Physics Theses and Dissertations
Permanent URI for this collectionhttp://hdl.handle.net/1903/2800
Browse
8 results
Search Results
Item Analyzing the Dynamics of Biological and Artificial Neural Networks with Applications to Machine Learning(2024) Srinivasan, Keshav; Girvan, Michelle; Biophysics (BIPH); Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The study of the brain has profoundly shaped the evolution of computational learning models and the history of neural networks. This journey began in the 1940s with Warren McCulloch and Walter Pitts’ groundbreaking work on the first mathematical model of a neuron, laying the foundation for artificial neural networks. The 1950s and 60s witnessed a significant milestone with Frank Rosenblatt’s development of the perceptron, showcasing the potential of neural networks for complex computational tasks. Since then, the field of neural networks has witnessed explosive growth, and terms like “Artificial Intelligence” and “Machine Learning” have become commonplace across diverse fields, including finance,medicine, and science. This dissertation explores the symbiotic parallels between neuroscience and machine learning, focusing on the dynamics of biological and artificial neural networks. We begin by examining artificial neural networks, particularly in predicting the dynamics of large, complex networks—a paradigm where traditional machine learning algorithms often struggle. To address this, we propose a novel approach utilizing a parallel architecture that mimics the network’s structure, achieving scalable and accurate predictions. Shifting our focus to biological neuronal networks, we delve into the theory of critical systems. This theory posits that the brain, when viewed as a complex dynamical system, operates near a critical point, a state ideal for efficient information processing. A key experimental observation of this type of criticality is neuronal avalanches—scale-free cascades of neuronal activity—which have been documented both in vitro (in neuronal cultures and acute brain slices) and in vivo (in the brains of awake animals). Recent advancements in experimental techniques, such as multi-photon imaging and genetically encoded fluorescent markers, allow for the measurement of activity in living organisms with unparalleled single-cell resolution. Despite these advances, significant challenges remain when only a fraction of neurons can be recorded with sufficient resolution, leading to inaccurate estimations of power-law relationships in size, duration, and scaling of neuronal avalanches. We demonstrate that by analyzing simulated critical neuronal networks alongside real 2-photon imaging data, temporal coarse-graining can recover the critical value of the mean size vs. duration scaling of neuronal avalanches, allowing for more accurate estimations of critical brain dynamics even from subsampled data. Finally, we bridge the gap between machine learning and neuroscience by exploring the concept of excitatory-inhibitory balance, a crucial feature of neuronal networks in the brain, within the framework of reservoir computing. We emphasize the stabilizing role of inhibition in reservoir computers (RCs), mirroring its function in the brain. We propose a novel inhibitory adaptation mechanism that allows RCs to autonomously adjust inhibitory connections to achieve a specific firing rate target, motivated by the firing rate homeostasis observed in biological neurons. Overall, this dissertation strives to deepen the ongoing collaboration between neuroscience and machine learning, fostering advancements that will benefit both fields.Item Unveiling secrets of brain function with generative modeling: Motion perception in primates & Cortical network organization in mice(2023) Vafaii, Hadi; Pessoa, Luiz; Butts, Daniel A; Physics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This Dissertation is comprised of two main projects, addressing questions in neuroscience through applications of generative modeling. Project #1 (Chapter 4) is concerned with how neurons in the brain encode, or represent, features of the external world. A key challenge here is building artificial systems that represent the world similarly to biological neurons. In Chapter 4, I address this by combining Helmholtz's “Perception as Unconscious Inference”---paralleled by modern generative models like variational autoencoders (VAE)---with the hierarchical structure of the visual cortex. This combination results in the development of a hierarchical VAE model, which I subsequently test for its ability to mimic neurons from the primate visual cortex in response to motion stimuli. Results show that the hierarchical VAE perceives motion similar to the primate brain. I also evaluate the model's capability to identify causal factors of retinal motion inputs, such as object motion. I find that hierarchical latent structure enhances the linear decodability of data generative factors and does so in a disentangled and sparse manner. A comparison with alternative models indicates the critical role of both hierarchy and probabilistic inference. Collectively, these results suggest that hierarchical inference underlines the brain's understanding of the world, and hierarchical VAEs can effectively model this understanding. Project #2 (Chapter 5) is about how spontaneous fluctuations in the brain are spatiotemporally structured and reflect brain states such as resting. The correlation structure of spontaneous brain activity has been used to identify large-scale functional brain networks, in both humans and rodents. The majority of studies in this domain use functional MRI (fMRI), and assume a disjoint network structure, meaning that each brain region belongs to one and only one community. In Chapter 5, I apply a generative algorithm to a simultaneous fMRI and wide-field calcium imaging dataset and demonstrate that the mouse cortex can be decomposed into overlapping communities. Examining the overlap extent shows that around half of the mouse cortical regions belong to multiple communities. Comparative analyses reveal that calcium-derived network structure reproduces many aspects of fMRI-derived network structure. Still, there are important differences as well, suggesting that the inferred network topologies are ultimately different across imaging modalities. In conclusion, wide-field calcium imaging unveils overlapping functional organization in the mouse cortex, reflecting several but not all properties observed in fMRI signals.Item UNCOVERING THE MOLECULAR BASIS OF ACTIVITY-DEPENDENT RETINOFUGAL SYNAPSE PLASTICITY(2023) Zhang, Chenghang; Speer, Colenso; Biophysics (BIPH); Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Activity-dependent synapse plasticity is important for the establishment of neuron wiring in the central nervous system, particularly in the context of sensory processing. In the visual system, image-forming and non-image-forming retinal input into the brain is a popular model for studying activity-dependent plasticity due to the well-characterized neural activity and bulk-level innervation pattern. However, investigation of synaptic connection during early development has been impeded by the limited resolution of conventional fluorescent microscopy or lack of profile tagging in electron microscopy (EM) images. To overcome these challenges, we employed volumetric STochastic Optical Reconstruction Microscopy, immunohistochemistry synaptic protein labeling, and anterograde retinal tract tracing to investigate the activity-dependent retinogeniculate and retinohypothalamic synapse plasticity. Through our findings, we uncover the developmental pattern of retinofugal innervation and shed light on the impact of spontaneous activity on retinal synapse maturation at the synaptic level. During the first postnatal week in mice, the dorsal lateral geniculate nucleus (dLGN) initially receives overlapping input from the two eyes before the binocular innervation segregated. The changes in individual synapse properties during the eye-specific segregation process have remained unknown. In Chapter 2, we uncovered eye-specific differences in presynaptic vesicle pool size and vesicle association with the active zone at the earliest stages of retinogeniculate refinement but found no evidence of eye-specific differences in subsynaptic domain number, size, or transsynaptic alignment across development. Genetic disruption of spontaneous retinal activity decreased retinogeniculate synapse density, delayed the emergence of eye-specific differences in vesicle organization, and disrupted subsynaptic domain maturation. These results suggest that activity-dependent eye-specific presynaptic maturation underlies synaptic competition in the mammalian visual system. The dLGN relays visual information from the retina to the visual cortex through parallel processing pathways. In adult mice, such processing is achieved through spatial clustering of several retinal ganglion cells (RGCs) boutons to integrate convergent or divergent visual information. It is unknown whether such RGC synapse clustering occurs during the early developmental stage. In Chapter 3, we identified a subset of complex retinogeniculate synapses with larger presynaptic vesicle pools and multiple AZs that simultaneously promote the clustering of like-eye synapses (synaptic stabilization) and prevent synapse formation from the opposite eye (synaptic punishment). In mutant mice with disrupted spontaneous retinal wave activity, complex synapses are formed but fail to drive eye-specific synaptic clustering and punishment. These results reveal the early formation of a unique synaptic subset that regulates activity-dependent eye-specific synaptic competition and may serve as substrates for later synapse clustering formation. A subset of RGCs that express the photopigment melanopsin (OPN4) innervate the suprachiasmatic nucleus (SCN), which serves as the central pacemaker responsible for controlling circadian rhythm in mammals. The function of OPN4 is important for SCN photoentrainment, but its impact on retinal synapse maturation during early development is unknown. In Chapter 4, we found that OPN4 plays an important role in retinal synapse formation and activation in the SCN during the early developmental stage. Loss of OPN4 leads to reduced retinal synapse density, and increased variability in the ratio of synapses with few or no docking vesicles, but has not effect on total vesicle pool volume. Meanwhile, the subsequent maturation of retinohypothalamic tract (RHT) synapses after the first