MR/PET from a MR-Jock’s Point of View: Self-inflicted Wounds, Synergies and Unique Opportunities

BIO: Dr. Fernando Boada received his B.Sc. from Universidad Simon Bolivar in Caracas Venezuela in 1985. After teaching physics courses for two years, Dr. Boada enrolled in the graduate studies program at Case Western Reserve University where he received his Master of Science and PhD in Physics in 1989 and 1990, respectively. Dr. Boada's doctoral work was performed under the supervision of Drs. E. Mark Haacke and William Tobocman, and dealt with the use of localized polynomial approximations for the inverse scattering reconstruction of ultrasonic data in one dimension. After one year of post-doctoral work in MRI at Case Western Reserve, Dr. Boada spent two years at the Massachusetts General Hospital and Harvard Medical School, first as a post-doctoral fellow and later as an Instructor in Radiology. Dr. Boada then joined the faculty at the University of Pittsburgh where he was Professor of Radiology and Bioengineering and the Director of the University of Pittsburgh MR Research Center until 2012. In 2012, Dr. Boada joined New York University to head the newly formed Center for Advanced Imaging Innovation and Research (CAI²R), which focuses on rapid development of translational imaging tools. Dr. Boada’s early work was focused on the development of Fast Three Dimensional imaging techniques with ultra-short TE for sodium MRI of stroke and cancer. He has also worked on k-space Trajectory Design, Parallel Imaging and Transmission, RF pulse design, MR Image Reconstruction and more recently MR/PET data acquisition/reconstruction and fast Diffusion Spectrum Imaging. Dr. Boada has been a member of the ISMRM for 29 years and has also served as charter member of several NIH study sections over the last twenty years. He was elected into the college of fellows of the American Institute for Medical and Biological Engineering. He is Senior Fellow of the ISMRM and a Distinguished Investigator of the Academy of Radiology Research. 

ABSTRACT: MR/PET is a truly synergistic hybrid imaging modality stemming for the truly simultaneous and independent nature of the data acquisition of  its PET and MRI subsystems. Initial usage of MR/PET workflow protocols were, however,  burdened with concerns related to the proper quantification of the signal. These are now much better understood and, theoretically, properly addressed through the use of time-of-flight technology. Despite these early setbacks, the simultaneous nature of the MR/PET data acquisition process has led to new opportunities for taking advantage of PET’s unique sensitivity for probing metabolic processes in vivo. Examples include: improving PET’s detection sensitivity through the use of joint MR/PET reconstruction, reduction of radiation exposure (90%) for standard-of-care Fluorodeoxyglucose (FDG) scans and motion correction of “dynamic” PET scans.

In this presentation, we will demonstrate two applications of MR/PET synergies. The first application will illustrate the use of MR-guided PET reconstruction as an effective approach for improving the detection sensitivity of PET scans in vivo. Specifically, we will illustrate how the removal of partial voluming effects using MR’s anatomical information improves the detection of hypometabolism (FDG) in the setting of Epilepsy. Implications for other neuro-degenerative disorders will also be illustrated through the use of PET tracers where signal-to-noise ratio and tissue uptake limits the attainable image quality via standard image reconstruction means. The second demonstration will illustrate an approach for the correction of cardiac-respiratory motion blur based on the use of self-refocused, multi-dimensional “coil fingerprints” modules. These fingerprints provide a generalized projection through the imaged object that allows motion tracking with a time resolution approaching the MR sequence’s TR. We have demonstrated, using in-vivo as well as motion phantom data, that these modules are compatible with both 3D and 2D MRI sequences and that their use allows the removal of large (60%) quantitative biases in PET scans due to multi-dimensional (i.e., cardiac and respiratory) motion blur.