Features selection in a predictive model for cardiac surgery-associated acute kidney injury

被引:0
作者
Li, Qian [1 ]
Shen, Jingjia [1 ]
Lv, Hong [1 ]
Chen, Yuye [1 ]
Zhou, Chenghui [1 ]
Shi, Jia [1 ]
机构
[1] Chinese Acad Med Sci & Peking Union Med Coll, Fuwai Hosp, Natl Ctr Cardiovasc Dis, Dept Anesthesiol,State Key Lab Cardiovasc Dis, 167 Beilishi Rd, Beijing 100037, Peoples R China
来源
PERFUSION-UK | 2025年 / 40卷 / 05期
基金
中国国家自然科学基金;
关键词
feature selection; machine learning; logistic regression; cardiac surgical procedure; acute kidney injury; RISK-FACTORS; VALIDATION; DERIVATION;
D O I
10.1177/02676591241289364
中图分类号
R5 [内科学];
学科分类号
1002 ; 100201 ;
摘要
Background Cardiac surgery-associated acute kidney injury (CSA-AKI) is related to increased morbidity and mortality. However, limited studies have explored the influence of different feature selection (FS) methods on the predictive performance of CSA-AKI. Therefore, we aimed to compare the impact of different FS methods for CSA-AKI.Methods CSA-AKI is defined according to the kidney disease: Improving Global Outcomes (KDIGO) criteria. Both traditional logistic regression and machine learning methods were used to select the potential risk factors for CSA-AKI. The area under the receiver operating characteristic curve (AUC) was used to evaluate the performance of the models. In addition, the importance matrix plot by random forest was used to rank the features' importance.Results A total of 1977 patients undergoing cardiac surgery at Fuwai hospital from December 2018 to April 2021 were enrolled. The incidence of CSA-AKI during the first postoperative week was 27.8%. We concluded that different enrolled numbers of features impact the final selected feature number. The more you input, the more likely its output with all FS methods. In terms of performance, all selected features by various FS methods demonstrated excellent AUCs. Meanwhile, the embedded method demonstrated the highest accuracy compared with the LR method, while the filter method showed the lowest accuracy. Furthermore, NT-proBNP was found to be strongly associated with AKI. Our results confirmed some features that previous studies have reported and found some novel clinical parameters.Conclusions In our study, FS was as suitable as LR for predicting CSA-AKI. For FS, the embedded method demonstrated better efficacy than the other methods. Furthermore, NT-proBNP was confirmed to be strongly associated with AKI.
引用
收藏
页码:1218 / 1228
页数:11
相关论文
共 36 条
[1]   Keratoconus Severity Classification Using Features Selection and Machine Learning Algorithms [J].
Aatila, Mustapha ;
Lachgar, Mohamed ;
Hamid, Hrimech ;
Kartit, Ali .
COMPUTATIONAL AND MATHEMATICAL METHODS IN MEDICINE, 2021, 2021
[2]   Improved predictive models for acute kidney injury with IDEA: Intraoperative Data Embedded Analytics [J].
Adhikari, Lasith ;
Ozrazgat-Baslanti, Tezcan ;
Ruppert, Matthew ;
Madushani, R. W. M. A. ;
Paliwal, Srajan ;
Hashemighouchani, Haleh ;
Zheng, Feng ;
Tao, Ming ;
Lopes, Juliano M. ;
Li, Xiaolin ;
Rashidi, Parisa ;
Bihorac, Azra .
PLOS ONE, 2019, 14 (04)
[3]   Risk Factors and Outcome of Acute Kidney Injury after Isolated CABG Surgery: a Prospective Cohort Study [J].
Amini, Shahram ;
Najafi, Mona Najaf ;
Karrari, Seyedeh Parissa ;
Mashhadi, Mohammadghasem Etemadi ;
Mirzaei, Sahereh ;
Tashnizi, Mohammad Abbasi ;
Moeinipour, Ali Asghar ;
Hoseinikhah, Hamid ;
Aazami, Mathias Hossain ;
Jafari, Mahdieh .
BRAZILIAN JOURNAL OF CARDIOVASCULAR SURGERY, 2019, 34 (01) :70-75
[4]   Multivariable prediction of renal insufficiency developing after cardiac surgery [J].
Brown, Jeremiah R. ;
Cochran, Richard P. ;
Leavitt, Bruce J. ;
Dacey, Lawrence J. ;
Ross, Cathy S. ;
MacKenzie, Todd A. ;
Kunzelman, Karyn S. ;
Kramer, Robert S. ;
Hernandez, Felix, Jr. ;
Helm, Robert E. ;
Westbrook, Benjamin M. ;
Dunton, Robert F. ;
Malenka, David J. ;
O'Connor, Gerald T. .
CIRCULATION, 2007, 116 (11) :I139-I143
[5]  
Chertow GM, 1997, CIRCULATION, V95, P878
[6]   A systematic review shows no performance benefit of machine learning over logistic regression for clinical prediction models [J].
Christodoulou, Evangelia ;
Ma, Jie ;
Collins, Gary S. ;
Steyerberg, Ewout W. ;
Verbakel, Jan Y. ;
Van Calster, Ben .
JOURNAL OF CLINICAL EPIDEMIOLOGY, 2019, 110 :12-22
[7]  
Guyon I., 2003, J MACH LEARN RES, V3, P1157, DOI DOI 10.1162/153244303322753616
[8]  
Hameed S.S., 2018, Int. J. Advance Soft Compu. Appl, V10, P90
[9]   Defining oliguria during cardiopulmonary bypass and its relationship with cardiac surgery-associated acute kidney injury [J].
Hori, D. ;
Katz, N. M. ;
Fine, D. M. ;
Ono, M. ;
Barodka, V. M. ;
Lester, L. C. ;
Yenokyan, G. ;
Hogue, C. W. .
BRITISH JOURNAL OF ANAESTHESIA, 2016, 117 (06) :733-740
[10]   Derivation and validation of a model to predict acute kidney injury following cardiac surgery in patients with normal renal function [J].
Hu, Penghua ;
Mo, Zhiming ;
Chen, Yuanhan ;
Wu, Yanhua ;
Song, Li ;
Zhang, Li ;
Li, Zhilian ;
Fu, Lei ;
Liang, Huaban ;
Tao, Yiming ;
Liu, Shuangxin ;
Ye, Zhiming ;
Liang, Xinling .
RENAL FAILURE, 2021, 43 (01) :1205-1213