xpertsystems/hccar010-sample

HC-CAR-010 — 卒中风险预测合成队列是一个完全合成的数据集，专为卒中风险预测建模设计，覆盖了初级和二级预防全流程。数据集包含500名患者的144个特征列，涵盖10个模块：人口统计学、心血管风险因素、心脏生物标志物、凝血与血液学、心房颤动（包括CHA₂DS₂-VASc和HAS-BLED评分）、颈动脉与血管成像、心脏成像、神经成像、验证风险评分（如Framingham 10年卒中风险、SCORE2、Charlson、RRE-90）以及急性卒中结局（包括NIHSS、mRS、Barthel指数、TOAST分类、溶栓、血栓切除术、复发和1年死亡率）。该数据集基于真实世界注册研究（如Framingham心脏研究、GBD 2022、MESA、IST-3等）和临床指南校准，适用于分析、机器学习建模、演示和教育用途，同时保持100%合成性，无真实患者数据或隐私风险。数据集以CSV格式提供，样本采用CC-BY-NC-4.0许可，完整产品需商业许可。

HC-CAR-010 — Stroke Risk Prediction Synthetic Cohort is a fully synthetic dataset designed for stroke risk prediction modeling, spanning the complete primary and secondary prevention pipeline. The dataset includes 500 patients × 144 columns across 10 modules: Demographics, Cardiovascular Risk Factors, Cardiac Biomarkers, Coagulation & Hematology, Atrial Fibrillation (with CHA₂DS₂-VASc and HAS-BLED scoring), Carotid & Vascular Imaging, Cardiac Imaging, Neuroimaging, Validated Risk Scores (e.g., Framingham 10-year stroke risk, SCORE2, Charlson, RRE-90), and Acute Stroke Outcomes (including NIHSS, mRS, Barthel index, TOAST classification, thrombolysis, thrombectomy, recurrence, and 1-year mortality). It is calibrated against real-world registries and guidelines (e.g., Framingham Heart Study, GBD 2022, MESA, IST-3) and is suitable for analytics, machine learning modeling, demos, and education, while remaining 100% synthetic with no real patient data or privacy risks. The dataset is provided in CSV format, with the sample under CC-BY-NC-4.0 license and the full product requiring a commercial license.