Bayesian¶
Bayesian methods use probabilistic frameworks to estimate source activity, incorporating prior knowledge about source distributions. They can provide uncertainty estimates and often yield sparser solutions than minimum norm methods.
This category contains 14 solvers.
Solvers¶
| Full Name | Solver ID | Description |
|---|---|---|
| Bayesian Compressive Sensing | BCS |
Sparse Bayesian inverse method based on Bayesian compressive sensing, using hierarchical priors to promote sparse sou... |
| Champagne (Sparse Bayesian Learning) | Champagne |
Sparse Bayesian learning method for M/EEG/EEG source imaging using Type-II maximum likelihood / evidence maximization... |
| Noise-Learning Champagne | Champagne-NL |
Champagne sparse Bayesian learning variant that jointly estimates per-channel noise parameters/covariance while updat... |
| Flex-Champagne | FLEX-CHAMPAGNE |
Flexible-extent Champagne variant using a multi-order diffusion-smoothed leadfield dictionary to jointly select sourc... |
| Flex-NL-Champagne | FLEX-NL-CHAMPAGNE |
Two-pass refined flexible-extent Champagne variant using a multi-order diffusion dictionary with Convexity/MM updates... |
| Gamma Maximum A Posteriori | Gamma-MAP |
Sparse Bayesian inverse solution using a Gamma hyperprior (ARD-style) within the unified Bayesian source imaging fram... |
| Gamma MAP with MSP Priors | Gamma-MAP-MSP |
Gamma-MAP sparse Bayesian inverse approach combined with MSP-style spatial smoothness/patch priors. |
| Multiple Sparse Priors | MSP |
Bayesian source imaging method that combines multiple spatial priors and estimates their hyperparameters with Restric... |
| OmniChampagne | OMNI-CHAMPAGNE |
Adaptive sparse Bayesian solver that selects between a dipole-only Champagne-style model and a multi-order patch-dict... |
| SubspaceSBL (SSM + NL-Champagne) | SUBSPACE-SBL |
Two-stage solver that detects sources with signal subspace matching and refines amplitudes/noise parameters using NL-... |
| Source Maximum A Posteriori | Source-MAP |
Sparse Bayesian inverse solution computing a source MAP estimate in the unified Bayesian source imaging framework. |
| Source MAP with MSP Priors | Source-MAP-MSP |
Source-MAP sparse Bayesian inverse approach augmented with MSP-style spatial priors/patch smoothing (conceptually rel... |
| Variational Bayes Sparse Bayesian Learning | VB-SBL |
Variational Bayesian ARD/SBL solver with Gamma hyperpriors, implemented with efficient sensor-space updates. |
| Coherent Maximum Entropy on the Mean | cMEM |
Maximum-entropy-on-the-mean approach using graphical models and parcel-wise optimization (data-driven parcellation) t... |