Ce manuscrit présente des travaux de recherche sur l’analyse et la détection des Electrogrammes Auriculaires Complexes Fractionnés (EACF). Dans une première partie faisant suite à une présentation des mécanismes et des signaux bioélelectriques de la Fibrillation Auriculaire (FA), les outils les plus couramment utilisés pour l’analyse des signaux EACF sont présentés. Des outils linéaires sont dans un premier temps appliqués aux signaux intracardiaques issus des procédures d’ablation de la FA par radiofréquence puis des outilsnon linéaires sont présentés et intégrés à un algorithme de détection des EACF. Ce dernier s’appuie sur la quantification des propriétés de récurrence des électrogrammes. Dans la seconde partie, la cellule et le tissu musculaire cardiaque sont détaillés puis simulés à l’aide de plusieurs modèles mathématiques. Ceux de FitzHugh Nagumo, Aliev Panfilov et Courtemanche Ramirez Nattel sont mis en oeuvre afin de reproduire les mécanismes de la FA évoqués dans la présentation de cette pathologie. L’acquisition des champs de potentiels est également reproduite à l’aide d’un modèle numérique de cathéter tel que celui utilisé lors des procédures. Les signaux temporels ainsi générés permettent de lier les activations spatiotemporelles au niveau du substrat aux motifs observables dans les EACF. Un modèle expérimental vient compléter la partie modélisation. Les cultures de cellules de rats nouveaux nés sur puces MEA (Micro Electrode Array) permettent de recréer des conditions de fibrillation et d’acquérir des potentiels extracellulaires. Là encore, les électrogrammes sont comparés aux signaux issus des simulations numériques ainsi qu’aux signaux cliniques. L’analyse des séquences de motifs via les trois types de modèles utilisés permet de rattacher les motifs observés dans les électrogrammes aux mécanismes se produisant au niveau du tissu cardiaque lors de la FA. Une analyse en temps réel permettrait de fournir au praticien des informations déterminantes lors de l’ablation concernant la nature et la localisation des sources d’arythmie / This manuscript presents research on the analysis and the detection of Complex Fractionated Atrial Electrograms (CFAE). In the first part, following a presentation over Atrial Fibrillation (AF) mechanisms and bioelectrical signals, the most commonly used tools for analyzing CFAE are presented. Linear tools are initially applied to signals from AF ablation procedures, then nonlinear tools are shown and integrated intoa CFAE detection algorithm. This one is based on the quantification of electrogram recurrence properties. In the second part, the cell and cardiac muscle tissue are described and simulated using mathematical models. Models such as FitzHugh Nagumo, Aliev Panfilov and Courtemanche Ramirez Nattel are implemented to reproduce the mechanisms of AF mentioned in the presentation of this disease. The acquisition of fields of potential is also reproduced using a numerical model of catheter as the one used during ablation process. Time signals thus generated are used to match the spatiotemporal activations at the substrate level with the patterns to be observed in CFAE. An experimental model completes the analysis. Cell cultures of newborn rats on MEA (Micro ElectrodeArray) can recreate fibrillation conditions and acquire extracellular potentials. Again, electrogramsare compared with signals from computer simulations and the clinical database signals. The analysisof pattern sequence via the three types of models can attach the observed patterns in electrograms with the mechanisms occurring at the cardiac tissue level during AF. Real-time analysis would allow the practitioner to receive critical information during ablation about the nature and the location of arrhythmia sources

This thesis presents results of investigations on fields and multipole expansion coefficients in axially symmetric (referred to as 3D)and two dimensional (2D)ion trapmass analysers. 3D mass analysers have a three-electrode geometry with two (electrically shorted) endcap electrodes and one central ring electrode. rf-only or rf/dc potential applied across the electrodes creates a linear trapping field in the central cavity of the mass analyser.2Dmass analysers have four longitudinal electrodes in which the opposite pairs of electrodes are electrically shorted. Here, rf-only or rf/dc potential applied across the pair of electrodes creates a linear trapping field and fragment ions of the analyte gas are trapped along the central axis of the mass analyser. Both these mass analysers have apertures machined on the electrodes (holes in case of 3D traps and slits in case of 2D traps) to permit entry of electrons for ionizing the analyte gas and for collection of destabilized fragment ions. This thesis is concerned with how these apertures influence the fields and multipole expansion coefficients within the traps. This thesis is divided into five chapters. Chapter 1 provides the background information which is required for the thesis. It begins with a description of the geometry of the 3D and the 2D mass analysers used in the present work.These include the quadrupole ion trap (QIT) and cylindrical ion trap (CIT) for 3D structures and the linear ion trap (LIT) and the rectilinear ion trap (RIT) for 2D structures. This is followed by a brief description of the numerical method, the boundary element method (BEM), used in the thesis. Also presented here