Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Brain and Cognitive Sciences, 2006. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Includes bibliographical references. / The ability to retain multiple items in short-term memory is fundamental for cognition, yet almost nothing is known about its neural basis. To explore the mechanisms underlying this ability, we trained two monkeys to remember a sequence of two images across a short delay. We then recorded the activity of neurons from the lateral prefrontal cortex during task performance. We found that the majority of neurons showed delay activity that depended on the identity of both images (a minority reflected just one image), and that activity related to a given combination of images was only partially predictable from each neuron's activity to individual images. A model to predict the resultant neural activity was tested. We also examined the effect of task demands on the neural representation of multiple images. Our first experiment showed that each of the two images in memory was represented with a certain strength, and that this strength was dependent on how long the image had been in memory; image strength decayed as time progressed. / (cont.) We found that changing the way that the memory of the images was reported, from a bar release to a sequence of eye movements, changed the relative strength of the image representations. In the eye-movement version of the task the strength of the representation of the image did not decay with time; in fact the strength of older images could even surpass the strength of newer images, depending on how frequently the tasks were switched. Further experiments showed that when the monkey switched between the two tasks individual neurons could turn their image coding on and off. We also found a substantial population of cells that directly represented the task that the animal was performing. / by Melissa R. Warden. / Ph.D.

Thesis: Ph. D. in Neuroscience, Massachusetts Institute of Technology, Department of Brain and Cognitive Sciences, 2016. / Cataloged from PDF version of thesis. Vita. / Includes bibliographical references (pages 121-134). / The roles played by cortical inhibitory neurons in experience-dependent plasticity and learning are not well understood. Here we evaluate the participation of parvalbumin-expressing (PV+) GABAergic neurons in two forms of experience-dependent modification of primary visual cortex (V1) in adult mice: ocular dominance (OD) plasticity resulting from monocular deprivation and stimulus-selective response potentiation (SRP) resulting from supplemental visual experience. These two forms of plasticity are triggered by different events but lead to a similar increase in visual cortical response. Both also require the NMDA class of glutamate receptor (NMDAR). However, we find that PV+ inhibitory neurons in V1 play a critical role in the expression of SRP and its behavioral correlate of familiarity recognition, but not in the expression of OD plasticity. Furthermore, NMDARs expressed within PV+ cells play a critical role in SRP, but not in the induction or expression of adult OD plasticity. We also explore the use of visual cortical plasticity paradigms to better understand the function of proteins implicated in autism spectrum disorders (ASDs) and schizophrenia. We find that NMDAR-dependent long-term depression (LTD) and deprived-eye depression in layer 4 of V1 require metabotropic glutamate receptor 5 (mGluR5) signaling during postnatal development. Additionally, schizophrenia-associated protein neurogranin overexpression in V1 disrupts juvenile ocular dominance plasticity. Finally, we evaluate SRP in two models of ASDs associated with excitatory/ inhibitory imbalance: Rett syndrome (RTT) and tuberous sclerosis complex (TSC). Surprisingly, mouse models of RTT and TSC exhibit abnormal SRP phenotypes, but in opposite directions. / by Eitan S. Kaplan. / Ph. D. in Neuroscience

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Brain and Cognitive Sciences, 1995. / Includes bibliographical references (leaves 118-125). / by Tracy S. Myers. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Brain and Cognitive Sciences, 2003. / Includes bibliographical references. / The large majority of inputs to primary visual cortex (V1) neurons arise from the dense local projections of neighboring neurons. Although several proposals have been made, it is not known what role these connections play in shaping the response properties of V1 neurons. It was reasoned that the influence of local inputs on orientation tuning should be different at different locations in the orientation map, because the available data suggest that the orientation composition of the local connections varies with location in the map. In particular, near pinwheel centers, neurons of varying orientation preferences are likely to be connected, whereas far from pinwheel centers, the local connections are likely to only connect similarly tuned neurons. To approach this issue, the responses of neurons at pinwheel centers, and in orientation domains have been compared. The subthreshold responses are found to have much broader orientation tuning near pinwheel centers, reflecting the broader orientation specificity of local connectivity. However, the broad subthreshold inputs are filtered out by the spike threshold and strong inhibition, such that spike tuning is similar at all locations in the orientation map. Spike tuning in pinwheel neurons is found to be sharp during the entire timecourse of the response, suggesting that the mechanism that sharpens the broad inputs is rapid and stable. The broad connectivity near pinwheel centers also leads to correlated firing between pairs of neurons with widely different orientation preferences. Thus, the local inputs to a V1 neuron depend on its location in the orientation map, but the inputs are filtered to produce sharp orientation tuning, regardless of the selectivity of the inputs. / by James M. Schummers. / Ph.D.

