Epileptic Pattern Recognition and Discovery of the Local Field Potential in Amygdala Kindling Process

Yu Lin Wang, Yin Lin Chen, Wen-Yu Su, Fu-Zen Shaw, Sheng-Fu Liang

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Epileptogenesis, which occurs in an epileptic brain, is an important focus for epilepsy. The spectral analysis has been popularly applied to study the electrophysiological activities. However, the resolution is dominated by the window function of the algorithm used and the sample size. In this report, a temporal waveform analysis method is proposed to investigate the relationship of electrophysiological discharges and motor outcomes with a kindling process. Wistar rats were subjected to electrical amygdala kindling to induce temporal lobe epilepsy. During the kindling process, different morphologies of afterdischarges (ADs) were found and a recognition method, using template matching techniques combined with morphological comparators, was developed to automatically detect the epileptic patterns. The recognition results were compared to manually labeled results, and 79%-91% sensitivity was found. In addition, the initial ADs (the first 10 s) of different seizure stages were specifically utilized for recognition, and an average of 85% sensitivity was achieved. Our study provides an alternative viewpoint away from frequency analysis and time-frequency analysis to investigate epileptogenesis in an epileptic brain. The recognition method can be utilized as a preliminary inspection tool to identify remarkable changes in a patient's electrophysiological activities for clinical use. Moreover, we demonstrate the feasibility of predicting behavioral seizure stages from the early epileptiform discharges.

Original languageEnglish
Article number7378295
Pages (from-to)374-385
Number of pages12
JournalIEEE Transactions on Neural Systems and Rehabilitation Engineering
Volume24
Issue number3
DOIs
Publication statusPublished - 2016 Mar 1

    Fingerprint

All Science Journal Classification (ASJC) codes

  • Neuroscience(all)
  • Biomedical Engineering
  • Computer Science Applications

Cite this