5TB患者影像检查数据集

本数据集是一个5TB患者影像检查数据集，由文本与图像多模态数据构成。数据均源于临床诊疗实践，涵盖近5年数千组全科患者在就诊过程中的影像和报告数据。该数据集不仅包括了DR、MR和CT三类医学影像检查信息并与报告进行对齐，还对疑似肿瘤患者进行标注。数据集已经过脱敏处理，不含个人信息。常用于医学影像辅助诊断任务的训练和研究。This 5TB patient imaging examination dataset consists of multimodal data combining text and image modalities. All data originates from clinical practice, covering thousands of sets of imaging and report data from general practice patients during their medical visits over the past five years. This dataset not only includes three types of medical imaging examination information (DR, MR, CT) aligned with their corresponding medical reports, but also provides annotations for patients suspected of having tumors. The dataset has been de-identified and contains no personal identifiable information (PII). It is commonly used for training and research in medical imaging-assisted diagnosis tasks.

搜集汇总

背景与挑战

背景概述

该数据集包含5TB临床医学影像与报告数据，涵盖近5年全科患者的DR/MR/CT检查记录及对应报告，含肿瘤疑似标注且经脱敏处理，适用于医学影像辅助诊断研究。