scholarly journals Machine learning approaches improve risk stratification for secondary cardiovascular disease prevention in multiethnic patients

Open Heart ◽  
2021 ◽  
Vol 8 (2) ◽  
pp. e001802
Author(s):  
Ashish Sarraju ◽  
Andrew Ward ◽  
Sukyung Chung ◽  
Jiang Li ◽  
David Scheinker ◽  
...  
Keyword(s):  
Machine Learning ◽  
Cardiovascular Disease ◽  
Risk Stratification ◽  
Secondary Prevention ◽  
Risk Score ◽  
Characteristic Curve ◽  
World Population ◽  
Learning Approaches ◽  
Cvd Risk ◽  
Improve Risk Stratification

ObjectivesIdentifying high-risk patients is crucial for effective cardiovascular disease (CVD) prevention. It is not known whether electronic health record (EHR)-based machine-learning (ML) models can improve CVD risk stratification compared with a secondary prevention risk score developed from randomised clinical trials (Thrombolysis in Myocardial Infarction Risk Score for Secondary Prevention, TRS 2°P).MethodsWe identified patients with CVD in a large health system, including atherosclerotic CVD (ASCVD), split into 80% training and 20% test sets. A rich set of EHR patient features was extracted. ML models were trained to estimate 5-year CVD event risk (random forests (RF), gradient-boosted machines (GBM), extreme gradient-boosted models (XGBoost), logistic regression with an L2 penalty and L1 penalty (Lasso)). ML models and TRS 2°P were evaluated by the area under the receiver operating characteristic curve (AUC).ResultsThe cohort included 32 192 patients (median age 74 years, with 46% female, 63% non-Hispanic white and 12% Asian patients and 23 475 patients with ASCVD). There were 4010 events over 5 years of follow-up. ML models demonstrated good overall performance; XGBoost demonstrated AUC 0.70 (95% CI 0.68 to 0.71) in the full CVD cohort and AUC 0.71 (95% CI 0.69 to 0.73) in patients with ASCVD, with comparable performance by GBM, RF and Lasso. TRS 2°P performed poorly in all CVD (AUC 0.51, 95% CI 0.50 to 0.53) and ASCVD (AUC 0.50, 95% CI 0.48 to 0.52) patients. ML identified nontraditional predictive variables including education level and primary care visits.ConclusionsIn a multiethnic real-world population, EHR-based ML approaches significantly improved CVD risk stratification for secondary prevention.

Abstract P352: Is Brain Natriuretic Peptide or Pro- Brain Natriuretic Peptide of value for Early Cardiovascular Disease Risk Stratification in asymptomatic Subjects?

Circulation ◽  
2014 ◽  
Vol 129 (suppl_1) ◽  
Author(s):  
Mahfouz El Shahawy ◽  
Omar Y El Shahawy ◽  
Miglena O Entcheva
Keyword(s):  
Cardiovascular Disease ◽  
Risk Stratification ◽  
Brain Natriuretic Peptide ◽  
Natriuretic Peptide ◽  
Risk Score ◽  
Left Ventricular ◽  
Structural Abnormality ◽  
Close Monitoring ◽  
Cvd Risk ◽  
Asymptomatic Subjects

