Paula Montesinos1,
2, Himanshu Bhat3, Guangping Dai4, Manuel Desco1,
2, Elfar Adalsteinsson5, Reza Nezafat6, David E.
Sosnovik7
1Departamento
de Bioingeniera e Ingeniera Aeroespacial, Universidad Carlos III de Madrid,
Madrid, Spain; 2Instituto de Investigacin Sanitaria Gregorio Maran
(IiSGM), Madrid, Spain; 3Siemens Medical Solutions, Charlestown,
MA, United States; 4Massachusetts General Hospital, Charlestown,
MA, United States; 5Martinos Center for Biomedical Imaging,
Massachusetts General Hospital, Harvard Medical School, Charlestown, MA, United States; 6Beth Israel
Deaconess Medical Center, Harvard Medical School, Boston, MA, United States; 7Martinos
Center for Biomedical Imaging, Massachusetts General Hospital, Harvard
Medical School, Charlestown, MA, United States
The requirements for the accurate classification of human coronary atherosclerotic plaque by MRI in vivo remain unknown. We aimed here to reconstruct a high resolution ex vivo ground truth dataset of human coronary atherosclerotic plaques with varying degrees of spatial resolution, SNR and motion. Our results indicate that a spatial resolution of 0.5 mm, SNR > 10 and motion correction in all 3 directions are needed for in vivo MRI of coronary plaque. We further show that the acquisition of the central 10-15% of k-space during a motion-free breathold could reduce artifacts even further.