Analytic Bounds on GAMLSS Model Variability of Normative White Matter Brain Charts.

Area of Science:

• Neuroimaging
• Developmental Neuroscience
• Biostatistics

Background:

• Brain charts provide normative models for quantitative neuroimaging, tracking brain development and identifying abnormalities.
• Extending brain charts to white matter microstructural and macrostructural features requires variance assessment for model stability.

Purpose of the Study:

• To implement an analytic approach for characterizing parameter variability in existing white matter brain charts.
• To empirically validate the analytic models against a bootstrapping approach across a wide age range (0.2 to 90 years).

Main Methods:

• Utilized generalized additive models for location, scale, and shape (GAMLSS) framework for brain chart creation.
• Implemented an analytic method to assess parameter variability (coefficient of variation - COV).
• Compared analytic COV estimates with empirical COV estimates derived from bootstrapping.

Main Results:

• Analytic coefficient of variation (COV) remained below 5% for ages > 0.25 years; empirical COV reached a maximum of 7% at 0.2 years.
• High agreement between analytic and empirical assessments (Pearson correlation coefficient of 0.776).
• Volume and surface area showed the largest average COV; analytic method identified axial diffusivity as having the smallest COV, while empirical method identified average length.

Conclusions:

• The analytic approach may overestimate model stability in white matter brain charts, particularly when COV is low.
• Empirical validation is crucial for assessing the stability of GAMLSS models used in neuroimaging.
• Both methods indicate that white matter volume and surface area are the most variable features.