Quantifying human gray matter microstructure using neurite exchange imaging (NEXI) and 300 mT/m gradients

Area of Science:

• Neuroimaging
• Biophysical modeling
• Human brain microstructure

Background:

• Biophysical models like Neurite EXchange Imaging (NEXI) are crucial for quantifying gray matter microstructure.
• NEXI requires multi-shell, multi-diffusion time data and was previously limited to preclinical animal studies.
• Translating advanced neuroimaging models to human in vivo applications is essential for clinical research.

Purpose of the Study:

• To translate the Neurite EXchange Imaging (NEXI) model for in vivo human cortical microstructure analysis.
• To compare the performance of different NEXI model variants, including those correcting for Rician noise.
• To assess the reliability and sensitivity of NEXI estimates in the human cortex.

Main Methods:

• Utilized a 3 T Connectom MRI system with high gradient strength (300 mT/m) for data acquisition.
• Acquired data across a wide range of b-values (0–7.5 ms/µm²) and diffusion times (20–49 ms).
• Compared four NEXI model variants (NEXI, NEXI_dot, NEXI_RM, NEXI_dot,RM) using synthetic and human in vivo data.

Main Results:

• The NEXI_RM variant, which corrects for Rician noise, demonstrated superior performance in human cortical data.
• NEXI framework showed sensitivity to individual differences in cortical microstructure.
• Scan-rescan variability was evaluated against between-subject variability to assess reliability.

Conclusions:

• The successful in vivo application of NEXI in the human cortex is a significant advancement.
• Rician noise correction is critical for accurate NEXI parameter estimation in human neuroimaging.
• NEXI offers new possibilities for studying neurodevelopment, aging, and neurodegenerative diseases.