are the Green’s function for 3D and 2D geometries. In the final section, the scope of the thesis is presented. Chapter 2 presents two approximate analytical expressions for nonlinear electric fields in the principal direction in axially symmetric (3D) and two dimensional (2D) ion trap mass analysers with apertures on the electrodes. Considered together (3D and 2D), we present composite approximations for the principal unidirectional nonlinear electric fields in these ion traps. The composite electric field E has the form E= EnoAperture + EdueToAperture where EnoAperture is the field within an imagined trap which is identical to the practical trap except that the apertures are missing; and where EdueToAperture is the field contribution due to apertures on the two trap electrodes. The field along the principal axis of the trap can in this way be well approximated for any aperture that is not too large. To derive EdueToAperture, classical results of electrostatics have been extended to electrodes with finite thickness and different aperture shapes. EnoAperture is a modified truncated multipole expansion for the imagined trap with no aperture. The first several terms in the multipole expansion are in principle exact (though numerically determined using the BEM), while the last term is chosen to match the field at the electrode. This expansion, once computed, works with any aperture in the practical trap. The composite field approximation for axially symmetric (3D) traps is checked for three geometries: the quadrupole ion trap (QIT), the cylindrical ion trap (CIT) and an arbitrary other trap. The approximation for 2D traps is verified using two geometries: the linear ion trap (LIT)and the rectilinear ion trap (RIT). In each case, for two aperture sizes (10% and 50% of the trap dimension), highly satisfactory fits are obtained. These composite approximations may be used in more detailed nonlinear ion dynamics studies than have been hitherto attempted. In Chapter 3we complement and complete the work presented in Chapter 2 by considering off-axis fields in the axially symmetric (3D) and the two dimensional (2D) ion traps whose electrodes have apertures. Our approximation has two parts. The first, EnoAperture, is the field obtained numerically for the trap under study with no apertures. We have used the boundary element method (BEM) for obtaining this field. The second part, EdueToAperture, is an analytical expression for the field contribution of the aperture. In EdueToAperture, aperture size is a free parameter. A key element in our approximation is the electrostatic field near an infinite thin plate with an aperture, and with different constant valued far field intensities on either side. Compact expressions for this field can be found using separation of variables, wherein the choice of coordinate system is crucial. This field is, in turn, used four times within our trap specific approximation. The off-axis field expressions for the 3D geometries were tested on the quadrupole ion trap (QIT) and the cylindrical ion trap (CIT), and the corresponding expressions for the 2D geometries were tested on the linear ion trap (LIT) and rectilinear ion trap (RIT). For each geometry, we have considered apertures which are 10%, 30% and 50% of the trap dimension. We have found that our analytical correction term EdueToAperture, though based on a classical small-aperture approximation, gives good results even for relatively large apertures. Chapter 4 presents approximate analytical expressions for estimating the variation in multipole expansion coefficients with the size of apertures in axially symmetric (3D) and two dimensional (2D) ion trap mass analysers. Following the approach adopted in Chapter 2 and Chapter 3 which focused on the role of apertures to fields within traps, here too, the analytical expression is a sum of two terms, An,noAperture, the multipole expansion coefficient for a trap with no apertures and An,dueToAperture, the multipole expansion coefficient contributed by the aperture. An,noAperture has been obtained numerically and An,dueToAperture is obtained from the nth derivative of the potential within the trap. The expressions derived have been tested on two 3D geometries and two 2D geometries. These include the quadrupole ion trap (QIT) and the cylindrical ion trap (CIT) for 3D geometries and the linear ion trap (LIT) and the rectilinear ion trap (RIT) for the 2D geometries. Multipole expansion coefficients A2 to A24, estimated by our analytical expressions were compared with the values obtained numerically (using the boundary element method) for aperture sizes varying up to 50% of the trap size. In all the plots presented, it is observed that our analytical expression for the variation of multipole expansion coefficients versus aperture size closely follows the trend of the numerical evaluations for the range of aperture sizes considered. The maximum relative percentage errors, which provide an estimate of the deviation of our values from those obtained numerically for each multipole expansion coefficient, are seen to be in the range of 10% to 15%. The leading multipole expansion coefficient, A2, however, is seen to be estimated very well by our expressions, with most values being within 1% of the numerically determined values, with larger deviations seen for the QIT and LIT only at larger aperture sizes. Chapter 5 presents a few concluding remarks.

