Comparative evaluation of machine learning models versus TIMI score in ST-segment-elevation myocardial infarction patients

Background: Risk stratification is an integral component of ST-segment-elevation myocardial infarction (STEMI) management practices. This study aimed to derive a machine learning (ML) model for risk stratification and identification of factors associated with in-hospital and 30-day mortality in pati...

Full description

Saved in:
Bibliographic Details
Main Authors: Mohit D. Gupta, Dixit Goyal, Shekhar Kunal, Manu Kumar Shetty, M.P. Girish, Vishal Batra, Ankit Bansal, Prashant Mishra, Mansavi Shukla, Vanshika Kohli, Akul Chadha, Arisha Fatima, Subrat Muduli, Anubha Gupta, Jamal Yusuf
Format: Article
Language:English
Published: Elsevier 2025-05-01
Series:Indian Heart Journal
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S0019483225000562
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850217767838416896
author Mohit D. Gupta
Dixit Goyal
Shekhar Kunal
Manu Kumar Shetty
M.P. Girish
Vishal Batra
Ankit Bansal
Prashant Mishra
Mansavi Shukla
Vanshika Kohli
Akul Chadha
Arisha Fatima
Subrat Muduli
Anubha Gupta
Jamal Yusuf
author_facet Mohit D. Gupta
Dixit Goyal
Shekhar Kunal
Manu Kumar Shetty
M.P. Girish
Vishal Batra
Ankit Bansal
Prashant Mishra
Mansavi Shukla
Vanshika Kohli
Akul Chadha
Arisha Fatima
Subrat Muduli
Anubha Gupta
Jamal Yusuf
author_sort Mohit D. Gupta
collection DOAJ
description Background: Risk stratification is an integral component of ST-segment-elevation myocardial infarction (STEMI) management practices. This study aimed to derive a machine learning (ML) model for risk stratification and identification of factors associated with in-hospital and 30-day mortality in patients with STEMI and compare it with traditional TIMI score. Methods: This was a single center prospective study wherein subjects >18 years with STEMI (n = 1700) were enrolled. Patients were divided into two groups: training (n = 1360) and validation dataset (n = 340). Six ML algorithms (Extra Tree, Random Forest, Multiple Perceptron, CatBoost, Logistic Regression and XGBoost) were used to train and tune the ML model and to determine the predictors of worse outcomes using feature selection. Additionally, the performance of ML models both for in-hospital and 30-day outcomes was compared to that of TIMI score. Results: Of the 1700 patients, 168 (9.88 %) had in-hospital mortality while 30-day mortality was reported in 210 (12.35 %) subjects. In terms of in-hospital mortality, Random Forest ML model (sensitivity: 80 %; specificity: 74 %; AUC: 80.83 %) outperformed the TIMI score (sensitivity: 70 %; specificity: 64 %; AUC:70.7 %). Similarly, Random Forest ML model (sensitivity: 81.63 %; specificity: 78.35 %; AUC: 78.29 %) had better performance as compared to TIMI score (sensitivity: 63.26 %; specificity: 63.91 %; AUC: 63.59 %) for 30-day mortality. Key predictors for worse outcomes at 30-days included mitral regurgitation on presentation, smoking, cardiogenic shock, diabetes, ventricular septal rupture, Killip class, age, female gender, low blood pressure and low ejection fraction. Conclusions: ML model outperformed the traditional regression based TIMI score as a risk stratification tool in patients with STEMI.
format Article
id doaj-art-0168fc7855c54d7b99698767f8c7ff6e
institution OA Journals
issn 0019-4832
language English
publishDate 2025-05-01
publisher Elsevier
record_format Article
series Indian Heart Journal
spelling doaj-art-0168fc7855c54d7b99698767f8c7ff6e2025-08-20T02:07:58ZengElsevierIndian Heart Journal0019-48322025-05-0177313314110.1016/j.ihj.2025.03.010Comparative evaluation of machine learning models versus TIMI score in ST-segment-elevation myocardial infarction patientsMohit D. Gupta0Dixit Goyal1Shekhar Kunal2Manu Kumar Shetty3M.P. Girish4Vishal Batra5Ankit Bansal6Prashant Mishra7Mansavi Shukla8Vanshika Kohli9Akul Chadha10Arisha Fatima11Subrat Muduli12Anubha Gupta13Jamal Yusuf14Department of Cardiology, Govind Ballabh Pant Institute of Post Graduate Medical Education and Research, Delhi, India; Corresponding author. Room no. 125, First floor, Academic block, Department of Cardiology, Govind Ballabh Pant Institute of Postgraduate Medical Education & Research, New Delhi, 110002, India.Department of Cardiology, Govind Ballabh Pant Institute of Post Graduate Medical Education and Research, Delhi, IndiaDepartment of Cardiology, ESIC Medical College and Hospital, Faridabad, Haryana, IndiaDepartment of Pharmacology, Maulana Azad Medical College, Delhi, IndiaDepartment of Cardiology, Govind Ballabh Pant Institute of Post Graduate Medical Education and Research, Delhi, IndiaDepartment of Cardiology, Govind Ballabh Pant Institute of Post Graduate Medical Education and Research, Delhi, IndiaDepartment of Cardiology, Govind Ballabh Pant Institute of Post Graduate Medical Education and Research, Delhi, IndiaDepartment of Cardiology, Govind Ballabh Pant Institute of Post Graduate Medical Education and Research, Delhi, IndiaDepartment of Cardiology, Govind Ballabh Pant Institute of Post Graduate Medical Education and Research, Delhi, IndiaDepartment of Cardiology, Govind Ballabh Pant Institute of Post Graduate