PhysioNet Apnea-ECG

该数据集包含了70名参与者的夜间心电信号记录，专家对这些记录进行了标注，区分了伴有呼吸暂停和没有呼吸暂停的心电信号段。记录被分为训练集和测试集，每分钟的心电段都被标注了呼吸暂停和正常状态。数据规模包括70名参与者，共70份记录，已发布的集合中有16,713个心电段，而未公开的集合中有16,946个心电段。该任务旨在通过心电分类来检测睡眠呼吸暂停。

This dataset contains overnight electrocardiogram (ECG) recordings from 70 participants, which were manually annotated by experts to differentiate between ECG segments with apnea and those without apnea. The recordings are divided into training and test sets, with each one-minute ECG segment labeled as either apneic or normal. In terms of data scale, there are 70 participants totaling 70 recordings: the publicly released subset contains 16,713 ECG segments, while the non-public subset has 16,946 ECG segments. The goal of this task is to detect sleep apnea via ECG classification.

搜集汇总

背景与挑战

背景概述

PhysioNet Apnea-ECG数据库包含70条记录，每条记录包含ECG信号、呼吸暂停注释和QRS注释，用于睡眠呼吸暂停检测研究。数据集分为学习集和测试集，部分记录还包含额外的呼吸相关信号，由德国Phillips-University的Thomas Penzel博士贡献。