ACDC 2017 dataset

整个ACDC数据集是根据在第戎大学医院获得的真实临床检查创建的。获取的数据完全匿名，并按照法国第戎医院当地伦理委员会制定的规定进行处理。我们的数据集涵盖了几个明确定义的病理，有足够的病例 (1) 正确训练机器学习方法和 (2) 清楚地评估从电影MRI获得的主要生理参数的变化 (特别是舒张期容积和射血分数)。数据集由如下所述的150检查 (全部来自不同的患者) 组成，该检查分为5个均匀分布的亚组 (4个病理加上1个健康受试者组)。此外，每位患者都附带以下附加信息: 体重，身高以及舒张期和收缩期时刻。

The entire ACDC dataset was developed from real clinical examinations acquired at the University Hospital of Dijon. All collected data is fully anonymized and processed in compliance with the regulations formulated by the local ethics committee of the Dijon University Hospital, France. This dataset covers several well-defined clinical pathologies, and includes sufficient patient cases to (1) properly train machine learning models, and (2) rigorously evaluate changes in major physiological parameters derived from cine MRI, particularly end-diastolic volume and ejection fraction. The dataset comprises 150 clinical examinations, all from unique patients, which are divided into 5 evenly balanced subgroups: 4 pathological cohorts plus 1 healthy control group. Furthermore, each patient is paired with the following supplementary information: body weight, height, as well as end-diastolic and end-systolic time points.

搜集汇总

背景与挑战

背景概述

ACDC 2017 dataset是一个包含150例患者MRI检查数据的医学影像数据集，涵盖4种病理和1组健康受试者，附带体重、身高等临床信息，适用于机器学习训练和生理参数评估。