Introduction: There has been much discussion in recent years about the use of Brain natriuretic peptide (BNP) or Pro-brain natriuretic peptide (Pro-BNP) for cardiovascular disease (CVD) risk stratification. The purpose of this study is to assess the association between the presence of abnormal BNP/pro BNP with Left ventricular Hypertrophy (LVH) in asymptomatic subjects with or without hypertension and their value for early CVD risk stratification. Methods: We evaluated 2230 subjects aged 23-80 years, who underwent screening for CVD risk using Early CVD Risk Score (ECVDRS), also known as Rasmussen Risk Score (RRS). ECVDRS consists of 10 non-invasive tests: large (C1) and small (C2) artery stiffness, blood pressure (BP) at rest and post mild exercise (BP PME), Carotid Intima Media Thickness (CIMT), abdominal aorta ultrasound, retinal photography, microalbuminuria, electrocardiogram (ECG), left ventricular ultrasound, and BNP/Pro-BNP. Results: Among the 2230 asymptomatic participants, we analyzed 933 who were treated by at least one or more cardiovascular medication for Diabetes, Hypertension (HTN) or Hyperlipidemia. Our hypothesis was that asymptomatic subjects with LVH have a significantly higher BNP/ProBNP values than those without LVH independent of HTN management. We used 3 Chi-square tests to test the independence between the two groups (with or without LVH) regarding their BNP/ProBNP status for the whole sample and further stratified by their HTN control. The findings in three tests were highly significant; the proportion of abnormal LVH was 0.17 and 0.39 respectively with normal vs elevated BNP/ProBNP for the whole sample, χ 2 (1, N = 933) = 36.693, p < 0.001. When we stratified by HTN control, with uncontrolled HTN the proportion of abnormal LVH was 0.25 and 0.46 respectively with the normal vs elevated BNP/ProBNP, χ 2 (1, N = 271) = 11.983, p =0.0005. Among participants with controlled HTN the proportion of abnormal LVH was 0.15 and 0.33 respectively with normal vs elevated BNP/ProBNP, χ 2 (1, N = 662) = 18.257, p < 0.001. Conclusions: Abnormal BNP/pro-BNP levels are significantly associated with abnormal LVH findings in asymptomatic subjects independent of their HTN control. Thus, ProBNP/BNP could be considered an important follow up test for early detection of serious CV structural abnormality like LVH during the management of patients with HTN regardless of the degree of their HTN control. This finding may justify a more aggressive treatment approach and close monitoring with BNP/ProBNP as a possible screener for CVD manifestations like LVH in HTN patients.

scholarly journals Dynamic Detection of Delayed Cerebral Ischemia Using Machine Learning

2020 ◽  
Author(s):  
Murad Megjhani ◽  
Kalijah Terilli ◽  
Ayham Alkhachroum ◽  
David J. Roh ◽  
Sachin Agarwal ◽  
...  
Keyword(s):  
Machine Learning ◽  
Cerebral Ischemia ◽  
Characteristic Curve ◽  
Delayed Cerebral Ischemia ◽  
Risk Scores ◽  
Support Vector ◽  
Model Parameters ◽  
Learning Approaches ◽  
Physiologic Data ◽  
Over Time

AbstractObjectiveTo develop a machine learning based tool, using routine vital signs, to assess delayed cerebral ischemia (DCI) risk over time.MethodsIn this retrospective analysis, physiologic data for 540 consecutive acute subarachnoid hemorrhage patients were collected and annotated as part of a prospective observational cohort study between May 2006 and December 2014. Patients were excluded if (i) no physiologic data was available, (ii) they expired prior to the DCI onset window (< post bleed day 3) or (iii) early angiographic vasospasm was detected on admitting angiogram. DCI was prospectively labeled by consensus of treating physicians. Occurrence of DCI was classified using various machine learning approaches including logistic regression, random forest, support vector machine (linear and kernel), and an ensemble classifier, trained on vitals and subject characteristic features. Hourly risk scores were generated as the posterior probability at time t. We performed five-fold nested cross validation to tune the model parameters and to report the accuracy. All classifiers were evaluated for good discrimination using the area under the receiver operating characteristic curve (AU-ROC) and confusion matrices.ResultsOf 310 patients included in our final analysis, 101 (32.6%) patients developed DCI. We achieved maximal classification of 0.81 [0.75-0.82] AU-ROC. We also predicted 74.7 % of all DCI events 12 hours before typical clinical detection with a ratio of 3 true alerts for every 2 false alerts.ConclusionA data-driven machine learning based detection tool offered hourly assessments of DCI risk and incorporated new physiologic information over time.

scholarly journals Machine learning approaches revealed metabolic signatures of incident chronic kidney disease in persons with pre-and type 2 diabetes

