Dataset of EEG recordings of pediatric patients with epilepsy based on the 10-20 system

# Dataset of EEG recordings containing HFO markings for 30 pediatric patients with epilepsy

## Summary
High-frequency oscillations in scalp EEG are promising non-invasive biomarkers of epileptogenicity. However, it is unclear how high-frequency oscillations are impacted by age in the pediatric population.
We recorded and processed the first 3 hours of sleep EEG data in 30 children and adolescents with focal or generalized epilepsy. We used an automated and clinically validated high-frequency oscillation detector to determine ripple rates (80-250 Hz) in bipolar channels. The software for the detection of HFOs is freely available at the GitHub repository (https://github.com/ZurichNCH/Automatic-High-Frequency-Oscillation-Detector). Furthermore HFO markings are also added in this database for the selected N3 intervals.

## Repository structure

### Main directory (hfo/)
Contains metadata files in the BIDS standard about the participants and the study. Folders are explained below.

### Subfolders
* hfo/sub-**/
Contains folders for each subject, named sub-<subject number> and session information.
* hfo/sub-**/ses-01/eeg
Contains the raw eeg data in .edf format for each subject. The duration is typically 3 hours, that was recorded in the beginning of the sleep. Details about the channels are given in the corresponding .tsv file.
* hfo/derivatives
Besides containingsubfolders for the raw data, there are two .json files. The events_description.json explains the meaning of the columns of the event description tsv files (in the subfolders).
The interval_description.json explains the meaning of the columns of the interval description tsv files (in the subfolders).

* hfo/derivatives/sub-**/ses-01/eeg/
Contains processed data for each subject. Based on the sleep annotations, first we identified the sleep stages. Then we cut 5 minutes data intervals from the N3 sleep stages. We applied bipolar referencing by considering all nearest neighbour chanels, thus resulting in 52 bipolar channels. Each run corresponds to one 5 minute data interval. The DataIntervals.tsv file provides information about how the various runs are related to the raw data by providing the start and end indeces. Besides the .edf and channel descriptor .tsv files there is an other .tsv file containing the detected candidate event details. Eg. sub-26_ses-01_task-hfo_run-01_events.tsv contains for subject 26 for the first processed data interval the event markings as indeces with additional features of this event described in the abovementioned events_description.json file.

## Related materials
The code for HFO detection is available at https://github.com/ZurichNCH/Automatic-High-Frequency-Oscillation-Detector

## Support
For questions on the dataset or the task, contact Johannes Sarnthein at [[e-mail]]([e-mail]).

# 收录30例癫痫儿科患者脑电记录及高频振荡（High-Frequency Oscillations, HFO）标记的数据集

## 数据集概况
头皮脑电图（scalp EEG）中的高频振荡（HFO）是颇具前景的致痫性无创生物标志物。然而目前尚不清楚儿科人群中的高频振荡如何受年龄影响。

本研究记录并处理了30例局灶性或全面性癫痫儿童及青少年的前3小时睡眠脑电数据。我们采用经临床验证的自动化高频振荡检测工具，对双极导联中的锐波（ripple，80~250Hz）发生率进行测算。该高频振荡检测软件可在GitHub仓库（https://github.com/ZurichNCH/Automatic-High-Frequency-Oscillation-Detector）免费获取。此外，本数据集还为选定的N3睡眠阶段添加了高频振荡标记。

## 仓库结构

### 主目录（hfo/）
包含遵循脑成像数据结构（Brain Imaging Data Structure, BIDS）标准的受试者及研究元数据文件，各子文件夹说明如下。

### 子文件夹
* hfo/sub-**/：为每位受试者创建的专属文件夹，命名格式为sub-<受试者编号>，并包含会话相关信息。
* hfo/sub-**/ses-01/eeg：存放每位受试者的.edf格式原始脑电数据，数据时长通常为3小时，采集于睡眠初始阶段。导联详情请参见对应.tsv文件。
* hfo/derivatives：除包含原始数据子文件夹外，还包含两个.json文件。其中events_description.json用于说明事件描述.tsv文件（位于各子文件夹中）的各列含义；interval_description.json用于说明区间描述.tsv文件（位于各子文件夹中）的各列含义。
* hfo/derivatives/sub-**/ses-01/eeg/：存放每位受试者的处理后数据。首先基于睡眠标注识别睡眠阶段，随后从N3睡眠阶段中截取5分钟的数据区间。我们采用双极参考法，以所有相邻导联作为参考，最终得到52个双极导联。每个run对应一段5分钟的数据区间。DataIntervals.tsv文件通过提供起始和结束索引，阐明了各run与原始数据的对应关系。除.edf文件和导联描述.tsv文件外，还包含一个.tsv文件，用于存储检测到的候选事件详情。例如，sub-26_ses-01_task-hfo_run-01_events.tsv文件包含了受试者26首个处理数据区间内的事件标记（以索引形式呈现），以及该事件的额外特征，相关说明详见前文提及的events_description.json文件。

## 相关资料
高频振荡检测代码可在https://github.com/ZurichNCH/Automatic-High-Frequency-Oscillation-Detector获取。

## 技术支持
若对本数据集或研究任务有疑问，请联系Johannes Sarnthein，邮箱为[e-mail]。

搜集汇总

背景与挑战

背景概述

该数据集包含30名癫痫儿科患者的EEG记录，基于10-20系统，重点研究高频振荡（HFO）作为癫痫活动的生物标志物。数据包括睡眠EEG的前3小时记录，并提供了HFO标记和自动检测工具，适用于研究儿科癫痫患者的年龄对HFO的影响。