Estudos teóricos dos mecanismos moleculares responsáveis pela formação e estabilidade de complexos moleculares vêm ganhando relevância pelas possibilidades práticas que oferecem, por exemplo, na compreensão de diversas doenças e no desenho racional de fármacos. Neste projeto, nossa ênfase está no estudo de complexos de proteínas, extraídos do banco de dados de proteínas (PDB), onde desenvolvemos ferramentas computacionais as quais permitem efetuar análises em duas direções: 1) efetuar previsões básicas, através do emprego de propriedades eletrostáticas de proteínas, em diferentes condições e níveis preditivos e 2) realização de um conjunto de análises estatísticas, como freqüência de contato, em busca de preditores de complexos de proteínas e identificar padrões de interação entre seus aminoácidos em função da distância de separação. Com base nos resultados obtidos por ambos os estudos, objetivamos quantificar as forças físicas envolvidas na formação dos complexos protéicos. O foco do projeto, a longo prazo, é prever o fenômeno da complexação através da fusão dessas duas linhas de estudos: preditor básico de complexos protéicos e análise do potencial estatístico entre os aminoácidos que formam o complexo. O presente projeto é concluído com a construção de portais web que disponibilizarão os resultados obtidos por nossos trabalhos bem como a possibilidade de qualquer usuário, efetuar consultas por propriedades de proteínas e/ou grupo de proteínas. / Theoretical studies of the molecular mechanisms responsible for the formation and stability of molecular complexes are gaining relevance for the practical possibilities that they offer, for example, in the understanding of diverse diseases and rational drug design. In this project, our emphasis is on the study of protein complexes, extracted from protein data bank (PDB). We have developed computational tools which allow to perform analyses in two directions: 1) to make basic complexation forecasts, through the use of electrostatics properties of proteins, in different conditions and predictive levels, and 2) to carry out a set of statistical analyses, as contacts frequency, in order to build up predictor of protein complexes and to identify patters of interactions between the amino acids as a function of their separation distance. Based on the results obtained on both studies, we aim quantify the physical forces involved in the formation of protein complexes. The focus of the project, in the long run, is to foresee the phenomenon of the protein complexes through the fusing of these two study lines: a coarse-grained predictor of protein complexes and analysis of the statistical potentials between the amino acids that form the complex. The present project is concluded with the construction of web services where we make available the results obtained on our works. This server also has the possibility to be used by any computer user, that wishes to perform search on protein and/or protein group properties

Introduction : Issus des données animales, les modèles neurotrophiniques du mécanisme d’action des médicaments antidépresseurs pourraient permettre d’identifier chez l’Homme des biomarqueurs prédictifs de la réponse et de la rémission sous antidépresseurs. Nous évaluons l’intérêt clinique, chez les patients souffrant de trouble dépressif caractérisé unipolaire, de 11 biomarqueurs : polymorphismes nucléotidiques simples (SNP) du Brain Derived Neurotrophic Factor (BDNF) et de son récepteur, le Récepteur Tyrosine-Kinase B (TRKB), taux plasmatiques de BDNF et volume hippocampique sur la réponse/rémission sous antidépresseurs. Méthode : Les données originales de ce travail sont issues de la cohorte METADAP. Il s’agit d’une cohorte, prospective, multicentrique incluant 624 patients présentant un épisode dépressif caractérisé dans le cadre d’un trouble dépressif caractérisé unipolaire et nécessitant l’introduction d’un traitement antidépresseur. Le traitement antidépresseur est prescrit de façon naturaliste (tous antidépresseurs commercialisés en France). Les patients sont évalués 1, 3 et 6 mois après l’introduction du traitement antidépresseur. Les biomarqueurs étudiés sont les polymorphismes Val66Met du BDNF et 8 SNP du TRKB et les dosages de BDNF plasmatiques. Une étude ancillaire est menée à partir de 63 patients ayant bénéficié d’Imagerie par Résonnance Magnétique cérébrale réalisée en pratique courante à l’inclusion de cette cohorte afin d’évaluer les volumes hippocampiques. Résultats : 1) Une revue de la littérature met en évidence une association entre la réponse aux antidépresseurs et 12 SNP du BDNF/TRKB sur 242 étudiés, ainsi qu’une association entre allèle Met du polymorphisme Val66Met du BDNF et meilleure réponse sous antidépresseurs chez les patients asiatiques. 2) Nos données ne mettent pas en évidence d’impact de 8 SNP du TRKB sur la réponse/rémission après traitement antidépresseur, mais un effet différentiel du Val66Met du BDNF selon la classe de traitement antidépresseur. 3) L’étude des dosages de BDNF plasmatiques n’est pas concluante. 