The Impact of Multiband and In-Plane Acceleration on White Matter Microstructure Analysis

Area of Science:

• Neuroimaging
• Diffusion MRI
• Biomarker Development

Background:

• Accelerated diffusion MRI (dMRI) is crucial for reducing scan times, especially for patient populations like those with mild cognitive impairment (MCI).
• The impact of acceleration techniques, including simultaneous multislice (SMS) and in-plane acceleration (IPA), on the reliability and effect sizes of dMRI measures is not fully understood.
• Understanding these impacts is vital for developing reliable neuroimaging biomarkers for neurological conditions.

Purpose of the Study:

• To investigate the effects of SMS and IPA on the reliability (intraclass correlations, ICCs) and effect sizes of diffusion tensor imaging (DTI) and neurite microstructure measures.
• To evaluate these effects across different acceleration levels (S1P1, S3P1, S3P2, S6P1, S6P2) and participant groups (MCI, healthy older adults, young adults).
• To provide guidance on balancing scan time reduction with data quality for biomarker development.

Main Methods:

• A test-retest study was conducted using diffusion MRI with varying levels of SMS and IPA.
• Diffusion tensor imaging (DTI) measures (FA, MD, AD, RD) and neurite orientation dispersion imaging (NODDI) measures (OD, ISOVF, ICVF) were evaluated.
• Scans were acquired with no acceleration (S1P1, ~20 min) to high acceleration (S6P2, <4 min), and ICCs and effect sizes were calculated.

Main Results:

• Reliability (ICCs) generally decreased with increasing acceleration, with S1P1, S3P1, and S3P2 showing good ICCs, S6P1 moderate-to-good, and S6P2 poor-to-moderate.
• In-plane acceleration did not improve ICCs in areas with high susceptibility distortion.
• Effect sizes in group comparisons (MCI vs. healthy older adults, young vs. healthy older adults) remained similar across S1P1, S3P1, S3P2, and S6P1, indicating preserved sensitivity.

Conclusions:

• Moderate SMS acceleration (SMS=3, S3P1) offers a significant reduction in scan time (one-third) with comparable reliability and effect sizes to no acceleration (S1P1).
• Higher acceleration levels (S6P1, S6P2) reduce reliability, particularly in challenging regions.
• Accelerated dMRI, especially with moderate SMS, is a viable tool for developing reliable biomarkers of neurological decline, balancing efficiency and data integrity.