Multi-Scale Brain Network Modeling by Integrating DCM and Biophysics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing brain function studies using different experimental technologies provide incomplete pictures due to varying spatio-temporal sampling rates, necessitating an approach to integrate these measurements for a comprehensive model.
Innovation Solution
Combining dynamic causal modeling (DCM) with biophysics modeling to integrate functional neuroimaging and electrophysiology data, using sequential model fitting to enhance modeling accuracy and generate a comprehensive brain neuronal circuitry model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple experimental technologies are used to study brain function, then measurement coverage and spatio-temporal resolution are improved, but data integration complexity and model comprehensiveness deteriorate
Solution Approach 1:
The patent combines multiple experimental technologies (fMRI, electrophysiology, optical imaging) into a unified multi-scale modeling framework. Different data modalities are integrated through a common computational model that reconciles their varying spatio-temporal resolutions, transforming complex multi-source data into a coherent representation of brain network function.
Solution Approach 2:
The patent introduces computational modeling as an intermediary layer between raw experimental data and biological interpretation. This mediator translates data from different measurement modalities with varying resolutions into a unified multi-scale model, resolving the complexity of direct data integration while preserving the unique advantages of each technology.
2Loss of information
If a comprehensive model of brain function is constructed, then understanding of brain network function is improved, but computational complexity and data processing requirements worsen
Solution Approach 1:
The patent segments the comprehensive brain function model into multiple scales (e.g., neural, network, system levels), each handled by specialized computational components. This segmentation allows the model to maintain comprehensiveness while managing computational complexity through modular processing of different spatial and temporal scales separately.
Solution Approach 2:
The patent adds a computational modeling dimension that bridges experimental measurements with biological interpretation. By introducing this intermediate computational layer, the system can process comprehensive multi-modal data without proportionally increasing computational burden, as the modeling framework provides structured organization and efficient computation pathways.
3Measurement precision
If sequential model fitting is used to improve modeling accuracy, then model precision is improved, but processing time and computational resources worsen
Solution Approach 1:
The patent performs preliminary model fitting at coarser scales or with simplified parameters before refining the model at finer scales. This preliminary action establishes a foundation that guides subsequent detailed fitting, reducing the total computational time required for sequential model fitting while maintaining high accuracy in the final model.
Solution Approach 2:
The patent implements iterative refinement cycles where the model is fitted, evaluated, and adjusted in periodic stages. Each cycle improves accuracy incrementally, allowing computational resources to be distributed efficiently across multiple processing stages rather than requiring all resources simultaneously, thus managing processing time while achieving high modeling precision.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Methods, systems, and devices, including computer programs encoded on a computer storage medium are provided for combined dynamic causal modeling and biophysics modeling of brain function. In particular, the disclosed methods of modeling brain function can be used to integrate brain function measurements by two or more methods, such as functional neuroimaging and electrophysiology. Sequential model fitting is used to improve modeling accuracy to generate a more comprehensive model of brain neuronal circuitry.