Student: Giorgia Milotta
Cardiovascular disease (CVD) remains the leading single cause of death worldwide despite improvements in prevention and advances in diagnosis and treatment.1 The major cause for CVD death is sudden coronary atherothrombosis due to plaque rupture and subsequent thrombus formation. Detection of coronary atherosclerosis and prediction of adverse myocardial remodeling remain challenging with current imaging techniques. Magnetic resonance imaging (MRI) is considered the gold standard for the assessment of cardiac anatomy, left ventricular (LF) function (CINE-MRI), myocardial viability (LGE-MRI), tissue characterization (T1 and T2 mapping) and lately also perfusion (MR-perfusion) due to its excellent soft tissue contrast, high spatial resolution and lack of ionizing radiation. However, MRI assessment of coronary lumen integrity and plaque burden/activity remains challenging. Nevertheless, MRI has shown great potential for coronary lumen2,3, plaque (with and without contrast agents)4-6 and thrombus/haemorrhage7,8 visualization. Especially coronary plaque and thrombus characterization could add invaluable prognostic information as shown by Noguchi et al.8 but still suffers from lack of robustness and unpredictable scan times due to the use of simplified respiratory motion gating and insufficient motion models. The major challenge of 3D cardiac MRI protocols is the need for improved motion correction to minimize motion artefacts and the need for image acceleration to shorten scan time.
To enable high-resolution 3D cardiac imaging with isotropic resolution and no need for breath holding, and to address the technical limitations of MR coronary artery imaging (lack of robustness, long scan times) we recently have developed an imaged based self-gated motion correction technique, also referred to as “image navigator” or iNAV, that has significantly improved image quality and shortened the overall scan time9. The proposed iNAV creates 2D self-navigation images by adding phase and frequency encoding gradients to the startup profiles of the high-resolution balanced steady-state free precession (SSFP) 3D CMRA sequence. Compared to 1D self-navigation approaches which are typically acquired as 1D projections (center k-space line of imaging sequence)10-12 of the field-of-view (FOV) and include both signal from static and moving tissue, the iNAV allows separating moving from static tissue using spatial encoding, and thus improves motion tracking as it allows for multidimensional and more complex motion correction. Similar to 1D self-navigation, the iNAV does not require any additional planning as its geometry is derived from the high-resolution 3D imaging sequence itself thereby improving ease of use and minimizing operator dependence. Registration of the 2D iNAV’s to a common reference (typically end-expiration) using template matching algorithms allows estimating the respiratory motion of the heart in foot-head and left-right direction (coronal view) or foot-head and anterior-posterior direction (sagittal view) and to correct for this motion with rigid and/or non-rigid motion compensation.
In this proposal we aim to devise, implement and validate an accelerated and respiratory self-navigated parametric 3D magnetic resonance imaging (MRI) iT2prep coronary sequence that enables the 1) simultaneous acquisition of co-registered coronary lumen and vessel wall images as well as 2) myocardial T2 maps, and 3) accelerated and respiratory self-navigated 3D black blood late gadolinium enhancement MRI sequence (BB-LGE) for comprehensive assessment of whole-heart coronary plaque and myocardial scar.
The specific objectives of the project are:
1) To devise and implement respiratory self-navigation for simultaneous 3D coronary lumen and vessel wall imaging (so called, iT2prep sequence) as well as myocardial T2 mapping to improve ease-of-use, reliability, scan time predictability and finally diagnostic image quality.
2) To implement undersampled 3D MRI reconstruction using parallel imaging and low-rank compressed sensing to speed up the acquisition of the parametric iT2prep sequence.
3) To devise and implement respiratory self-navigation for 3D black blood late gadolinium enhancement MRI sequence (so called, BB-LGE) to improve ease-of-use, reliability, scan time predictability and finally diagnostic image quality.
4) To implement undersampled 3D MRI reconstruction using parallel imaging and low-rank compressed sensing to speed up the acquisition of BB-LGE sequence.
If time permits, both sequences may be implemented as dual phase acquisition to provide systolic and diastolic datasets for assessment of left ventricular function and to select the best dataset for coronary and myocardial visualisation.