4) Concernant les volumes hippocampiques, notre méta-analyse montre que des volumes hippocampiques moindres prédisent une moindre réponse/rémission après traitement antidépresseur. 5) Concernant les liens entre les biomarqueurs étudiés, nous ne mettons pas en évidence d’association. Conclusion : Sur les 11 biomarqueurs étudiés, seuls 2 pourraient présenter une utilité en pratique clinique. Si nos travaux étaient répliqués, le polymorphisme Val66Met du BDNF et le volume hippocampique pourraient conduire à orienter le choix des antidépresseurs dans le traitement des épisodes dépressifs caractérisés. Malgré une littérature cohérente chez l’Animal, nous n’avons pas mis en évidence, dans l’échantillon étudié, de lien entre les biomarqueurs génétiques étudiés et les volumes hippocampiques. Nous poursuivons ce travail d’évaluation des biomarqueurs neurotrophiniques et neurogéniques avec des méthodes d’évaluations nouvelles : séquençage nouvelle génération pour la génétique et imagerie multimodale (acquisition répétée d’IRM structurelle, fonctionnelle et de diffusion) de l’hippocampe. Nous évaluerons également de nouveaux biomarqueurs. / Introduction: developed with Animal preclinical approachs, neurtrophinic and neurogenic models of antidepressant mechanism of action lead to identify biomarkers in Human which could be predict antidepressant response and remission in depressed patients. We assess the clinical benefit of 11 biomarkers in depressed patients: Brain Derived Neurotrophic Factor (BDNF) and its receptor Tyrosine Receptor -Kinase B (TRKB), Plasma BDNF and Hippocampal volumes to predict antidepressant response/remission. Methods: The original research data of this work are from METADAP cohort. It is a prospective, multicentric cohort including 624 patients with a diagnosis of major depressive disorder and a current major depressive episode at the start of the index antidepressant treatment. Antidepressant treatment is prescribed in naturalistic conditions (all commercialized antidepressant in France). Patient are assessed 1, 3 and 6 months after the start of antidepressant treatment. Studied biomarkers are BDNF Val66Met polymorphism, 8 TRKB SNP and plasma BDNF. Ancillary study are done with 63 patients which benefit in clinical practice of Magnetic Resonnance (MRI) at the inclusion of the cohort. Results: 1) A review of literature reports associations between antidepressant efficacy and 12 BDNF/TRKB SNP on 242 studied SNP and an association with Met allele of Val66Met BDNF polymorphism and a best antidepressant efficacy in Asian patients. 2) Our original data show no impact of 8 TRKB SNP on antidepressant response remission but a differential effect of Val66Met BDNF polymorphism depending on antidepressant treatment class. 3) Plasma BDNF study is not conclusive. 4) Concerning hippocampal volumes, our meta-analysis show that smaller hippocampal volumes predict lower response/remission rate after antidepressant treatment. 5) No association is found between studied biomarkers. Conclusion: 2 of the 11 studied biomarkers could be useful in clinical practice. After replication of our results, Val66Met polymorphism could lead to personalized antidepressant prescription in major depressive disorder. Although the animal prelinical littérature appar strong, we dont report association between genetic biomarker and hippocampal volume in ours ample. We will assess neurotrophinic and neurogenic biomarkers with new methods: next generation sequencing for genetic, multimodal imaging (repeated structural, functional and diffusion MRI) of hippocampus. We also will assess new biomarkers.

Este trabalho tem por objetivo estudar e analisar a influência do parâmetro de barreira e de seu fator de correção no processo de convergência dos métodos de pontos interiores primal-dual, primal-dual barreira modificada e primal-dual barreira modificada com as técnicas preditor-corretor e Newton composto. A grande motivação para o desenvolvimento desta pesquisa está relacionada com a busca de métodos eficientes para resolver problemas de otimização de programação não-linear, existentes na área de engenharia elétrica mais especificamente na operação de sistemas elétricos de potência. Esses métodos foram aplicados a um problema de programação não-linear e aos sistemas elétricos de três e de trinta barras para analisar a sensibilidade em relação ao parâmetro de barreira e ao seu fator de correção. / This work has for objective to study and to analyze the influence of the barrier parameter and its correction factor in the convergence process of the methods primal-dual interior point, primal-dual modified barrier and primal-dual barrier modified with the techniques predictor-corrector and composed Newton. The great motivation for the development of this research is related with the search of efficient methods to solve nonlinear programming optimization problems, existent in the area of electric engineering more specifically in the operation of power systems. Those methods were applied to a nonlinear programming problem and the electric systems of three and thirty buses to analyze the sensibility in relation to the barrier parameter and its correction factor.

