Skip to content

Minimum Norm

Minimum norm methods estimate source activity by finding the solution with the smallest norm (typically L2) that explains the measured data. These methods are computationally efficient and provide smooth source estimates.

This category contains 15 solvers.

Solvers

Full Name Solver ID Description
MNE with Basis Functions BF-MNE Minimum-norm inverse using a reduced basis (e.g., geometric-informed basis functions) to parameterize the source space.
Backus–Gilbert BG Resolution-optimizing linear inverse method that trades off spatial resolution versus noise amplification using Backu...
EPIFOCUS EPIFOCUS Epileptic focus localization method that uses source-wise normalization/weighting to promote focal solutions.
Local Auto-Regressive Average LAURA Spatially weighted minimum-norm inverse using local neighborhood constraints (LAURA) to encourage physiologically pla...
Low Resolution Electromagnetic Tomography LORETA Smoothness-constrained minimum-norm inverse that penalizes spatial roughness via a Laplacian operator on the source s...
Minimum Current Estimate (L1) MCE Sparse (L1) distributed inverse, typically solved via iterative shrinkage/thresholding to promote focal source estima...
Minimum Norm Estimate MNE Classic L2 minimum-norm inverse solution. Estimates source currents by minimising the L2 norm of the source distribut...
Mixed-Norm Estimate (L1/L2) MxNE Mixed-norm (L1/L2) inverse that promotes group sparsity across time by applying an L1 penalty over sources and an L2 ...
S-MAP (Quadratic + Spatial MAP) S-MAP MAP-style quadratic inverse with spatial regularization via a graph Laplacian over the source space.
Standardized Shrinking LORETA-FOCUSS SSLOFO Hybrid method combining sLORETA and reweighted (FOCUSS-style) updates with iterative source-space shrinking to obtain...
Graph Total Variation (Huber) TV Iteratively reweighted graph-TV (edge-preserving) regularizer on the source-space mesh adjacency.
Dynamic Statistical Parametric Mapping dSPM Noise-normalized minimum-norm inverse (MNE) that yields statistical maps by scaling the MNE estimate with an estimate...
Exact Low Resolution Electromagnetic Tomography eLORETA LORETA-family inverse that iteratively estimates source weights to achieve exact (zero-error) localization under idea...
Standardized Low Resolution Electromagnetic Tomography sLORETA Standardized (variance-normalized) LORETA/MNE-type inverse designed to reduce localization bias by normalizing each s...
Weighted Minimum Norm Estimate wMNE Minimum-norm inverse with depth/weighting to reduce superficial bias by scaling the source prior or leadfield columns.