AF Recurrence Dataset
收藏资源简介:
该数据集由巴斯克大学和巴斯克大学医院的研究人员创建,旨在预测心房颤动(AF)的复发。数据集包含1508名患者的文档化AF发作记录,并使用自然语言处理技术将结构化和非结构化临床数据相结合,以生成描述患者健康状况的表格数据。数据集通过将结构化电子健康记录(EHR)数据与自由文本出院报告相结合,克服了传统临床评分在预测AF复发方面的局限性。该数据集可用于评估传统临床评分、机器学习模型和大型表格模型(LTM)的预测性能,并探索性别和年龄对AF复发预测的影响。
This dataset was developed by researchers from the University of the Basque Country and the University of the Basque Country Hospital, with the primary objective of predicting the recurrence of atrial fibrillation (AF). It includes documented AF episode records from 1508 patients, and leverages natural language processing (NLP) techniques to integrate both structured and unstructured clinical data, thereby generating tabular data that characterizes patients' health status. By combining structured electronic health record (EHR) data with free-text discharge reports, this dataset addresses the limitations of traditional clinical scoring systems in predicting AF recurrence. Furthermore, it can be utilized to assess the predictive performance of traditional clinical scoring systems, machine learning models, and large tabular models (LTM), as well as to investigate the effects of gender and age on AF recurrence prediction.




