Integration of feature vectors from raw laboratory, medication and procedure names improves the precision and recall of models to predict postoperative mortality and acute kidney injury.
Integration of feature vectors from raw laboratory, medication and procedure names improves the precision and recall of models to predict postoperative mortality and acute kidney injury.
复制标题
整合来自原始化验室、药物和手术名称的特征向量可提高预测术后死亡率和急性肾损伤模型的精确度和召回率。
DOI：
10.1038/s41598-022-13879-7
复制
发表时间：
2022-06-17
影响因子：
4.6
通讯作者：
Halperin, Eran
中科院分区：
文献类型：
--
作者：
Hofer, Ira S.;Kupina, Marina;Laddaran, Lori;Halperin, Eran
成功使用机器学习（ML）预测各种围手术期结果的文章通常只使用临床医生选择的有限数量的特征。我们假设，与临床医生选择的较有限的特征集相比，利用患者化验结果、药物和手术名称的广泛特征集的技术将提高性能。化验结果的特征向量包括从 39 项化验中提取的共 702 个特征，药物包括 126 种常用药物的二进制标志，手术名称使用 Word2Vec 软件包创建长度为 100 的向量。共训练了九个模型：基线特征、三种数据类型各一个、基线+每种数据类型、（所有特征，然后使用特征缩减算法训练所有特征。在两种结果中，包含所有特征的模型（模型 8）（死亡率 ROC-AUC 94.32 ± 1.01，PR-AUC 36.80 ± 5.10 AKI ROC-AUC 92.45 ± 0.64，PR-AUC 76.22 ± 1.95）优于仅包含特征子集的模型。利用大量临床数据的特征化技术可以提高围手术期预测模型的性能。
Manuscripts that have successfully used machine learning (ML) to predict a variety of perioperative outcomes often use only a limited number of features selected by a clinician. We hypothesized that techniques leveraging a broad set of features for patient laboratory results, medications, and the surgical procedure name would improve performance as compared to a more limited set of features chosen by clinicians. Feature vectors for laboratory results included 702 features total derived from 39 laboratory tests, medications consisted of a binary flag for 126 commonly used medications, procedure name used the Word2Vec package for create a vector of length 100. Nine models were trained: baseline features, one for each of the three types of data Baseline + Each data type, (all features, and then all features with feature reduction algorithm. Across both outcomes the models that contained all features (model 8) (Mortality ROC-AUC 94.32 ± 1.01, PR-AUC 36.80 ± 5.10 AKI ROC-AUC 92.45 ± 0.64, PR-AUC 76.22 ± 1.95) was superior to models with only subsets of features. Featurization techniques leveraging a broad away of clinical data can improve performance of perioperative prediction models.
登录
查看更多内容
影响因子：
8.8
作者：
Lee CK;Hofer I;Gabel E;Baldi P;Cannesson M
通讯作者：
Cannesson M
影响因子：
3.9
作者：
Lee HC;Yoon SB;Yang SM;Kim WH;Ryu HG;Jung CW;Suh KS;Lee KH
通讯作者：
Lee KH
影响因子：
168.9
作者：
Pearse, Rupert M.;Moreno, Rui P.;Bauer, Peter;Pelosi, Paolo;Metnitz, Philipp;Spies, Claudia;Vallet, Benoit;Vincent, Jean-Louis;Hoeft, Andreas;Rhodes, Andrew
通讯作者：
Rhodes, Andrew
影响因子：
8.8
作者：
Hatib, Feras;Jian, Zhongping;Cannesson, Maxime
通讯作者：
Cannesson, Maxime
影响因子：
5.8
作者：
Kursa, Miron B.;Rudnicki, Witold R.
通讯作者：
Rudnicki, Witold R.