subsequent regression of previously acquired motor and speech skills in girls. In human and animal models, it is clear syndromic phenotypes are dynamic over time but phenotyping regression over time in animal models has remained elusive. Lack of established timelines to study the molecular, cellular, and behavioral features of regression in female …

Solving The Cable Equation, A Second-Order Time Dependent Pde For Non-Ideal Cables With Action Potentials In The Mammalian Brain Using Kss Methods, Nirmohi Charbe

Master's Theses

In this thesis we shall perform the comparisons of a Krylov Subspace Spectral method with Forward Euler, Backward Euler and Crank-Nicolson to solve the Cable Equation. The Cable Equation measures action potentials in axons in a mammalian brain treated as an ideal cable in the first part of the study. We shall subject this problem to the further assumption of a non-ideal cable. Assume a non-uniform cross section area along the longitudinal axis. At the present time, the effects of torsion, curvature and material capacitance are ignored. There is particular interest to generalize the application of the PDEs including and …

Destined Failure, Chengjun Pan

Masters Theses

I attempt to examine the complex structure of human communication, explaining why it is bound to fail. By reproducing experienceable phenomena, I demonstrate how they can expose communication structure and reveal the limitations of our perception and symbolization.I divide the process of communication into six stages: input, detection, symbolization, dictionary, interpretation, and output. In this thesis, I examine the flaws and challenges that arise in the first five stages. I argue that reception acts as a filter and that understanding relies on a symbolic system that is full of redundancies. Therefore, every interpretation is destined to be a deviation.

Neural Correlates Of Post-Traumatic Brain Injury (Tbi) Attention Deficits In Children, Meng Cao

Dissertations

Traumatic brain injury (TBI) in children is a major public health concern worldwide. Attention deficits are among the most common neurocognitive and behavioral consequences in children post-TBI which have significant negative impacts on their educational and social outcomes and compromise the quality of their lives. However, there is a paucity of evidence to guide the optimal treatment strategies of attention deficit related symptoms in children post-TBI due to the lack of understanding regarding its neurobiological substrate. Thus, it is critical to understand the neural mechanisms associated with TBI-induced attention deficits in children so that more refined and tailored strategies can …

Deep Hybrid Modeling Of Neuronal Dynamics Using Generative Adversarial Networks, Soheil Saghafi

Dissertations

Mechanistic modeling and machine learning methods are powerful techniques for approximating biological systems and making accurate predictions from data. However, when used in isolation these approaches suffer from distinct shortcomings: model and parameter uncertainty limit mechanistic modeling, whereas machine learning methods disregard the underlying biophysical mechanisms. This dissertation constructs Deep Hybrid Models that address these shortcomings by combining deep learning with mechanistic modeling. In particular, this dissertation uses Generative Adversarial Networks (GANs) to provide an inverse mapping of data to mechanistic models and identifies the distributions of mechanistic model parameters coherent to the data.

Chapter 1 provides background information on …

Computing Brain Networks With Complex Dynamics, Anca R. Radulescu

Biology and Medicine Through Mathematics Conference

No abstract provided.