Este trabalho descreve as pesquisas e desenvolvimentos em instrumentação eletrônica computacional, realizados para viabilizar experiências na área de neurobiofísica, tendo como objetivos principais a geração de estímulos visuais para invertebrados e a captação de sinais eletrofisiológicos gerados por sistemas biológicos sensoriais submetidos a estímulos. Trata-se de um conjunto de equipamentos que, operando de maneira integrada, são capazes de fornecer e sincronizar estímulos, realizar a aquisição dos dados de sinais neurais a serem utilizados para controle e análise em experiências in vivo\" nos estudos da visão de invertebrados no Laboratório de Neurobiofísica - DipteraLab do IFSC. A integração desta instrumentação eletrônica visa facilitar a sua utilização durante os experimentos, permitindo o acompanhamento das aquisições de dados neurais, viabilizando a realização de experimentos com alterações dos estímulos através de realimentação em tempo real. / This work describes the research and development of computational instrumentation to be used in experimental neurobiophysics. The developed electronic modules operate in an integrated manner and are used to generate visual stimuli for invertebrates and capture electrophysiological signals generated by biological systems subjected to sensory stimuli. They are able to provide synchronized stimuli and perform data acquisition of neural signals events to be used for control and analysis of vision experiments with invertebrates at the Laboratory of Neurobiophysics Dipteralab Laboratory, at the IFSC. The integration of electronic instrumentation facilitate its use during experiments allowing, through its monitoring capabilities of the neural data acquisition, the realization of experiments with real time stimuli changes through feedback. The possibility to perform pre-analyses of neural responses in behavioral closed loop experiments is also implemented.

Coronavírus bovino (BCoV) é o agente causador de doença, tanto entérica como respiratória em bovinos, mas até agora existem controvérsias sobre a relação genealógica entre as amostras de BCoV em diferentes tecidos. Neste estudo, amostras de fezes e secreções nasais de 14 vacas de um mesmo rebanho apresentando simultaneamente disenteria epizoótica e doença respiratória foram estudados quanto a presença de BCoV. As amostras virais detectadas tiveram tanto o gene de espícula (S) como o gene hemaglutinina-esterase (HE) parcialmente sequenciados. Para o gene HE, foram obtidas 12 sequências de secreções nasais e 12 de amostras de fezes e para o gene S, foram obtidas 14 sequências de secreções nasais e 12 de amostras de fezes, com 100% de identidade nucleotídica para cada gene para as amostras deste estudo. Estes resultados apresentam algumas divergências com estudos anteriores os quais relatam que linhagens diferentes de BCoV podem ser esperados em casos de disenteria e doença respiratória em vacas, pois linhagens com sequências idênticas dos genes S e HE podem não mostrar diferenças em relação tropismo pelos diferentes tecidos. Sequências completas de duas amostras brasileiras de BCoV mostram que o já descrito padrão filogeográfico baseado no sequenciamento do gene S parcial foi mantido, foram encontradas substituições de aminoácidos específicos. / Bovine coronavirus (BCoV) is the causative agent of both enteric and respiratory disease in cattle, but hitherto there were some controversy on the genealogic relationship amongst strains from these different tissues. In this study, samples of feces and nasal secretions of 14 cows from a same herd simultaneously presenting epizootic dysentery and respiratory disease were screened for BCoV and the strains detected had both the spike (S) and hemagglutinin-esterase (HE) genes partially sequenced. For HE gene, 12 sequences from nasal secretions and 12 from fecal samples were obtained and for S gene, 14 sequences from nasal secretions and 12 from fecal samples were obtained, with 100% nucleotide identities for each gene for the strains of this study. These results have some disagreements with previous reports which try to put forward that divergent BCoV strain should be expected in cases of dysentery and respiratory disease in cows, showing that strain with identical S and HE sequences might show no differences in tropisms. Complete S gene sequences of two Brazilian BCoV strains show that the already described phylogeographic pattern based on partial S gene is sustained, though specific amino acids subtitutions are found.

