Multicenter Acute Ischemic Stroke, MRI and Clinical Text Dataset

随着全球中风风险的增加，对中风临床症状和临床决策的研究已成为尤其有必要。高质量、大规模、足够丰富的临床文本数据是开展临床文本研究的基础。中风。我们收集了 5788 名急性缺血性中风患者的多模态 MRI 数据集，据我们所知，这是最大的中风数据集，包含详细和完整的临床文本数据。该数据集的发布对于促进旨在实现分析方法多样化的人工智能模型的发展。这些模型探索一系列目前依赖人类的可学习的标准化或重复性任务，包括与疾病相关的评分计算，如 Aspect 评分、预后预测、治疗方案选择、病变标记和病变分割。

Against the backdrop of rising global stroke risk, research on clinical symptoms and clinical decision-making for stroke has become increasingly imperative. High-quality, large-scale and sufficiently diverse clinical text data serves as the foundational resource for clinical text-related research. We have collected a multimodal MRI dataset involving 5788 patients with acute ischemic stroke. To the best of our knowledge, this is the largest stroke dataset that includes detailed and complete clinical text data. The release of this dataset will significantly promote the development of AI models aimed at diversifying analytical methods. These models cover a series of learnable standardized or repetitive tasks that currently rely on human execution, including disease-related scoring calculations such as the Aspect score, prognosis prediction, treatment plan selection, lesion labeling, and lesion segmentation.