Este trabalho tem por objetivo estudar e analisar a influência do parâmetro de barreira e de seu fator de correção no processo de convergência dos métodos de pontos interiores primal-dual, primal-dual barreira modificada e primal-dual barreira modificada com as técnicas preditor-corretor e Newton composto. A grande motivação para o desenvolvimento desta pesquisa está relacionada com a busca de métodos eficientes para resolver problemas de otimização de programação não-linear, existentes na área de engenharia elétrica mais especificamente na operação de sistemas elétricos de potência. Esses métodos foram aplicados a um problema de programação não-linear e aos sistemas elétricos de três e de trinta barras para analisar a sensibilidade em relação ao parâmetro de barreira e ao seu fator de correção. / This work has for objective to study and to analyze the influence of the barrier parameter and its correction factor in the convergence process of the methods primal-dual interior point, primal-dual modified barrier and primal-dual barrier modified with the techniques predictor-corrector and composed Newton. The great motivation for the development of this research is related with the search of efficient methods to solve nonlinear programming optimization problems, existent in the area of electric engineering more specifically in the operation of power systems. Those methods were applied to a nonlinear programming problem and the electric systems of three and thirty buses to analyze the sensibility in relation to the barrier parameter and its correction factor.

Les douleurs rachidiennes représentent un problème de santé important aux impacts socio-économique majeurs. Les rachialgies débutent tôt, durant l’enfance, leur prévalence augmente avec l’âge, particulièrement durant la période pubertaire, pour atteindre un niveau similaire à celui de l’âge adulte aux alentours de 20 ans. Malgré la fréquence de cette condition, les facteurs prédictifs du développement du premier épisode de douleur rachidienne ou d’un nouvel épisode sont très peu connus. Les travaux constitutifs de cette thèse ont pour objectif d’explorer différents facteurs prédictifs potentiels de douleurs rachidiennes chez les enfants et les adolescents. Une première revue de littérature a montré qu’il existe une association entre la puberté et les rachialgies. Les critères de causalité, décrits par Bradford-Hill, comme la force de l’association, la consistance entre les études et la relation dose-effet sont également présents. Cependant, il n’est pas encore possible de se prononcer sur le critère essentiel de la temporalité. La deuxième revue de littérature a permis de mettre en évidence une association positive entre l’endurance musculaire en extension et les douleurs rachidiennes, alors qu’aucune association n’a été mise en évidence entre la force musculaire en extension du tronc et les douleurs rachidiennes. Les données disponibles de la littérature ne permettent pas de ce prononcer concernant l’association entre les rachialgies et les capacités aérobies. La troisième étude, une étude prospective de suivi des enfants durant une année, n’a pas mis en évidence d’association entre les capacités aérobies, la composition corporelle et les douleurs au niveau du rachis. Deux facteurs sont donc associés aux douleurs rachidiennes, la puberté et une faible endurance musculaire. Les recherches dans ce domaine doivent êtres poursuivies pour déterminer l’existence d’un lien de causalité de ces facteurs sur les douleurs du rachis. / Spinal pain is a major health problem with socio-economic impacts. Back pain starts early, during childhood, the prevalence increases with age, particularly during puberty, to reach a similar level to that found at adulthood, around the age of 20. Despite the high prevalence of this condition, predictors for the first episode or for a recurrent episode are not known. The aim of this thesis is to explore different potential predictors of back pain during childhood and adolescence. A first literature review showed an association between puberty and back pain. The causality criteria described by Bradford- Hill, e.g. strength of the association, consistency across studies and dose -response are also present. Unfortunately,, it is not yet possible to conclude whether there is a temporal aspect as well; for temporality being the most important criterion. The second literature review showed a positive association between muscular endurance in extension and back pain, whereas no association was found between muscle strength in trunk extension and spinal pain. The available literature does not allow us to conclude about the potential association between aerobic capacity and back pain. The third study is a prospective study, in which children were followed during ten months. This study did not reveal any association between aerobic capacity, body composition and back pain. In conclusion, two factors are associated with back pain, puberty and low muscle endurance. Future research in this area should focus on the causal relationship between these factors and back pain.

