Myocardial Radiomics and Mechanics in the Pathology and Prognosis of Cardiovascular Disease

PROJECT SUMMARY Though knowledge advances have been made in identification and treatment of risk factors, cardiovascular disease (CVD) remains the leading cause of death in the U.S., and has been worldwide for the past 15 years. The mechanisms of three leading sources of CVD morbidity, ischemic coronary artery disease, heart failure, and atrial fibrillation, appear rooted in myocardial pathophysiology of the left ventricle (LV), suggesting that deeper phenotyping of the LV may provide insight into the diseases. The detection of earliest forms of LV dysfunction has been challenging: traditional measures of LV structure and function associated with poor outcomes, such as mass and ejection fraction, reflect advanced stages of disease. However, recent advances in quantitative imaging of the LV myocardium using cardiovascular magnetic resonance (CMR) images may bridge this gap. Myocardial texture analysis is a type of ‘radiomics’, the computation of myocardial pixel intensity and patterns, which has shown the ability to differentiate pathological patterns in LV hypertrophy. Additionally, analysis of myocardial mechanics including strain and torsion have shown prognostic value in evaluation of cardiomyopathies. Thus, increasing evidence indicates roles for these technological advances in imaging analysis to evaluate pathological LV remodeling in subclinical cardiovascular disease in the general population. We hypothesize that application of these novel imaging analytics, correlated with biology, subclinical disease phenotyping, and outcomes, will enable more granular insight into subclinical LV structural and functional changes predating overt CVD and its forms. The Framingham and Jackson Heart Studies, community-based cohorts of whites and African Americans with longitudinal follow up, offer the opportunity to gain insight in CVD development through integration of advanced secondary imaging analysis with detailed and broad phenotyping and genotyping, and clinical end-points. Thus, the objectives of our proposal are threefold: 1) to first identify myocardial texture analysis and mechanics patterns associated with prevalent CVD and risk factors, 2) to define the biological and genetic underpinnings of these phenotypes, and 3) to understand their inter-association between structure and function, and their joint relations with long-term prognosis. Execution of our Aims will elucidate novel patterns, determinants, and prognosis of underlying early and progressive myocardial remodeling and dysfunction in CVD in a large bi-racial cohort. Ultimately, our goal is for knowledge gained from this study to advance phenotyping of LV remodeling groups, methods of cardiovascular risk stratification, and development of therapies for patient care.