← 返回

Development and Validation of the AI-HeartAge Model in Framingham and UK Biobank

基于人工智能的心脏年龄模型在Framingham心脏研究和UK Biobank中的开发与验证

期刊: Hypertension 日期: 2026-03-19 PMID: 41853838 DOI: 10.1161/HYPERTENSIONAHA.125.26209 浏览: 92
作者: Timothy J Korzinski, Benjamin R Prescott, David J Hamel-Sellman, Vasanthi Xanthakis, Lucas L Cooper, Naomi M Hamburg, Chunming Wei, Emelia J Benjamin, Ramachandran S Vasan, Gary F Mitchell
Korzinski, T.J., Prescott, B.R., Hamel-Sellman, D.J., Xanthakis, V., Cooper, L.L., Hamburg, N.M., Wei, C., Benjamin, E.J., Vasan, R.S., & Mitchell, G.F. (2026). Development and Validation of the AI-HeartAge Model in Framingham and UK Biobank. Hypertension. https://doi.org/10.1161/HYPERTENSIONAHA.125.26209
Korzinski TJ, Prescott BR, Hamel-Sellman DJ, Xanthakis V, Cooper LL, Hamburg NM, et al. Development and Validation of the AI-HeartAge Model in Framingham and UK Biobank. Hypertension. 2026; doi: 10.1161/HYPERTENSIONAHA.125.26209
Korzinski TJ, Prescott BR, Hamel-Sellman DJ, et al. Development and Validation of the AI-HeartAge Model in Framingham and UK Biobank[J]. Hypertension. 2026. DOI: 10.1161/HYPERTENSIONAHA.125.26209.
@article{korzinski2026,
  author = {Timothy J Korzinski and Benjamin R Prescott and David J Hamel-Sellman and Vasanthi Xanthakis and Lucas L Cooper and Naomi M Hamburg and Chunming Wei and Emelia J Benjamin and Ramachandran S Vasan and Gary F Mitchell},
  title = {Development and Validation of the AI-HeartAge Model in Framingham and UK Biobank},
  journal = {Hypertension},
  year = {2026},
  doi = {10.1161/HYPERTENSIONAHA.125.26209},
  note = {PMID: 41853838},
}
TY  - JOUR
AU  - Timothy J Korzinski
AU  - Benjamin R Prescott
AU  - David J Hamel-Sellman
AU  - Vasanthi Xanthakis
AU  - Lucas L Cooper
AU  - Naomi M Hamburg
AU  - Chunming Wei
AU  - Emelia J Benjamin
AU  - Ramachandran S Vasan
AU  - Gary F Mitchell
TI  - Development and Validation of the AI-HeartAge Model in Framingham and UK Biobank
T2  - Hypertension
PY  - 2026
DO  - 10.1161/HYPERTENSIONAHA.125.26209
AN  - PMID:41853838
ER  - 

摘要

动脉压力波形形状传达了左心室和主动脉之间相互作用的信息,可提供生物学心脏年龄和心血管疾病风险的估计。人工智能心脏年龄(AI-HA)通过平均两个卷积神经网络的预测结果来估计,这些网络使用未校准的动脉张力测量或光电容积脉搏波波形作为输入,预测二尖瓣环组织多普勒e'和s'峰值速度。模型使用Framingham心脏研究参与者压力波形和超声心动图测量数据开发(N=6916名参与者,38174个波形,56%女性,平均年龄61±12岁)。在FHS基线桡动脉波形保留集(N=7018,54%女性,年龄50±16岁)和UK Biobank参与者(N=67986,53%女性,年龄57±8岁)中验证了AI-HA。在FHS(最长10年随访,148例心衰和331例CVD事件)中,AI-HA与心衰发病(HR 2.09)和CVD(HR 1.52)显著相关。在UK Biobank中同样观察到显著关联。AI-HA是一种新型且易于获取的左心室功能和心衰风险测量指标。

AI 智能解读

相关文献

返回分类: 心血管预防 查看原文 (DOI)
已选择 0 篇文献