Optimized Diffusion Imaging for Brain Structural Connectome Analysis.

Area of Science:

• Neuroimaging
• Diffusion Magnetic Resonance Imaging (dMRI)
• Structural Connectomics

Background:

• High Angular Resolution Diffusion Imaging (HARDI) is crucial for human brain structural connectome analysis.
• Acquiring dense q-space samples in HARDI leads to lengthy scanning times and logistical issues.
• Accurate estimation of the structural connectome relies on efficient dMRI data acquisition.

Purpose of the Study:

• To develop a statistical method for optimal q-space direction selection in HARDI.
• To enable accurate estimation of the local diffusion function from sparse dMRI observations.
• To reduce scanning time and logistical challenges in HARDI data acquisition.

Main Methods:

• A statistical approach was proposed to optimally select q-space directions.
• Leveraged historical dMRI data to compute prior distributions for local diffusion variability.
• Mapped priors to subject-specific coordinates to guide q-space sample selection.
• Applied the method to Human Connectome Project data and mild cognitive impairment datasets.

Main Results:

• The proposed method demonstrated significant advantages over existing HARDI frameworks in simulations.
• Structural brain networks were recovered with high fidelity using only 15-20 q-space samples.
• Results were comparable to those obtained using 60+ diffusion directions with conventional methods.

Conclusions:

• The developed statistical method effectively reduces HARDI sampling requirements.
• Accurate structural connectome estimation is achievable with significantly fewer diffusion directions.
• This approach offers a more efficient and practical solution for HARDI-based neuroimaging studies.