MAPL: Tissue microstructure estimation using Laplacian-regularized MAP-MRI and its application to HCP data.

Area of Science:

• Neuroimaging
• Biophysics
• Medical Physics

Background:

• Diffusion MRI is crucial for studying white matter microstructure.
• Accurate estimation of microstructural features from diffusion MRI data remains a challenge.
• Existing methods struggle with noisy and sparsely sampled data.

Purpose of the Study:

• To introduce and validate a novel analytical regularization technique for Mean Apparent Propagator (MAP)-MRI.
• To improve the reconstruction of Ensemble Average Propagator (EAP) and Orientation Distribution Function (ODF).
• To enhance the accuracy and reduce variance in estimating microstructural parameters using multi-compartment models.

Main Methods:

• Proposed MAPL: analytically regularizing MAP-MRI coefficient estimation using the Laplacian of the reconstructed signal.
• Compared MAPL against state-of-the-art functional basis approaches (original MAP-MRI, modified Spherical Polar Fourier - mSPF).
• Utilized MAPL as a preprocessing step for multi-compartment models (Axcaliber, NODDI) to estimate axon diameter and dispersion.

Main Results:

• MAPL outperformed original MAP-MRI and mSPF in signal fitting and EAP/ODF reconstruction on phantom data.
• MAPL preprocessing significantly reduced parameter estimation variance in human Corpus Callosum data.
• Correlations between Fractional Anisotropy (FA) and estimated microstructural parameters (axon diameter, dispersion) were significant.

Conclusions:

• MAPL provides a robust and accurate method for estimating white matter microstructure from diffusion MRI.
• Using MAPL as a preprocessing step enhances the reliability of multi-compartment model parameter estimation.
• This work contributes to overcoming challenges in understanding brain white matter microstructure.