In vivo measurement of the myelin structure in the human brain using MRI is significant for a wide variety of scientific and clinical neuroimaging applications including brain development and neurological diseases such as multiple sclerosis, Alzheimer's diseases, and amyotrophic lateral sclerosis. Direct imaging of myelin using MRI is difficult because the extremely short T2 of myelin protons causes the signal to decay to zero before it can be acquired. Therefore, in routine clinical MRI protocols, methods that indirectly measure myelin structure using water protons that interact with the myelin in the lipid-protein bilayer structure are used. These methods include diffusion-weighted imaging, magnetization transfer and multi-exponential fitting. These indirect measurement methods sometimes can be ambiguous in distinguishing signal from the myelin water and that from the intracellular and extracellular water. Recent studies show the prospect of direct myelin MRI (DM-MRI) using ultra-short echo times (< 50 s). The DM-MRI contrast is generated using an inversion pulse to suppress signal from long T2 species such as water and fat, half-pulse excitations and radial readouts for the ultra-short echo acquisition, and subtraction of data acquired at two different echoes to produce a myelin image with high morphological contrast. The drawbacks of DM-MRI are the long scan time (~5 min for a 128x128 slice) due to the idling time in inversion and the radial acquisition and long reconstruction time from the non-Cartesian radial acquisition. These drawbacks prohibit DM-MRI from being used routinely for whole brain clinical and scientific scans. The proposed work will use simultaneous multi-slice (SMS) parallel imaging methods to significantly increase the imaging speed of DM-MRI. Multiple slices will be simultaneously inverted using a power independent radio frequency (RF) pulse method and then simultaneously excited using a multi-slice half RF pulse. Phase cycling will be applied to each radial line to control aliasing for a robust SMS reconstruction. To reduce the time and complexity of the reconstruction, a data-driven reconstruction and a vectorized gridding method that fully uses the parallel processing power of modern CPUs are proposed. The final result will be a whole brain DM-MRI method with a scan time of approximately five minutes that will be practical for clinical and research protocols on conventional MRI scanners.

Public Health Relevance

Direct in vivo imaging of myelin in the human brain using MRI has a significant impact for clinical and basic scientific research. The main obstacle in whole brain myelin imaging is the prohibitively long scan times. The aim of the proposed research is to increase the imaging speed of direct myelin MRI using parallel imaging and simultaneous multi-slice acquisitions for whole brain scanning. We will also develop a parallel imaging reconstruction that is fully integrated on the scanner.

Agency
National Institute of Health (NIH)
Institute
National Institute of Biomedical Imaging and Bioengineering (NIBIB)
Type
Exploratory/Developmental Grants (R21)
Project #
5R21EB020760-02
Application #
9261522
Study Section
Biomedical Imaging Technology B Study Section (BMIT-B)
Program Officer
Liu, Guoying
Project Start
2016-05-01
Project End
2018-02-28
Budget Start
2017-03-01
Budget End
2018-02-28
Support Year
2
Fiscal Year
2017
Total Cost
$159,187
Indirect Cost
$46,687
Name
University of Hawaii
Department
Internal Medicine/Medicine
Type
Schools of Medicine
DUNS #
965088057
City
Honolulu
State
HI
Country
United States
Zip Code
96822
Rettenmeier, Christoph; Maziero, Danilo; Qian, Yongxian et al. (2018) A circular echo planar sequence for fast volumetric fMRI. Magn Reson Med :
Song, Hao; Ruan, Dan; Liu, Wenyang et al. (2017) Respiratory motion prediction and prospective correction for free-breathing arterial spin-labeled perfusion MRI of the kidneys. Med Phys 44:962-973