2020 ◽  
Author(s):  
Ada Admin ◽  
Jialing Huang ◽  
Cornelia Huth ◽  
Marcela Covic ◽  
Martina Troll ◽  
...  
Keyword(s):  
Machine Learning ◽  
Type 2 Diabetes ◽  
Chronic Kidney Disease ◽  
Kidney Disease ◽  
Characteristic Curve ◽  
Predictive Performance ◽  
Population Based ◽  
Learning Approaches ◽  
Clinical Variables

Early and precise identification of individuals with pre-diabetes and type 2 diabetes (T2D) at risk of progressing to chronic kidney disease (CKD) is essential to prevent complications of diabetes. Here, we identify and evaluate prospective metabolite biomarkers and the best set of predictors of CKD in the longitudinal, population-based Cooperative Health Research in the Region of Augsburg (KORA) cohort by targeted metabolomics and machine learning approaches. Out of 125 targeted metabolites, sphingomyelin (SM) C18:1 and phosphatidylcholine diacyl (PC aa) C38:0 were identified as candidate metabolite biomarkers of incident CKD specifically in hyperglycemic individuals followed during 6.5 years. Sets of predictors for incident CKD developed from 125 metabolites and 14 clinical variables showed highly stable performances in all three machine learning approaches and outperformed the currently established clinical algorithm for CKD. The two metabolites in combination with five clinical variables were identified as the best set of predictors and their predictive performance yielded a mean area value under the receiver operating characteristic curve of 0.857. The inclusion of metabolite variables in the clinical prediction of future CKD may thus improve the risk prediction in persons with pre- and T2D. The metabolite link with hyperglycemia-related early kidney dysfunction warrants further investigation.

Cardiovascular Disease Prediction from Electrocardiogram by using Machine Learning Method: A Snapshot from the Subjects of the Malaysian Cohort

2020 ◽  
Author(s):  
Nazrul Anuar Nayan ◽  
Hafifah Ab Hamid ◽  
Mohd Zubir Suboh ◽  
Noraidatulakma Abdullah ◽  
Rosmina Jaafar ◽  
...  
Keyword(s):  
Machine Learning ◽  
Cardiovascular Disease ◽  
Nearest Neighbor ◽  
Total Cholesterol Level ◽  
Population Based ◽  
Support Vector ◽  
K Nearest Neighbor ◽  
Cvd Risk ◽  
Linear Discriminant ◽  
Artificial Neural Network Ann

Abstract Background: Cardiovascular disease (CVD) is the leading cause of deaths worldwide. In 2017, CVD contributed to 13,503 deaths in Malaysia. The current approaches for CVD prediction are usually invasive and costly. Machine learning (ML) techniques allow an accurate prediction by utilizing the complex interactions among relevant risk factors. Results: This study presents a case–control study involving 60 participants from The Malaysian Cohort, which is a prospective population-based project. Five parameters, namely, the R–R interval and root mean square of successive differences extracted from electrocardiogram (ECG), systolic and diastolic blood pressures, and total cholesterol level, were statistically significant in predicting CVD. Six ML algorithms, namely, linear discriminant analysis, linear and quadratic support vector machines, decision tree, k-nearest neighbor, and artificial neural network (ANN), were evaluated to determine the most accurate classifier in predicting CVD risk. ANN, which achieved 90% specificity, 90% sensitivity, and 90% accuracy, demonstrated the highest prediction performance among the six algorithms. Conclusions: In summary, by utilizing ML techniques, ECG data can serve as a good parameter for CVD prediction among the Malaysian multiethnic population.

scholarly journals Antidepressant Use and the Risk of Major Adverse Cardiovascular Events in Patients Without Known Cardiovascular Disease: A Retrospective Cohort Study

2020 ◽  
Vol 11 ◽  
Author(s):  
Ha Young Jang ◽  
Jae Hyun Kim ◽  
Yun-Kyoung Song ◽  
Ju-Young Shin ◽  
Hae-Young Lee ◽  
...  
Keyword(s):  
Cardiovascular Disease ◽  
Risk Score ◽  
Cardiovascular Events ◽  
Major Adverse Cardiovascular Events ◽  
Cvd Risk ◽  
Hazard Ratios ◽  
Increased Risk ◽  
History Of ◽  
Antidepressant Use ◽  
Cox Proportional Hazard Regression

