Low Latency Real-Time Seizure Detection Using Transfer Deep Learning

Abstract

Scalp electroencephalogram (EEG) signals inherently have a low signal-to-noise ratio due to the way the signal is electrically transduced. Temporal and spatial information must be exploited to achieve accurate detection of seizure events. Most popular approaches to seizure detection using deep learning do not jointly model this information or require multiple passes over the signal, which makes the systems inherently non-causal. In this paper, we exploit both simultaneously by converting the multichannel signal to a grayscale image and using transfer learning to achieve high performance. The proposed system is trained end-to-end with only very simple pre-and postprocessing operations which are computationally lightweight and have low latency, making them conducive to clinical applications that require real-time processing. We have achieved a performance of 42.05% sensitivity with 5.78 false alarm per 24 hours on the development dataset of v1.5.2 of the Temple University Hospital Seizure Detection Corpus. On a single core CPU operating at 1.7 GHz, the system runs faster than real-time (0.58 xRT), uses 16 Gbytes of memory, and has a latency of 300 msec.

Multi-center Assessment of CNN-Transformer with Belief Matching Loss for Patient-independent Seizure Detection in Scalp and Intracranial EEG

“…Moreover, when the 2D models were trained and evaluated on the CHB-MIT dataset, we obtained the worst results thus far, with PRE lower than 5% for all cases. These numerical results are in line with many 2D models reported in the literature 20,28,29,[57][58][59] .…”

Deep learning methods for diagnosis of epilepsy from electroencephalograms

“…Moreover, when the 2D models were trained and evaluated on the CHB-MIT dataset, we obtained the worst results thus far, with PRE lower than 5% for all cases. These numerical results are in line with many 2D models reported in the literature [293,298,299,355,369,370].…”

Self-Supervised Learning for Anomalous Channel Detection in EEG Graphs: Application to Seizure Analysis

“…We compare our proposed EEG-CGS with two streams of DL-based methods 1 : (1) DL models in the EEG time-series and/or spectrograms domain, including EEGNet (Lawhern et al 2018), EEG-TL (Khalkhali et al 2021), Dense-CNN, LSTM and CNN-LSTM (Tang et al 2021); and (2) DL models in the EEG graph domain (Tang et al 2021). Unlike our method, these DL models make use of the seizure data and their corresponding labels in the training phase.…”

“…Although many studies have proposed deep learning (DL) based models for automated seizure detection (Saab et al 2020;Shoeibi et al 2021;Abdelhameed and Bayoumi 2021;Thuwajit et al 2021;Khalkhali et al 2021;Rashed-Al-Mahfuz et al 2021;Mahajan, Somaraj, and Sameer 2021;Saichand et al 2021;Shen et al 2022;Gao et al 2022), several challenges still remain unsolved. First, these studies train their proposed model in a supervised approachi.e.…”