eeg-source

Estimate cortical and dipole EEG sources with MNE-Python inverse modeling.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/dengzhe-hou/auto-eeg-analysis --skill eeg-source
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: eeg-source
Source: https://github.com/dengzhe-hou/auto-eeg-analysis/tree/main/skills/eeg-source
Command: npx skills add https://github.com/dengzhe-hou/auto-eeg-analysis --skill eeg-source

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you estimate where in the brain scalp EEG signals originate, converting sensor-level activity into cortical/dipole source estimates for scientific reporting and analysis.

Core Features & Use Cases

  • Computes a forward model (BEM) and inverse solution using MNE-Python (dSPM, sLORETA, eLORETA, or MNE) or an LCMV beamformer option, producing time-resolved source estimates.
  • Supports template (fsaverage) and individual-head workflows with guardrails for FreeSurfer requirements and coregistration constraints.
  • Generates ROI time courses from standard cortical atlases (aparc, aparc.a2009s, HCPMMP1) for downstream stats/connectivity and writes parameter provenance for auditability.
  • Produces reusable artifacts and documentation, including platform checks, forward/inverse outputs, and parameter summaries you can cite in a methods write-up.

Quick Start

Ask the agent to run source localization on your epoched EEG data using the frozen analysis plan and fsaverage head model to generate dSPM source estimates and ROI time courses.

Frequently Asked Questions about eeg-source

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I estimate cortical EEG sources from epoched data using MNE-Python?

You can use LCMV beamforming or MNE-Python inverse solutions like dSPM and sLORETA to estimate cortical EEG sources from epoched FIF files, generating time-resolved intracranial activity estimates for evoked conditions.

How do I compute ROI time courses from EEG source localization results?

Compute ROI time courses from EEG source localization results by extracting source estimates mapped onto standard cortical atlases like aparc, aparc.a2009s, and HCPMMP1, producing time series data ready for downstream statistics and connectivity analysis.

Do I need FreeSurfer to build a BEM forward model for individual head models?

Yes, you need FreeSurfer to build a multi-layer BEM forward model for individual head models, as strict guardrails enforce correct FreeSurfer availability and coregistration constraints for accurate individualized EEG source localization.

Can I use fsaverage template head models instead of individual head models for EEG source localization?

Yes, fsaverage template head models are supported for EEG source localization alongside individual head models, allowing you to compute BEM forward models and inverse solutions without individual FreeSurfer reconstructions.

What files and environment setup do I need to run dSPM source localization on my EEG data?

To run dSPM source localization, you need a frozen ANALYSIS_PLAN.md, existing epochs-stage FIF files, and an ENVIRONMENT.json file, ensuring strict parameter provenance and correct BEM layering for the inverse modeling pipeline.

What is the difference between dSPM, sLORETA, eLORETA, and LCMV beamforming for EEG source localization?

dSPM, sLORETA, and eLORETA are MNE-Python inverse modeling approaches that estimate cortical sources from a forward model, whereas LCMV beamforming is a spatial filtering alternative; both produce time-resolved source estimates for evoked-condition averages.