African-Neurodata-Research-Lab-ANR-LAB/ANR-EEG-Analysis-Pipeline

Reproducible EEG preprocessing, analysis, and reporting workflows developed by African NeuroData Research Lab (ANR). # ANR EEG Analysis Pipeline **African NeuroData Research Lab (ANR)** A reproducible, MNE-Python-powered EEG research workflow developed by the African NeuroData Research Lab. The pipeline is designed around a simple idea: **record EEG → understand the data → check acquisition quality → preprocess → analyze frequency content → export reproducible results** The primary input format is the CSV produced by the ANR Muse EEG Recorder, while common MNE-compatible formats are also supported. --- ## Start here ### I want to analyze EEG without installing anything Open the guided Google Colab notebook: **`notebooks/ANR_EEG_Colab.ipynb`** The notebook is deliberately **step by step**. Run one cell, inspect the result, read the explanation, and then continue. It guides you through: 1. installing the latest ANR pipeline from this GitHub repository; 2. uploading your EEG recording; 3. loading the recording into MNE; 4. inspecting channels and recording information; 5. viewing raw EEG; 6. running technical QC; 7. selecting preprocessing settings; 8. preprocessing the EEG; 9. viewing PSD; 10. calculating relative band power; 11. reviewing event markers; 12. exporting research outputs; 13. generating an HTML report; 14. downloading the complete result package. ### I want to understand the repository first Read the documentation in this order: 1. **Getting Started** 2. **Repository Navigation Guide** 3. **Outputs and Interpretation** --- ## ANR Muse EEG input The ANR Muse EEG Recorder CSV is the first-class input format. Expected EEG columns: ```text TP9_uV AF7_uV AF8_uV TP10_uV ``` The loader also uses: ```text timestamp_ms iso_time event_marker ``` when available. The pipeline converts Muse values from microvolts to volts for MNE and converts ANR event markers into MNE annotations. --- ## Current v1 workflow ```text ANR Muse EEG CSV ↓ Input validation ↓ MNE Raw object ↓ Technical QC ↓ Notch filtering ↓ 1–40 Hz band-pass ↓ Optional re-referencing ↓ Welc …