scholarly journals Machine-learning to Improve Prediction of Mortality following Acute Myocardial Infarction: An Assessment in the NCDR-Chest Pain-Myocardial Infarction Registry

2019 ◽  
Author(s):  
Rohan Khera ◽  
Julian Haimovich ◽  
Nate Hurley ◽  
Robert McNamara ◽  
John A Spertus ◽  
...  
Keyword(s):  
Machine Learning ◽  
Myocardial Infarction ◽  
Acute Myocardial Infarction ◽  
Logistic Regression ◽  
Chest Pain ◽  
High Risk ◽  
Hospital Mortality ◽  
Gradient Descent ◽  
Machine Learning Methods ◽  
C Statistic

ABSTRACTIntroductionAccurate prediction of risk of death following acute myocardial infarction (AMI) can guide the triage of care services and shared decision-making. Contemporary machine-learning may improve risk-prediction by identifying complex relationships between predictors and outcomes.Methods and ResultsWe studied 993,905 patients in the American College of Cardiology Chest Pain-MI Registry hospitalized with AMI (mean age 64 ± 13 years, 34% women) between January 2011 and December 2016. We developed and validated three machine learning models to predict in-hospital mortality and compared the performance characteristics with a logistic regression model. In an independent validation cohort, we compared logistic regression with lasso regularization (c-statistic, 0.891 [95% CI, 0.890-0.892]), gradient descent boosting (c-statistic, 0.902 [0.901-0.903]), and meta-classification that combined gradient descent boosting with a neural network (c-statistic, 0.904 [0.903-0.905]) with traditional logistic regression (c-statistic, 0.882 [0.881-0.883]). There were improvements in classification of individuals across the spectrum of patient risk with each of the three methods; the meta-classifier model – our best performing model - reclassified 20.9% of individuals deemed high-risk for mortality in logistic regression appropriately as low-to-moderate risk, and 8.2% of deemed low-risk to moderate-to-high risk based consistent with the actual event rates.ConclusionsMachine-learning methods improved the prediction of in-hospital mortality for AMI compared with logistic regression. Machine learning methods enhance the utility of risk models developed using traditional statistical approaches through additional exploration of the relationship between variables and outcomes.

AN EFFICIENT MACHINE LEARNING MODEL FOR PREDICTION OF ACUTE MYOCARDIAL INFARCTION

2020 ◽  
Vol 13 ◽  
Author(s):  
Dhilsath Fathima.M ◽  
S. Justin Samuel ◽  
R. Hari Haran
Keyword(s):  
Machine Learning ◽  
Myocardial Infarction ◽  
Acute Myocardial Infarction ◽  
Logistic Regression ◽  
Decision Tree ◽  
Learning Model ◽  
Training Dataset ◽  
Data Set ◽  
Machine Learning Model ◽  
Proposed Model

Aim: This proposed work is used to develop an improved and robust machine learning model for predicting Myocardial Infarction (MI) could have substantial clinical impact. Objectives: This paper explains how to build machine learning based computer-aided analysis system for an early and accurate prediction of Myocardial Infarction (MI) which utilizes framingham heart study dataset for validation and evaluation. This proposed computer-aided analysis model will support medical professionals to predict myocardial infarction proficiently. Methods: The proposed model utilize the mean imputation to remove the missing values from the data set, then applied principal component analysis to extract the optimal features from the data set to enhance the performance of the classifiers. After PCA, the reduced features are partitioned into training dataset and testing dataset where 70% of the training dataset are given as an input to the four well-liked classifiers as support vector machine, k-nearest neighbor, logistic regression and decision tree to train the classifiers and 30% of test dataset is used to evaluate an output of machine learning model using performance metrics as confusion matrix, classifier accuracy, precision, sensitivity, F1-score, AUC-ROC curve. Results: Output of the classifiers are evaluated using performance measures and we observed that logistic regression provides high accuracy than K-NN, SVM, decision tree classifiers and PCA performs sound as a good feature extraction method to enhance the performance of proposed model. From these analyses, we conclude that logistic regression having good mean accuracy level and standard deviation accuracy compared with the other three algorithms. AUC-ROC curve of the proposed classifiers is analyzed from the output figure.4, figure.5 that logistic regression exhibits good AUC-ROC score, i.e. around 70% compared to k-NN and decision tree algorithm. Conclusion: From the result analysis, we infer that this proposed machine learning model will act as an optimal decision making system to predict the acute myocardial infarction at an early stage than an existing machine learning based prediction models and it is capable to predict the presence of an acute myocardial Infarction with human using the heart disease risk factors, in order to decide when to start lifestyle modification and medical treatment to prevent the heart disease.

