eeg-preprocess

Preprocess raw EEG recordings into ICA-ready data with filtering and artifact removal.

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Converts messy raw EEG recordings into a clean, standardized, preprocessed dataset by filtering, line-noise removal, bad-channel detection, interpolation, and re-referencing so downstream ERP/TFR/ICA steps become reliable and reproducible.

Core Features & Use Cases

  • Canonical EEG preprocessing pipeline: filter → bad-channel detection (RANSAC/PREP-style options) → interpolation → re-reference → optional resampling, with order preserved to reduce methodological bias.
  • Backend-validated execution: verifies the computation environment via ENVIRONMENT.json and uses MNE-Python for deterministic preprocessing.
  • Study-ready outputs: writes per-subject preprocessed FIF files plus structured summaries for auditing and methods reporting.
  • Use Case: Starting a new EEG study from raw recordings, especially when you need paper-grade preprocessing logs that later skills can verify.

Quick Start

Run eeg-preprocess in your project folder by pointing it at the study directory containing DATASET_BRIEF.md, ENVIRONMENT.json, and raw/ files.

Frequently Asked Questions about eeg-preprocess

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

FAQPage Schema
What is the canonical order for EEG preprocessing before running ICA?

The canonical EEG preprocessing order applies filtering, line-noise removal, bad-channel detection, interpolation, and re-referencing before optional resampling to produce ICA-ready data.

How do I batch preprocess raw EEG recordings in MNE-Python?

Batch EEG preprocessing in MNE-Python requires pointing the pipeline at a study directory containing raw files, DATASET_BRIEF.md, and ENVIRONMENT.json to validate the backend and process multiple subjects.

Does EEG preprocessing support .bdf, .edf, and .set raw data formats?

EEG preprocessing supports raw formats including .bdf, .edf, .set, .fif, and .vhdr, reading them from the raw/ directory to standardize diverse recordings into clean FIF outputs.

How do I detect and interpolate bad channels in raw EEG data?

Bad channel detection in raw EEG data uses RANSAC or PREP-style options to identify noisy channels, which are then interpolated and re-referenced during the preprocessing pipeline.

Do I need an environment configuration file to start EEG preprocessing?

Yes, an ENVIRONMENT.json file is mandatory to validate the computation environment and ensure MNE-Python executes deterministic, reproducible preprocessing steps across all subjects.

Why does my EEG preprocessing pipeline need a dataset brief file?

A DATASET_BRIEF.md file is required as a mandatory project input to define study parameters, ensuring the preprocessing pipeline generates accurate per-subject summaries for downstream auditing.