La Résonance Quadrupolaire Nucléaire (RQN) est une technique spectroscopique appliquée à l’état solide et qui ne nécessite aucun champ magnétique statique contrairement à la Résonance Magnétique Nucléaire (RMN). Elle permet de détecter, les molécules qui contiennent un noyau quadrupolaire (tel que l'azote-14). La RQN de l’azote-14 suscite un intérêt particulier car elle s'avère utile dans la détection de drogues et d’explosifs, qui contiennent, pour la plupart, des molécules azotées. Cette thèse porte en premier lieu sur l’amélioration de la qualité spectrale en résonance quadrupolaire (RQN) de l’azote-14. Ainsi, nous nous sommes préoccupés des signaux transitoires (ringing) qui suivent l’impulsion radiofréquence (rf) et qui masquent en partie le signal RQN. Nous avons d’abord montré, par une étude théorique et expérimentale, que les mesures à basse puissance (< 3Watts) étaient possibles, ce qui nous a permis de diminuer l’amplitude du ringing. De plus, grâce à ce mode de fonctionnement (basse puissance), nous avons pu proposer un circuit original de commutation du facteur de qualité (Q-switch) basé sur un commutateur rf CMOS et qui a permis de diminuer la durée du ringing. La seconde partie traite de l’influence d’un champ magnétique statique faible amplitude et des divers paramètres expérimentaux en RQN de l’azote-14. Dans le cas où le tenseur de gradient de champ électrique est de symétrie axiale, nous avons observé des élargissements, voire des doublets Zeeman. En revanche, dans le cas général (absence de symetrie), nous avons mis en évidence la possibilité définir le tenseur de gradient de champ électrique à partir de la forme de raie d’une seule transition au lieu de deux transitions (sans champ magnétique statique). Ce dernier résultat pourrait permettre de diminuer de moitié le temps nécessaire à la caractérisation d’une nouvelle molécule azotée. / Nuclear quadrupole resonance (NQR) is a solid-state spectroscopic technique which doesn't require a static magnetic field in contrast to Nuclear Magnetic Resonance (NMR). It is used to detect molecules containing quadrupole nuclei (such as nitrogen-14). Nitrogen-14 quadrupole resonance is of particular interest because it is useful in the detection of drugs and explosives, which contain mostly of nitrogen molecules. In this thesis we have primarily focused on improving the spectral quality in nitrogen-14 quadrupole resonance (NQR) . Thus, we were concerned with the transient signals ( ringing ) following the radiofrequency (rf) pulse and which partly overwhelm the NQR signal. First, we demonstrated by theoretical and experimental study that low power measurements (< 3Watts ) were possible, allowing us to reduce the ringing amplitude. In addition, through this mode ( low power ), we have proposed an original quality factor switching circuit (Q -switch ) based on rf CMOS switch and which has reduced the ringing duration. The second part deals with the effect of a weak static magnetic field and various experimental parameters in nitrogen-14 quadrupole resonance. In the case of an axially symmetric electric field gradient tensor, we observed broadening or Zeeman doublets. By contrast, in the general case (no symmetry), we demonstrated the possibility to define the electric field gradient tensor from the line shape of a single transition instead of two transitions ( without static magnetic field ). This result reduces by half the time required for the characterization of a new nitrogen molecule.

