This thesis deals with machine learning techniques for the extraction of structure and the analysis of the vertebrate olfactory pathway based on related methods. Some of its main contributions are summarized below. We have performed a systematic investigation for classification in biomedical images with the goal of recognizing a material in these images by its texture. This investigation included (i) different measures for evaluating the importance of image descriptors (features), (ii) methods to select a feature set based on these evaluations, and (iii) classification algorithms. Image features were evaluated according to their estimated relevance for the classification task and their redundancy with other features. For this purpose, we proposed a framework for relevance and redundancy measures and, within this framework, we proposed two new measures. These were the value difference metric and the fit criterion. Both measures performed well in comparison with other previously used ones for evaluating features. We also proposed a Hopfield network as a method for feature selection, which in experiments gave one of the best results relative to other previously used approaches. We proposed a genetic algorithm for clustering and tested it on several realworld datasets. This genetic algorithm was novel in several ways, including (i) the use of intra-cluster distance as additional optimization criterion, (ii) an annealing procedure, and (iii) adaptation of mutation rates. As opposed to many conventional clustering algorithms, our optimization framework allowed us to use different cluster validation measures including those which do not rely on cluster centroids. We demonstrated the use of the clustering algorithm experimentally with several cluster validity measures as optimization criteria. We compared the performance of our clustering algorithm to that of the often-used fuzzy c-means algorithm on several standard machine learning datasets from the University of California/Urvine (UCI) and obtained good results. The organization of representations in the brain has been observed at several stages of processing to spatially decompose input from the environment into features that are somehow relevant from a behavioral or perceptual standpoint. For the perception of smells, the analysis of such an organization, however, is not as straightforward because of the missing metric. Some studies report spatial clusters for several combinations of physico-chemical properties in the olfactory bulb at the level of the glomeruli. We performed a systematic study of representations based on a dataset of activity-related images comprising more than 350 odorants and covering the whole spatial array of the first synaptic level in the olfactory system. We found clustered representations for several physico-chemical properties. We compared the relevance of these properties to activations and estimated the size of the coding zones. The results confirmed and extended previous studies on olfactory coding for physico-chemical properties. Particularly of interest was the spatial progression by carbon chain that we found. We discussed our estimates of relevance and coding size in the context of processing strategies. We think that the results obtained in this study could guide the search into olfactory coding primitives and the understanding of the stimulus space. In a second study on representations in the olfactory bulb, we grouped odorants together by perceptual categories, such as floral and fruity. By the application of the same statistical methods as in the previous study, we found clustered zones for these categories. Furthermore, we found that distances between spatial representations were related to perceptual differences in humans as reported in the literature. This was possibly the first time that such an analysis had been done. Apart from pointing towards a spatial decomposition by perceptual dimensions, results indicate that distance relationships between representations could be perceptually meaningful. In a third study, we modeled axon convergence from olfactory receptor neurons to the olfactory bulb. Sensory neurons were stimulated by a set of biologically-relevant odors, which were described by a set of physico-chemical properties that covaried with the neural and glomerular population activity in the olfactory bulb. Convergence was mediated by the covariance between olfactory neurons. In our model, we could replicate the formation of glomeruli and concentration coding as reported in the literature, and further, we found that the spatial relationships between representational zones resulting from our model correlated with reported perceptual differences between odor categories. This shows that natural statistics, including similarity of physico-chemical structure of odorants, can give rise to an ordered arrangement of representations at the olfactory bulb level where the distances between representations are perceptually relevant. / <p>QC 20120224</p>

Les informations olfactives sont connues pour leur capacité à induire des comportements moteurs spécifiques. En dépit de nombreuses observations comportementales chez les vertébrés, on ne connaît toujours pas les mécanismes et les voies nerveuses qui sous-tendent ces phénomènes de transformation olfacto-locomotrices. Chez la lamproie, des travaux récents ont permis de décrire cette voie, et les mécanismes responsables de la transformation des entrées olfactives en activité locomotrice (Derjean et al., 2010). Cette voie prend origine dans la partie médiane du bulbe olfactif, et envoie des projections vers le tubercule postérieur, une région qui se trouve dans le diencéphale. De là, les neurones projettent directement vers la Région Locomotrice Mésencéphalique, connue pour envoyer des connexions vers les neurones réticulospinaux, et activer la locomotion. L’objectif de cette étude était d’établir si l’ensemble des neurones réticulospinaux répond aux stimulations olfactives. Pour ce faire, nous avons utilisé sur une préparation de cerveau isolé de lamproie des techniques d’électrophysiologie et d’imagerie calcique. La stimulation électrique des nerfs olfactifs, de la région médiane du bulbe olfactif ou du tubercule postérieur a provoqué une activation de toutes les cellules réticulospinales qui se retrouvent