Thesis: Ph. D., Massachusetts Institute of Technology, Department of Brain and Cognitive Sciences, 2016. / Cataloged from PDF version of thesis. / Includes bibliographical references. / Incoming sensory information from all modalities, with the exception of olfaction, synapses in the thalamus on the way to neocortex. This sensory relay is uniquely positioned to act as a gate, to determine which inputs from the periphery are processed by the neocortex. A key 'guardian' of the gate may be the thalamic reticular nucleus (TRN). The TRN is a primary source of GABAergic input to thalamic relay nuclei. The TRN projects directly to the rest of thalamus, generating feedforward and feedback inhibition. It is therefore positioned to mediate forebrain function, and specifically the computations of the neocortex-thalamic loop. Accordingly, failures of the normal dynamics of the TRN are prominent in disease Thalamocortical and corticothalamic projections synapse within this nucleus, and it is subject to a variety of neuromodulatory influences. Depolarization of TRN neurons, and their subsequent firing, is driven by a variety of sources on a range of time scales. The TRN receives excitatory inputs ranging from single spikes to sustained tonic firing to bursting in thalamic relay neurons or layer 6 of neocortex. The temporal dynamics of these inputs, and their spatial organization, can drive different types of firing behavior in TRN. Layer 6 cells form strong synapses in the TRN and even sparse activity in this layer would be predicted to drive substantial inhibition in vivo. Primary thalamocortical relay projections branch into the TRN on their way to sensory cortices, and the nature of this excitatory input reflects the functional modes of the relay nuclei. Inputs include tonic firing that reflects high fidelity to peripheral input, as well as extended bouts of bursting, similar to that seen in TRN itself. In sum, a variety of inputs can excite TRN neurons on different time scales. Understanding how these different patterns may regulate excitability in general, and burst activity specifically, is key to understanding thalamocortical function. The Moore laboratory previously showed that TRN activation could modulate firing and bursting in relay neurons, and induce spindles in the neocortex. In these experiments, the activity of TRN cells during stimulation could only be inferred from downstream effects on spiking and spindle rhythms. Characterizing responses within TRN using a similar stimulation protocol provided a more complete view of the circuit activity underlying this evoked behavior. In Chapter 2 I provided optogenetic input while characterizing multi-unit responses in the TRN and well-sorted single units. I found that longer duration activation drove enhanced bursting and decreased latency to bursting. I also discovered two new -types of cell responses, a more sensitive 'non-linear' cell type that was prone to sustained responses and to bursting, and a more 'linearly' responsive cell class that fired in direct proportion to the duration of stimulation. These findings provide direct predictions as to the behavior of TRN neurons in response to a range of natural depolarizing inputs, and a guide for the optical control of this key structure in studies of network function and behavior. As indicated by the availability of neuromodulatory inputs to TRN, and its apparent role in basic state changes such as sleep and wakefulness, long-term shifts in its depolarization are also likely essential to normal brain function. Optogenetics has rapidly become a standard technique in systems neuroscience, and its genetic specificity and rapid development of new compounds has revolutionized our ability to causally manipulate neural circuits. While the use of light to drive cellular reactions brings a number of advantages when compared to electrical stimulation, there are still many limitations, especially in vivo. Light delivery through tissue is problematic in the intact brain, so targeting deep structures relies on implanted fiber optics and/or LEDs. These methods are not ideal for illuminating large or irregularly-shaped regions without using high light intensity or large arrays of invasive devices. I have been key in inventing a new approach using bioluminescent light to drive optogenetic responses ('BL-OG'). This approach leverages the variety of light sensitive molecules and bioluminescent emitters while providing a means of chemical control. BL-OG combines the cell-type specificity of conventional optogenetics with the potential for noninvasive, system-wide activation. In Chapter 3, I review both this new method and some of my contributions to its realization, specifically demonstrating its functionality in the TRN in vitro and in vivo. / by Bryan T. Higashikubo. / Ph. D.

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Brain and Cognitive Sciences, 2014. / Cataloged from PDF version of thesis. / Includes bibliographical references. / In the core of this thesis I test and confirm the hypothesis that separate classes of interneurons respond differentially to sensory stimulation independent of volitional or other top-down control on the part of the animal. I also test and confirm the hypothesis that, based only on bottom-up sensory stimulation the activity of two major classes of interneurons (adapting Parvalbumin positive and facilitating Somatostatin positive interneurons) predominates during different phases of what corresponds to natural sensing cycles in a behaving rodent. These questions are addressed using an in vivo mouse model with intrinsically fluorescent, but differentiable, interneuron populations combined with 2-photon imaging, Ca²+-sensitive dyes. Anesthesia and electrical control of facial muscles allowed for naturalistic stimulation without the confounds presented by volitional whisking and unknown top-down or behavioral states. Additional chapters in this thesis focus on ancillary work related to computational modeling of neural systems and systems' level perspectives on maturation and disease. / by Ethan M. Skowronski-Lutz. / Ph. D.

Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Brain and Cognitive Sciences, 1998. / Includes bibliographical references (p. 45-46). / The ferret dorsal lateral geniculate nucleus (dLGN) undergoes two periods of retinal afferent segregation during postnatal development. The first establishes the eye-specific laminae, A and A 1, and the second establishes ON/OFF sublaminae within laminae A and At. In contrast to eyespecific segregation, which seems to rely only on presynaptic activity, ON /OFF sublamination requires both pre- and postsynaptic activity. Because of its dependence on postsynaptic (relay cell) activity, sub lamination may be influenced by extra retinal inputs which alter relay cell excitability. We have examined the postnatal development of cholinergic brainstem inputs to the dLGN to determine whether these inputs arrive in time to influence sublamination and whether the cholinergic innervation is present in laminar zones (i.e., whether it might target relay cells). Choline acetyltransferase (ChA1) immunoreactivity is not detected in the dLGN until a few days after the second postnatal week Gust after sublamination begins), at which time it can be seen in both A and C laminae. ChA T labeling increases in intensity until two days before the end of the fourth postnatal week (when ON/OFF sublamination is complete), when it drops dramatically throughout the dLGN . ChAT labeling returns a few days later, but appears in the interlaminar and intersublaminar zones instead of within the A and C laminae. However, the pattern of ChA T labeling reverses once more, so that in the adult, ChA T labeling appears in the A and C laminae and is relatively absent from interlaminar zones. Acetylcholinesterase (AChE) labeling in the dLGN shows a similar ontogenetic pattern and time course. Retrngrade labeling of brainstem cholinergic nuclei demonstrates that these inputs are in place in the dLGN after the second postnatal week. Thus, cholinergic inputs to A and A 1 laminae of the ferret dLG N do arrive in time to influence ON/OFF sublamination. / by Sherri Lynn Hitz. / S.M.

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Brain and Cognitive Sciences, 2018. / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 129-140). / In many situations, from economists predicting unemployment rates to chemists estimating fuel safety, individuals have differing opinions or predictions. We consider the wisdom-of-the-crowd problem of aggregating the judgments of multiple individuals on a single question, when no outside information about their competence is available. Many standard methods select the most popular answer, after correcting for variations in confidence. Using a formal model, we prove that any such method can fail even if based on perfect Bayesian estimates of individual confidence, or, more generally, on Bayesian posterior probabilities. Our model suggests a new method for aggregating opinions: select the answer that is more popular than people predict. We derive theoretical conditions under which this new method is guaranteed to work, and generalize it to questions with more than two possible answers. We conduct empirical tests in which respondents are asked for both their own answer to some question and their prediction about the distribution of answers given by other people, and show that our new method outperforms majority and confidence-weighted voting in a range of domains including geography and trivia questions, laypeople and professionals judging art prices, and dermatologists evaluating skin lesions. We develop and evaluate a probabilistic generative model for crowd wisdom, including applying it across questions to determine individual respondent expertise and comparing it to various Bayesian hierarchical models. We extend our new crowd wisdom method to operate on domains where the answer space is unknown in advance, by having respondents predict the most common answers given by others, and discuss performance on a cognitive reflection test as a case study of this extension. / by John Patrick McCoy. / Ph. D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Brain and Cognitive Sciences, 1994. / Includes bibliographical references (leaves 180-191). / by Sherif Maher Botros. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Brain and Cognitive Sciences, 2009. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Cataloged from student submitted PDF version of thesis. / Includes bibliographical references. / The elucidation of the biological bases of a complex trait like human language proceeds from identification of precise behavioral phenotypes to investigation of the underlying genes. The human behavioral parts of this dissertation focus on understanding the reasons for children's overuse of definite article 'the', to refer to one of several objects in a context set, as opposed to the unique established referent. Competing theories argue the deficit is either in children's semantic computational knowledge (of uniqueness/maximality), or in their pragmatic/social awareness/theory-of-mind development. Experiments in this dissertation focused on children's comprehension and interpretation of the indefinite and definite determiners, as well as 'that', anaphors 'another' and 'same', and free relative clauses. The results in this thesis suggest that in typically developing (TD) children the late acquisition of determiner 'the' is due to the late maturation of the semantic principle of maximality. Children with autism spectrum disorders (ASD) and with Williams syndrome (WS) either manifested an adult-like competence, an absence of manifestation of knowledge, or a pattern found in TD younger children (where 'that' is understood better than 'the' as referring to the salient unique referent) -- indicating delay of development of the language faculty, but no deviance. This suggests that the observed deficits in ASD and WS pattern with those in TD, and hence are also semantic in nature. The mouse neurogenetic part of this dissertation investigates whether the GTF2I family of genes, causal to WS behavioral phenotype, also contributes to WS cortical development. / (cont.) By overexpression of Gtf2i and Gtf2ird1 in the mouse neocortex via in utero electroporation, their effects on laminar patterning and cell morphogenesis during brain development are characterized. The present results suggest that these genes can synergistically contribute to the abnormal neocortical development in WS, and thereby could contribute to language deficits in WS. Beyond posing an explanatory challenge to linguistic theories, the research comparing typical and atypical development sheds light on the mechanisms of language development and impairment, and provides endophenotypic descriptions of ASD and WS, which are crucial for elucidating not only genetics of neurodevelopmental disorders, but also the genetic basis of the human language faculty. / by Nadezhda N. Modyanova. / Ph.D.

