Une des caractéristiques les plus singulières de l’activité corticale est son degré élevé de variabilité. Ma thèse dedoctorat s’est focalisée sur l’étude de (i) l’irrégularité des intervalles entre potentiels d’action (PAs)successivement émis par un neurone, et (ii) la variabilité dans l’évolution temporelle de l’activité d’un ensemblede neurones. Premièrement, j’ai étudié l’irrégularité des neurones enregistrés dans le cortex moteur de singesmacaques performant une tâche d’estimation du temps et de préparation à l’action. J’ai montré que l’irrégularitén’est pas un paramètre libre de l’activité neuronale, contrairement au taux de PAs, mais est déterminée par lescontraintes structurelles des réseaux neuronaux. Deuxièmement, j’ai utilisé le modèle de Markov caché (MMC)pour analyser l’activité d’ensembles de neurones enregistrés dans plusieurs aires corticales, sensorielles etmotrices, de singes exécutant une tâche de discrimination tactile. J’ai montré que les processus sensoriels etdécisionnels sont distribués dans plusieurs aires corticales. Les résultats suggèrent que l’action et la décision surlaquelle elle est basée sont reliées par une cascade d’évènements non stationnaires et stochastiques. Finalement,j’ai utilisé le MMC pour caractériser l’activité spontanée d’un ensemble de neurones du cortex préfrontal d’unrat. Les résultats montrèrent que l’alternance entre les états UP et DOWN est un processus stochastique etdynamique. La variabilité apparaît donc aussi bien pendant l’activité spontanée que pendant le comportementactif et semble être contrainte par des facteurs structurels qui, à leur tour, contraignent le mode d’opération desréseaux neuronaux. / A hallmark of cortical activity is its high degree of variability. The present work focused on (i) the variability ofintervals between spikes that single neurons emit, called spike time irregularity (STI), and (ii) the variability inthe temporal evolution of the collective neuronal activity. First, I studied the STI of macaque motor corticalneurons during time estimation and movement preparation. I found that although the firing rate of the neuronstransmitted information about these processes, the STI of a neuron is not flexible and is determined by thebalance of excitatory and inhibitory inputs. These results were obtained by means of an irregularity measure thatI compared to other existing measures. Second, I analyzed the neuronal ensemble activity of severalsomatosensory and motor cortical areas of macaques during tactile discrimination. I showed that ensembleactivity can be effectively described by the Hidden Markov Model (HMM). Both sensory and decision-makingprocesses were distributed across many areas. Moreover, I showed that decision-related changes in neuronalactivity rely on a noise-driven mechanism and that the maintenance of the decision relies on transient dynamics,subtending the conversion of a decision into an action. Third, I characterized the statistics of spontaneous UP andDOWN states in the prefrontal cortex of a rat, using the HMM. I showed that state alternation is stochastic andthe activity during UP states is dynamic. Hence, variability is prominent both during active behavior andspontaneous activity and is determined by structural factors, thus rending it inherent to cortical organization andshaping the function of neural networks.