Aims: Conflicting data exist on whether an association exists between antidepressants and the risk of major adverse cardiovascular events (MACEs) in patients with depression. This may be due to the use of various study designs and residual or unmeasured confounding. We aimed to assess the association between antidepressant use and the risk of MACEs while considering various covariates, including severity of depression and the cardiovascular disease (CVD) risk score.Methods: Patients newly diagnosed with depression with no history of ischemic heart disease and stroke were followed-up from 2009 to 2015. We conducted Cox proportional hazard regression analysis to estimate hazard ratios (HRs) for each antidepressant for MACE risk.Result: We followed-up (median, 4.4 years) 31,830 matched patients with depression (15,915 antidepressant users and 15,915 non-users). In most patients (98.7%), low-dose tricyclic antidepressants (TCAs) were related with a significantly increased risk of MACEs [adjusted HR = 1.20, 95% confidence interval (CI) = 1.03–1.40]. Duration response relationship showed a gradually increasing HR from 1.15 (95% CI = 0.98–1.33; &lt;30 days of use) to 1.84 (95% CI = 1.35–2.51; ≥365 days of use) (p for trend &lt;0.01). High Korean atherosclerotic CVD risk score (≥7.5%) or unfavorable lifestyle factors (smoking, alcohol intake, and exercise) were significantly associated with MACEs.Conclusion: Even at low doses, TCA use was associated with MACEs during primary prevention. Longer duration of TCA use correlated with higher HR. Careful monitoring is needed with TCA use in patients with no known CVD history.

Risk results from screening for a high cardiovascular disease risk by means of traditional risk factor measurement or coronary artery calcium scoring in the ROBINSCA trial

2020 ◽  
Vol 41 (Supplement_2) ◽  
Author(s):  
C Van Der Aalst ◽  
S.J.A.M Denissen ◽  
M Vonder ◽  
J.-W.C Gratema ◽  
H.J Adriaansen ◽  
...  
Keyword(s):  
Cardiovascular Disease ◽  
Coronary Artery ◽  
High Risk ◽  
Risk Score ◽  
Coronary Artery Calcium ◽  
Preventive Treatment ◽  
Low Risk ◽  
Funding Source ◽  
Cvd Risk ◽  
Increased Risk

Abstract Aims Screening for a high cardiovascular disease (CVD) risk followed by preventive treatment can potentially reduce coronary heart disease (CHD)-related morbidity and mortality. ROBINSCA (Risk Or Benefit IN Screening for CArdiovascular disease) is a population-based randomized controlled screening trial that investigates the effectiveness of CVD screening in asymptomatic participants using the Systematic COronary Risk Evaluation (SCORE) model or Coronary Artery Calcium (CAC) scoring. This study describes the distributions in risk and treatment in the ROBINSCA trial. Methods and results Individuals at expected elevated CVD risk were randomized (1:1:1) into the control arm (n=14,519; usual care); screening arm A (n=14,478; SCORE, 10-year fatal and non-fatal risk); or screening arm B (n=14,450; CAC scoring). Preventive treatment was largely advised according to current Dutch guidelines. Risk and treatment differences between the screening arms were analysed. 12,185 participants (84.2%) in arm A and 12,950 (89.6%) in arm B were screened. 48.7% were women, and median age was 62 (InterQuartile Range 10) years. SCORE screening identified 45.1% at low risk (SCORE&lt;10%), 26.5% at intermediate risk (SCORE 10–20%), and 28.4% at high risk (SCORE≥20%). According to CAC screening, 76.0% were at low risk (Agatston&lt;100), 15.1% at high risk (Agatston 100–399), and 8.9% at very high risk (Agatston≥400). CAC scoring significantly reduced the number of individuals indicated for preventive treatment compared to SCORE (relative reduction women: 37.2%; men: 28.8%). Conclusion We showed that compared to risk stratification based on SCORE, CAC scoring classified significantly fewer men and women at increased risk, and less preventive treatment was indicated. ROBINSCA flowchart Funding Acknowledgement Type of funding source: Public grant(s) – EU funding. Main funding source(s): Advanced Research Grant

scholarly journals Cardiovascular Disease Prediction from Electrocardiogram by Using Machine Learning

