EEG-Based Epileptic Seizure Detection Dataset (CHSZ)

The CHSZ dataset contains electroencephalography (EEG) recordings collected from 27 pediatric patients, aged from 3 months to 10 years, at Wuhan Children’s Hospital, Tongji Medical College, Huazhong University of Science and Technology, China. The original sampling rate was either 500 Hz or 1000 Hz. Each patient experienced between one and six seizure events. The onset and offset of each seizure were annotated by clinical experts. According to these annotations, each segmented EEG trial was assigned a binary label for seizure detection, where 0 denotes non-seizure and 1 denotes seizure.

The original EEG recordings were acquired using 19 unipolar electrodes placed according to the international 10–20 system. Based on these recordings, 18 bipolar channels were derived: Fp2-F4, F4-C4, C4-P4, P4-O2, Fp1-F3, F3-C3, C3-P3, P3-O1, Fp2-F8, F8-T4, T4-T6, T6-O2, Fp1-F7, F7-T3, T3-T5, T5-O1, Fz-Cz, and Cz-Pz. Each bipolar EEG channel was preprocessed using a 50 Hz notch filter and a 0.5–50 Hz bandpass filter. The continuous EEG recordings were then segmented into non-overlapping 4-second trials.

The processed data are organized in the form of (N, T, C), where N is the number of trials for a subject, T is the number of time samples in each trial, and C is the number of channels. Since the original sampling rate was either 500 Hz or 1000 Hz, T is 2000 or 4000 accordingly. The number of channels, C, is 18, corresponding to the derived bipolar montage.

For users who require a unified sampling rate, the EEG trials can be downsampled to 500 Hz using the resample function provided in the MNE package, following the preprocessing strategy adopted in our previous studies [1], [2]. 

If you use this dataset in your research, please cite [1], which is most directly related to this dataset. [2] also presents some interesting usage of this dataset. 

References

[1] Z. Wang, W. Zhang, S. Li, X. Chen, and D. Wu, “Unsupervised domain adaptation for cross-patient seizure classification,” Journal of Neural Engineering, vol. 20, no. 6, p. 066002, 2023.

[2] Z. Wang, S. Li, and D. Wu, “Canine EEG helps human: Cross-species and cross-modality epileptic seizure detection via multi-space alignment,” National Science Review, vol. 12, no. 6, p. nwaf086, 2025.