Computationally efficient neural network classifiers for next generation closed loop neuromodulation therapy -- a case study in epilepsy

TL;DR

This paper presents a compact neural network classifier designed for real-time seizure detection in epilepsy, offering comparable accuracy to traditional methods but with reduced latency and computational complexity suitable for implantable devices.

Contribution

Introduces a minimalistic neural network architecture for seizure classification that balances accuracy and computational efficiency for neuromodulation systems.

Findings

Neural network classifier achieves similar accuracy to filter-based methods.

Reduces detection latency compared to classical classifiers.

Maintains low complexity suitable for implantable hardware.

Abstract

This work explores the potential utility of neural network classifiers for real-time classification of field-potential based biomarkers in next-generation responsive neuromodulation systems. Compared to classical filter-based classifiers, neural networks offer an ease of patient-specific parameter tuning, promising to reduce the burden of programming on clinicians. The paper explores a compact, feed-forward neural network architecture of only dozens of units for seizure-state classification in refractory epilepsy. The proposed classifier offers comparable accuracy to filter classifiers on clinician-labelled data, while reducing detection latency. As a trade-off to classical methods, the paper focuses on keeping the complexity of the architecture minimal, to accommodate the on-board computational constraints of implantable pulse generator systems.