Made available in DSpace on 2014-07-29T15:07:37Z (GMT). No. of bitstreams: 1 Dissertacao Reginaldo Jacovetti.pdf: 647335 bytes, checksum: 8f3872d3d6f0b90e4cc497b5edf0f833 (MD5) Previous issue date: 2012-02-27 / The Brazilian farmers make use of vary forages plants to be conserved in the ensilage form, for supplementation of the animals in the period of fall of pastures. However, forage millet presents as a possible alternative, in view of its potential of production and possibility of culture as off-season cultivation. The objective of this work was to evaluate the production and the morph-anatomical composition of millet compared to some forages, to characterize the fermentative standard of the produced ensilages, as well as, estimate the cost of production and the potential of milk production in function of the dry substance of the produced ensilages. The experiment was divided in three phases: (1) Quantitative Analysis: anatomical composition, agronomics parameters, production of dry mass of the used forages plants for the silage process; (2) Qualitative Analysis: characterization of the fermentative process, through the opening of the experimental in different times (3, 7, 15, 30, 60 and 120 days) after the silage incubation, proceeding itself the effluent evaluations of loss of and gases beyond the determination and chemical composition and fraction of protein of the ensilages; (3) Economic Analysis: survey of the costs of production of the ensilage of millet in relation to other traditional forages cultures. Millet presented the greater stand with 237,000 plants/ha and the maize without spike with the minor stand a 58,300 plants/ha. Already the production of green mass (GM) of sorghum was raised with 60,400 kg GM/ha, followed for the integral maize with 50,200 kg/ha. The production of dry mass (DM) of the integral maize was of 20,000 kg DM/ha and, of millet, was of 9,700 kg DM/ha. About protein fraction, they not had differences (P<0.05) for the fraction A in the ensilage of millet in the different times of opening of the silos. In the economic analysis, the maize presented a cost of lesser production in function of its bigger income for area. Although millet to present potential of similar production to the maize, the production accomplishes was lesser, what comparatively it would make impracticable its use. So that this either made possible economically is necessary that the raised production either more. / Os pecuaristas do Brasil dispõem de várias plantas forrageiras para serem conservadas na forma de silagem para suplementação dos animais no período de escassez de pastagens. Contudo, o milheto forrageiro é uma possível alternativa, tendo em vista o seu potencial de produção e possibilidade de cultivo como safrinha. O objetivo deste trabalho foi avaliar o milheto comparando com forrageiras para silagem (milho integral, milho sem espiga, sorgo e cana-de-açúcar) sob aspectos de produção e composição morfo-anatômica, caracterização do padrão fermentativo das silagens produzidas nos tempos de abertura dos mini-silos em 3, 7, 15, 30, 60 e 120 dias, assim como, estimar o custo de produção e o potencial de produção de leite em função da matéria seca das silagens produzidas. O experimento foi dividido em três fases: (1) Análise Quantitativa: composição anatômica, parâmetros agronômicos, produção de massa seca das plantas forrageiras utilizadas para o processo de ensilagem; (2) Análise Qualitativa: caracterização do processo fermentativo, através da abertura dos mini-silos em diferentes tempos (3, 7, 15, 30, 60, 120 dias) após a ensilagem, procedendo-se as avaliações de perda de efluentes e gases, além da determinação e composição bromatológica e fracionamento de proteína das silagens; (3) Análise Econômica: levantamento dos custos de produção da silagem de milheto em relação a outras culturas forrageiras tradicionais. O milheto apresentou o maior estande com 237 mil plantas/ha e o milho sem espiga com o menor estande 58,3 mil plantas/ha. Já a produção de massa verde (MV) do sorgo foi a mais elevada com 60,4 ton/ha, seguido pelo milho integral com 50,2 t/ha. A produção de matéria seca (MS) do milho integral foi de 20 ton MS/ha e, a do milheto, foi de 9,7 ton MS/ha. Quanto ao fracionamento de proteína, não houve diferenças (P<0,05) para a fração A na silagem de milheto nos diferentes tempos de abertura dos silos. Na análise econômica, o milho apresentou custo de produção menor em função de seu maior rendimento por área. Apesar do milheto apresentar potencial de produção semelhante ao milho, a produção efetiva foi inferior, o que comparativamente inviabilizaria sua utilização. Para que esta seja viabilizada economicamente é necessário que a produção seja mais elevada.

