Non-contrast 3D T1p Mapping for Myocardial Fibrosis Quantification of Pediatric Cardiomyopathy Patients

点击翻译按钮获取中文摘要

PROJECT SUMMARY The development of myocardial fibrosis is associated with nearly all forms of pediatric heart disease including hypertrophic cardiomyopathy, congenital heart disease, diastolic dysfunction, arrhythmia, myocarditis, and sudden cardiac death. Despite the pervasive nature of myocardial fibrosis, the current technology available to detect fibrosis is suboptimal for studying pediatric cardiomyopathy. Cardiac MRI (CMR) is the gold standard noninvasive screening tool to detect both diffuse and focal fibrosis, through extracellular volume (ECV) and late gadolinium enhancement (LGE) imaging, respectively. Unfortunately, both ECV and LGE CMR require the administration of a gadolinium-based contrast agent (GBCA), which accumulates in the brain even when renal function is normal, including in children. In addition, traditional CMR requires subjects to hold their breath for accurate imaging. However, many pediatric patients cannot adequately hold their breath and so are put under general anesthesia (GA), which is not ideal as GA poses an additional health risk and significant financial cost. Furthermore, the current 2D techniques for fibrosis imaging have insufficient spatial resolution, and thus are only able to acquire data in sections of the left ventricle (6-10 mm thick) of the heart, completely missing fibrosis information in the right ventricle (3-5 mm thick), which is known to be the substrate for some tachycardia arrhythmias. Therefore, breathing, T1ρ mapping is a promising non-contrast CMR technique that can be used to detect both focal and diffuse myocardial fibrosis. Despite its enormous potential for assessment of myocardial fibrosis in pediatric patients, cardiac T1ρ mapping suffers from several technical limitations: (a) poor spatial resolution, (b) long scan time (up to 18 min), and (c) undeveloped pipeline for clinical integration. Additionally, the volumetric cardiac T1ρ mapping sequences that have been developed have only been tested on adult patients, and in very few subjects (n < 15). Therefore, in this study, I seek to address these limitations of 3D cardiac T1ρ mapping by (1) using innovative k-space sampling with deep learning for achieving unprecedented image quality with acceptable scan and reconstruction time, (2) implementing deep learning to automate image analysis and fibrosis quantification to make the information readily accessible for patient care, and (3) scanning a large population of pediatric patients to make this the most comprehensive T1ρ mapping study to date. there is a strong need to develop a non-contrast, free- volumetric imaging test for detecting fibrosis in pediatric patients.