Reduced encoding diffusion spectrum imaging implemented with a bi-Gaussian model.

Area of Science:

• Biomedical Imaging
• Neuroimaging
• Diffusion MRI

Background:

• Diffusion spectrum imaging (DSI) characterizes tissue microstructure via 3-D water diffusion spectra.
• Conventional DSI requires long acquisition times and high q-values, leading to long echo times and low signal-to-noise ratios on clinical scanners.

Purpose of the Study:

• To introduce reduced-encoding DSI (RE-DSI) with a bi-Gaussian diffusion model to overcome limitations of conventional DSI.
• To enable efficient and accurate mapping of complex fiber microstructures using clinical MRI systems.

Main Methods:

• Developed RE-DSI utilizing a bi-Gaussian extrapolation kernel applied to reduced q-space sampling data.
• Validated RE-DSI accuracy using a crossing phantom model and a manganese-enhanced rat model.
• Assessed fiber orientation estimation errors and acquisition time reduction in human studies.

Main Results:

• RE-DSI demonstrated minimal errors in fiber orientation estimation, comparable to the noise limit.
• The method significantly reduced acquisition time for resolving complex fiber orientations.
• RE-DSI successfully fulfilled high q-value requirements with reduced sampling.

Conclusions:

• RE-DSI offers a viable solution for rapid and accurate diffusion spectrum imaging on clinical MRI scanners.
• The bi-Gaussian diffusion model effectively reconstructs 3-D diffusion probability density functions from undersampled data.
• This technique facilitates broader clinical application of advanced diffusion MRI analysis for tissue microstructure.