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Brain and Cognitive Sciences, February 2017. / Cataloged from PDF version of thesis. "September 2016." / Includes bibliographical references. / Neocortex learns predictive models of sensory input, allowing mammals to anticipate future events. A fundamental component of this process is the comparison between expected and actual sensory input, and the layered architecture of neocortex is presumably central to this computation. In this thesis, I examine the role of laminar differences, and specifically the role of layer 6 (L6) in the encoding and perception of stimuli that deviate from previous patterns. In awake mice, layer 4 neurons encode current stimulus deviations with a predominantly monotonic, faithful encoding, while neurons in layer 2/3 encode history dependent change signals with heterogeneous receptive fields. Corticothalamic (CT) cells in Layer 6 respond sparsely, but faithfully encode stimulus identity. Weak optogenetic drive of L6 CT cells disrupted this encoding in layer 6 without affecting overall firing rates. This manipulation also caused layer 2/3 to represent only current stimuli. In a head-fixed stimulus detection task, small stimulus deviations typically make stimuli more detectable, and the L6 manipulation removed this effect, without affecting detection of non-changing stimuli. Analogously, in free sensory decision making behavior, the manipulation selectively impaired perception of deviant stimuli, without affecting basic performance. In contrast, stronger L6 drive reduced sensory gain and impaired tactile sensitivity. These results show an explicit laminar encoding of stimulus changes, and that L6 can play a role in the perception of sensory changes by modulating responses depending on previous, or expected input. This finding provides a new perspective on how the layered cortical architecture can implement computations on hierarchical models of the world. / by Jakob Voigts. / Ph. D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Brain and Cognitive Sciences, 1990. / Includes bibliographical references (p. 185-201). / by David J. Bennett. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Brain and Cognitive Sciences, 2011. / Cataloged from PDF version of thesis. / Includes bibliographical references (p. 121-130). / The visual system is a dynamic entity whose response properties depend on context and experience. In this thesis, I examine how the brain changes as we learn to see - what changes occur during the onset of recognition, in the mature visual system on the one hand, and in a developmentally nascent one, on the other? Working with normal adults, I focus on the processes that underlie the interpretation of images as meaningful entities. This interpretation is greatly facilitated by prior information about a stimulus. What are the neural sites that exhibit experience dependent changes? Using multivariate decoding techniques, I find pervasive evidence of such changes throughout the visual system. Critically, cortical regions previously implicated in such learning are not the same loci as sites of increased information. Examining the temporal mechanisms of recognition, I identify the perceptual state transitions corresponding to the onset of meaning in an observed image. Furthermore, decoding techniques reveal the flow of information during this 'eureka moment.' I find feedback processing when a degraded image is first meaningfully interpreted, and then a rapid transition into feed-forward processing for more coherent images. Complementing the studies with mature subjects, my work with developmentally nascent observers explores the genesis of visual interpretation. What neural changes accompany the earliest stages of visual learning? I show that children treated for congenital blindness exhibit significant cortical re-organization after sight onset, in contrast to the classical notion of a critical period for visual plasticity. The specific kind of reorganization suggests that visual experience enhances information coding efficiency in visual cortex. Additionally, I present evidence of rapid development of functionally specialized cortical regions. Overall, the thesis presents two complementary perspectives on the genesis of visual meaning. The results help advance our understanding of how short-term experience, as well as developmental history, shapes our interpretation of the complex visual world. / by Scott Gorlin. / Ph.D.

Thesis (Ph. D. in Neuroscience)--Massachusetts Institute of Technology, Dept. of Brain and Cognitive Sciences, 2013. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Cataloged from student-submitted PDF version of thesis. / Includes bibliographical references. / The hippocampus is required for the formation, but not storage, of long-term episodic memories. During memory formation, however, the hippocampus is not a lone actor; rather it works in concert with various structures across the brain. The mechanisms by which diverse populations of cells are coordinated for the formation of a single, coherent memory remain unknown. This thesis is an investigation of the temporal and hemispheric structure of replay events. The timing of replay is investigated at the levels of hippocampal sharp-wave ripples and multi-unit activity. We found that, during sleep, ripples generation is modulated by a 10-15Hz rhythm. We also observed this rhythm in the multi-unit firing rate of hippocampal neurons. Next we investigated and quantified the level of coordination between the hippocampal during replay events. Using bilateral hippocampal recordings from several rats during spatial navigation and subsequent sleep epochs, we directly compared the activity of these two spatially isolated networks at the level of the local field potential and the information encoded by the two neuronal populations. We found that the neural activity of the two hippocampi was highly correlated in some aspects but not others. As previously reported in the mouse, we found that, in the rat, sharp-wave ripples were simultaneously generated spontaneously in both hippocampi and that, although the intrinsic frequencies of ripple oscillations were correlated bilaterally, the phases of the individual ripple wavelets were not. Finally, we found that information encoded by both hippocampal ensembles is highly correlated during replay events. / by Stuart Pope Layton. / Ph.D.in Neuroscience