dans les quatre noyaux réticulaires (ARRN : Noyau Réticulaire Rhombencéphalique Antérieur; MRN : Noyau Réticulaire Mésencéphalique; MRRN : Noyau Réticulaire Rhombencéphalique Moyen; PRRN : Noyau Réticulaire Rhombencéphalique Postérieur). Seule la partie médiane du bulbe olfactif est impliquée dans le passage de l’information olfactive vers les neurones réticulospinaux. Nous avons aussi découvert que le blocage des récepteurs GABAergiques dans la partie médiane du bulbe olfactif augmentait les réponses olfactives de façon considérable dans les cellules réticulospinales. Nous avons montré ainsi qu’il existe un tonus inhibiteur impliqué dans la dépression modulatrice de la voie olfacto-locomotrice. Ce travail a permis de montrer que la stimulation des afférences sensorielles olfactives active simultanément l’ensemble des populations de neurones réticulospinaux qui commandent la locomotion. De plus, il existerait un tonus inhibiteur GABAergique, au niveau de la partie médiane du bulbe olfactif, responsable d’une dépression modulatrice dans la voie olfacto-locomotrice. / Olfactory inputs are known for their ability to induce specific motor behaviors. Despite numerous behavioral observations in vertebrates, the mechanisms and the neural pathways underlying the olfactory-locomotor transformation are still unknown. In lamprey, recent studies have described this pathway and the mechanism underlying the transformation of olfactory input into a locomotor activity (Derjean et al., 2010). This pathway originates in the medial part of the olfactory bulb, sends projections to the posterior tuberculum, a diencephalic region. From there, the neurons project directly to the mesencephalic locomotor region that is known to send projections to the reticulospinal neurons to activate locomotion. Using lamprey brain preparation, electrophysiology and calcium imaging, the aim of this study was to establish whether all reticulospinal neurons respond to olfactory stimuli. Electrical stimulation of the olfactory nerves, the medial part of the olfactory bulb or the posterior tuberculum activates all reticulospinal cells in the four reticular nuclei (ARRN: Anterior rhombencephalic reticular nucleus; MRN: middle mesencephalic reticular nucleus; MRRN: middle rhombencephalic reticular nucleus; PRRN: posterior rhombencephalic reticular nucleus). The medial part of the olfactory bulb is the only region that is implicated in transmitting the olfactory information to reticulospinal neurons. We also discovered that when blocking the GABAergic receptors in the medial part of the olfactory bulb, the reticulospinal neurons have a stronger response to olfactory stimulation. Thus we showed that a tonic inhibition is involved in the modulating depression of the olfacto-locomotor pathway. Altogether, this work shows that stimulation of the olfactory sensory inputs activates simultaneously the entire population of reticulospinal neurons that control locomotion. In addition, there is a GABAergic tonic inhibition at the level of the medial part of the olfactory bulb that causes a modulating depression in the olfacto-locomotor pathway.

Anxiety disorders are the most prevalent psychiatric conditions within primary care, affecting up to 29% of people across their lifetime. Generalized Anxiety disorder (GAD) is frequently comorbid with Major Depressive Disorder (MDD), resulting in greater functional impairment. Gut microbiota have been shown to modulate brain chemistry and function, possibly also playing a role in the genesis of anxiety. Bacteria are also able to produce, or interact with the host metabolism of neuroactive substances, including classical neurotransmitters and trace amines, like octopamine, which although found in trace concentrations in the mammalian brain, can affect CNS function. Specifically, trace amines can affect catecholamine release, reuptake and biosynthesis, and modulate dopamine and serotonin metabolism. We investigated whether microbiota from patients with GAD with no signs of immune activation can alter behaviour in gnotobiotic mice and whether this is accompanied by changes in metabolites within the gastrointestinal tract. Germ-free NIH Swiss mice (n=35) were colonized with microbiota from either a GAD patient (n=18) with severe anxiety, comorbid depression, and low serum and fecal octopamine, or an age and sex-matched healthy control (HC) (n=17). Three weeks post- colonization, mouse behaviour was assessed by standard psychometric tests. Emotionality z-scores were calculated to provide a robust integrated behavioural assessment. Microbiota profiles were assessed by 16S rRNA based Illumina, fecal β-defensin-3 level was measured by ELISA. After sacrifice, mouse brain BDNF and GDNF expression was assessed by immunofluorescence, and gene expression in the hippocampus, amygdala, and olfactory bulbs was assessed by Nanostring. Stool and cecum metabolites were measured in all colonized mice by multisegment injection-capillary electrophoresis-mass spectrometry (MSI-CE-MS). There were no differences in fecal β-defensin levels between mice colonized with GAD microbiota as compared to mice colonized with HC microbiota. However, GAD mice exhibited greater anxiety and depressive-like behavior compared to HC mice in the digging and tail suspensions tests. Behavioural z-scoring across all six standard psychometric tests showed a significant increase in group emotionality score means of GAD-colonized mice compared to HC-colonized mice. Mice colonized with microbiota from a GAD patient had distinct bacterial profiles from mice colonized with HC microbiota. Compared to HC mice, GAD mice had lower levels of dopamine, octopamine and acetylcholine in cecum contents. Furthermore, GAD mice had higher expression of BDNF in the amygdala, lower expression of BDNF in the hippocampus, and lower expression of GDNF in the midbrain. GAD mice also had lower expression of CCR2 in the hippocampus, higher Cnlp/CAMP in the amygdala and olfactory bulb, and higher Nfkb1 in the olfactory bulb compared to HC mice. Our results suggest that microbiota from a selected patient with GAD has the ability to induce anxiety and depressive-like behavior, by mechanisms independent of immune system, likely by altered production of biogenic amines and neurotransmitters. / Thesis / Master of Science (MSc)