Le cerveau et la moelle épinière adultes sont câblés pour traiter l’information sensorielle et la transformer en patrons d’activité cohérents qui forment la base de la perception du monde extérieur et de la motricité. Ces connexions très précises, loin d’être complètement établies à la naissance, sont affinées par l’activité électrique. Des activités spontanées (AS) peuvent être générées en l’absence de stimulation extrinsèque au système considéré. Au niveau de la moelle épinière, les AS sont responsables des mouvements spontanés et des tremblements myocloniques observés chez tous les mammifères immatures y compris chez l’humain. Elles constituent en quelque sorte les premiers pas du réseau locomoteur, ce qui signifie d’une part, qu’elles apparaissent avant l’émergence d’une activité de type locomotrice, et d’autre part, qu’elles participent à la mise en place et au raffinement d’un réseau moteur fonctionnel. La locomotion est générée par des réseaux de neurones localisés au niveau de la moelle épinière lombaire. Ces neurones constituent le générateur de rythme locomoteur (ou CPG, pour « central pattern generator »). Contrairement aux AS, l’activité locomotrice est déclenchée par une stimulation du réseau. Le patron locomoteur de décharges rythmiques est alterné de chaque côté de la moelle spinale, et entre les racines lombaires qui innervent des muscles extenseurs et fléchisseurs. / The mature brain and spinal cord are precisely wired to process sensory information into coherent patterns of activity that form the basis of our perception and motor behaviors. This precise wiring is not fully developed at birth. The pattern of connections that emerges during prenatal development only roughly approximates the final wiring. This initially coarse pattern of connections is subsequently refined by activity-dependent mechanisms that match precisely the presynaptic neurons to their appropriate target cells. In spinal cord, spontaneous activity (SA) is responsible for spontaneous limb movements and myoclonic twitching observed in all immature mammals, including human babies. SA can be seen as the first steps of the locomotor network since its participates in the development of the locomotor system.Locomotion is produced by neural networks located in the spinal cord (Central Pattern Generators (CPGs). Activation of CPGs, evokes a fictive locomotor pattern consisting of alternation between the motor bursts on the left and right sides of the spinal cord, as well as alternation between flexor and extensor bursts on the same side. Operation of neural networks depends on the balance between excitation and inhibition. At early stages of development, neuronal assemblies are hyperexcitable mainly because of GABA and glycine, the major inhibitory neurotransmitters in adults,that are depolarizing. GABA and glycine action depends on the intracellular concentration of chloride ([Cl-]i) which is finely regulated by specific cation-chloride co-transporters, called KCC2 and NKCC1.

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第18124号 / 医博第3844号 / 新制||医||1001(附属図書館) / 30982 / 京都大学大学院医学研究科医学専攻 / (主査)教授 金子 武嗣, 教授 大森 治紀, 教授 渡邉 大 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

No description available.

No description available.

L'activité électrique du coeur est initiée par la génération spontanée de potentiels d'action venant des cellules pacemaker du noeud sinusal (SN). Toute dysfonction au niveau de cette région entraîne une instabilité électrique du coeur. La majorité des patients souffrant d'un noeud sinusal déficient nécessitent l'implantation chirurgicale d'un pacemaker électronique; cependant, les limitations de cette approche incitent à la recherche d'une alternative thérapeutique. La base moléculaire des courants ioniques jouant un rôle crucial dans l'activité du noeud sinusal sont de plus en plus connues. Une composante importante de l'activité des cellules pacemakers semble être le canal HCN, responsable du courant pacemaker If. Le facteur T-box 3 (Tbx3), un facteur de transcription conservé durant le processus de l'évolution, est nécessaire au développement du système de conduction cardiaque. De précédentes études ont démontré que dans différentes lignées cellulaires le Phorbol 12-myristate 13-acetate (PMA) active l'expression du gène codant Tbx3 via des réactions en cascade partant de la protéine kinase C (PKC). L'objectif principal de cette étude est de tester si le PMA peut augmenter la fréquence et la synchronisation de l'activité spontanée du pacemaker biologique en culture. Plus précisément, nous avons étudié les effets de l'exposition chronique au PMA sur l'expression du facteur de transcription Tbx3, sur HCN4 et l'activité spontanée chez des monocouches de culture de myocytes ventriculaires de rats néonataux (MVRN). Nos résultats démontrent que le PMA augmente significativement le facteur transcription de Tbx3 et l'expression ARNm de HCN4, favorisant ainsi l'augmentation du rythme et de la stabilité de l'activité autonome. De plus, une diminution significative de la vitesse de conduction a été relevée et est attribuée à la diminution du couplage intercellulaire. La diminution de la vitesse de conduction pourrait expliquer l'effet négatif du PMA sur la synchronisation de l'activité autonome du pacemaker biologique. Ces résultats ont été confirmés par un modèle mathématique multicellulaire suggérant que des fréquences et résistances intercellulaires plus élevée pourraient induire une activité plus stable et moins synchrone. Cette étude amène de nouvelles connaissances très importantes destinées à la production d'un pacemaker biologique efficient et robuste. / The normal heartbeat is initiated by the spontaneous generation of action potentials in pacemaker cells of the sinoatrial node (SAN) region. Dysfunction of this region leads to electrical instability of the heart. The majority of the patients with sinus node dysfunction require surgical implantation of electronic pacemaker devices; however, limitations of this therapeutic approach lead to a need to search for alternatives. To date, the molecular basis of the ionic currents which play pivotal role in SAN action potential has been discovered. It is thought that an important component of the pacemaker cells are HCN channels, responsible for the funny current (If) in the SAN. Meanwhile, T-box factor 3 known as an evolutionary conserved transcription factors is necessary for development of the conduction system. In previous studies, it has been shown that Phorbol 12-myristate 13-acetate (PMA) activates Tbx3 gene expression in a PKC-dependent manner in several cell lines. The main objective of this study is to test if PMA can increase the frequency and synchronization of spontaneous activity of cultured biopacemakers. More precisely, we studied the effects of chronic exposure to PMA on the expression of the Tbx3 transcription factor and HCN4 in neonatal rat ventricular myocytes monolayers and how spontaneous activity was altered. Our results show that PMA significantly increases the Tbx3 transcription factor and HCN4 mRNA expression favoring an increased in the rate and spatial-temporal stability of the spontaneous activity. In addition, a significant decrease in conduction velocity was found that is attributed to decrease electrical intercellular coupling of the cells. The decrease in the conduction velocity could explain the negative effect PMA has on synchronization of spontaneous activity of the biopacemaker. These findings are confirmed by a multicellular mathematical model implying that faster frequency and higher intercellular resistance of the pacemaker cells may lead to a more stable and less synchronous activity. This study provides important new knowledge to produce efficient and robust biological pacemakers.