Several targeting algorithms are developed and analyzed for possible future use onboard a spacecraft. Each targeter is designed to determine the appropriate propulsive burn for translunar injection to obtain desired orbital parameters upon arrival at the moon. Primary design objectives are to minimize the computational requirements for each algorithm but also to ensure reasonable accuracy, so that the algorithm’s errors do not force the craft to conduct large mid-course corrections. Several levels of accuracy for dynamical models are explored, the convergence range and speed of each algorithm are compared, and the possible benefits of the Broyden and trust-region targeters are evaluated. These targeters provide a proof of concept for the feasibility of a translunar injection targeting algorithm. Anticipating some future improvements, these algorithms could serve as a viable alternative to uploading ground-based targeting solutions and bypass the problems of delays and disruptions in communication, enabling the craft to conduct a translunar injection burn autonomously.

Sleep is a basic need as important as physical fitness and good nutrition. Without enough sleep, we will create a sleep debt and experience sleepiness. Sleepiness can be defined as the inability to stay awake, a condition that has become a health problem in our 24-hour-7-day-a-week society. Estimates suggest that up to one-third of the population suffers from excessive sleepiness. Among other interactions, sleepiness affects our performance, increasing the risk of being involved in accidents. A considerable portion of work related accidents and injuries are related to sleepiness resulting in large costs for the individuals and society. Professional drivers are one example of workers who are at risk of sleepiness related accidents. Up to 40% of heavy truck accidents could be related to sleepiness. A better knowledge about reliable indicators and predictors of sleepiness is important in preventing sleepiness related accidents. This thesis investigates both objective and subjective indicators of sleepiness, how these relate to each other, and how their pattern changes over time. The indicators investigated were electroencephalography, heart rate variability, simple reaction time, head movement, and subjective ratings of sleepiness (Study I-IV). In Study V, a questionnaire study was conducted with professional drivers in northern Sweden. This study mainly deals with predictors of sleepiness. When subjects were sleep deprived both objective and subjective ratings indicated a rapid increase in sleepiness during the first hour of the test followed by a levelling off. This change in pattern was evident for all the indicators except heart rate and heart rate variability. On the other hand, HRV was correlated with the increase of EEG parameters during the post-test sleep period. The changes in pattern of the indicators included in the thesis are analysed in the perspective of temporal patterns and relationships. Of the tested indicators, a subjective rating of sleepiness with CR-10 was considered to be the most reliable indicator of sleepiness. Of the investigated predictors of sleepiness, prior sleep habits were found to be strongly associated to sleepiness and the sleepiness related symptoms while driving. The influences of driving conditions and individual characteristics on sleepiness while driving were lower. A multidisciplinary approach when investigating and implementing indicators and predictors of sleepiness is important. In addition to their actual relations to the development of sleepiness, factors such as technical and practical limitations, work, and individual and situational needs must be taken into account.

