A large scale 12-lead electrocardiogram database for arrhythmia study

这个新成立的12导联心电图 (ECG) 信号研究数据库是在查普曼大学，绍兴市人民医院 (浙江大学医学院绍兴医院) 和宁波市第一医院的主持下创建的。它旨在使科学界能够对心律不齐和其他心血管疾病进行新的研究。某些类型的心律失常，例如房颤，对公共卫生，生活质量和医疗支出有明显的负面影响。作为一种非侵入性测试，ECG是检测这些情况的主要且至关重要的诊断工具。但是，这种做法会产生大量数据，对这些数据进行分析需要人类专家花费大量时间和精力。现代机器学习和统计工具可以在高质量的大数据上进行培训，以实现自动化诊断准确性的卓越水平。因此，我们收集并传播了这个新颖的数据库，该数据库包含45,152例患者的12导联心电图，其采样率为500  Hz，具有多种常见节律和其他心血管疾病的特征，所有这些都由专业专家标记。该数据集可用于设计，比较和微调针对心律失常和其他心血管疾病的研究中的新的和经典的统计和机器学习技术。

This newly established 12-lead electrocardiogram (ECG) signal research database was developed under the auspices of Chapman University, Shaoxing People's Hospital (Shaoxing Hospital, Zhejiang University School of Medicine) and Ningbo First Hospital. It is intended to enable the scientific community to conduct novel research on arrhythmias and other cardiovascular diseases. Certain types of arrhythmias, such as atrial fibrillation, exert significant negative impacts on public health, quality of life and healthcare expenditures. As a non-invasive test, ECG serves as the primary and critically important diagnostic tool for detecting these conditions. However, such examinations generate large volumes of data, and analyzing such data demands considerable time and effort from human experts. Modern machine learning and statistical tools, trained on high-quality large-scale datasets, can achieve exceptional levels of automated diagnostic accuracy. Accordingly, we have collected and disseminated this novel database, which contains 12-lead ECG recordings from 45,152 patients at a sampling rate of 500 Hz, featuring signatures of multiple common cardiac rhythms and other cardiovascular diseases, all annotated by professional specialists. This dataset can be used to design, compare and fine-tune both novel and classic statistical and machine learning techniques in research focused on arrhythmias and other cardiovascular diseases.

搜集汇总

背景与挑战

背景概述

这是一个大规模12导联心电图数据库，专为心律失常和心血管疾病研究设计。它包含45,152例患者的ECG信号，采样率为500 Hz，由专业专家标记，覆盖多种常见节律和疾病。该数据集旨在支持机器学习和统计方法的开发，以提升自动化诊断的准确性，发布于2022年，由多家医疗机构合作创建。