Breakfast & Science Seminar 53

Abstract

Purpose: Cardiac magnetic resonance (CMR) Multitasking enables simultaneous cine imaging and T1-T2 mapping without ECG-gating or breath-holding. However, clinical translation is limited by the long computation times required for iterative Bloch-equation-based T1-T2 fitting, particularly in 3D. This study proposes and evaluates an extended phase graph (EPG) dictionary matching approach to accelerate T1-T2 quantification for a 3D Multitasking stack-of-stars sequence.

Methods: To evaluate the proposed dictionary matching approach, a simulation framework was developed for 3D Multitasking based on the XCAT digital phantom by assigning tissue-specific T1 and T2 values, converting these to MR signal evolution using the 3D sequence timing and EPG simulations, and generating k-space data by Fourier transform and interpolation along the Multitasking readout trajectory. An EPG-based dictionary for T1-T2-B1+ was created and applied in a two-step T1-T2 quantification process: initial T2/B1+ matching followed by T1 refinement. This approach was compared with the original Bloch-based fitting in three scenarios: ground truth phantom images, reconstructed phantom images, and in five healthy volunteers. Quantification times were recorded for both methods. Septal regions-of-interests were analyzed to extract mean T1-T2 values and within-segment standard deviations. Mean absolute errors (MAE) were reported for phantom studies, and a Wilcoxon signed-rank test assessed in-vivo measurement differences.

Results: Both methods produced accurate T1 and T2 values in the phantom (MAE T1: 2.78, MAE T2: 1.58). Minor loss of T1-T2 precision was incurred in the reconstructed phantom and in-vivo data (MAE T1: 79.9, MAE T2: 2.83), but the EPG approach was not inferior to Bloch fitting. In vivo, the two methods did not yield statistically different T1 and T2 values (p > 0.3), but the EPG approach reduced T2 quantification time by 96% and T1 quantification time by 23%.

Conclusion: EPG dictionary matching significantly reduces T1-T2 quantification time without compromising accuracy or precision.

About the Speaker

Katerina Eyre-Yerly is a postdoctoral fellow specializing in cardiac MRI research. She earned her Bachelor of Science in Biochemistry from the University of Toronto, followed by a Master’s degree in Biomedical Engineering from the University of Göttingen, where she focused on the development of cardiac vessel grafts. She later completed her PhD in Experimental Medicine at McGill University, contributing to the clinical translation of novel cardiac MR imaging techniques through research studies. During her doctoral work, she collaborated on the Multitasking framework developed at Cedars-Sinai and on oxygen-sensitive MRI techniques pioneered in Montreal.

In her current role, Katerina is advancing cardiac imaging on a low-field MRI system, with a strong focus on oxygen-sensitive imaging methods. She is also leading efforts to establish and manage a multi-centre, multi-vendor clinical trial utilizing the free-running framework sequence on a 1.5T MRI scanner. Her work bridges basic science and clinical practice, fostering collaboration between researchers and clinicians to improve patient outcomes.