The aim of this work is to develop and validate a continuum model for the simulation of the thermomechanical response of a shape memory polymer (SMP). Rather than integral type viscoelastic model, the approach here is based on the idea of two inter-penetrating networks, one which is permanent and the other which is transient together with rate equations for the time evolution of the transient network. We find that the activation stress for network breakage and formation of the material controls the gross features of the response of the model, and exhibits a "thermal Bauschinger effect". The model developed here is similar to a thermoviscoelastic model, and is developed with an eye towards ease of numerical solutions to boundary value problems. The primary hypothesis of this model is that the hysteresis of temperature dependent activation-stress plays a lead role in controlling its main response features. Validation of this hypothesis is carried out for the uniaxial response from the experimental data available in the literature for two different SMP samples: shape memory polyurethane and Veriflex, to show the control of the evolution of the temperature sensitive activation stress on the response. We extend the validated 1D model to a three dimensional small strain continuum SMP model and carry out a systematic parameter optimization method for the identification of the activation stress coefficients, with different weights given to different features of the response to match the parameters with experimental data. A comprehensive parametric study is carried out, that varies each of the model material and loading parameters, and observes their effect on design-relevant response characteristics of the model undergoing a thermomechanical cycle. We develop "response charts" for the response characteristics: shape fixity, shape recovery and maximum stress rise during cooling, to give the designer an idea of how the simultaneous variation of two of the most influential material parameters changes a specific response parameter. To exemplify the efficacy of the model in practical applications, a thermoviscoelastic extension of a beam theory model will be developed. This SMP beam theory will account for activation stress governed inelastic response of a SMP beam. An example of a three point bend test is simulated using the beam theory model. The numerical solution is implemented by using an operator split technique that utilizes an elastic predictor and dissipative corrector. This algorithm is validated by using a three-point bending experiment for three different material cases: elastic, plastic and thermoplastic response. Time step convergence and mesh density convergence studies are carried out for the thermoviscoelastic FEM model. We implement and study this model for a SMP beam undergoing three-point bending strain recovery, stress recovery and cyclic thermomechanical loading. Finally we develop a thermodynamically consistent finite continuum model to simulate the thermomechanical response of SMPs. The SMP is modeled as an isotropic viscoplastic material where thermal changes govern the evolution of the activation stress of the material. The response of the SMP in a thermomechanical cycle is modeled as a combination of a rubbery and a glassy element in series. Using these assumptions, we propose a specific form for the Helmholtz potential and the rate of dissipation. We use the technique of upper triangular decomposition for developing the constitutive equations of the finite strain SMP model. The resulting model is implemented in an ODE solver in MATLAB, and solved for a simple shear problem. We study the response of the SMP model for shear deformation as well as cyclic shear deformation at different initial temperatures. Finally, we implement the thermomechanical cycle under shear deformations and study the behavior of the model.

L’utilisation des mesures d’isolement et de contentions en milieu psychiatrique intrahospitalier se produit fréquemment en réponse à des comportements agressifs et continue de soulever la controverse. À cet égard, de plus en plus d’études tendent à démontrer que le personnel soignant travaillant sur ces unités est influencé par plusieurs facteurs de nature différente, notamment la perception de l’agressivité, quand vient le temps de prendre une décision quant à l’utilisation (ou non) de ces mesures coercitives. Méthodologie : Plus de trois cents membres du personnel soignant travaillant en milieu psychiatrique intrahospitalier ont été recrutés dans huit établissements psychiatriques du Québec. Dans un premier temps, un questionnaire leur a été distribué afin de mettre en relief les différents facteurs (individuels et organisationnels) ayant un impact sur l’utilisation des mesures coercitives. Simultanément, l’analyse factorielle de la version française de deux échelles permettant de mesurer la perception de l’agressivité en milieu hospitalier (le MOAS et le POAS) a été faite. Résultats : Un modèle final multivarié a démontré que le type d’unité psychiatrique, l’expression de la colère et de l’agressivité parmi les membres de l’équipe de soins, la perception de la fréquence de gestes autoagressifs et la perception de mesures de sécurité insuffisantes dans le milieu de travail étaient des prédicteurs indépendants de l’utilisation de procédures d’isolement et de contentions. L’analyse factorielle a pour sa part mis en évidence une structure à 4 facteurs pour le MOAS et à 3 facteurs pour le POAS, conformément à ce que l’on retrouvait dans la littérature scientifique. Conclusion : Ces résultats soulignent l’importance des facteurs organisationnels par rapport aux facteurs individuels dans l’utilisation des mesures coercitives en psychiatrie et la nécessité d’évaluer les perceptions quant à l’agressivité et à la sécurité chez le personnel soignant. En comprenant mieux les phénomènes qui amènent leur utilisation, il sera possible de trouver des alternatives aux mesures d’isolement et de contentions et ainsi réduire le recours à ces dernières. / Psychiatric staff can have both positive and negative attitudes towards inpatient aggression. Different attitudes towards aggression can have a substantial influence on how such behaviour is dealt with on psychiatric wards. In this manner, seclusion and restraint are frequently used measures for managing violent behaviour. However, their use raises several concerns. Method: Over three hundred staff members were recruited from eight psychiatric hospitals in the province of Quebec. First of all, an examination was conducted from the staff perspective of the organizational and staff factors that may be associated with increased recourse to seclusion and restraint in psychiatric wards. Simultaneously, factorial analysis of the French version of two scales used to measure staff attitude towards institutional violence and aggression (MOAS and POAS) was completed. Results: The final multivariate model showed that the following factors independently predict to greater use of seclusion and restraint: the type of hospital ward; greater expression of anger and aggression among staff members; perceptions of the frequency of incidents of physical aggression against the self; and the perception of insufficient protection measures in the workplace. Factor analyses revealed a four factor structure for the MOAS and a three factor structure for the 12-item POAS, which is similar to what is found in recent scientific literature in North American and European countries. Conclusion: These findings underscore the importance of evaluating a variety of factors, including perceptions of safety and violence, when exploring the management of aggression and violence on psychiatric wards and the reasons seclusion and restraint measures are used. These findings represent the first stage of a research program of the multidisciplinary group to whom the author is associated, aimed at reducing recourse to seclusion and restraint in Quebec psychiatric services.