postnatal week shows diminished reliance on OPN4 function and further compensates for the early defects in the absence of OPN4. This study reveals a moderate influence of OPN4 on early RHT synapse development and sheds light on the role of photopigment in regulating SCN synapse plasticity. This dissertation introduces a novel approach using super-resolution fluorescent imaging in the thalamus and hypothalamus tissue. Our work has yielded insights into the activity-dependent maturation in synapse properties and spatial distribution in the dLGN, as well as the impact of OPN4 on retinohypothalamic synapses in the SCN. By revealing the synapse development at the molecular level, our study demonstrates presynaptic mechanisms that underlie activity-dependent retinal synapse plasticity during the early developmental stage. Furthermore, our application of super-resolution fluorescent microscopy highlights its potential as a valuable tool for future in situ studies on brain development.Item MULTISCALE MEASUREMENTS OF ELECTRICAL & MECHANICAL CELLULAR DYNAMICS(2023) Alvarez, Phillip; Losert, Wolfgang; Biophysics (BIPH); Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This dissertation focuses on the study and measurement of coupled electrical and mechanical responses in mammalian cells, tissues, and organs. Cellular biophysics often studies forces and their impact on biochemical pathways. These forces can be electrical, resulting in neuronal action potentials or cardiac cell contractions, or mechanical, driving e.g., a cell’s ability to recognize physical probing or surface texture. These forces and their responses, though, are frequently coupled through interlinked cellular mechanisms which result in emergent responses that take both electrical and mechanical signals into account. One challenge in capturing these emergent responses is that they occur on multiple scales, from the intracellular scale to the organ scale, limiting the ability of commercial microscopes to image these responses simultaneously. In this work I use surface texture, optical imaging, and multiscale-capable image analysis algorithms across these scales to elicit and measure electrical and mechanical responses. To image emergent responses from electrical and mechanical coupling, I developed two custom microscopes that can image at multiple length scales and timescales simultaneously. The Multiscale Microscope can capture slow intracellular mechanical dynamics concurrently with fast tissue scale electrical dynamics, while the BEAMM microscope links fast tissue scale electrical dynamics with both intracellular mechanical dynamics and slower organ-scale mechanical and electrical responses. Finally, I describe ongoing and future studies which exploit these new capabilities for multiscale measurements of electrical and mechanical dynamics.Item Spatiotemporal Dynamics and Functional Organization of Auditory Cortex Networks(2021) Bowen, Zac; Kanold, Patrick O; Losert, Wolfgang; Biophysics (BIPH); Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The sensory cortices of the brain are highly complex systems that are uniquely adapted to reliably process any encountered sensory stimulus. Sensory stimuli such as sound are encoded in large populations of neurons that exhibit some functional organization in the cortex. For example, the auditory cortex has a characteristic organization of sound frequency by which neuronal responses are organized. However, this organization is a broad approximation of more complex and diverse functional properties of individual neurons. Furthermore, on a finer temporal scale, the moment-to-moment activity dynamics of populations of neurons are incredibly complex. Numerous studies have shown that spatiotemporal cascades of co-active neurons organize as neuronal avalanches possessing certain characteristics such as size, duration, and shape that fit the parameters of a critical system. Nevertheless, it remains that the exact manner in which neuronal populations encode information is still not fully understood. This dissertation makes use of neuroimaging data acquired with 2-photon calcium imaging of the auditory cortex in awake mice to investigate the spatiotemporal and functional organization of active neuronal populations in auditory cortex at a range of temporal and spatial scales. I aimed to gain a deeper understanding into how neuronal population dynamics and the underlying network organization contribute to sound encoding in auditory cortex. I studied input and associative layers of auditory cortex (L4 and L2/3) in a mouse model with normal hearing and another with age-related hearing loss due to loss of proper cochlear function to high-frequency sound. L4 and L2/3 contained populations of neurons with a large diversity in functional properties, though diversity was reduced in the hearing loss model due to paucity of high frequency tuned neurons. Despite the diverse tuning in both, similarly responding neurons tended to be co-localized in cortical space. I found that this result extended to volumetric samples of L2/3 where large populations of neurons contained a functional network architecture indicative of small-world topology. Furthermore, I demonstrated that L4 and L2/3 contain ensembles of co-active neurons indicative of