scholarly journals Using machine-learning risk prediction models to triage the acuity of undifferentiated patients entering the emergency care system: a systematic review

2020 ◽  
Vol 4 (1) ◽  
Author(s):  
Jamie Miles ◽  
Janette Turner ◽  
Richard Jacques ◽  
Julia Williams ◽  
Suzanne Mason
Keyword(s):  
Machine Learning ◽  
Systematic Review ◽  
Neural Networks ◽  
Logistic Regression ◽  
Emergency Care ◽  
Model Performance ◽  
Machine Learning Methods ◽  
C Statistic ◽  
Emergency Care System ◽  
Care System

Abstract Background The primary objective of this review is to assess the accuracy of machine learning methods in their application of triaging the acuity of patients presenting in the Emergency Care System (ECS). The population are patients that have contacted the ambulance service or turned up at the Emergency Department. The index test is a machine-learning algorithm that aims to stratify the acuity of incoming patients at initial triage. This is in comparison to either an existing decision support tool, clinical opinion or in the absence of these, no comparator. The outcome of this review is the calibration, discrimination and classification statistics. Methods Only derivation studies (with or without internal validation) were included. MEDLINE, CINAHL, PubMed and the grey literature were searched on the 14th December 2019. Risk of bias was assessed using the PROBAST tool and data was extracted using the CHARMS checklist. Discrimination (C-statistic) was a commonly reported model performance measure and therefore these statistics were represented as a range within each machine learning method. The majority of studies had poorly reported outcomes and thus a narrative synthesis of results was performed. Results There was a total of 92 models (from 25 studies) included in the review. There were two main triage outcomes: hospitalisation (56 models), and critical care need (25 models). For hospitalisation, neural networks and tree-based methods both had a median C-statistic of 0.81 (IQR 0.80-0.84, 0.79-0.82). Logistic regression had a median C-statistic of 0.80 (0.74-0.83). For critical care need, neural networks had a median C-statistic of 0.89 (0.86-0.91), tree based 0.85 (0.84-0.88), and logistic regression 0.83 (0.79-0.84). Conclusions Machine-learning methods appear accurate in triaging undifferentiated patients entering the Emergency Care System. There was no clear benefit of using one technique over another; however, models derived by logistic regression were more transparent in reporting model performance. Future studies should adhere to reporting guidelines and use these at the protocol design stage. Registration and funding This systematic review is registered on the International prospective register of systematic reviews (PROSPERO) and can be accessed online at the following URL: https://www.crd.york.ac.uk/PROSPERO/display_record.php?ID=CRD42020168696 This study was funded by the NIHR as part of a Clinical Doctoral Research Fellowship.

scholarly journals Biochemical Markers of Acute Myocardial Infarction: Strategies for Improving Their Clinical Usefulness

1998 ◽  
Vol 35 (3) ◽  
pp. 393-399 ◽  
Author(s):  
Ian A Katz ◽  
Les Irwig ◽  
John D Vinen ◽  
Lyn March ◽  
Lindsay E Wyndham ◽  
...  
Keyword(s):  
Myocardial Infarction ◽  
Acute Myocardial Infarction ◽  
Logistic Regression ◽  
Chest Pain ◽  
Troponin T ◽  
Mass Measurement ◽  
World Health ◽  
Likelihood Ratios ◽  
Significant Independent Predictor ◽  
Activity Measurements

