Optimal Spike-based Encoding (Analog to Spike - ASC) Methods for Wearable EEG

alternate text

Motivation:

Wearable devices for health monitoring are required to perform real-time, energy-efficient data processing and communication to carry out proper interventions. Neuromorphic front-ends offer a promising solution by encoding collected analog data into sparse, event-driven spike representations, enabling low-power always-on operation [1]. A critical application of such devices is epileptic seizure detection [2], where robust, real-time monitoring can transform traditional clinical practices and there is an urgent need for reliable wearable detectors. This project aims to leverage a novel adaptive Asynchronous Delta Modulator (adaADM) to encode EEG signals into spikes and classify seizure events with minimal latency [3]. The ultimate goal is to validate and optimize a neuromorphic biomedical signal processing pipeline that balances sparse encoding, high sensitivity, and low false detection rate.

The research will be conducted at the Institute of Neuroinformatics (INI, Y55, Irchel Campus ETH/UZH), in the Neuromorphic Cognitive Systems (NCS) group of Prof. Dr. Giacomo Indiveri. The work will be supervised by Olympia Gallou, Dr. Saptarshi Ghosh, and Dr. Elisa Donati.

 

Objectives:

  1. 1. Data Preprocessing

  2. 2. Spike Encoding: Utilize adaADM to encode preprocessed data into sparse spike trains, and tune hyperparameters to preserve maximal information.

  3. 3. Feature Extraction + Classification: Investigate one of the two classification problems: (i) seizure vs. non-seizure detection, and (ii) seizure type classification.

  4. 4. Comparative Analysis: Evaluate which sensor setups (e.g., full EEG cap vs. reduced) yield the best trade-off between performance and sparsity.

  5. 5. SNN Design: Develop a Spiking Neural Network (SNN) for real-time anomaly detection, validated through simulation and potentially deployed on neuromorphic hardware.

 

Learning Outcomes:

  • - Learn about EEG signal processing and neuromorphic interfaces

  • - Gain experience in applying Neuromorphic Front-ends for real-time anomaly detection paradigms

  • - Explore challenges in deploying neuromorphic systems for wearable applications

 

Type of Work (%):

  • - 10% Literature Study: Review relevant research on wearable EEG, seizure detection, neuromorphic processing, and sensor fusion.

  • - 40% Encoding: Contribute to the development of the adaptive encoding toolbox - for different physiological signals.

  • - 40% Design and Implementation of Classification Algorithm: Explore ML classifier designs or optimize SNNs for binary classification (ictal - interictal states).

  • - 10% Validation: Perform a comparative analysis of different electrode configurations and evaluate system performance through simulations and/or hardware tests.

 

Prerequisites:

  • - Interest in signal processing, computational neuroscience, and neuroinformatics.

  • - Basic programming skills (Python strongly preferred).

  • - Familiarity with neural dynamics, dynamical systems, and machine learning concepts.

  • - Motivation to work with real-world noisy data and test neuromorphic systems in practical conditions.

This project can be tailored to both semester and master’s thesis projects, offering a valuable opportunity to contribute to cutting-edge neuromorphic solutions for wearable health and BCI applications.

 

References

[1] M. Sharifshazileh, K. Burelo, J. Sarnthein, and G. Indiveri, “An electronic neuromorphic system for real-time detection of high-frequency oscillations (HFOs) in intracranial EEG,” Nature Communications, vol. 12, no. 1, pp. 1–14, 2021.

[2] Gallou, Olympia, et al. "Online Epileptic Seizure Detection in Long-term iEEG Recordings Using Mixed-signal Neuromorphic Circuits." IEEE Biomedical Circuits and Systems (BIOCAS), (BioCAS), 2024.

[3] M. Sharifshazileh and G. Indiveri, “An adaptive event-based data converter for always-on biomedical applications at the edge,” International Symposium on Circuits and Systems, (ISCAS), 2023.

 

Contact

Olympia Gallou olympia (at) ini.uzh.ch
Dr. Saptarshi Ghosh, sapta (at) ini.uzh.ch

Dr. Elisa Donati, elisa@ini.uzh.ch

back
back