DDXPlus
收藏资源简介:
该数据集名为DDXPlus,提供了差异诊断和病理严重度信息,特点在于其包含了分类和多项选择证据,以促进更高效的医患互动。该数据集涵盖了49种病理和223个证据点,其中包括110种症状和113个前置条件。此外,该数据集被划分为三个子集:训练集、验证集和测试集,支持除二元证据之外的多样的证据类型,从而允许更广泛的互动范围。在规模上,该数据集拥有超过一百万个合成患者数据用于训练,大约各有140,000个数据用于验证和测试。其任务旨在支持自动诊断和证据采集。
This dataset, named DDXPlus, provides differential diagnosis and pathological severity information, and features categorical and multiple-choice evidence to facilitate more efficient doctor-patient interactions. It covers 49 pathologies and 223 evidence points, including 110 symptoms and 113 pre-existing conditions. Furthermore, the dataset is divided into three subsets: training, validation, and test sets, and supports diverse evidence types beyond binary evidence, enabling a broader range of interaction scenarios. In terms of scale, the dataset contains over one million synthetic patient samples for training, with approximately 140,000 samples each for validation and testing. The tasks supported by this dataset are designed to facilitate automatic diagnosis and evidence acquisition.

- 1DDXPlus: A New Dataset For Automatic Medical Diagnosis · 2022年



