A Two-Layer LSTM Deep Learning Model for Epileptic Seizure Prediction

Shiva Maleki Varnosfaderani, Rihat Rahman, Nabil J. Sarhan, Levin Kuhlmann, Eishi Asano, Aimee Luat, Mohammad Alhawari

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

5 Scopus citations

Abstract

We propose an efficient seizure prediction model based on a two-layer LSTM using the Swish activation function. The proposed structure performs feature extraction based on the time and frequency domains and uses the minimum distance algorithm as a post-processing step. The proposed model is evaluated on the Melbourne dataset and achieves the highest Area Under Curve (AUC) score of 0.92 and the lowest False Positive Rate (FPR) of 0.147 compared to previous work while having sensitivity and accuracy of 86.8 and 85.1, respectively. The proposed system has a low number of trainable parameters, and thus reducing the complexity of resource-constrained applications.

Original languageEnglish
Title of host publication2021 IEEE 3rd International Conference on Artificial Intelligence Circuits and Systems, AICAS 2021
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781665419130
DOIs
StatePublished - Jun 6 2021
Externally publishedYes
Event3rd IEEE International Conference on Artificial Intelligence Circuits and Systems, AICAS 2021 - Washington, United States
Duration: Jun 6 2021Jun 9 2021

Publication series

Name2021 IEEE 3rd International Conference on Artificial Intelligence Circuits and Systems, AICAS 2021

Conference

Conference3rd IEEE International Conference on Artificial Intelligence Circuits and Systems, AICAS 2021
Country/TerritoryUnited States
CityWashington
Period06/6/2106/9/21

Keywords

  • Deep learning
  • LSTM
  • Melbourne dataset
  • classification
  • epilepsy seizure prediction
  • feature extraction
  • iEEG

Fingerprint

Dive into the research topics of 'A Two-Layer LSTM Deep Learning Model for Epileptic Seizure Prediction'. Together they form a unique fingerprint.

Cite this