2020 ◽  
Vol 16 (07) ◽  
pp. 34
Author(s):  
Nayan Nazrul Anuar ◽  
Ab Hamid Hafifah ◽  
Suboh Mohd Zubir ◽  
Abdullah Noraidatulakma ◽  
Jaafar Rosmina ◽  
...  
Keyword(s):  
Machine Learning ◽  
Cardiovascular Disease ◽  
Nearest Neighbor ◽  
Total Cholesterol Level ◽  
Population Based ◽  
Support Vector ◽  
K Nearest Neighbor ◽  
Cvd Risk ◽  
Linear Discriminant ◽  
Artificial Neural Network Ann

<p>Cardiovascular disease (CVD) is the leading cause of deaths worldwide. In 2017, CVD contributed to 13,503 deaths in Malaysia. The current approaches for CVD prediction are usually invasive and costly. Machine learning (ML) techniques allow an accurate prediction by utilizing the complex interactions among relevant risk factors. This study presents a case–control study involving 60 participants from The Malaysian Cohort, which is a prospective population-based project. Five parameters, namely, the R–R interval and root mean square of successive differences extracted from electrocardiogram (ECG), systolic and diastolic blood pressures, and total cholesterol level, were statistically significant in predicting CVD. Six ML algorithms, namely, linear discriminant analysis, linear and quadratic support vector machines, decision tree, k-nearest neighbor, and artificial neural network (ANN), were evaluated to determine the most accurate classifier in predicting CVD risk. ANN, which achieved 90% specificity, 90% sensitivity, and 90% accuracy, demonstrated the highest prediction performance among the six algorithms. In summary, by utilizing ML techniques, ECG data can serve as a good parameter for CVD prediction among the Malaysian multiethnic population.</p>

scholarly journals Assessment of medium-term cardiovascular disease risk after Japan’s 2011 Fukushima Daiichi nuclear accident: a retrospective analysis

BMJ Open ◽  
2017 ◽  
Vol 7 (12) ◽  
pp. e018502 ◽  
Author(s):  
Haruka Toda ◽  
Shuhei Nomura ◽  
Stuart Gilmour ◽  
Masaharu Tsubokura ◽  
Tomoyoshi Oikawa ◽  
...  
Keyword(s):  
Risk Factors ◽  
Cardiovascular Disease ◽  
Risk Score ◽  
Disease Risk ◽  
Cardiovascular Disease Risk ◽  
Multivariate Regression Analysis ◽  
Nuclear Accident ◽  
Medium Term ◽  
Cvd Risk ◽  
Fukushima Daiichi

ObjectiveTo assess the medium-term indirect impact of the 2011 Fukushima Daiichi nuclear accident on cardiovascular disease (CVD) risks and to identify whether risk factors for CVD changed after the accident.ParticipantsResidents aged 40 years and over participating in annual public health check-ups from 2009 to 2012, administered by Minamisoma city, located about 10 to 40 km from the Fukushima Daiichi nuclear plant.MethodsThe sex-specific Framingham CVD risk score was considered as the outcome measure and was compared before (2009–2010) and after the accident (2011–2012). A multivariate regression analysis was employed to evaluate risk factors for CVD.ResultsData from 563 individuals (60.2% women) aged 40 to 74 years who participated in the check-ups throughout the study period was analysed. After adjusting for covariates, no statistically significant change was identified in the CVD risk score postaccident in both sexes, which may suggest no obvious medium-term health impact of the Fukushima nuclear accident on CVD risk. The risk factors for CVD and their magnitude and direction (positive/negative) did not change after the accident.ConclusionsThere was no obvious increase in CVD risks in Minamisoma city, which may indicate successful management of health risks associated with CVD in the study sample.

Sign in / Sign up

Export Citation Format

Share Document