Classification based on event in survival machine learning analysis of cardiovascular disease cohort

BMC Cardiovascular Disorders

Research

Nawzad Muhammed Ahmed¹  Shokh Mukhtar Ahmad² 

[1] Department of Statistics and Informatics, College of Administration and Economics, Sulaymaniyah University, Sulaymaniyah, Kurdistan, Iraq;Department of Statistics and Informatics, College of Administration and Economics, Sulaymaniyah University, Sulaymaniyah, Kurdistan, Iraq;Department of Medical Laboratory, Komar University of Science and Technology Science, Sulaymaniyah, Kurdistan, Iraq;

关键词: Survival analysis;    Machine learning;    Logistic regression;    SVM;    Tree descent;    Random forest.;   

DOI  :  10.1186/s12872-023-03328-2

 received in 2023-03-11, accepted in 2023-05-31,  发布年份 2023

来源: Springer

【 摘 要 】

The aim of this study is to assess the effectiveness of supervised learning classification models in predicting patient outcomes in a survival analysis problem involving cardiovascular patients with a significant cured fraction. The sample comprised 919 patients (365 females and 554 males) who were referred to Sulaymaniyah Cardiac Hospital and followed up for a maximum of 650 days between 2021 and 2023. During the research period, 162 patients (17.6%) died, and the cure fraction in this cohort was confirmed using the Mahler and Zhu test (P < 0.01). To determine the best patient status prediction procedure, several machine learning classifications were applied. The patients were classified into alive and dead using various machine learning algorithms, with almost similar results based on several indicators. However, random forest was identified as the best method in most indicators, with an Area under ROC of 0.934. The only weakness of this method was its relatively poor performance in correctly diagnosing deceased patients, whereas SVM with FP Rate of 0.263 performed better in this regard. Logistic and simple regression also showed better performance than other methods, with an Area under ROC of 0.911 and 0.909 respectively.

【 授权许可】

CC BY   
© The Author(s) 2023

【 图 表 】

41116_2023_37_Article_IEq139.gif

41116_2023_37_Article_IEq137.gif

41116_2023_37_Article_IEq118.gif

42004_2023_919_Article_IEq16.gif

41116_2023_37_Article_IEq117.gif

41116_2023_37_Article_IEq116.gif

41116_2023_37_Article_IEq89.gif

41116_2023_37_Article_IEq72.gif