Materiały źródłowe

• #1 Use of machine learning to identify risk factors for coronary artery disease | PLOS One

  https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0284103

  Coronary artery disease (CAD) is the leading cause of death in both developed and developing nations. The objective of this study was to identify risk factors for coronary artery disease through machine-learning and assess this methodology. Univariate logistic models, with CAD as the outcome, were used to identify covariates that were associated with CAD. The machine learning model XGBoost was used due to its prevalence within the literature as well as its increased predictive accuracy in healthcare prediction. Machine learning models can effectively predict coronary artery disease using demographic, laboratory, physical exam, and lifestyle covariates and identify key risk factors. The greatest predictors for coronary artery disease included age, total cholesterol, total platelets, and family history of a heart attack. Machine learning models can effectively predict coronary artery disease using demographic, laboratory, physical exam, and lifestyle covariates. Age, total cholesterol, total platelets, and family history of heart attack are the strongest predictors of coronary artery disease.

• #1 Prediction of mortality from coronary heart disease among diverse populations: is there a common predictive function? | Heart

  https://heart.bmj.com/content/88/3/222

  Objectives: To examine the generalisability of multivariate risk functions from diverse populations in three contexts: ordering risk, magnitude of relative risks, and estimation of absolute risk. […] Main outcome measure: Death from coronary heart disease (CHD). […] Results: The analysis included 105 420 men and 56 535 women 3574 years of age and free of CHD at baseline from 16 observational studies with a total of 27 analytical groups. […] The area under the receiver operating characteristic curve (AUC) was used to judge the ability of the multivariate risk function to order risk correctly. […] The prediction of absolute risk was not very accurate in most of the cases when a model derived from one study was applied to a different study. […] Conclusions: When considered qualitatively, the major risk factors are associated with CHD mortality in a diverse set of populations. However, when considered quantitatively, there was significant heterogeneity in all three aspects: ordering risk, magnitude of relative risks, and estimation of absolute risk.

• #1 Prediction of mortality from coronary heart disease among diverse populations: is there a common predictive function?

  https://pmc.ncbi.nlm.nih.gov/articles/PMC1767330/

  Objectives: To examine the generalisability of multivariate risk functions from diverse populations in three contexts: ordering risk, magnitude of relative risks, and estimation of absolute risk. […] Main outcome measure: Death from coronary heart disease (CHD). […] The prediction of absolute risk was not very accurate in most of the cases when a model derived from one study was applied to a different study. […] When considered qualitatively, the major risk factors are associated with CHD mortality in a diverse set of populations. However, when considered quantitatively, there was significant heterogeneity in all three aspects: ordering risk, magnitude of relative risks, and estimation of absolute risk. […] The absolute risk was overestimated when applying the northern European model to southern European populations and vice versa.

• #1 Cardiovascular Event Prediction by Machine Learning: The Multi-Ethnic Study of Atherosclerosis

  https://pmc.ncbi.nlm.nih.gov/articles/PMC5640485/

  Machine learning may be useful to characterize cardiovascular risk, predict outcomes and identify biomarkers in population studies. […] To test the ability of random survival forests (RF), a machine learning technique, to predict six cardiovascular outcomes in comparison to standard cardiovascular risk scores. […] Age was the most important predictor for all-cause mortality. Fasting glucose levels and carotid ultrasonography measures were important predictors of stroke. Coronary artery calcium score was the most important predictor of coronary heart disease and all atherosclerotic cardiovascular disease combined outcomes. […] Machine learning in conjunction with deep phenotyping improve prediction accuracy in cardiovascular event prediction in an initially asymptomatic population. […] The RF technique performed better than established risk scores with increased prediction accuracy (decreased Brier score by 1025%).

• #1 Integrative proteomic analyses across common cardiac diseases yield mechanistic insights and enhanced prediction | Nature Cardiovascular Research

  https://www.nature.com/articles/s44161-024-00567-0

  The proteins most strongly associated with incident coronary artery disease were growth differentiation factor 15 (GDF15; HR 1.31) and matrix metalloproteinase-12 (MMP12; HR 1.29). […] To gain insights into biological pathways associated with the identified proteins, we carried the 820 observed proteindisease associations forward for pathway enrichment analysis using the Gene Ontology resource. […] Genetic analyses suggested causal or mediating roles either protective or deleterious for a substantial proportion of biomarkers identified in observational analyses. […] Finally, we constructed sparse protein-based risk scores that improved the prediction of cardiac disease development in the general population.

• #1 Adverse Outcome Prediction of Iron Deficiency in Patients with Acute Coronary Syndrome

  https://www.mdpi.com/2218-273X/8/3/60

  Iron deficiency strongly predicted non-fatal MI and cardiovascular mortality with a hazard ratio (HR) of 1.52 (95% confidence interval (CI) 1.03-2.26; p = 0.037) adjusted for age, sex, hypertension, smoking status, diabetes, hyperlipidemia, body-mass-index (BMI) […] This association remained significant (HR 1.73 (95% CI 1.07–2.81; p = 0.026)) after an additional adjustment for surrogates of cardiac function and heart failure severity (N-terminal pro B-type natriuretic peptide, NT-proBNP), for the size of myocardial necrosis (troponin), and for anemia (hemoglobin). […] Our data showed that iron deficiency is strongly associated with adverse outcome in acute coronary syndrome.