We investigated the early diagnostic utility, including incremental value, of the serum cardiac markers creatine kinase (CK), CK-MB (mass and activity measurements), cardiac troponin T, and myoglobin in the diagnosis of acute myocardial infarction (AMI) in patients presenting to a major teaching hospital with chest pain and non-diagnostic electrocardiographs (ECG). The reference diagnosis of acute myocardial infarction was made by a single, independent cardiologist using World Health Organization criteria. CK and CK-MB mass were the only significant predictors of AMI at presentation to the Emergency Department. Logistic regression analysis revealed that CK did not significantly predict ( P = 0·23) myocardial infarction once CK-MB mass was in the model. Using test results on follow up, in addition to presentation CK-MB mass, change in CK-MB mass was the only other significant independent predictor of AMI. Likelihood ratios for various levels of the significant markers in the logistic regression are given. In conclusion, CK-MB mass measurement was the only useful serum cardiac marker for the diagnosis of AMI in patients presenting with chest pain with non-diagnostic ECGs.

Development of Prediction Models for Acute Myocardial Infarction at Prehospital Stage with Machine Learning Based on a Nationwide Database (Preprint)

2021 ◽  
Author(s):  
Arom Choi ◽  
Min Joung Kim ◽  
Ji Min Sung ◽  
Hyuk-Jae Chang ◽  
Jayong Lee ◽  
...  
Keyword(s):  
Machine Learning ◽  
Myocardial Infarction ◽  
Acute Myocardial Infarction ◽  
Logistic Regression ◽  
Statistical Methods ◽  
Prediction Models ◽  
National Database ◽  
Type Model ◽  
Prehospital Management ◽  
Extreme Gradient Boosting

BACKGROUND Since acute myocardial infarction (AMI) is a leading cause of mortality worldwide, the accurate evaluation of risk factors of AMI at prehospital stage provides appropriate prehospital management and rapid transportation to the most appropriate hospital for treatment. Prediction of AMI derived from national database can accelerate early recognition and timely management to improve the survival rate. OBJECTIVE This study was conducted to develop and compare the efficacy of models for the prediction of AMI at the prehospital stage based on variables identified from a nationwide systematic emergency medical service (EMS) registry using conventional statistical methods and machine learning algorithms. METHODS From among patients transferred from the public EMS to emergency departments in Korea from January 2016 to December 2018, the patients aged >15 years in the EMS cardiovascular registry were enrolled. Two datasets were constructed according to the hierarchical structure of the EMS cardiovascular registry. For each dataset, several predictive models for AMI were derived and compared using conventional statistical methods and machine learning. RESULTS In total, 184,577 patients (Dataset 1) in the EMS cardiovascular registry were included in the final analysis. Among them, 72,439 patients (Dataset 2) were suspected to have AMI at the prehospital stage (as assessed by paramedics). Between the models derived using the conventional logistic regression method, the B-type model incorporated AMI-specific variables from the A-type model, and exhibited a superior discriminative ability (P = 0.02). The models that used extreme gradient boosting and multilayer perceptron yielded a higher predictive performance than the model derived based on conventional logistic regression for all analyses that used both datasets. Each machine learning algorithm yielded different classification lists regarding the 10 most important features. CONCLUSIONS This study demonstrates that prediction models, which use nationwide prehospital data and are developed with appropriate structures, can improve the identification of patients who need timely AMI management.

scholarly journals A comparison of logistic regression models with alternative machine learning methods to predict the risk of in-hospital mortality in emergency medical admissions via external validation

2018 ◽  
Vol 26 (1) ◽  
pp. 34-44 ◽  
Author(s):  
Muhammad Faisal ◽  
Andy Scally ◽  
Robin Howes ◽  
Kevin Beatson ◽  
Donald Richardson ◽  
...  
Keyword(s):  
Machine Learning ◽  
Logistic Regression ◽  
Hospital Mortality ◽  
Receiver Operating Characteristic Curve ◽  
Operating Characteristic ◽  
Characteristic Curve ◽  
External Validation ◽  
Learning Methods ◽  
Machine Learning Methods ◽  
Operating Characteristic Curve

