人工智能培训

搜索

深度学习论文:重症监护病房患者预测模型的概括性(Generalizability of predictive models for intensive care

[复制链接]
kuhuhgiu 发表于 2018-12-7 11:36:35 | 显示全部楼层 |阅读模式
kuhuhgiu 2018-12-7 11:36:35 118 0 显示全部楼层
深度学习论文:重症监护病房患者预测模型的概括性(Generalizability of predictive models for intensive care unit patients)大量研究已经考虑为临床数据创建预测模型;然而,现有的许多文献都报道了仅使用单一数据源的结果。在这项工作中,我们评估了公共可用的eICU协作研究数据库中的模型的性能。我们展示了使用许多不同中心的交叉验证在新中心提供了合理的模型性能估计。我们进一步表明,跨中心培训的单一模型可以很好地转移到不同的医院,甚至比使用医院特定数据重新培训的模型也要好。我们的研究结果将多中心数据集用于模型开发,并突出了医院之间数据共享的需求,以最大限度地提高模型性能。
A large volume of research has considered the creation of predictive modelsfor clinical data;however, much existing literature reports results using onlya single source of data.In this work, we evaluate the performance of modelstrained on the publicly-available eICU Collaborative Research Database.We showthat cross-validation using many distinct centers provides a reasonableestimate of model performance in new centers.We further show that a singlemodel trained across centers transfers well to distinct hospitals, evencompared to a model retrained using hospital-specific data.Our resultsmotivate the use of multi-center datasets for model development and highlightthe need for data sharing among hospitals to maximize model performance.深度学习论文:重症监护病房患者预测模型的概括性(Generalizability of predictive models for intensive care unit patients) Paa0U0qYrqJQT76j.jpg
URL地址:https://arxiv.org/abs/1812.02275     ----pdf下载地址:https://arxiv.org/pdf/1812.02275    ----深度学习论文:重症监护病房患者预测模型的概括性(Generalizability of predictive models for intensive care unit patients)
回复

使用道具 举报

您需要登录后才可以回帖 登录 | 立即注册

本版积分规则 返回列表 发新帖

kuhuhgiu当前离线
新手上路

查看:118 | 回复:0

快速回复 返回顶部 返回列表