Part I of the Epilepsy EEG Classification Project
Classification of CHB-MIT epilepsy data set based on CNN
The data uses Boston EEG data
Dataset introduction
The CHB-MIT dataset is the EEG signals from Boston Children's Hospital included in the MIT EEG database. This dataset consists of scalp EEG recordings from children with treatment-resistant epilepsy. A total of 23 records from 22 subjects were included (chb21 was recorded again after 1.5 years from the same female subject as chb01, and chb24 was added to the database in December 2010).
All EEGs are collected using 10-20 international standard electrode positions, EEGs are recorded using 18/23 leads, and the sampling frequency and resolution are 256Hz and 16bit, respectively
The entire data set includes a total of 967.85 hours of continuous scalp EEG recordings, including 178 epileptic seizure records, each chb file contains 9-42 records, and the duration of a single record is 1 hour
The CHB-MIT dataset is available through PhysioNet,
https://archive.physionet.org/physiobank/database/chbmit/
The CHB-MIT file directory information is shown in the figure.