CQ500

我们回顾性地收集了一个数据集，其中包含来自各个中心的313,318个头部ct扫描及其临床报告。该数据集的一部分 (Qure25k数据集) 用于验证，其余用于开发算法。此外，从不同的中心收集了两个批次B1和B2的数据集 (CQ500数据集)，以临床验证算法。

We retrospectively collected a dataset consisting of 313,318 head CT scans and their corresponding clinical reports from multiple medical centers. A subset of this dataset, designated as the Qure25k dataset, was used for validation, while the remaining portion was employed for algorithm development. Furthermore, two batches of datasets (B1 and B2) collected from distinct centers, which collectively form the CQ500 dataset, were utilized for the clinical validation of the developed algorithm.