Related Experiment Video
Updated: Jun 9, 2026
13:32
Recording Human Electrocorticographic (ECoG) Signals for Neuroscientific Research and Real-time Functional Cortical Mapping
Published on: June 26, 2012
Reproducible testing for embedded BCIs: a demultiplexing PCB and acquisition system for EEG signal emulation
Daniel Enériz1, Diego Antolín1, Nicolás Medrano1
1Group of Power Electronics and Microelectronics (GEPM), Aragon Institute for Engineering Research (I3A), University of Zaragoza, Zaragoza, Spain.
Hardwarex
|June 8, 2026
Summary
We developed DEEGMUX, an affordable hardware system for testing Brain-Computer Interface (BCI) machine learning models on edge devices. It ensures high-fidelity signal emulation, preserving classification accuracy for real-world BCI deployment.
Area of Science:
- Biomedical Engineering
- Machine Learning
- Neuroscience
Background:
- Validating Brain-Computer Interface (BCI) machine learning models on edge devices is difficult due to the cost of EEG equipment and limitations of simulations.
- Existing methods often fail to replicate real-world electronic signal characteristics crucial for accurate BCI performance.
Purpose of the Study:
- Introduce the DEEGMUX, a low-cost, open-source hardware system for high-fidelity, hardware-in-the-loop (HIL) testing of EEG classification algorithms.
- Enable rigorous validation of BCI models against realistic electronic inputs on resource-constrained edge devices.
Main Methods:
- The DEEGMUX system includes an EEG Demultiplexer Board and an EEG Acquisition and Processing Board with an ADS1299 ADC and Arduino Nano 33 BLE.
- Real EEG datasets (e.g., PhysioNet Motor Imagery) were used to generate precisely timed electronic signals for hardware-in-the-loop testing.
- Signal fidelity was characterized using Mean Squared Error and Signal-to-Noise Ratio.
Main Results:
- The DEEGMUX system demonstrated high signal fidelity with a Mean Squared Error of 1.7·10-10 V2 and a 16 dB Signal-to-Noise Ratio.
- An EEGNet motor imagery classifier showed a negligible accuracy difference (-0.3 ± 5%) when evaluated on hardware-emulated signals compared to original data.
- This confirms the preservation of classification-relevant features throughout the emulation process.
Conclusions:
- The DEEGMUX offers a scalable, reproducible, and affordable platform for testing edge-deployed BCI models.
- It facilitates the transition from simulation to robust real-world BCI applications by providing realistic electronic inputs.
- Accelerates the development and deployment of reliable Brain-Computer Interfaces on edge devices.
