Role of Data-driven Regional Growth Model in Shaping Brain Folding Patterns

Area of Science:

• Neuroscience
• Computational Biology
• Developmental Biology

Background:

• Brain surface morphology is vital for function and dysfunction.
• Computational modeling aids understanding of early brain folding mechanisms.
• Regional variations in brain tissue growth are observed, but their role in cortical development is unclear.

Purpose of the Study:

• To investigate how regional cortical growth influences brain folding patterns using computational simulation.
• To develop and apply machine learning-assisted growth models for simulating cortical development.

Main Methods:

• Developed regional cortical growth models using machine learning-assisted symbolic regression.
• Utilized longitudinal MRI data from 735 pediatric subjects (29 post-menstrual weeks to 24 months).
• Simulated cortical development with anatomically realistic models and quantified folding patterns.

Main Results:

• Regional growth models produced more accurate brain folding patterns than uniform growth models.
• Growth magnitude was the primary factor influencing folding, with trajectory having a minor effect.
• Multi-region models better captured folding intricacies than single-region models.

Conclusions:

• Incorporating regional growth heterogeneity is essential for accurate brain folding simulations.
• This approach can improve early diagnosis and treatment of cortical malformations and neurodevelopmental disorders.
• Findings highlight the importance of regional variations in understanding brain development.