L’utilisation de matrice de portes logiques reconfigurables (FPGA) est une des seules solutionspour traiter des flux de plusieurs 100 MÉchantillons/seconde en temps-réel. Toutefois, ce typede composant présente une grande difficulté de mise en oeuvre : au delà d’un type langage spécifique,c’est tout un environnement matériel et une certaine expérience qui sont requis pourobtenir les traitements les plus efficaces. Afin de contourner cette difficulté, de nombreux travauxont été réalisés dans le but de proposer des solutions qui, partant d’un code écrit dans unlangage de haut-niveau, vont produire un code dans un langage dédié aux FPGAs. Nos travaux,suivant l’approche d’assemblage de blocs et en suivant la méthode du skeleton, ont visé à mettreen place un logiciel, nommé CoGen, permettant, à partir de codes déjà développés et validés,de construire des chaînes de traitements en tenant compte des caractéristiques du FPGA cible,du débit entrant et sortant de chaque bloc pour garantir l’obtention d’une solution la plus adaptéepossible aux besoins et contraintes. Les implémentations des blocs de traitements sont soitgénérés automatiquement soit manuellement. Les entrées-sorties de chaque bloc doivent respecterune norme pour être exploitable dans l’outil. Le développeur doit fournir une descriptionconcernant les ressources nécessaires et les limitations du débit de données pouvant être traitées.CoGen fournit à l’utilisateur moins expérimenté une méthode d’assemblage de ces blocsgarantissant le synchronisme et cohérence des flux de données ainsi que la capacité à synthétiserle code sur les ressources matérielles accessibles. Cette méthodologie de travail est appliquéeà des traitements sur des flux vidéos (seuillage, détection de contours et analyse des modespropres d’un diapason) et sur des flux radio-fréquences (interrogation d’un capteur sans-fils parméthode RADAR, réception d’un flux modulé en fréquence, et finalement implémentation deblocs de bases pour déporter le maximum de traitements en numérique). / Using Field Programmable Gate Arrays (FPGA) is one of the very few solution for real time processingdata flows of several hundreds of Msamples/second. However, using such componentsis technically challenging beyond the need to become familiar with a new kind of dedicateddescription language and ways of describing algorithms, understanding the hardware behaviouris mandatory for implementing efficient processing solutions. In order to circumvent these difficulties,past researches have focused on providing solutions which, starting from a description ofan algorithm in a high-abstraction level language, generetes a description appropriate for FPGAconfiguration. Our contribution, following the strategy of block assembly based on the skeletonmethod, aimed at providing a software environment called CoGen for assembling various implementationsof readily available and validated processing blocks. The resulting processing chainis optimized by including FPGA hardware characteristics, and input and output bandwidths ofeach block in order to provide solution fitting best the requirements and constraints. Each processingblock implementation is either generated automatically or manually, but must complywith some constraints in order to be usable by our tool. In addition, each block developer mustprovide a standardized description of the block including required resources and data processingbandwidth limitations. CoGen then provides to the less experienced user the means to assemblethese blocks ensuring synchronism and consistency of data flow as well as the ability to synthesizethe processing chain in the available hardware resources. This working method has beenapplied to video data flow processing (threshold, contour detection and tuning fork eigenmodesanalysis) and on radiofrequency data flow (wireless interrogation of sensors through a RADARsystem, software processing of a frequency modulated stream, software defined radio).

La propriété de creusement spectral, que l’on retrouve dans certains cristaux dopés aux ions de terres rares refroidis à basse température offre des possibilités prometteuses pour le traitement analogique de signaux radiofréquence. En effet, celle-ci permet de programmer des fonctions de traitement dans le spectre d’absorption du cristal.Partant des premières démonstrations de principe d’un analyseur spectral radiofréquence large bande instantanée, l’objectif est d’en améliorer les performances, ce qui requiert une modélisation précise de l’interaction laser-matière et de l’ensemble des perturbations inhérentes à la montée en maturité technologique du dispositif. Nous avons par conséquent développé un modèle et cherché à étendre son domaine de validité pour qu’il s’applique à un maximum de protocoles.Nous l’avons ensuite appliqué à un matériau en particulier, à savoir un cristal de Tm³ ⁺ :YAG. Après avoir effectué une série de mesures des différentes caractéristiques intrinsèques à ce cristal, nous avons choisi un protocole d’application relativement proche de celui de l’analyseur spectral que nous cherchons à optimiser, à savoir la création de réseaux spectraux large bande. La comparaison des résultats expérimentaux à ceux de notre modèle nous a permis de démontrer sa validité.Nous avons finalement appliqué notre modèle au cas précis de l’analyseur spectral radiofréquence. Les simulations nous ont permis de déterminer théoriquement comment optimiser ses performances, en particulier la dynamique, avec des paramètres réalistes, atteignables expérimentalement. / The Spectral Hole Burning property, found in some rare-earth ion-doped crystals at low temperature is particularly relevant for analogic processing of radiofrequency signals. Indeed, it enables processing functions to be programmed in the crystal’s absorption spectrum.Starting with the first demonstrations of a wideband radiofrequency spectrum analyser, we aim at improving its performances, which requires an accurate modelling of the light-matter interaction and all the perturbations arising from the upgrade in TRL (Technology Readiness Level). Therefore, we have developed a model and extended its validity domain to a broad variety of SHB-based protocols.We applied this model on a particular material, namely a Tm³ ⁺:YAG crystal. After measuring experimentally the relevant intrinsic parameters of this crystal, we applied our model to a protocol which is quite similar to the one of the spectrum analyser we aim at optimizing, namely the engraving of wideband spectral gratings. The comparison of our experimental results to the simulations from our model proved its validity.Finally we applied it to the exact case of the radiofrequency spectrum analyser. With the simulations, we determined how to improve its performances, and particularly increase the dynamic range with realistic experimental parameters.

Catheter ablation is a unique method for cardiac arrhythmias treatment based on creating spot lesions of radiofrequency energy using intracardiac ablation catheter. Transmurality of these lesions should be achieved for long-term therapeutic treatment effect. The degree of myocardial damage depth is reflected in the parameters of sensed intracardiac ECG. From research studies is evident that the change in morphology and deflection of ablation signal play a key role in the assessment of transmurality. This thesis consists of theoretical introduction to electrophysiology, proposal and realization of methods for objective assessment of transmurality by processing signals in time, frequency and time – frequency domain. The end of this diploma thesis includes the statistical analysis of these clinical dat and comparison of methods.

No description available.

Mikrochips zur Tierkennzeichnung bestehen aus verschiedenen metallischen Materialien. Diese treten in der Magnetresonanztomographie in Wechselwirkung mit den elektromagnetischen Feldern. So verursachen die ferromagnetischen Materialen der Mikrochips gravierende fokale Bildstörungen. Diese Suszeptibilitätsartefakte können die Beurteilbarkeit der Halsregion erheblich einschränken. Ziel der Studie war, den Einfluss des Sequenztyps auf die Größe des Artefakts zu untersuchen und herauszufinden, welche Möglichkeiten zur Artefaktreduktion bei Veränderung bestimmter Sequenzparameter bestehen. Zusätzlich sollte geklärt werden, wie groß der Abstand zwischen Spinalkanal und Mikrochip mindestens sein muss, um spinale Strukturen beurteilen