Thesis: Ph. D. in Neuroscience, Massachusetts Institute of Technology, Department of Brain and Cognitive Sciences, 2016. / Cataloged from PDF version of thesis. Page 262 blank. / Includes bibliographical references. / As we interact with the environment, our senses are constantly bombarded with information. Neurons in the visual cortex have to transform these complex inputs into robust and parsimonious neural codes that effectively guide behavior. The ability of neurons to efficiently convey information is, however, limited by intrinsic and shared variability. Despite this limitation, neurons in primary visual cortex (V1) are able to respond with high fidelity to relevant stimuli. My thesis proposes that high fidelity encoding can be achieved by dynamically increasing trial-to-trial response reliability. In particular, in this thesis, I use the mouse primary visual cortex (V1) as a model to understand how reliable coding arises, and why it is important for visual perception. Using a combination of novel experimental and computational techniques, my thesis identifies three main factors that can modulate intrinsic variability. My first goal was to understand the extrinsic, stimulus-dependent, factors responsible for reliably coding (Chapter 3). Natural scenes contain unique statistical properties that could be leveraged by the visual cortex for efficient coding. Thus, the first aim is to elucidate how image statistics modulate reliable coding in V1. To this end, I developed a novel noise masking procedure that allowed us to specifically perturb the spectral content of natural movies without altering the edges. Using high-speed twophoton calcium imaging in mice, I discovered that movies with stronger spatial correlations are more reliably processed by V1 neurons than movies lacking these correlations. In particular, perturbing spatial correlations in the movie dynamically altered the structure of interneuronal correlations. Movies with more naturalistic correlations typically recruited large neuronal ensembles that were weakly noise correlated. Using computational modeling, I discovered that these ensembles were able reduce shared noise through divisive normalization. Together, these findings demonstrate that natural scene statistics dynamically recruit neuronal ensembles to ensure reliable coding. Microcircuits of inhibitory interneurons lie at the heart of all cortical computations. It has been proposed that these interneurons are responsible for reliable spiking by controlling the temporal window over which synaptic inputs are integrated. However, no study has yet conclusively investigated the role of different interneuron subtypes. Thus, my second goal was to establish how natural scenes are reliably encoded by dissecting the inhibitory mechanisms underlying reliable coding (Chapter 4). Specifically, I investigated the role of somatostatin-expressing dendrite targeting interneurons (SST) and parvalbumin-expressing soma targeting interneurons (PV), which are known to provide distinct forms of inhibition onto pyramidal neurons. Using a novel combination of dual-color calcium imaging and optogenetic manipulation, I have discovered that the SST->PV inhibitory circuit plays a crucial role in modulating pyramidal cell reliability. In particular, by transiently suppressing PV neurons, SST neurons are able to route inhibition rapidly from the soma to the dendrites. Strong dendritic inhibition allows noisy inputs to be filtered out by the dendrites, while weaker somatic inhibition allows these inputs to be integrated to produce reliable spikes. In agreement with these results, I found that selectively deleting MeCP2 from these interneurons resulted in unreliable visual processing and other circuit-specific deficits, which are commonly observed in Rett Syndrome (Chapter 5). These results underscore the importance of intact inhibitory microcircuits in reliable processing. Finally, my goal was to determine why reliable coding is necessary for visual processing (Chapter 6). To this end, I trained head-fixed mice to perform a natural movie discrimination task. Mice were able to learn how to discriminate between two movies after a short training period. By perturbing the amplitude spectrum of these movies, I discovered that mice used structural information in the phase spectrum to discriminate between the different movies. This suggests that mice also use similar strategies as higher mammals for scene recognition. Inspired by this result, we trained mice on a harder target categorization task, where mice had to identify the movies from an ensemble that were more similar to the target movie to gain a water reward. We developed this movie ensemble by blending together the phase spectrum of a target and non-target movie at different fractions. Optically activating SST neurons in V1 improved the ability of mice to correctly identify "target-like" movies. This increase in behavioral performance correlated well with an increase in V1 coding reliability. Thus, reliable codes are a prerequisite for accurate visual perception. Taken together, this work bridges the gap between cells, circuits and behavior, and provides mechanistic insight into how complex visual stimuli are encoded with high fidelity in the visual cortex. / by Rajeev V. Rikhye. / Ph. D. in Neuroscience