This thesis describes the research underpinning and the development of a cross platform application for the analysis of simultaneously recorded multi-dimensional spike trains. These spike trains are believed to carry the neural code that encodes information in a biological brain. A number of statistical methods already exist to analyse the temporal relationships between the spike trains. Historically, hundreds of spike trains have been simultaneously recorded, however as a result of technological advances recording capability has increased. The analysis of thousands of simultaneously recorded spike trains is now a requirement. Effective analysis of large data sets requires software tools that fully exploit the capabilities of modern research computers and effectively manage and present large quantities of data. To be effective such software tools must; be targeted at the field under study, be engineered to exploit the full compute power of research computers and prevent information overload of the researcher despite presenting a large and complex data set. The Visualisation Studio application produced in this thesis brings together the fields of neuroscience, software engineering and information visualisation to produce a software tool that meets these criteria. A visual programming language for neuroscience is produced that allows for extensive pre-processing of spike train data prior to visualisation. The computational challenges of analysing thousands of spike trains are addressed using parallel processing to fully exploit the modern researcher’s computer hardware. In the case of the computationally intensive pairwise cross-correlation analysis the option to use a high performance compute cluster (HPC) is seamlessly provided. Finally the principles of information visualisation are applied to key visualisations in neuroscience so that the researcher can effectively manage and visually explore the resulting data sets. The final visualisations can typically represent data sets 10 times larger than previously while remaining highly interactive.

Patterned neuronal activity driven by the sensory environment plays a key role in the development of precise synaptic connectivity in the brain. It is well established that the action potentials (‘spikes’) generated by individual neurons are crucial to this developmental process. A neuron’s spiking activity is jointly determined by its synaptic inputs and its intrinsic excitability. It is therefore important to ask how a neuron develops these attributes, and whether the emergence of spiking might itself be governed by activity-dependent processes. In this thesis, I address these questions in the retinotectal system of Xenopus laevis. First, I investigate the extent to which visuospatial information is available to the developing retinotectal system. I show that the eyes of developing Xenopus larvae are hyperopic at the onset of vision, but rapidly grow towards correct vision. Despite its imperfect optics, the Xenopus eye is able to generate spatially restricted activity on the retina, which is evident in the spatial structure of the receptive fields (RFs) of tectal neurons. Using a novel method to map the visually driven spiking output and synaptic inputs of the same tectal neuron in vivo, I show that neuronal spiking activity closely follows the spatiotemporal profile of glutamatergic inputs. Next, I characterise a population of neurons in the developing optic tectum that does not fire action potentials, despite receiving visually evoked glutamatergic and γ-aminobutyric acid (GABA)ergic synaptic inputs. A comparison of visually spiking and visually non-spiking neurons reveals that the principal reason these neurons are ‘silent’ is that they lack sufficient glutamatergic synaptic excitation. In the final section of the thesis, I investigate whether visually driven activity can play a role in the ‘unsilencing’ of these silent neurons. I show that non-spiking tectal neurons can be rapidly converted into spiking neurons through a visual conditioning protocol. This conversion is associated with a selective increase in glutamatergic input and implicates a novel, spike-independent form of synaptic potentiation. I provide evidence that this novel plasticity process is mediated by GABAergic inputs that are depolarising during early development, and can act in synergy with N-methyl-D-aspartate receptors (NMDARs) to strengthen immature glutamatergic synapses. Consistent with this, preventing the depolarising effects of GABA or blocking NMDARs abolished the activity-dependent unsilencing of tectal neurons. These results therefore support a model in which GABAergic and glutamatergic transmitter systems function synergistically to enable a neuron to recruit the synaptic excitation it needs to develop sensory-driven spiking activity. This represents a transition with important consequences for both the functional output and the activity-dependent development of a neuron.

