2020/07/22 by Ramy Hussein, Soojin Lee, Hussein, Ramy +5
Computer Science · Medicine · Neuroscience · #Blind Source Separation Techniques #EEG and Brain-Computer Interfaces #Epilepsy research and treatment #FOS: Electrical engineering #Image and Video Processing (eess.IV) #Signal Processing (eess.SP) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2007.11716
openalex publication_date 2020/07/22 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
Accurate prediction of epileptic seizures has remained elusive, despite the\nmany advances in machine learning and time-series classification. In this work,\nwe develop a convolutional network module that exploits Electroencephalogram\n(EEG) scalograms to distinguish between the pre-seizure and normal brain\nactivities. Since these scalograms have rectangular image shapes with many more\ntemporal bins than spectral bins, the presented module uses "semi-dilated\nconvolutions" to create a proportional non-square receptive field. The proposed\nsemi-dilated convolutions support exponential expansion of the receptive field\nover the long dimension (image width, i.e. time) while maintaining high\nresolution over the short dimension (image height, i.e., frequency). The\nproposed architecture comprises a set of co-operative semi-dilated\nconvolutional blocks, each block has a stack of parallel semi-dilated\nconvolutional modules with different dilation rates. Results show that our\nproposed solution outperforms the state-of-the-art methods, achieving seizure\nprediction sensitivity scores of 88.45% and 89.52% for the American Epilepsy\nSociety and Melbourne University EEG datasets, respectively.\n