zu können. In das Untersuchungsgut der Studie gingen die Kadaver von 26 Katzen und 2 Hunden ein. An einem 0,5-Tesla-MRT wurde für verschiedene Sequenztypen (SE-Sequenzen, TSE-Sequenzen, GRE-Sequenzen) und Kombinationen modifizierter Sequenzparameter (Echozeit (TE), Voxelgröße, Ausleserichtung) das Ausmaß der Artefakte ermittelt. Berechnet wurde der Flächeninhalt des Artefakts (cm2). Dieser wurde dann als prozentualer Anteil zur Fläche des Halsquerschnitts angegeben. Diese Berechnung erfolgte für alle untersuchten Einstellungen an transversalen Aufnahmen. Eine ergänzende computertomografische Untersuchung dienste dazu, die Distanz zwischen Spinalkanalund Mikrochip zu messen. Die Untersuchungen der Studie haben gezeigt, dass TSE-Sequenzen wegen ihrer geringeren Artefaktanfälligkeit den SE- und GRE-Sequenzen vorgezogen werden sollten. Besonders kleine Artefakte konnten bei einer T1-TSE-Sequenz mit kleiner TE (10 ms) und kleiner Voxelgröße (große Akquisitionsmatrix von 256 x 256 Pixel, kleines Field of View (FOV) von 160 mm, geringe Schichtdicke (ST) von 2 mm) erreicht werden. Durch Anpassung der Kodierrichtung war es möglich, die Form und Richtung des Artefaktes zu beeinflussen. Lag das Zentrum des Mikrochips näher als 19 mm von der Mitte des Wirbelkanals entfernt, ließen sich auch mit dieser optimierten Sequenz die spinalen Strukturen auf Höhe des Mikrochips nicht beurteilen. Die Größe und Form der Suszeptibilitätsartefakte konnten durch die Wahl des Sequenztyps und Modifikation von Sequenzparametern verändert werden. Dies ist besonders bei kleinen Tieren von Bedeutung. Bei diesen kann es aufgrund der geringen Distanz zwischen Mikrochip und Wirbelsäule zur Beeinträchtigung der MR-Bildauswertung kommen. Eine T1-gewichtete TSE-Sequenz mit kleiner Echozeit (10 ms) und kleiner Voxelgröße (Akquisitionsmatrix 256 x 256 Pixel, FOV 160 mm, ST 2 mm) bietet bei 0,5 Tesla das größte Potenzial zur Artefaktreduktion.

This thesis is concerned about ion-molecular reactions at low temperatures, which are important the fully understand the chemical evolution in interstellar medium. For realization of experimental part of thesis has been used the apparatus of 22-pole radiofrequency ion trap, which allows study the rate constant of reactions at temperatures 10 - 100 K. Thesis contains measuring results of reaction NH+ + H → N+ + H2, which follows the previous study of reaction N+ + H2 → NH+ + H on the same apparatus, description of the apparatus and the general introduction.

Purpose: To evaluate the efficiency of a multimodality approach consisting of transcatheter arterial chemoembolization (TACE) and radiofrequency ablation (RFA) as bridging therapy for patients with hepatocellular carcinoma (HCC) awaiting orthotopic liver transplantation (OLT) and to evaluate the histopathological response in explant specimens. Materials and Methods: Between April 2001 and November 2011, 36 patients with 50 HCC nodules (1.4–5.0 cm, median 2.8 cm) on the waiting list for liver transplantation were treated by TACE and RFA. The drop-out rate during the follow-up period was recorded. The local efficacy was evaluated by histopathological examination of the explanted livers. Results: During a median follow-up time of 29 (4.0–95.3) months the cumulative drop-out rate for the patients on the waiting list was 0, 2.8, 5.5, 11.0, 13.9 and 16.7% at 3, 6, 12, 24, 36 and 48 months, respectively. 16 patients (with 26 HCC lesions) out of 36 (44.4%) were transplanted by the end of study with a median waiting list time of 13.7 (2.5–37.8) months. The histopathological examination of the explanted specimens revealed a complete necrosis in 20 of 26 HCCs (76.9%), whereas 6 (23.1%) nodules showed viable residual tumor tissue. All transplanted patients are alive at a median time of 29.9 months. Imaging correlation showed 100% specificity and 66.7% sensitivity for the depiction of residual or recurrent tumor. Conclusion: We conclude that TACE combined with RFA could provide an effective treatment to decrease the drop-out rate from the OLT waiting list for HCC patients. Furthermore, this combination therapy results in high rates of complete tumor necrosis as evaluated in the histopathological analysis of the explanted livers. Further randomized trials are needed to demonstrate if there is a benefit in comparison with a single-treatment approach. / Dieser Beitrag ist mit Zustimmung des Rechteinhabers aufgrund einer (DFG-geförderten) Allianz- bzw. Nationallizenz frei zugänglich.