sLORETA Source Weighting for EEG Point Source Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for analyzing electrophysiological signals, such as EEG and MEG, face challenges in uniquely determining the distribution of current sources due to limited sensors, unknown noise, and more unknown values than known values, leading to ill-posed inverse problems, which hinder the exact localization of point sources.
Innovation Solution
A method involving the use of sLORETA transformation and Source Weighting techniques to compute a current density vector field, where the sLORETA technique localizes point sources correctly and the Source Weighting method utilizes a diagonal weighting matrix to enhance the representation of physiological activity, implemented in computer software for data transformation and visualization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current methods (minimum-norm least-squares) are used to compute current sources from electrophysiological measurements, then a unique solution can be obtained through regularization, but the localization of point sources is inaccurate due to the ill-posed inverse problem
Solution Approach 1:
The patent transforms the inverse problem by changing the parameter being optimized. Instead of directly minimizing the L2-norm of current sources (minimum-norm least-squares), the invention uses a different objective function that incorporates both the data misfit and a constraint on the current source distribution. This parameter transformation allows accurate point source localization while maintaining computational feasibility through iterative optimization rather than direct analytical solutions.
2Device complexity
If the number of sensors is limited, then the measurement system is simpler and more practical, but the current source distribution cannot be computed uniquely due to more unknown values than known values
Solution Approach 1:
The patent applies local quality by making different parts of the solution space have different properties. The weighting matrix W is constructed to give different weights to different brain regions based on prior knowledge and data characteristics. This allows the system to obtain reliable current source estimates in specific regions of interest even with limited sensors, by locally adapting the solution strategy to the characteristics of each brain region.
Solution Approach 2:
The invention performs preliminary action by pre-computing the lead field matrix A and the weighting matrix W before solving the inverse problem. The weighting matrix is constructed using prior information about source covariances and noise characteristics. This preliminary preparation transforms the ill-posed problem into a better-conditioned problem that can be solved reliably with limited sensors.
3Ease of operation
If measurement noise is present and unknown, then the measurement system remains practical for real-world applications, but the current source computation becomes ill-posed and cannot be solved uniquely
Solution Approach 1:
The patent implements feedback by using the measured data characteristics to adaptively construct the weighting matrix W. The noise covariance matrix is estimated from the data itself, and this information feeds back into the inverse solution process through the weighting matrix. This feedback mechanism allows the system to automatically adjust to the actual noise conditions present in the measurements, improving accuracy without requiring precise prior knowledge of noise levels.
Data Source
Figure 1
Figure 2(a)~2(b)
Figure 3(a)~3(d)
AI summary
The invention provides a method and apparatus for acquisition and analysis of data that displays a linear relationship or can be transformed into a linearized relationship, such as electrophysiological signal data from sensors such as those suitable for EEG, MEG, ECG and the like. The method, which can be implemented in computer software, includes computing a current density vector field by solving the related unweighted linear inverse problem, pre-processing the current density vector field using an sLORETA transformation, computing a diagonal weighting matrix so that its entries are determined by a monotonically increasing function of their corresponding values in the sLORETA method outputs, and computing the current density vector field by solving the related weighted linear inverse problem. The outputs of the method can be stored in computer files for display on suitable monitors.