Conductive hearing loss (otitis media) in young children can effect speech and language development. However, little is known about the effects of conductive loss on neural activity in the auditory system. Hypothesis: Conductive hearing loss will change resting activity levels at the inner hair cell synapse, and lead to auditory deprivation of central auditory pathways. A conductive loss was produced by blocking the ear canals in mice. Resting neural activity patterns were quantified in brainstem and midbrain using c-fos immuno-labelling. Experimental subjects were compared to normal hearing controls and subjects with cochlear ablation. Conductive loss subjects showed a trend in reduction in c-fos labelled cells in cochlear nucleus and the central nucleus of inferior colliculus compared to normal controls. Results seen in this study may indicate the influence of conductive hearing loss on the developing auditory brain during early postnatal years when the system is highly plastic.

Conductive hearing loss (otitis media) in young children can effect speech and language development. However, little is known about the effects of conductive loss on neural activity in the auditory system. Hypothesis: Conductive hearing loss will change resting activity levels at the inner hair cell synapse, and lead to auditory deprivation of central auditory pathways. A conductive loss was produced by blocking the ear canals in mice. Resting neural activity patterns were quantified in brainstem and midbrain using c-fos immuno-labelling. Experimental subjects were compared to normal hearing controls and subjects with cochlear ablation. Conductive loss subjects showed a trend in reduction in c-fos labelled cells in cochlear nucleus and the central nucleus of inferior colliculus compared to normal controls. Results seen in this study may indicate the influence of conductive hearing loss on the developing auditory brain during early postnatal years when the system is highly plastic.

The goal of this work is to improve the robustness and generalization of deep learning models, using a similar approach to the unsupervised "innate learning" strategy in visual development. A series of research studies are presented to demonstrate how an unsupervised machine learning efficient coding approach can create filters similar to the receptive fields of the primary visual cortex (V1) in the brain, and these filters are capable of pretraining convolutional neural networks (CNNs) to enable faster training times and higher accuracy with less dependency on the source data. Independent component analysis (ICA) is used for unsupervised feature extraction as it has shown success in both applied machine learning and modeling biological neural receptive fields. This pretraining applies equally well to various forms of visual input, including natural color images, black and white, binocular, and video to drive the V1-like Gabor filters in the brain. For efficient processing of typical visual scenes, the filters that ICA produces are developed by encoding natural images. These filters are then used to initialize the kernels in the first layer of a CNN to train on the CIFAR-10 dataset to perform image classification. Results show that the ICA initialization for a custom made CNN produces models with a test accuracy up to 12% better than the standard model in the first 10 epochs, which for specific accuracy thresholds reduces the number of training epochs by approximately 40% (to reach 60% accuracy) and 50% (to reach 70% accuracy). Additionally, this pre-training results in marginally higher accuracy even after extensive training over 50 epochs. This proposed method of unsupervised machine learning to pre-train weights in deep learning improves both training time and accuracy, which is why it is observed in biological networks and is finding increased application in applied deep learning.

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal.