Skip to Main content Skip to Navigation

Développements méthodologiques pour l’IRM de diffusion cardiaque

Abstract : Diffusion magnetic resonance imaging (MRI) is a non-invasive imaging modality that allows the assessment of microarchitecture in biological tissues. Knowing the arrangement of myocardial fiber is important to fully understand cardiac electrophysiology in patients, and therefore to improve the treatment of complex arrythmia such as ventricular tachycardia. Nevertheless, efforts have to be made to apply diffusion MRI to moving organs because of the various artefacts created by displacements. Macroscopic motion, including cardiac contraction and breathing is in the range of a few millimeters, whereas diffusion motion is at the micrometer scale. To deal with cardiac motion, a cardiac-triggered diffusion-weighted spin-echo echo planar imaging sequence can be used with first and second order motion-compensated diffusion-encoding gradients. However, these sequences are very sensitive to the phase of the cardiac cycle chosen to perform the acquisition. Moreover, the ventricle coverage is often limited to a few slices due to the long scan time required to have sufficient SNR. Considering these limitations, the thesis has been separated in two parts with the following objectives: to implement a subject-specific cardiac synchronization and to set up an efficient protocol that performs a full coverage of the left-ventricle with isotropic resolution. The subject-specific cardiac synchronisation has been implemented using a real-time phase contrast (RTPC) sequence which is sensitive to motion. RTPC sequence provides knowledge on the variability in duration of cardiac phases. By modelling the cardiac variability, it was possible to study the quality of the diffusion measurement as a function of the cardiac phase. In regards of the results, the sequence can be valuable to optimize and/or adapt the trigger delay to target the diastasis. This phase is the quiescent phase of the cardiac cycle but also the most variable as a function of heart-rate. The online reconstruction of RTPC sequence has been successfully implemented to facilitate the deployment of the sequence to adapt the trigger delay for future clinical research, possibly with other applications. The focus of the second part of the work was to retrieve tri-dimensional information about myocardial fiber organisation with a full coverage of left ventricle. This full coverage is mandatory to simulate cardiac electrophysiology. A bulk motion-corrected super-resolution approach to cDTI has been proposed to improve spatial resolution without significant cost on SNR. Therefore, the total scan time can be reduced by this method. Evaluation in a numerical heart phantom and in a physical helicoidal phantom, have shown the improvement of spatial resolution and diffusion tensor estimation in a general case (except when the underlying geometry is highly favorable to low resolution images). This is of particular interest when there are significant changes in orientation of principal diffusion direction along the slice direction, according to the global geometry of the heart. The feasibility of super-resolution in-vivo cDTI on healthy volunteers has shown the applicability of the approach.
Complete list of metadata
Contributor : Thèses UL Connect in order to contact the contributor
Submitted on : Tuesday, September 28, 2021 - 8:45:48 AM
Last modification on : Saturday, October 16, 2021 - 11:18:08 AM
Long-term archiving on: : Wednesday, December 29, 2021 - 6:08:04 PM


Files produced by the author(s)


  • HAL Id : tel-03356358, version 1


Anne-Lise Le Bars. Développements méthodologiques pour l’IRM de diffusion cardiaque. Imagerie. Université de Lorraine, 2021. Français. ⟨NNT : 2021LORR0089⟩. ⟨tel-03356358⟩



Record views


Files downloads