Medical Education and Research, Delhi, IndiaDepartment of Cardiology, Govind Ballabh Pant Institute of Post Graduate Medical Education and Research, Delhi, IndiaDepartment of Cardiology, Govind Ballabh Pant Institute of Post Graduate Medical Education and Research, Delhi, IndiaDepartment of Cardiology, Govind Ballabh Pant Institute of Post Graduate Medical Education and Research, Delhi, IndiaDepartment of Electronics and Communications Engineering, Indraprastha Institute of Information Technology, Delhi, IndiaDepartment of Cardiology, Govind Ballabh Pant Institute of Post Graduate Medical Education and Research, Delhi, IndiaBackground: Risk stratification is an integral component of ST-segment-elevation myocardial infarction (STEMI) management practices. This study aimed to derive a machine learning (ML) model for risk stratification and identification of factors associated with in-hospital and 30-day mortality in patients with STEMI and compare it with traditional TIMI score. Methods: This was a single center prospective study wherein subjects >18 years with STEMI (n = 1700) were enrolled. Patients were divided into two groups: training (n = 1360) and validation dataset (n = 340). Six ML algorithms (Extra Tree, Random Forest, Multiple Perceptron, CatBoost, Logistic Regression and XGBoost) were used to train and tune the ML model and to determine the predictors of worse outcomes using feature selection. Additionally, the performance of ML models both for in-hospital and 30-day outcomes was compared to that of TIMI score. Results: Of the 1700 patients, 168 (9.88 %) had in-hospital mortality while 30-day mortality was reported in 210 (12.35 %) subjects. In terms of in-hospital mortality, Random Forest ML model (sensitivity: 80 %; specificity: 74 %; AUC: 80.83 %) outperformed the TIMI score (sensitivity: 70 %; specificity: 64 %; AUC:70.7 %). Similarly, Random Forest ML model (sensitivity: 81.63 %; specificity: 78.35 %; AUC: 78.29 %) had better performance as compared to TIMI score (sensitivity: 63.26 %; specificity: 63.91 %; AUC: 63.59 %) for 30-day mortality. Key predictors for worse outcomes at 30-days included mitral regurgitation on presentation, smoking, cardiogenic shock, diabetes, ventricular septal rupture, Killip class, age, female gender, low blood pressure and low ejection fraction. Conclusions: ML model outperformed the traditional regression based TIMI score as a risk stratification tool in patients with STEMI.http://www.sciencedirect.com/science/article/pii/S0019483225000562Artificial intelligenceAcute coronary syndromeRisk stratification
spellingShingle Mohit D. Gupta
Dixit Goyal
Shekhar Kunal
Manu Kumar Shetty
M.P. Girish
Vishal Batra
Ankit Bansal
Prashant Mishra
Mansavi Shukla
Vanshika Kohli
Akul Chadha
Arisha Fatima
Subrat Muduli
Anubha Gupta
Jamal Yusuf
Comparative evaluation of machine learning models versus TIMI score in ST-segment-elevation myocardial infarction patients
Indian Heart Journal
Artificial intelligence
Acute coronary syndrome
Risk stratification
title Comparative evaluation of machine learning models versus TIMI score in ST-segment-elevation myocardial infarction patients
title_full Comparative evaluation of machine learning models versus TIMI score in ST-segment-elevation myocardial infarction patients
title_fullStr Comparative evaluation of machine learning models versus TIMI score in ST-segment-elevation myocardial infarction patients
title_full_unstemmed Comparative evaluation of machine learning models versus TIMI score in ST-segment-elevation myocardial infarction patients
title_short Comparative evaluation of machine learning models versus TIMI score in ST-segment-elevation myocardial infarction patients
title_sort comparative evaluation of machine learning models versus timi score in st segment elevation myocardial infarction patients
topic Artificial intelligence
Acute coronary syndrome
Risk stratification
url http://www.sciencedirect.com/science/article/pii/S0019483225000562
work_keys_str_mv AT mohitdgupta comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients
AT dixitgoyal comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients
AT shekharkunal comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients
AT manukumarshetty comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients
AT mpgirish comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients
AT vishalbatra comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients
AT ankitbansal comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients
AT prashantmishra comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients
AT mansavishukla comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients
AT vanshikakohli comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients
AT akulchadha comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients
AT arishafatima comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients
AT subratmuduli comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients
AT anubhagupta comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients
AT jamalyusuf comparativeevaluationofmachinelearningmodelsversustimiscoreinstsegmentelevationmyocardialinfarctionpatients