We compare the performance of logistic regression with several alternative machine learning methods to estimate the risk of death for patients following an emergency admission to hospital based on the patients’ first blood test results and physiological measurements using an external validation approach. We trained and tested each model using data from one hospital ( n = 24,696) and compared the performance of these models in data from another hospital ( n = 13,477). We used two performance measures – the calibration slope and area under the receiver operating characteristic curve. The logistic model performed reasonably well – calibration slope: 0.90, area under the receiver operating characteristic curve: 0.847 compared to the other machine learning methods. Given the complexity of choosing tuning parameters of these methods, the performance of logistic regression with transformations for in-hospital mortality prediction was competitive with the best performing alternative machine learning methods with no evidence of overfitting.

scholarly journals Assessment of the CHA2DS2-VASc score in predicting ischemic stroke and death in patients with acute myocardial infarction without atrial fibrillation

2021 ◽  
Vol 22 (Supplement_1) ◽  
Author(s):  
A Marques ◽  
A Briosa ◽  
AR Pereira ◽  
S Alegria ◽  
J Grade Santos ◽  
...  
Keyword(s):  
Myocardial Infarction ◽  
Atrial Fibrillation ◽  
Acute Myocardial Infarction ◽  
Ischemic Stroke ◽  
Chest Pain ◽  
Hospital Mortality ◽  
Hospital Stay ◽  
Rank Test ◽  
Log Rank Test ◽  
Previous Therapy

Abstract Funding Acknowledgements Type of funding sources: None. onbehalf on behalf of the investigators of the Portuguese Registry of Acute Coronary Syndromes Introduction The CHA2DS2-VASc score is used in clinical practice to stratify the risk of stroke in patients (pts) with atrial fibrillation (AF). Its usefulness in the population of pts with acute myocardial infarction without AF is not well known. Objectives To investigate whether CHA2DS2-VASc predicts ischemic stroke and death during hospital stay in pts with acute myocardial infarction without known AF. To determine independent predictors of ischemic stroke in this population. Methods A multicentre, retrospective study was performed during 01/10/2010-04/09/2019 period, and included all pts admitted due to acute myocardial infarction. Pts with previous AF, AF rhythm in the electrocardiogram at admission or AF during hospital stay were excluded. Statistical analysis with Kaplan-Mayer and Cox regression was applied. Results Of 29851 pts admitted with acute myocardial infarction, were included in our study 19218 pts (74% male, mean age of 65 ± 14 years).  During hospital stay, 78 (0.4%) pts had an ischemic stroke and 462 (2.4%) pts died.  The event-free survival analysis showed significant differences according to the CHA2DS2-VASc score at admission (log rank test p = 0.015 for ischemic stroke; log rank test p < 0.001 for in-hospital mortality). (Figure)  The CHA2DS2-VASc score demonstrated a good predictive accuracy for in-hospital mortality (area under the ROC curve 0.69; 95% CI 0.67-0.72; p < 0.001). The area under the ROC curve indicates that the CHA2DS2-VASc score performed modestly for ischemic stroke (0.62; 95% CI 0.56-0.68; p < 0.001).  In univariate analysis, the factors that were positively associated with ischemic stroke during hospital stay were CHA2DS2-VASc, absence of previous therapy with statin, time between cardiac symptoms and hospital admission, absence of chest pain, Killip-Kimball class, cardiorespiratory arrest, complete left ventricular block and left ventricle ejection fraction <50% (p < 0.05).  After multivariate analysis, CHA2DS2-VASc≥3 (HR 2.25; 95% CI 1.37-3.71; p = 0.001), absence of chest pain (HR 3.17; CI 1.44-6.14, p < 0.001) and previous therapy with statin (HR 0.39; 95% CI 0.22-0.67; p = 0.001) were independent predictors of ischemic stroke. Conclusion  Among patients with acute myocardial infarction without known atrial fibrillation, the CHA2DS2-VASc score was associated with risk of ischemic stroke and death during hospital stay. This score may be useful for estimating the risk of stroke and in-hospital mortality in these population without known atrial fibrillation. Abstract Figure.

