mne_sensitivity_map
Overview
mne_sensitivity_map computes a sensitivity map from a forward solution. The map shows how sensitive each MEG/EEG sensor configuration is to sources at each location on the cortical surface, and is useful for evaluating source space coverage.
This is a C++ port of the original MNE-C tool by Matti Hämäläinen.
Usage
mne_sensitivity_map [options]
Options
| Option | Description |
|---|---|
--fwd <file> | Forward solution FIFF file |
--out <file> | Output sensitivity map (text format) |
--method <name> | Method: norm (column norms, default) or svd (leading singular value) |
--help | Print help |
--version | Print version |
Description
Sensitivity maps quantify the strength of the forward solution at each source location. Two methods are available:
- norm: Computes the Euclidean norm of the forward solution columns for each source. This gives a measure of signal strength.
- svd: Computes the leading singular value of the forward matrix at each source, providing a measure of the best-detectable component.
Workflow Context
Sensitivity maps are useful for assessing the quality of a forward model before performing source localization. Regions with low sensitivity may produce unreliable source estimates. The map can be used to:
- Verify that the sensor array provides adequate coverage of the cortex.
- Identify regions where source localization results should be interpreted with caution.
- Compare different sensor configurations.
Example
# Compute sensitivity map using column norms
mne_sensitivity_map --fwd sam-meg-fwd.fif --out sensitivity.txt
# Use SVD method
mne_sensitivity_map --fwd sam-meg-fwd.fif --out sensitivity_svd.txt --method svd