Yitong Yang
BME PhD Defense Presentation

Date: 2026-03-30
Time: 10:00 AM - 12:00 PM
Location / Meeting Link: HSRBII N100 (https://emory.zoom.us/j/91002600925)

Committee Members:
John Oshinski, PhD (advisor); Deqiang Qiu, PhD; Eva Dyer, PhD; Matthias Stuber, PhD; Matthias Chung, PhD


Title: Optimizing 5D free-running Cardiac MRI for rapid assessment of respiratory-related cardiac diseases

Abstract:
Cardiac magnetic resonance (CMR) can play a pivotal role in characterizing cardiac anatomy, function, and hemodynamics in the setting of myocardial infarction, heart failure, and congenital heart disease. However, CMR is currently under-utilized due to the complexity of scan planning and need for breath-holding or external gating which results in unpredictable scan times or poor patient compliance. A free-running 5D CMR technique has previously developed that can be easily planned and is independent of external gating. The fixed-time whole-heart coverage of the 5D acquisition provides 3D image volumes with isotropic resolution resolved in multiple cardiac phases. However, the current reconstruction from the acquired free-running raw data to the 5D image takes hours since an iterative algorithm is needed to solve the reconstruction problem. This long reconstruction time restricts free-running CMR from online display of the 5D image sets and therefore hampers routine clinical usage. Since free-running CMR provides a respiratory dimension in the image, it also opens the possibility to investigate into the respiratory variability of cardiac function and hemodynamics that may be important for diagnosis, treatment planning, and prognosis. In this thesis, the specific aims are: 1) develop and apply numerical and machine learning-based techniques to streamline the reconstruction of free-running CMR for functional analysis; 2) investigate respiratory variability of hemodynamics in right heart function analysis in volunteers and pediatric patients with congenital heart disease.