scholarly journals Current Status and Hospital-Level Differences in Care and Outcomes of Patients With Acute Non-ST-Segment Elevation Myocardial Infarction in China: Insights From China Acute Myocardial Infarction Registry

2022 ◽  
Vol 8 ◽  
Author(s):  
Qinghao Zhao ◽  
Haiyan Xu ◽  
Xuan Zhang ◽  
Yunqing Ye ◽  
Qiuting Dong ◽  
...  
Keyword(s):  
Myocardial Infarction ◽  
Acute Myocardial Infarction ◽  
High Risk ◽  
Hospital Mortality ◽  
Mortality Risk ◽  
County Level ◽  
Hospital Level ◽  
High Risk Patients ◽  
Risk Patients ◽  
Province Level

BackgroundWith the growing burden of non-ST-elevation myocardial infarction (NSTEMI), developing countries face great challenges in providing equitable treatment nationwide. However, little is known about hospital-level disparities in the quality of NSTEMI care in China. We aimed to investigate the variations in NSTEMI care and patient outcomes across the three hospital levels (province-, prefecture- and county-level, with decreasing scale) in China.MethodsData were derived from the China Acute Myocardial Infarction Registry on patients with NSTEMI consecutively registered between January 2013 and November 2016 from 31 provinces and municipalities throughout mainland China. Patients were categorized according to the hospital level they were admitted to. Multilevel generalized mixed models were fitted to examine the relationship between the hospital level and in-hospital mortality risk.ResultsIn total, 8,054 patients with NSTEMI were included (province-level: 1,698 patients; prefecture-level: 5,240 patients; county-level: 1,116 patients). Patients in the prefecture- and county-level hospitals were older, more likely to be female, and presented worse cardiac function than those in the province-level hospitals (P <0.05). Compared with the province-level hospitals, the rate of invasive strategies was significantly lower in the prefecture- and county-level hospitals (65.3, 43.3, and 15.4%, respectively, P <0.001). Invasive strategies were performed within the guideline-recommended timeframe in 25.4, 9.7, and 1.7% of very-high-risk patients, and 16.4, 7.4, and 2.4% of high-risk patients in province-, prefecture- and county-level hospitals, respectively (both P <0.001). The use of dual antiplatelet therapy in the county-level hospitals (87.2%) remained inadequate compared to the province- (94.5%, P <0.001) and prefecture-level hospitals (94.5%, P <0.001). There was an incremental trend of in-hospital mortality from province- to prefecture- to county-level hospitals (3.0, 4.4, and 6.9%, respectively, P-trend <0.001). After stepwise adjustment for patient characteristics, presentation, hospital facilities and in-hospital treatments, the hospital-level gap in mortality risk gradually narrowed and lost statistical significance in the fully adjusted model [Odds ratio: province-level vs. prefecture-level: 1.23 (0.73–2.05), P = 0.441; province-level vs. county-level: 1.61 (0.80–3.26), P = 0.182; P-trend = 0.246].ConclusionsThere were significant variations in NSTEMI presentation and treatment patterns across the three hospital levels in China, which may largely explain the hospital-level disparity in in-hospital mortality. Quality improvement initiatives are warranted, especially among lower-level hospitals.

