A computational framework for characterizing normative development of structural brain connectivity in the perinatal...

Area of Science:

• Neuroscience
• Medical Imaging
• Computational Biology

Background:

• Accurate assessment of perinatal brain structural connectivity is crucial for understanding neurodevelopment.
• Estimating the structural connectome from diffusion MRI is computationally challenging, especially during rapid perinatal development.
• Lack of normative connectivity trends hinders the reliable assessment of normal and abnormal brain development.

Approach:

• Developed a computational framework utilizing spatio-temporal atlases to establish baseline trends in perinatal brain connectivity.
• Applied the framework to diffusion MRI data from 169 subjects aged 33-45 postmenstrual weeks.
• Investigated the impact of different connection weighting strategies and evaluated various connectivity metrics.

Key Points:

• The proposed framework successfully reveals clear and robust trends in structural connectivity development during the perinatal period.
• Connection weighting based on neurite density demonstrated more consistent developmental trends compared to other methods.
• Global efficiency, local efficiency, and characteristic path length showed more consistent trends than other evaluated connectivity metrics.

Conclusions:

• The spatio-temporal atlas-based framework provides a reliable method for charting normative structural connectome development in the perinatal stage.
• Findings highlight the importance of neurite density for consistent connectivity assessment and identify key metrics for tracking perinatal brain development.
• This work establishes crucial baselines for identifying deviations in brain development during this critical period.