Synthetic EEG Signal Generator
Synthetic EEG Signal Generator is a Python-based tool designed to simulate electroencephalographic (EEG) signals, focusing on epileptogenic morphologies such as spike-element graphs, slow-wave, and slow-wave-peak patterns.
The project enables:
- Controlled signal generation simulating frequency bands, background activity, and epileptiform events.
- Noise injection and reproducibility features to support large-scale datasets.
- Visualizer for analyzing both generated and user-uploaded EEG signals (TXT, CSV, EDF).
- Evaluation with EEGLAB confirming accuracy and consistency of simulated signals.
Impact
Supports neurological research, diagnosis, and AI/ML models requiring robust EEG datasets — particularly in the study of epilepsy.
🔑 Key Technologies
- Programming & Analysis: Python, MATLAB
- Signal Processing: Synthetic EEG generation, noise modeling
- Evaluation: EEGLAB analysis for signal validation
- Formats: TXT, CSV, EDF
📂 Publications & Conferences
2024 — IEEE CCE Conference
- CCE 2024 — 21st International Conference on Electrical Engineering, Computing Science and Automatic Control
Mexico City, Oct 23–25, 2024- IEEE Xplore Publication
- DOI:
10.1109/CCE62852.2024.10770922 - Contributors: José-Emmanuel Vázquez-Galán (Main Author), Blanca Tovar-Corona, Laura-Ivoone Garay-Jiménez, Martín-Arturo Silva-Ramírez
2024 — Research in Computing Science (RCS) Journal
- Open-access publication by CIC-IPN (ISSN 1870-4069)
- RCS PDF
- RCS Journal Page
- Contributors:
- José-Emmanuel Vázquez-Galán — Investigation, Software, Writing & Editing
- Blanca Tovar-Corona — Supervision, Investigation
- Laura-Ivoone Garay-Jiménez — Supervision, Investigation
🌍 Future Directions
Pending web hosting, the generator will expand accessibility to researchers and clinicians, supporting:
- Education and training tools for neuroscience students.
- Benchmark datasets for AI/ML models in biomedical signal processing.
- Collaborative research into synthetic approaches for neurological disease modeling.
Notes & Learnings (click to expand)
- Synthetic signals provide scalable datasets without patient privacy concerns.
- Reproducibility and noise injection are crucial for training ML models.
- Open-access publication ensures broader visibility and collaborative adoption.