P5014The NPAC score for predicting survival after incident acute myocardial infarction

2019 ◽  
Vol 40 (Supplement_1) ◽  
Author(s):  
K H Li ◽  
J Ho ◽  
Z Xu ◽  
I Lakhani ◽  
G Bazoukis ◽  
...  
Keyword(s):  
Myocardial Infarction ◽  
Acute Myocardial Infarction ◽  
Logistic Regression ◽  
Deep Learning ◽  
Learning Model ◽  
Peripheral Vascular ◽  
Short Term ◽  
C Statistic ◽  
Short Term Mortality ◽  
Deep Learning Model

Abstract Background Risk stratification in acute myocardial infarction (AMI) is important for guiding clinical management. Current risk scores are mostly derived from clinical trials with stringent patient selection. We aimed to establish and evaluate a composite scoring system to predict short-term mortality after index episodes of AMI, independent of electrocardiography (ECG) pattern, in a large real-world cohort. Methods Using electronic health records, patients admitted to our regional teaching hospital (derivation cohort, n=2127) and an independent tertiary care center (validation cohort, n=1276) with index acute myocardial infarction between January 2013 and December 2017 as confirmed by principal diagnosis and laboratory findings, were identified retrospectively. Results Univariate logistic regression was used as the primary model to identify potential contributors to mortality. Stepwise forward likelihood ratio logistic regression revealed that neutrophil-to-lymphocyte ratio, peripheral vascular disease, age, and serum creatinine (NPAC) were significant predictors for 90-day mortality (Hosmer-Lemeshow test, P=0.21). Each component of the NPAC score was weighted by beta-coefficients in multivariate analysis. The C-statistic of the NPAC score was 0.75, which was higher than the conventional Charlson's score (C-statistic=0.63). Application of a deep learning model to our dataset improved the accuracy of classification with a C-statistic of 0.81. Multivariate binary logistic regression Variable β Adjusted Odds ratio (95% CI) P-value Points Age ≥65 years 1.304 3.68 (2.63–5.17) <0.001 2 Peripheral vascular disease 1.109 3.03 (1.52–6.04) 0.002 2 NLRt ≥9.51 1.100 2.73 (2.12–3.51) <0.001 1 Creatinine≥109 μmol/L 1.003 3.00 (2.35–3.85) <0.001 2 NPAC deep learning model Conclusions The NPAC score comprised of four items from routine laboratory parameters and basic clinical information and can facilitate early identification of cases at risk of short-term mortality following index myocardial infarction. Deep learning model can serve as a gate-keeper to provide more accurate prediction to facilitate clinical decision making.

scholarly journals Applications of machine learning to undifferentiated chest pain in the emergency department: A systematic review

PLoS ONE ◽  
2021 ◽  
Vol 16 (8) ◽  
pp. e0252612
Author(s):  
Jonathon Stewart ◽  
Juan Lu ◽  
Adrian Goudie ◽  
Mohammed Bennamoun ◽  
Peter Sprivulis ◽  
...  
Keyword(s):  
Machine Learning ◽  
Myocardial Infarction ◽  
Systematic Review ◽  
Acute Myocardial Infarction ◽  
Emergency Department ◽  
Chest Pain ◽  
Risk Stratification ◽  
Clinical Decision ◽  
Risk Of Bias ◽  
Current Risk

