Spectral Density Analysis With Logarithmic Regression Dependent Gaussian Mixture Model For Epilepsy Classification
COMPUTATIONAL VISION AND BIO-INSPIRED COMPUTING(2020)
摘要
One of the serious disorders causing seizures in the neurology is the Epilepsy. The formation of the attacks happens due to the unusual activities of the neurons. The EEG is utilized in the effective observation of the brain abnormalities. The EEG can effectively analyze the different sorts of the status of the physiological in the brain and can provide valuable data about any neurological disorder. Therefore, EEG is quite a powerful diagnostic tool for analyzing many neurological disorders like epilepsy, dementia, paralysis, sleep disorders etc. As the EEG recordings for epileptic patients are quite long, the most important features based on the Power Spectral Density (PSD) are extracted using the Logarithmic Regression dependent Gaussian Mixture Model to know the risk of epilepsy. Results showed that when PSD is classified with Logarithmic Regression Gaussian Mixture Model, an appropriate accuracy of 95.835% in the classification along with an appropriate Performance Index of 91.58% is acquired.
更多查看译文
关键词
Epilepsy, EEG, PSD, Gaussian Mixture
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络