Low - Field MR Fingerprinting for Myocardial Tissue Characterization in Patients with Cardiac Implantable Electronic Devices

About This Grant

PROJECT ABSTRACT/SUMMARY Over two million Americans have a cardiac implantable electronic device (CIED) to manage arrhythmias, often secondary to cardiomyopathies, and prevent sudden cardiac death. Magnetic Resonance Imaging (MRI) is a powerful tool for non-invasive detection of myocardial tissue abnormalities—including inflammation, edema, and fibrosis—using late gadolinium enhancement (LGE) imaging and quantitative T1, T2, and extracellular volume fraction (ECV) mapping. However, MRI remains widely underutilized in patients with CIEDs, as metallic components in these devices distort the magnetic field and create off-resonance artifacts that severely degrade image quality. Although 16% of patients with CIEDs will develop a clinical indication for cardiac MRI at some point, they are 40% less likely to undergo an MRI exam than non-CIED patients, leaving clinicians to rely on alternative modalities that lack detailed tissue characterization capabilities. This project aims to address this unmet need by developing MRI technology for robust myocardial tissue characterization specifically tailored for CIED patients, providing quantitative tissue property maps (T1, T2, proton density, and ECV) and multi-contrast LGE images from a single imaging platform. Our solution is based on Magnetic Resonance Fingerprinting (MRF), an innovative framework that measures temporal changes in magnetization (“fingerprints”) to achieve rapid multiparametric mapping. We propose a novel cardiac MRF technique that is separately optimized for conventional 1.5T scanners, given that most CIED patients are currently imaged at this field strength, as well as emerging low-field 0.55T scanners, which offer inherent advantages for imaging near metallic implants, such as reduced off-resonance artifacts. In Aim 1, we will develop the proposed technique in parallel for 1.5T and 0.55T by integrating robust data collection strategies, such as center-out radial sampling and wideband excitation pulses, with a physics-informed deep learning reconstruction that further suppresses off-resonance artifacts and residual motion. We will evaluate accuracy, precision, and repeatability in phantoms and healthy subjects with an externally positioned CIED scanned at both field strengths compared to conventional cardiac mapping methods. In Aim 2, we will assess MRF for native and post-contrast mapping in cardiomyopathy patients with CIEDs at 0.55T and 1.5T and validate methods for detecting myocardial fibrosis, given its prognostic importance in many conditions. To this end, we will generate synthetic multi-contrast (bright- and dark-blood) LGE images from post-contrast MRF maps, which we expect to simplify the exam and enhance detection of fibrosis compared to conventional bright-blood wideband LGE scans. Furthermore, we will develop a clustering algorithm that directly analyzes tissue property maps to identify fibrosis, serving as a semi-automated and operator- independent alternative to LGE. This project has the potential to have a significant and immediate impact on public health by expanding access to advanced MRI techniques for myocardial tissue characterization in CIED patients, who are at high risk of adverse cardiovascular events yet are underserved by current MRI technology.