Quantifying the brain's sheet structure with normalized convolution.

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Biophysics

Background:

• The hypothesis of 2D sheet-like brain pathways, analogous to book pages, remains debated due to a lack of quantitative validation.
• Previous support relied on qualitative diffusion MRI (dMRI) tractography, which is limited by noise and reconstruction challenges.
• A quantitative measure is essential to rigorously assess the prevalence and nature of these sheet structures in the brain.

Purpose of the Study:

• To develop a robust quantitative method for assessing brain pathway sheet structures.
• To overcome the limitations of existing tractography-based approaches for analyzing sheet-like neural organization.
• To provide a tool for validating the sheet structure hypothesis across various neuroimaging data.

Main Methods:

• Proposed a novel method to estimate the Lie bracket, a quantitative indicator of sheet structure.
• Utilized normalized convolution to compute derivatives of fiber orientation density function peaks, avoiding tractography reconstruction.
• Validated the method using simulated data and experimental dMRI datasets from different species and acquisition parameters.

Main Results:

• The novel Lie bracket estimation method demonstrated increased robustness against noise and missing orientation peaks compared to prior techniques.
• The method successfully quantified sheet structure support across diverse dMRI datasets (species, scanners, resolutions, etc.).
• The approach is adaptable for directional data from modalities beyond dMRI, broadening its applicability.

Conclusions:

• The developed method provides a reliable quantitative tool for investigating brain pathway sheet structures.
• This quantitative approach facilitates robust validation of the sheet structure hypothesis in neuroscience.
• The method's versatility supports its use in diverse neuroimaging research and cross-validation efforts.