La fibrillation auriculaire (FA) est une arythmie touchant les oreillettes. En FA, la contraction auriculaire est rapide et irrégulière. Le remplissage des ventricules devient incomplet, ce qui réduit le débit cardiaque. La FA peut entraîner des palpitations, des évanouissements, des douleurs thoraciques ou l’insuffisance cardiaque. Elle augmente aussi le risque d'accident vasculaire. Le pontage coronarien est une intervention chirurgicale réalisée pour restaurer le flux sanguin dans les cas de maladie coronarienne sévère. 10% à 65% des patients qui n'ont jamais subi de FA, en sont victime le plus souvent lors du deuxième ou troisième jour postopératoire. La FA est particulièrement fréquente après une chirurgie de la valve mitrale, survenant alors dans environ 64% des patients. L'apparition de la FA postopératoire est associée à une augmentation de la morbidité, de la durée et des coûts d'hospitalisation. Les mécanismes responsables de la FA postopératoire ne sont pas bien compris. L'identification des patients à haut risque de FA après un pontage coronarien serait utile pour sa prévention. Le présent projet est basé sur l'analyse d’électrogrammes cardiaques enregistrées chez les patients après pontage un aorte-coronaire. Le premier objectif de la recherche est d'étudier si les enregistrements affichent des changements typiques avant l'apparition de la FA. Le deuxième objectif est d'identifier des facteurs prédictifs permettant d’identifier les patients qui vont développer une FA. Les enregistrements ont été réalisés par l'équipe du Dr Pierre Pagé sur 137 patients traités par pontage coronarien. Trois électrodes unipolaires ont été suturées sur l'épicarde des oreillettes pour enregistrer en continu pendant les 4 premiers jours postopératoires. La première tâche était de développer un algorithme pour détecter et distinguer les activations auriculaires et ventriculaires sur chaque canal, et pour combiner les activations des trois canaux appartenant à un même événement cardiaque. L'algorithme a été développé et optimisé sur un premier ensemble de marqueurs, et sa performance évaluée sur un second ensemble. Un logiciel de validation a été développé pour préparer ces deux ensembles et pour corriger les détections sur tous les enregistrements qui ont été utilisés plus tard dans les analyses. Il a été complété par des outils pour former, étiqueter et valider les battements sinusaux normaux, les activations auriculaires et ventriculaires prématurées (PAA, PVA), ainsi que les épisodes d'arythmie. Les données cliniques préopératoires ont ensuite été analysées pour établir le risque préopératoire de FA. L’âge, le niveau de créatinine sérique et un diagnostic d'infarctus du myocarde se sont révélés être les plus importants facteurs de prédiction. Bien que le niveau du risque préopératoire puisse dans une certaine mesure prédire qui développera la FA, il n'était pas corrélé avec le temps de l'apparition de la FA postopératoire. Pour l'ensemble des patients ayant eu au moins un épisode de FA d’une durée de 10 minutes ou plus, les deux heures précédant la première FA prolongée ont été analysées. Cette première FA prolongée était toujours déclenchée par un PAA dont l’origine était le plus souvent sur l'oreillette gauche. Cependant, au cours des deux heures pré-FA, la distribution des PAA et de la fraction de ceux-ci provenant de l'oreillette gauche était large et inhomogène parmi les patients. Le nombre de PAA, la durée des arythmies transitoires, le rythme cardiaque sinusal, la portion basse fréquence de la variabilité du rythme cardiaque (LF portion) montraient des changements significatifs dans la dernière heure avant le début de la FA. La dernière étape consistait à comparer les patients avec et sans FA prolongée pour trouver des facteurs permettant de discriminer les deux groupes. Cinq types de modèles de régression logistique ont été comparés. Ils avaient une sensibilité, une spécificité et une courbe opérateur-receveur similaires, et tous avaient un niveau de prédiction des patients sans FA très faible. Une méthode de moyenne glissante a été proposée pour améliorer la discrimination, surtout pour les patients sans FA. Deux modèles ont été retenus, sélectionnés sur les critères de robustesse, de précision, et d’applicabilité. Autour 70% patients sans FA et 75% de patients avec FA ont été correctement identifiés dans la dernière heure avant la FA. Le taux de PAA, la fraction des PAA initiés dans l'oreillette gauche, le