A spatially distributed computational model of brain cellular metabolism

Area of Science:

• Computational neuroscience
• Biophysics
• Mathematical biology

Background:

• Spatially lumped models simplify complex brain tissue but may obscure crucial spatial dynamics.
• Understanding brain cellular metabolism requires accounting for the intricate interplay of diffusion, cellular activity, and vascular networks.

Purpose of the Study:

• To develop a 3D spatially distributed model of brain cellular metabolism.
• To investigate the impact of synaptic activity locus relative to capillaries on model behavior.
• To analyze the relationship between spatially distributed and lumped models.

Main Methods:

• Developed a novel multi-domain reaction-diffusion formulation for brain tissue, incorporating separate mitochondria.
• Employed model reduction techniques to relate distributed and lumped models.
• Utilized a 1D radial model to illustrate spatial resolution effects and analyze parameter changes.

Main Results:

• Spatially lumped models are consistent spatially averaged approximations of the distributed model.
• Varying synaptic activity locus relative to capillaries significantly alters lumped model parameters.
• Loss of spatial resolution in modeling leads to inherent uncertainty in lumped models.

Conclusions:

• The location of synaptic activity relative to capillaries is a critical factor in brain metabolism modeling.
• Parameter estimation from data is significantly affected by spatial resolution limitations.
• Spatially distributed models offer superior insights into brain dynamics compared to lumped models.