Free water elimination tractometry for aging brains.

Area of Science:

• Neuroimaging
• Biomedical Engineering
• Gerontology

Background:

• Diffusion-weighted magnetic resonance imaging (dMRI) tractometry assesses brain connections but faces reliability challenges in aging brains due to increased white matter free water.
• Existing tractometry methods may be less accurate in older adults because they do not account for free water content.

Purpose of the Study:

• To evaluate the impact of computational free water elimination (FWE) on the reliability and accuracy of dMRI tractometry in an aging population.
• To demonstrate that FWE improves tractometry analysis across various pipeline stages and generalizes to different data types.

Main Methods:

• Applied computational free water elimination (FWE) to dMRI data from a cohort of 339 older adults (aged 66-103).
• Assessed improvements in reliability using split-half comparisons at multiple tractometry stages: fiber orientation distribution function estimation, pathway delineation, and tissue property assessment.
• Validated FWE's impact on predictive accuracy by assessing cross-validated predictions of clinician-rated Fazekas scores.
• Tested generalizability using sub-sampled multi-b-value datasets to simulate single-b-value data.

Main Results:

• FWE significantly enhanced tractometry reliability by up to ~37% across all analyzed stages in the aging cohort.
• FWE improved the accuracy of tractometry, leading to more precise cross-validated predictions of Fazekas scores.
• The benefits of FWE were demonstrated to be applicable to both multi-b-value and single-b-value dMRI datasets.

Conclusions:

• Accounting for free water is crucial for reliable and accurate dMRI tractometry, particularly in aging brains.
• FWE offers substantial improvements in analyzing white matter integrity in older adults, enhancing clinical and research applications.
• Open-source software for FWE is provided to facilitate its adoption in diverse neuroimaging studies.