Background Chest pain is amongst the most common reason for presentation to the emergency department (ED). There are many causes of chest pain, and it is important for the emergency physician to quickly and accurately diagnose life threatening causes such as acute myocardial infarction (AMI). Multiple clinical decision tools have been developed to assist clinicians in risk stratifying patients with chest. There is growing recognition that machine learning (ML) will have a significant impact on the practice of medicine in the near future and may assist with diagnosis and risk stratification. This systematic review aims to evaluate how ML has been applied to adults presenting to the ED with undifferentiated chest pain and assess if ML models show improved performance when compared to physicians or current risk stratification techniques. Methods and findings We conducted a systematic review of journal articles that applied a ML technique to an adult patient presenting to an emergency department with undifferentiated chest pain. Multiple databases were searched from inception through to November 2020. In total, 3361 articles were screened, and 23 articles were included. We did not conduct a metanalysis due to a high level of heterogeneity between studies in both their methods, and reporting. The most common primary outcomes assessed were diagnosis of acute myocardial infarction (AMI) (12 studies), and prognosis of major adverse cardiovascular event (MACE) (6 studies). There were 14 retrospective studies and 5 prospective studies. Four studies reported the development of a machine learning model retrospectively then tested it prospectively. The most common machine learning methods used were artificial neural networks (14 studies), random forest (6 studies), support vector machine (5 studies), and gradient boosting (2 studies). Multiple studies achieved high accuracy in both the diagnosis of AMI in the ED setting, and in predicting mortality and composite outcomes over various timeframes. ML outperformed existing risk stratification scores in all cases, and physicians in three out of four cases. The majority of studies were single centre, retrospective, and without prospective or external validation. There were only 3 studies that were considered low risk of bias and had low applicability concerns. Two studies reported integrating the ML model into clinical practice. Conclusions Research on applications of ML for undifferentiated chest pain in the ED has been ongoing for decades. ML has been reported to outperform emergency physicians and current risk stratification tools to diagnose AMI and predict MACE but has rarely been integrated into practice. Many studies assessing the use of ML in undifferentiated chest pain in the ED have a high risk of bias. It is important that future studies make use of recently developed standardised ML reporting guidelines, register their protocols, and share their datasets and code. Future work is required to assess the impact of ML model implementation on clinical decision making, patient orientated outcomes, and patient and physician acceptability. Trial registration International Prospective Register of Systematic Reviews registration number: CRD42020184977.

scholarly journals Development and Validation of Interpretable Machine Learning for Stroke Occurrence in Older, Community Chinese Dwellers

2021 ◽  
Author(s):  
Yafei Wu ◽  
Zhongquan Jiang ◽  
Shaowu Lin ◽  
Ya Fang
Keyword(s):  
Machine Learning ◽  
Older Adults ◽  
Logistic Regression ◽  
High Risk ◽  
Prediction Models ◽  
Curve Analysis ◽  
Learning Methods ◽  
Machine Learning Methods ◽  
Interpretable Machine Learning ◽  
Extreme Gradient Boosting

Abstract Background: Prediction of stroke based on individuals’ risk factors, especially for a first stroke event, is of great significance for primary prevention of high-risk populations. Our study aimed to investigate the applicability of interpretable machine learning for predicting a 2-year stroke occurrence in older adults compared with logistic regression.Methods: A total of 5960 participants consecutively surveyed from July 2011 to August 2013 in the China Health and Retirement Longitudinal Study were included for analysis. We constructed a traditional logistic regression (LR) and two machine learning methods, namely random forest (RF) and extreme gradient boosting (XGBoost), to distinguish stroke occurrence versus non-stroke occurrence using data on demographics, lifestyle, disease history, and clinical variables. Grid search and 10-fold cross validation were used to tune the hyperparameters. Model performance was assessed by discrimination, calibration, decision curve and predictiveness curve analysis.Results: Among the 5960 participants, 131 (2.20%) of them developed stroke after an average of 2-year follow-up. Our prediction models distinguished stroke occurrence versus non-stroke occurrence with excellent performance. The AUCs of machine learning methods (RF, 0.823[95% CI, 0.759-0.886]; XGBoost, 0.808[95% CI, 0.730-0.886]) were significantly higher than LR (0.718[95% CI, 0.649, 0.787], p<0.05). No significant difference was observed between RF and XGBoost (p>0.05). All prediction models had good calibration results, and the brier score were 0.022 (95% CI, 0.015-0.028) in LR, 0.019 (95% CI, 0.014-0.025) in RF, and 0.020 (95% CI, 0.015-0.026) in XGBoost. XGBoost had much higher net benefits within a wider threshold range in terms of decision curve analysis, and more capable of recognizing high risk individuals in terms of predictiveness curve analysis. A total of eight predictors including gender, waist-to-height ratio, dyslipidemia, glycated hemoglobin, white blood cell count, blood glucose, triglycerides, and low-density lipoprotein cholesterol ranked top 5 in three prediction models.Conclusions: Machine learning methods, especially for XGBoost, had the potential to predict stroke occurrence compared with traditional logistic regression in the older adults.

Sign in / Sign up

Export Citation Format

Share Document