Joint 6D k-q Space Compressed Sensing for Accelerated High Angular Resolution Diffusion MRI.

Area of Science:

• Medical Imaging
• Neuroscience
• Biophysics

Background:

• Diffusion Tensor Imaging (DTI) assumes Gaussian diffusion, limiting its ability to characterize complex white matter micro-structure.
• High Angular Resolution Diffusion Imaging (HARDI) offers greater precision but requires extensive data, leading to long scan times.
• Compressed Sensing (CS) is a technique to reconstruct signals from fewer samples, but existing CS diffusion MRI (CS-dMRI) methods do not fully leverage k-space information redundancy.

Purpose of the Study:

• To develop a novel framework, 6-Dimensional Compressed Sensing diffusion MRI (6D-CS-dMRI), for reconstructing diffusion MRI signals and the Ensemble Average Propagator (EAP).
• To apply compressed sensing in the full 6D k-q space, integrating information from both k-space and q-space for improved reconstruction.
• To reconstruct the diffusion signal and EAP in continuous spaces from sub-sampled data.

Main Methods:

• Proposed a 6-Dimensional Compressed Sensing diffusion MRI (6D-CS-dMRI) framework.
• Sub-sampled data in both 3D k-space and 3D q-space.
• Reconstructed the diffusion signal and EAP in continuous spaces.

Main Results:

• 6D-CS-dMRI achieved excellent reconstruction of diffusion signals and EAP.
• Demonstrated low root-mean-square error (RMSE) compared to full sampling.
• Utilized 11 times fewer samples (3-fold reduction in k-space, 3.7-fold reduction in q-space) than full Diffusion Spectrum Imaging (DSI) sampling.

Conclusions:

• 6D-CS-dMRI is the first method to apply compressed sensing in the full 6D k-q space for diffusion MRI.
• This approach significantly reduces the number of required measurements, shortening scan times.
• The framework enables precise characterization of white matter micro-structure with substantially less data.