pNN50, le temps de conduction auriculo-ventriculaire, et la corrélation entre ce dernier et le rythme cardiaque étaient les variables de prédiction communes à ces deux modèles. / Atrial fibrillation (AF) is an abnormal heart rhythm (cardiac arrhythmia). In AF, the atrial contraction is rapid and irregular, and the filling of the ventricles becomes incomplete, leading to reduce cardiac output. Atrial fibrillation may result in symptoms of palpitations, fainting, chest pain, or even heart failure. AF is an also an important risk factor for stroke. Coronary artery bypass graft surgery (CABG) is a surgical procedure to restore the perfusion of the cardiac tissue in case of severe coronary heart disease. 10% to 65% of patients who never had a history of AF develop AF on the second or third post CABG surgery day. The occurrence of postoperative AF is associated with worse morbidity and longer and more expensive intensive-care hospitalization. The fundamental mechanism responsible of AF, especially for post-surgery patients, is not well understood. Identification of patients at high risk of AF after CABG would be helpful in prevention of postoperative AF. The present project is based on the analysis of cardiac electrograms recorded in patients after CABG surgery. The first aim of the research is to investigate whether the recordings display typical changes prior to the onset of AF. A second aim is to identify predictors that can discriminate the patients that will develop AF. Recordings were made by the team of Dr. Pierre Pagé on 137 patients treated with CABG surgery. Three unipolar electrodes were sutured on the epicardium of the atria to record continuously during the first 4 post-surgery days. As a first stage of the research, an automatic and unsupervised algorithm was developed to detect and distinguish atrial and ventricular activations on each channel, and join together the activation of the different channels belonging to the same cardiac event. The algorithm was developed and optimized on a training set, and its performance assessed on a test set. Validation software was developed to prepare these two sets and to correct the detections over all recordings that were later used in the analyses. It was complemented with tools to detect, label and validate normal sinus beats, atrial and ventricular premature activations (PAA, PVC) as well as episodes of arrhythmia. Pre-CABG clinical data were then analyzed to establish the preoperative risk of AF. Age, serum creatinine and prior myocardial infarct were found to be the most important predictors. While the preoperative risk score could to a certain extent predict who will develop AF, it was not correlated with the post-operative time of AF onset. Then the set of AF patients was analyzed, considering the last two hours before the onset of the first AF lasting for more than 10 minutes. This prolonged AF was found to be usually triggered by a premature atrial PAA most often originating from the left atrium. However, along the two pre-AF hours, the distribution of PAA and of the fraction of these coming from the left atrium was wide and inhomogeneous among the patients. PAA rate, duration of transient atrial arrhythmia, sinus heart rate, and low frequency portion of heart rate variability (LF portion) showed significant changes in last hour before the onset of AF. Comparing all other PAA, the triggering PAA were characterized by their prematurity, the small value of the maximum derivative of the electrogram nearest to the site of origin, as well as the presence of transient arrhythmia and increase LF portion of the sinus heart rate variation prior to the onset of the arrhythmia. The final step was to compare AF and Non-AF patients to find predictors to discriminate the two groups. Five types of logistic regression models were compared, achieving similar sensitivity, specificity, and ROC curve area, but very low prediction accuracy for Non-AF patients. A weighted moving average method was proposed to design to improve the accuracy for Non-AF patient. Two models were favoured, selected on the criteria of robustness, accuracy, and practicability. Around 70% Non-AF patients were correctly classified, and around 75% of AF patients in the last hour before AF. The PAA rate, the fraction of PAA initiated in the left atrium, pNN50, the atrio-ventricular conduction time, and the correlation between the latter and the heart rhythm were common predictors of these two models.