Neural Field Simulation via Cortical Spherical Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for simulating human brain phenomena, such as epilepsy, face challenges with low spatial resolution in brain models, leading to less accurate and computationally demanding results, particularly in applications like epilepsy simulation, tumor effects, and brain stimulation.
Innovation Solution
The method involves transforming the cortical surface of the brain into a spherical domain, decomposing the neural field using Fourier transforms, and recomposing it to achieve higher spatial resolution simulations while reducing computational burden, allowing for faster and more accurate results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If low spatial resolution modeling is used in the Virtual Brain, then computational resources required are reduced, but simulation accuracy deteriorates
Solution Approach 1:
The patent replaces the traditional finite difference time domain (FDTD) mechanical computation method with a neural field equation approach. This substitution transforms the problem from solving complex electromagnetic field equations to solving a simplified neural field equation that captures essential brain dynamics while requiring fewer computational resources, thus resolving the contradiction between computational efficiency and simulation accuracy.
Solution Approach 2:
The patent changes the fundamental parameters of the simulation model by transitioning from high-resolution anatomical models with detailed tissue properties to a neural field model that uses aggregated neural population parameters. This parameter transformation enables simulations to run with reduced computational burden while maintaining biological realism through the neural field equations that describe collective neural behavior.
2Measurement precision
If high spatial resolution modeling is used in the Virtual Brain, then simulation accuracy is improved, but computational resources required increase significantly
Solution Approach 1:
The patent replaces the computationally intensive FDTD method with a neural field equation approach that achieves comparable or superior accuracy for neural population dynamics while requiring fraction of the computational resources. The neural field equations naturally capture spatial resolution at the scale of neural populations without needing to model every individual neuron or detailed tissue structure.
Solution Approach 2:
The patent extracts and focuses only on the essential neural population dynamics from the complex brain tissue physics, removing unnecessary computational details while retaining the core phenomena of interest. By taking out only the relevant neural field variables and their evolution equations, the model achieves high spatial resolution for neural activity without the computational burden of modeling all underlying biological processes.
3Reliability
If traditional FDTD method is used for brain simulations, then detailed electromagnetic field dynamics are captured, but simulation speed decreases
Solution Approach 1:
The patent substitutes the FDTD electromagnetic field simulation with a neural field equation model that captures the essential dynamics of neural populations. This substitution maintains reliability for studying neural activity patterns, seizure propagation, and brain network dynamics while achieving simulation speeds orders of magnitude faster by avoiding the computationally expensive electromagnetic field calculations.
Solution Approach 2:
The patent segments the complex electromagnetic field problem into distinct neural population dynamics that can be modeled independently using neural field equations. By dividing the brain into neural populations and modeling their interactions through the neural field framework, the system captures essential dynamics without needing to resolve all electromagnetic field details, thus improving simulation speed while maintaining reliability for neural dynamics studies.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach significantly enhances simulation accuracy and speed, achieving results up to 1000 times faster than existing methods by efficiently handling high-resolution neural field simulations, effectively addressing the limitations of low spatial resolution in brain modeling.
Implementation Method 1
decomposing the neural field using Fourier transforms, and recomposing it to achieve higher spatial resolution simulations
Data Source
AI summary
The method of simulating a human brain neural field in a computerized platform modelling various zones of a human brain and connectivity between the zones includes: providing the computerized platform modelling the various zones of the human brain and connectivity between the zones; acquiring three-dimensional anatomical structural imaging data of a folded surface of a cortex of a brain of a human patient; personalizing the computerized platform according to the structural data; providing an equation describing a spatiotemporal evolution of the neural field and loading the equation in the computerized platform; performing a projection of the surface of the cortex of the brain of the patient on a spherical surface domain; simulating the neural field in the spherical domain; and translating the simulated neural field in the spherical domain in the cortical domain.