In diesem Forschungsprojekt bauen wir auf Erkenntnisse eines vorgängig bewilligten SNF Gesuches, wo wir ein gängiges Paradigma in der Herzbildgebung mittels Magnetresonanztomographie auf den Kopf gestellt haben. Anstatt die Herz- und Atembewegung durch Synchronisierung mit den bildgebenden Mess Sequenzen zu unterdrücken, hatten wir vorgeschlagen diese Bewegungen zuzulassen und mit modernen mathematischen Methoden herauszurechnen. Das führte dazu, dass die Patientenvorbereitung massiv vereinfacht wurde und dass das schlagende Herz mittels eines einzigen Knopfdrucks abgebildet werden kann. Eine komplizierte Planung der Schichtlagen ist nun nicht mehr nötig. Diese Errungenschaft wird es ermöglichen, dass Herz MRI auch außerhalb von universitären Zentren und in geographischen Regionen wo hochspezialisiertes Fachpersonal fehlt, als nichtinvasive Diagnosemethoden verbreitet werden kann. In diesem Gesuch wollen wir diese Methode nun gezielt verfeinern um kleinste anatomische Strukturen des Herzens darzustellen, um Gewebeveränderungen des Herzmuskels infolge einer Erkrankung genauestens zu beschreiben, und um Blutfluss in kleinen und grossen Gefässen des Herzens als zusätzliche Information zur verbesserten Diagnose und Therapieplanung zur Verfügung zu stellen. Hierbei werden wir auch künstliche Intelligenz auf verschiedenen Etappen dieser Forschungsreise einsetzen, und das Ziel ist es diese neuen Methoden in Zusammenarbeit mit klinischem Fachpersonal gezielt zu Prüfen und zu verfeinern.
Background and Rationale: The SNF generously funded our research project 173129 'A Paradigm Shift in Magnetic Resonance Imaging of the Heart: 5D Imaging - Sample Now and Ask Questions Later' where we developed a new cardiac MRI paradigm in which ECG lead placement, complex scan planning, repeated breath-holding, and time in-efficient acquisitions have been replaced by a free-running image acquisition, where more anatomical and functional information from the heart is collected per unit time than was previously possible, where one single mouse click suffices to initiate a scan, and where a fully flexible and retrospective 3D interrogation of the heart’s anatomy and function is enabled. FRF is so general in its formalism and modular in its design, that it now provides us with the opportunity to leverage its untapped potential in multiple ways: We can now start exploring its boundaries for improved resolution, we can begin to extend its capabilities with new contrast mechanisms for quantitative tissue characterization, and we are in a position to add flow measurements to expand the perimeter of FRF and to research its opportunities. This will critically be supported by the new PilotTone (PT) hardware that surveys physiological motion at a very high frequency conducive of further improved motion suppression as part of FRF.Overall Objectives: To build on the unique capabilities of FRF and extend it with new hardware and software in the pursuit of providing more detailed, better, and new quantitative information about the heart, its anatomy, its function, its tissues, and its vessels. We will rigorously and mechanistically research the boundaries and opportunities of the thus-extended paradigm, and translate it to the clinical setting together with Prof. Jürg Schwitter's team for validation and gold standard comparisons.Specific Aims & Methods to be Used: We will build on infrastructure, tools, and methods available to us and also developed during the original grant period (173129), on new strategic hirings of a mathematician (Dr. Aurélien Bustin), on new collaborations entered in the domain of artificial intelligence (Dr. Jonas Ricardi), and on a tried and tested interdisciplinary collaboration with Prof. Schwitter and his team at CHUV. Both on the acquisition and reconstruction side, new and critical innovation will be added in support of the overall objectives. On the acquisition side, we will extend our free-running pulse sequence with new concepts in support of improved resolution (Aim 1), added contrast mechanisms (Aim 2), and sensitive flow measurements (Aim 3) that add extra dimensions to FRF. Logically, and on the reconstruction side, we will extend the capabilities of our existing reconstruction engine to accommodate this new dimensionality, and we will augment it with new concepts to better constrain motion and improve resolution. By leveraging new hardware and software opportunities available at CHUV, we will rigorously research, translate, and validate this new free-running concept using the following Specific Aims: Specific Aim 1: Equip FRF with advanced acquisition and reconstruction methodology to push its resolution boundaries and test the hypothesis that the visibility of anatomical details is significantly improved in vivo. Specific Aim 2: Integrate, optimize, test, and validate an artificial intelligence (AI)-assisted radiofrequency (RF) pulse sequence design, RF phase cycling methods, and flexible pre-pulse concepts as part of FRF to accommodate new contrast dimensions for tissue characterization.Specific Aim 3: Expand FRF for flow measurements that are sensitive to the high dynamic range of blood-flow in the chambers and vessels of the heart, equip the reconstruction with extra dimensions to accommodate regularization across the different velocity encodes, optimize the parameter ranges, and validate in vivo. Expected Results & Impact for the Field: At the crossroads of prior success and new innovation, new hardware, new strategic collaborations, and new hirings, we will leverage existing FRF technology and logically and mechanistically develop it from a framework that provides information about anatomy and function of the heart only into a framework that offers much more quantitative and detailed information per unit time about the heart, its function, its anatomy, its tissues, and its vessels than was previously possible. Through this research, we will gain in-depth knowledge about new data acquisition and reconstruction strategies in cardiac MRI, and about the optimization of their respective components. We will learn for which parameter ranges the methods will work successfully, and we will also learn at which point success meets failure. We will gain insight into the performance of these concepts in gold standard comparisons, and will therefore likely be able to expand the perimeter of knowledge of both data acquisition and reconstruction paradigms in cardiovascular MRI. Provided that this concerns uncharted territory, our findings will contribute to filling a knowledge gap. FRF equipped with a higher resolution, new, needle-free contrast mechanisms, and a higher dynamic range for blood flow measurements, will critically enable future research and discovery and directly support the improved management of cardiovascular disease, the leading cause of death in industrialized nations. Finally, and given that this augmented FRF considerably improves the ease-of-use for the operator as it remains a 1-click solution, dissemination outside of highly specialized academic centers will be facilitated in support of a broader impact for the field.
Last updated:01.03.2022