critical dynamics in both the absence and presence of a stimulus. Finally, I developed software that facilitates real-time quantification of neuronal populations during an experiment which opens the door for novel closed-loop experiment design. This dissertation provides several avenues for further investigation into neuronal population coding and dynamics, functional network topology, and provides the groundwork for closed-loop experimental design.Item STABILITY AND SCALING OF NEURONAL AVALANCHES AND THEIR RELATIONSHIP TO NEURONAL OSCILLATIONS(2019) Miller, Stephanie Regina; Roy, Rajarshi; Biophysics (BIPH); Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The generation of cortical dynamics in awake mammals is not yet fully understood. However, it is known that neurons leverage distinct organizational schemes to achieve behavior and cognitive function, and that this precise spatiotemporal organization may go awry in illness. In 2003, a form of scale-free synchrony termed “neuronal avalanches” was first observed by Beggs & Plenz in cultured cortical tissue and later confirmed in rodents, nonhuman primates, and humans. In this dissertation, we draw from monkey and rodent studies to demonstrate that neuronal avalanches capture key features of neural population activity and constitute a robust and stable (e.g. self-organized) indicator of balanced excitation and inhibition in cortical networks. We also show for the first time that neuronal avalanches and oscillations co-exist in frontal cortex of nonhuman primates and identify the avalanche temporal shape as a biomarker predicated upon critical systems theory. Finally, we present progress towards characterizing altered avalanche dynamics in a developmental mouse model for schizophrenia using 2-photon calcium imaging in awake animals.Item LARGE-SCALE NEURAL NETWORK MODELING: FROM NEURONAL MICROCIRCUITS TO WHOLE-BRAIN COMPLEX NETWORK DYNAMICS(2018) Liu, Qin; Anlage, Steven; Horwitz, Barry; Physics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Neural networks mediate human cognitive functions, such as sensory processing, memory, attention, etc. Computational modeling has been proved as a powerful tool to test hypothesis of network mechanisms underlying cognitive functions, and to understand better human neuroimaging data. The dissertation presents a large-scale neural network modeling study of human brain visual/auditory processing and how this process interacts with memory and attention. We first modeled visual and auditory objects processing and short-term memory with local microcircuits and a large-scale recurrent network. We proposed a biologically realistic network implementation of storing multiple items in short-term memory. We then realized the effect that people involuntarily switch attention to salient distractors and are difficult to distract when attending to salient stimuli, by incorporating exogenous and endogenous attention modules. The integrated model could perform a number of cognitive tasks utilizing different cognitive functions by only changing a task-specification parameter. Based on the performance and simulated imaging results of these tasks, we proposed hypothesis for the neural mechanism beneath several important phenomena, which may be tested experimentally in the future. Theory of complex network has been applied in the analysis of neuroimaging data, as it provides a topological abstraction of the human brain. We constructed functional connectivity networks for various simulated experimental conditions. A number of important network properties were studied, including the scale-free property, the global efficiency, modular structure, and explored their relations with task complexity. We showed that these network properties and their dynamics of our simulated networks matched empirical studies, which verifies the validity and importance of our modeling work in testing neural network hypothesis.Item PRINCIPLES OF INFORMATION PROCESSING IN NEURONAL AVALANCHES(2011) Yang, Hongdian; Roy, Rajarshi; Plenz, Dietmar; Chemical Physics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)How the brain processes information is poorly understood. It has been suggested that the imbalance of excitation and inhibition (E/I) can significantly affect information processing in the brain. Neuronal avalanches, a type of spontaneous activity recently discovered, have been ubiquitously observed in vitro and in vivo when the cortical network is in the E/I balanced state. In this dissertation, I experimentally demonstrate that several properties regarding information processing in the cortex, i.e. the entropy of spontaneous activity, the information transmission between stimulus and response, the diversity of synchronized states and the discrimination of external stimuli, are optimized when the cortical network is in the E/I balanced state, exhibiting neuronal avalanche dynamics. These experimental studies not only support the hypothesis that the cortex operates in the critical state, but also suggest that criticality is a potential principle of information processing in the cortex. Further, we study the interaction structure in population neuronal dynamics, and discovered a special structure of higher order interactions that are inherent in the neuronal dynamics.