Visual Inspection of scalp EEG by machine for seizure detection

カテゴリ: 部門大会

論文No: TC1-17

グループ名: 【C】平成30年電気学会電子・情報・システム部門大会プログラム

発行日: 2018/09/05

タイトル(英語): Visual Inspection of scalp EEG by machine for seizure detection

著者名: Emami Ali(The University of Tokyo),Kunii Naoto(The University of Tokyo),Matsuo Takeshi(Tokyo Metropolitan Neurological Hospital),Matsuo Takeshi(National Institute of Information and Communications Technology),Kawai Kensuke(Jichi Medical University),Takahashi Hirokazu(The University of Tokyo)

著者名(英語): Ali Emami|Naoto Kunii|Takeshi Matsuo|Takeshi Matsuo|Kensuke Kawai|Hirokazu Takahashi

キーワード: Convolutional Neural Networks|Seizure Detection|Deep learning|Scalp electroencephalogram|Epileptic seizure

要約(日本語): We explored an image-based seizure detection by applying convolutional neural networks to long-term EEG that included epileptic seizures. After filtering, EEG data was divided into short segments based on a given time window and converted into plot EEG images, each of which was classified by convolutional neural networks as ‘seizure’ or ‘non-seizure’. These resultant labels by convolutional neural networks were then used to design a clinically practical index for seizure detection. The median of detected seizure rate by minutes was 100% by convolutional neural networks, which was higher than 73.3% by BESA and 81.7% by Persyst. The false alarm of convolutional neural networks seizure detection was issued at 0.2 per hour, which appears acceptable for clinical practice.

PDFファイルサイズ: 936 Kバイト