ECG in High Intensity Exercise Dataset

此处提供的数据是从瑞士洛桑联邦理工学院 (EPFL) 嵌入式系统实验室 (ESL) 与洛桑大学运动科学研究所 (ISSUL) 合作收集的更大数据集中提取的.在这个数据集中，我们报告了在高强度运动期间用于分析 R 峰值检测算法的提取片段。实验方案 实验方案如下。 22 名受试者使用防毒面具在自行车测力计上进行心肺最大运动测试。使用 BIOPAC 系统测量单导联心电图 (ECG)。记录最初的 3 分钟休息时间。在此基线之后，受试者开始根据他们的健康水平以 60W 或 90W 的功率骑自行车。然后，循环测力计的功率每 3 分钟增加 30W，直到筋疲力尽（根据最大摄氧量或 VO2max）。最后，生理学专家根据肺部数据（氧气量和二氧化碳量）评估了所谓的通气阈值和最大摄氧量。提取数据集的描述 数据集的特征如下： 我们只报告了用于分析的 22 个受试者中的 20 个，因为对于两个受试者的信号过于损坏或不完整。具体来说，受试者 5 和 12 被丢弃。 ECG 信号以 500 Hz 采样，然后以 250 Hz 下采样。在最大 10 mV 处测量原始 ECG 信号。然后，将它们按比例缩小 1000 倍，因此数据以 uV 表示。对于每个受试者，从心电图记录中提取 5 个 20 秒的片段，并根据最大运动测试的不同阶段（即所谓的第二通气阈值或 VT2 之前和之后、VO2max 之前和中间、以及在筋疲力尽后的恢复期间）代表不同强度的体力活动。 seg1 --> [VT2-50,VT2-30] seg2 --> [VT2+60,VT2+80] seg3 --> [VO2max-50,VO2max-30] seg4 --> [VO2max-10,VO2max+ 10] seg5 --> [VO2max+60,VO2max+80] R 峰值位置在所有段中手动注释，并由洛桑大学医院 CHUV 的医生进行审查。由于输入信号存在问题，因此无法注释对象 9 的仅分段 5。因此，提取的段总数为 20 * 5 - 1 = 99。 提取数据集的格式 数据集分为两个主要文件夹： 文件夹 `ecg_segments/` 包含以两种格式保存的心电图信号，`.csv`和`.mat`。该文件夹包括原始（`ecg_raw`）和处理过的（`ecg`）信号。该处理由形态滤波和相对能量非滤波方法组成，以增强 R 峰值。 `.csv` 文件只包含信号，而 `.mat` 文件包括信号、最大压力测试中的时间向量、采样频率和信号幅度的单位（uV，正如我们之前提到的）。文件夹“manual_annotations/”包含带注释的 R 峰的样本索引，格式为“.csv”。注释是在处理后的信号上完成的。

The data provided herein is extracted from a larger dataset collected in collaboration between the Embedded Systems Laboratory (ESL) at École Polytechnique Fédérale de Lausanne (EPFL), Switzerland, and the Institute of Sports Sciences of the University of Lausanne (ISSUL). In this dataset, we report extracted segments used for analyzing R-peak detection algorithms during high-intensity exercise.

Experimental Protocol
The experimental protocol is as follows. Twenty-two subjects performed a cardiopulmonary maximal exercise test on a bicycle ergometer using a respiratory mask. A single-lead electrocardiogram (ECG) was measured using a BIOPAC system. The initial 3-minute rest period was recorded. Following this baseline, subjects began cycling at a power of 60W or 90W based on their fitness level. Subsequently, the power output of the bicycle ergometer was increased by 30W every 3 minutes until exhaustion (defined as peak oxygen uptake, or VO2max). Finally, physiological experts evaluated the so-called ventilatory threshold and VO2max based on pulmonary data (oxygen and carbon dioxide volumes).

Dataset Description
The characteristics of the dataset are as follows:
We only report data from 20 of the 22 subjects included in the analysis, as the signals from two subjects were excessively corrupted or incomplete. Specifically, Subjects 5 and 12 were excluded. ECG signals were sampled at 500 Hz and then downsampled to 250 Hz. Raw ECG signals were measured at a maximum of 10 mV, then scaled down by a factor of 1000, such that the data is expressed in microvolts (uV).
For each subject, five 20-second segments were extracted from the ECG recordings, representing physical activity of varying intensities corresponding to different stages of the maximal exercise test: namely, before and after the second ventilatory threshold (VT2), before and at VO2max, and during the recovery period immediately following exhaustion. The segments are defined as:
seg1 --> [VT2-50, VT2-30]
seg2 --> [VT2+60, VT2+80]
seg3 --> [VO2max-50, VO2max-30]
seg4 --> [VO2max-10, VO2max+10]
seg5 --> [VO2max+60, VO2max+80]
R-peak positions were manually annotated across all segments and reviewed by a physician from the Lausanne University Hospital (CHUV). Due to issues with the input signal, only Segment 5 for Subject 9 could not be annotated. Thus, the total number of extracted segments is 20 * 5 - 1 = 99.

Dataset Format
The dataset is divided into two main folders:
The folder `ecg_segments/` contains ECG signals saved in two formats: `.csv` and `.mat`. This folder includes both raw (`ecg_raw`) and processed (`ecg`) signals. The processing consists of morphological filtering and a relative energy unfiltering method to enhance R-peaks. The `.csv` files only contain the signal data, while the `.mat` files include the signal, the time vector from the maximal stress test, the sampling frequency, and the unit of signal amplitude (uV, as noted previously). The folder `manual_annotations/` contains the sample indices of the annotated R-peaks in `.csv` format. Annotations were performed on the processed signals.

搜集汇总

背景与挑战

背景概述

该数据集包含20名受试者在高强度自行车运动测试期间的99个ECG片段(每个20秒)，覆盖运动不同阶段，包含原始和处理后的信号以及手动标注的R峰位置，采样率250Hz，主要用于R峰检测算法研究。