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Brain and Cognitive Sciences, 2016. / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 68-77). / Humans are able to simultaneously perceive the world as discrete, distinct "objects", as well as regions of statistical regularity, or "textures". This is evident in the way we describe our perceptual world. A street is made up of concrete and asphalt "stuff", while the people and dogs walking on it are the "things" that make use of it. Both of these types of representation, however, are derived from the same sensory input, and thus there must exist transformations that map one to the other. A complete model of perception must account for these transformations. I study the representations that lie at the interface of object and texture perception in vision, focusing on utilizing the intrinsically impaired perception in the periphery to disambiguate the predictions of different models. I find that many seemingly separate perceptual phenomena in crowding can be better understood as different aspects of a single underlying model. I extend this to the study of material perception, and find that considering images of materials as visual textures can explain human's ability to recognize materials in the periphery. Furthermore, I examine how the limitations of peripheral vision affects the perception of visual designs, namely webpages. / by Shaiyan Keshvari. / Ph. D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Brain and Cognitive Sciences, 2002. / Includes bibliographical references (p. 145-150). / The work presented in this thesis was conducted with two aims in mind. The first was to understand where speakers prefer to place intonational boundaries in language production. The second was to understand where listeners prefer to hear boundaries in language comprehension. / by Duane G. Watson. / Ph.D.

Thesis (Ph.D. in Computational Cognitive Science)--Massachusetts Institute of Technology, Dept. of Brain and Cognitive Sciences, 2002. / Includes bibliographical references (leaves 72-[77]). / This thesis proposes a miethodology for the design of man-machine interfaces by combining top-down and bottom-up processes in vision. From a computational perspective, we propose that the scientific-cognitive question of combining top-down and bottom-up knowledge is similar to the engineering question of labeling a training set in a supervised learning problem. We investigate these questions in the realm of facial analysis. We propose the use of a linear morphable model (LMM) for representing top-down structure and use it to model various facial variations such as mouth shapes and expression, the pose of faces and visual speech (visemes). We apply a supervised learning method based on support vector machine (SVM) regression for estimating the parameters of LMMs directly from pixel-based representations of faces. We combine these methods for designing new, more self-contained systems for recognizing facial expressions, estimating facial pose and for recognizing visemes. / by Vinay P. Kumar. / Ph.D.in Computational Cognitive Science

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. / "February 2009." / Includes bibliographical references. / The visually evoked potential (VEP) is a local field potential (LFP) evoked in visual cortex in response to visual stimuli. Unlike extracellular single unit recordings, which allow us to probe the function of single spiking cells acutely, the chronic VEP technique gives us insight into ensemble synaptic activity. However, while action potentials are easily interpreted as the output of the recorded neuron, LFPs are difficult to interpret because they may reflect the sum of activity occurring at or beyond the site of recording. The goal of this study was to use the current source density (CSD) method to derive information about synaptic activity occurring at the site of recording and to determine how this activity relates to the concurrent LFP. The mouse has recently become a widely-used experimental model for studying the mechanisms of plasticity and there has been an increase in the use of VEP recordings to study experience-dependent changes in mouse primary visual cortex (V1). These studies typically focus on changes occurring in the layer 4 VEP after a variable period of visual deprivation. Layer 4 of mouse V1 receives heavy direct input from the lateral geniculate nucleus. This initial input is followed by strict hierarchical connectivity from cortical layer 4 to superficial layers 2/3 and from 2/3 to deep layers 5/6. Using a method for silencing cortical activity without affecting geniculate input activity in conjunction with CSD analyses, we found that the laminar flow of activity in mouse V1 in response to various grating stimuli was consistent with the anatomical connectivity going from layer 4 ?? 