We study how to perform effective Bayesian inference in high-dimensional sparse Factor Analysis models with a zero-norm, sparsity-inducing prior on the model parameters. Such priors represent a methodological ideal, but Bayesian inference in such models is usually regarded as impractical. We test this view. After empirically characterising the properties of existing algorithmic approaches, we use techniques from statistical mechanics to derive a theory of optimal learning in the restricted setting of sparse PCA with a single factor. Finally, we describe a novel `Dense Message Passing' algorithm (DMP) which achieves near-optimal performance on synthetic data generated from this model.DMP exploits properties of high-dimensional problems to operate successfully on a densely connected graphical model. Similar algorithms have been developed in the statistical physics community and previously applied to inference problems in coding and sparse classification. We demonstrate that DMP out-performs both a newly proposed variational hybrid algorithm and two other recently published algorithms (SPCA and emPCA) on synthetic data while it explains at least the same amount of variance, for a given level of sparsity, in two gene expression datasets used in previous studies of sparse PCA.A significant potential advantage of DMP is that it provides an estimate of the marginal likelihood which can be used for hyperparameter optimisation. We show that, for the single factor case, this estimate exhibits good qualitative agreement both with theoretical predictions and with the hyperparameter posterior inferred by a collapsed Gibbs sampler. Preliminary work on an extension to inference of multiple factors indicates its potential for selecting an optimal model from amongst candidates which differ both in numbers of factors and their levels of sparsity.

Spiking neural networks (SNN) are dynamic models of biological neurons, that communicates with event-based signals called spikes. SNN that reproduce observed properties of biological senses like vision are developed to better understand how such systems function, and to learn how more efficient sensor systems can be engineered. A branching parameter describes the average probability for spikes to propagate between two different neuron populations. The adaptation of branching parameters towards critical values is known to be important for maximizing the sensitivity and dynamic range of SNN. In this thesis, a recently proposed SNN model for visual feature learning and pattern recognition known as the E-I Net model is studied and extended with a critical branching mechanism. The resulting modified E-I Net model is studied with numerical experiments and two different types of sensory queues. The experiments show that the modified E-I Net model demonstrates critical branching and power-law scaling behavior, as expected from SNN near criticality, but the power-laws are broken and the stimuli reconstruction error is higher compared to the error of the original E-I Net model. Thus, on the basis of these experiments, it is not clear how to properly extend the E-I Net model properly with a critical branching mechanism. The E-I Net model has a particular structure where the inhibitory neurons (I) are tuned to decorrelate the excitatory neurons (E) so that the visual features learned matches the angular and frequency distributions of feature detectors in visual cortex V1 and different stimuli are represented by sparse subsets of the neurons. The broken power-laws correspond to different scaling behavior at low and high spike rates, which may be related to the efficacy of inhibition in the model.

