MP-SeizNet: A multi-path CNN Bi-LSTM Network for seizure-type classification using EEG

Open Webpage

Hezam Albaqami, Ghulam Mubashar Hassan, Amitava Datta

Published: 01 Jan 2023, Last Modified: 14 Nov 2024Biomed. Signal Process. Control. 2023EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•A novel DL network for seizure-type classification, consisting of CNN and Bi-LSTM•The proposed MP-SeizNet was trained on different representations of EEG data.•MP-SeizNet uses wavelet-based features and raw signals for seizure-type classification•The achieved F1-scores are 87.6% across patients and 98.1% across seizures, using TUSZ 1.5.2.