Real-time sub-milliwatt epilepsy detection implemented on a spiking neural network edge inference processor

• Published in: Computers in biology and medicine Vol. 183; p. 109225
• Main Authors: Li, Ruixin, Zhao, Guoxu, Muir, Dylan Richard, Ling, Yuya, Burelo, Karla, Khoe, Mina, Wang, Dong, Xing, Yannan, Qiao, Ning
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.12.2024; Elsevier Limited
• Subjects: Convulsions & seizures; Datasets; Diagnosis; EEG; Electroencephalogram; Electroencephalography - methods; Energy requirements; Epilepsy; Epilepsy - diagnosis; Epilepsy - physiopathology; Firing pattern; Humans; Internal Medicine; Latency; Microprocessors; Neural networks; Neural Networks, Computer; Neuromorphic processor; Other; Portable equipment; Power consumption; Power management; Real time; Seizure detection; Seizures; Signal processing; Signal Processing, Computer-Assisted; Spiking; Spiking neural network; Synapses; Ultra-low power; Wearable technology; Electroencephalogram; Seizure detection; Neuromorphic processor; Spiking neural network; Ultra-low power
• ISSN: 0010-4825; 1879-0534; 1879-0534
• DOI: 10.1016/j.compbiomed.2024.109225

Analyzing electroencephalogram (EEG) signals to detect the epileptic seizure status of a subject presents a challenge to existing technologies aimed at providing timely and efficient diagnosis. In this study, we aimed to detect interictal and ictal periods of epileptic seizures using a spiking neural network (SNN). Our proposed approach provides an online and real-time preliminary diagnosis of epileptic seizures and helps to detect possible pathological conditions. To validate our approach, we conducted experiments using multiple datasets. We utilized a trained SNN to identify the presence of epileptic seizures and compared our results with those of related studies. The SNN model was deployed on Xylo, a digital SNN neuromorphic processor designed to process temporal signals. Xylo efficiently simulates spiking leaky integrate-and-fire neurons with exponential input synapses. Xylo has much lower energy requirements than traditional approaches to signal processing, making it an ideal platform for developing low-power seizure detection systems. Our proposed method has a high test accuracy of 93.3% and 92.9% when classifying ictal and interictal periods. At the same time, the application has an average power consumption of 87.4 μW (IO power) + 287.9 μW (compute power) when deployed to Xylo. Our method demonstrates excellent low-latency performance when tested on multiple datasets. Our work provides a new solution for seizure detection, and it is expected to be widely used in portable and wearable devices in the future. •Novel Approach for Epilepsy Detection: New epilepsy detection method uses SNNs to analyze EEG signals accurately.•Real-Time Streaming Signal Analysis with SNNs: Real-time SNN analysis enhances efficiency, lowering latency for clinical use.•Network Deployment on Neuromorphic Chip Xylo: Neuromorphic chip Xylo deployment saves energy compared to traditional chips.