2/3 ?? 5/6. To determine if the layer 4 VEP is indeed reflecting synaptic activity occurring in layer 4, we applied the CSD method to field potentials recorded from mouse V1. Our results indicate that changes in the layer 4 VEP strongly and significantly covaries with changes in layer 4 current sink activity suggesting that the layer 4 VEP is indeed reflecting local layer 4 synaptic activity. / (cont.) This layer 4 activity is likely due to direct geniculate input since it persisted after intracortical activity was blocked. If the layer 4 VEP reflects synaptic activity due to direct geniculo-cortical input and if this input is carrying information about the visual world then we would expect the VEP to change as the parameters of the stimuli vary. Indeed the binocular-driven VEP broadened in shape as we increased the spatial frequency (SF) of grating stimuli. Using CSD analyses, we were able to trace the transformations of the layer 4 VEP waveform to changes happening in layer 4 current sinks and layer 4 current sinks were in turn affected by events in deep layers. Specifically, increasing SF of the grating stimuli led to a reduction of current sink activity in deep layers and this unmasked prolonged current sink activity in layer 4. This prolonged layer 4 current sink activity persisted after cortical silencing suggesting that it is likely due to late-onset direct geniculate input. We suggest that late-onset activity from the ipsilateral-eye may be unmasked with increasing SF. VEPs have been used extensively in the clinical and laboratory setting to determine visual acuity in humans as well as anaesthetized animals. If the layer 4 VEP is to be a useful measure of visual function in awake head-fixed mice, VEP-assessed visual acuity and contrast sensitivity should be consistent with behaviorally-assessed measures. We found that VEP-assessed visual acuity agreed with previous behaviorally-assessed acuity; however, VEP-assessed contrast-sensitivity values were slightly higher. One of the reasons why inbred laboratory mice are becoming increasingly useful in Neuroscience is because individual mice are genetically identical and any behavioral variability should be experience-driven. While this is true for mice within a given strain, it is not true between strains since strains are genetically different. Therefore, it is crucial to understand how strain differences in genes affects neural activity before comparing results from different strains. / (cont.) To this end, we compared the VEP response of two commonly used laboratory mouse strains: C57BL/6 and 129/Sv and found important differences in the VEP waveform which may translate into differences in visual function. Specifically, our data suggest that 129/Sv mice may have better acuity than C57BL/6 mice. The advent of molecular engineering tools is another reason why the mouse has become the preferred model system for studying the cellular and molecular mechanisms underlying behavioral and physiological phenomena. Genetically modified mice are routinely screened for behavioral deficits using tasks such as the Morris watermaze -- test for spatial navigation which assumes that the mice have functional vision. In order to remove the experimental confound of vision, the layer 4 VEP can be used to assay the visual function of mice prior to behavioral experimentation. Using the VEP technique, we determined the visual function of Shank1-/- mice to be normal in response to low SF gratings but impaired in response to high SF gratings. Shank1-/- mice were not impaired in the eight-arm radial maze task - another test of spatial navigation suggesting that low SF vision may be sufficient for performing this task. Taken together, this study demonstrates that the VEP is an interpretable and useful recording technique which can be combined with CSD analysis to determine the laminar activity patterns which underlie visual function in the awake mouse. / by Rahmat Muhammad. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Brain and Cognitive Sciences, 1993. / Includes bibliographical references (v. 2, leaves 408-416). / by Gary Fred Marcus. / Ph.D.