Inférence et apprentissage dans les réseaux de neurones profonds nécessitent une grande quantité de calculs qui, dans beaucoup de cas, limite leur intégration dans les environnements limités en ressources. Les réseaux de neurones évènementiels de type « spike » présentent une alternative aux réseaux de neurones artificiels classiques, et promettent une meilleure efficacité énergétique. Cependant, entraîner les réseaux spike demeure un défi important, particulièrement dans le cas où l’apprentissage doit être exécuté sur du matériel de calcul bio-inspiré, dit matériel neuromorphique. Cette thèse constitue une étude sur les algorithmes d’apprentissage et le codage de l’information dans les réseaux de neurones spike.A partir d’une règle d’apprentissage bio-inspirée, nous analysons quelles propriétés sont nécessaires dans les réseaux spike pour rendre possible un apprentissage embarqué dans un scénario d’apprentissage continu. Nous montrons qu’une règle basée sur le temps de déclenchement des neurones (type « spike-timing dependent plasticity ») est capable d’extraire des caractéristiques pertinentes pour permettre une classification d’objets simples comme ceux des bases de données MNIST et N-MNIST.Pour dépasser certaines limites de cette approche, nous élaborons un nouvel outil pour l’apprentissage dans les réseaux spike, SpikeGrad, qui représente une implémentation entièrement évènementielle de la rétro-propagation du gradient. Nous montrons comment cette approche peut être utilisée pour l’entrainement d’un réseau spike qui est capable d’inférer des relations entre valeurs numériques et des images MNIST. Nous démontrons que cet outil est capable d’entrainer un réseau convolutif profond, qui donne des taux de reconnaissance d’image compétitifs avec l’état de l’art sur les bases de données MNIST et CIFAR10. De plus, SpikeGrad permet de formaliser la réponse d’un réseau spike comme celle d’un réseau de neurones artificiels classique, permettant un entraînement plus rapide.Nos travaux introduisent ainsi plusieurs mécanismes d’apprentissage puissants pour les réseaux évènementiels, contribuant à rendre l’apprentissage des réseaux spike plus adaptés à des problèmes réels. / Inference and training in deep neural networks require large amounts of computation, which in many cases prevents the integration of deep networks in resource constrained environments. Event-based spiking neural networks represent an alternative to standard artificial neural networks that holds the promise of being capable of more energy efficient processing. However, training spiking neural networks to achieve high inference performance is still challenging, in particular when learning is also required to be compatible with neuromorphic constraints. This thesis studies training algorithms and information encoding in such deep networks of spiking neurons. Starting from a biologically inspired learning rule, we analyze which properties of learning rules are necessary in deep spiking neural networks to enable embedded learning in a continuous learning scenario. We show that a time scale invariant learning rule based on spike-timing dependent plasticity is able to perform hierarchical feature extraction and classification of simple objects of the MNIST and N-MNIST dataset. To overcome certain limitations of this approach we design a novel framework for spike-based learning, SpikeGrad, which represents a fully event-based implementation of the gradient backpropagation algorithm. We show how this algorithm can be used to train a spiking network that performs inference of relations between numbers and MNIST images. Additionally, we demonstrate that the framework is able to train large-scale convolutional spiking networks to competitive recognition rates on the MNIST and CIFAR10 datasets. In addition to being an effective and precise learning mechanism, SpikeGrad allows the description of the response of the spiking neural network in terms of a standard artificial neural network, which allows a faster simulation of spiking neural network training. Our work therefore introduces several powerful training concepts for on-chip learning in neuromorphic devices, that could help to scale spiking neural networks to real-world problems.

In the past decade, characterizing spiking neuron models has been extensively researched as an essential issue in computational neuroscience. In this thesis, we examine the estimation problem of two different neuron models. In Chapter 2, We propose a modified Izhikevich model with an adaptive threshold. In our two-stage estimation approach, a linear least squares method and a linear model of the threshold are derived to predict the location of neuronal spikes. However, desired results are not obtained and the predicted model is unsuccessful in duplicating the spike locations. Chapter 3 is focused on the parameter estimation problem of a multi-timescale adaptive threshold (MAT) neuronal model. Using the dynamics of a non-resetting leaky integrator equipped with an adaptive threshold, a constrained iterative linear least squares method is implemented to fit the model to the reference data. Through manipulation of the system dynamics, the threshold voltage can be obtained as a realizable model that is linear in the unknown parameters. This linearly parametrized realizable model is then utilized inside a prediction error based framework to identify the threshold parameters with the purpose of predicting single neuron precise firing times. This estimation scheme is evaluated using both synthetic data obtained from an exact model as well as the experimental data obtained from in vitro rat somatosensory cortical neurons. Results show the ability of this approach to fit the MAT model to different types of reference data.

No description available.