• #1 Risk prediction for coronary heart disease by a genetic risk score – results from the Heinz Nixdorf Recall study | BMC Medical Genetics | Full Text

  https://bmcmedgenet.biomedcentral.com/articles/10.1186/s12881-020-01113-y

  A Genetic risk score for coronary artery disease (CAD) improves the ability of predicting coronary heart disease (CHD). It is unclear whether i) the use of a CAD genetic risk score is superior to the measurement of coronary artery calcification (CAC) for CHD risk assessment and ii) the CHD risk assessment using a CAD genetic risk score differs between men and women. […] Our results suggest that the CAD genetic risk score could be useful for CHD risk prediction, at least in men belonging to the higher genetic risk group, but it does not outbalance the value of CT-based quantification of CAC which works independently on both men and women and allows better risk stratification in both the genders. […] The important findings of our study are i) the CAD genetic risk score was associated with incident CHD, showing stronger effect in men with no effect observed in women, ii) the effect of genetic risk score did not alter even after adjusting for CAC, iii) the CAD genetic risk score was associated with incident CHD only in the group with presence of CAC, showing stronger effect in men and no effect in women and iv) CAD genetic risk score was associated with CAC with stronger effect observed in men.

• #1 Machine learning-based long-term outcome prediction in patients undergoing percutaneous coronary intervention – Liu – Cardiovascular Diagnosis and Therapy

  https://cdt.amegroups.org/article/view/69917/html

  Traditional prognostic risk assessment in patients with coronary artery disease undergoing percutaneous coronary intervention (PCI) is based on a limited selection of clinical and imaging findings. […] Machine learning (ML) models improved the prediction of long-term all-cause mortality in patients with coronary artery disease before PCI. […] The performance of the RF model was better than that of the other models, providing a meaningful stratification of long-term outcomes in coronary patients following PCI.

• #1 Predictive model and risk analysis for coronary heart disease in people living with HIV using machine learning | BMC Medical Informatics and Decision Making | Full Text

  https://bmcmedinformdecismak.biomedcentral.com/articles/10.1186/s12911-024-02511-5

  This study aimed to construct a coronary heart disease (CHD) risk-prediction model in people living with human immunodeficiency virus (PLHIV) with the help of machine learning (ML) per electronic medical records (EMRs). […] The LightGBM model exhibited the highest AUC (0.849; 95% CI, 0.8140.883). Additionally, the SHAP plot per the LightGBM depicted that age, heart failure, hypertension, glucose, serum creatinine, indirect bilirubin, serum uric acid, and amylase can help identify PLHIV who were at a high or low risk of developing CHD. […] This study developed a CHD risk prediction model for PLHIV utilizing ML techniques and EMR data. The LightGBM model exhibited improved comprehensive performance and thus had higher reliability in assessing the risk predictors of CHD. […] The mortality rate of myocardial infarction is 1.5 to 1.7 times that of the general population, and the average age of death is about 48 years old, which is far lower than that of the general population.

• #1 Development of a multidimensional prediction model for long-term prognostic risk in patients with acute coronary syndromes after percutaneous coronary intervention: A retrospective observational cohort study | PLOS One

  https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0318445

  The aim of this study is to examine the critical variables that impact the long-term prognosis of patients with acute coronary syndrome (ACS) after percutaneous coronary intervention (PCI) and to create a multidimensional predictive risk assessment model that can serve as a theoretical basis for accurate cardiac rehabilitation. […] We found white blood cell count (WBC) (OR: 4.110) and the effective average number of daily steps (ANS) (OR: 2.689) as independent prognostic risk factors for acute myocardial infarction (AMI). The independent risk factors for unstable angina prognosis were white blood cell count (OR: 6.257), VO2 at anaerobic threshold (OR: 4.294), and effective autonomic nervous system function (OR: 4.097). […] This study developed a multimodal predictive model that integrates the inflammatory response after onset, physical performance and exercise tolerance before discharge, and daily activity after discharge to predict the long-term prognosis of patients with ACS. The multidimensional model is more effective than the single-factor model for assessing risk in ACS patients.

• #1 Predictive model and risk analysis for coronary heart disease in people living with HIV using machine learning | BMC Medical Informatics and Decision Making | Full Text

  https://bmcmedinformdecismak.biomedcentral.com/articles/10.1186/s12911-024-02511-5

  The risk of CVD among PLHIV is 2.16 times than that in the general population. […] The mortality rate among people living with HIV (PLHIV) is 1.6 per 1000 people and has been observed to increase annually. […] Given that CVD has become the first cause of non-AIDS death among PLHIV, the management of CVD in PLHIV should be given focus to control the death rate. […] In PLHIV, the risk factors are similar but more severe than traditional CVD. […] Even after controlling for traditional CVD risk factors, people with HIV are twice as likely to develop CVD as those without HIV. […] Current risk factors fail to assist predicting potential CVD risk in PLHIV, so a more accurate predictive indicator is well positioned to be discovered. […] The LightGBM model demonstrated superior performance, achieving the highest AUC of 0.849 (95% CI, 0.8140.883), whereas the decision tree exhibited the lowest AUC one (0.753; 95% CI, 0.7040.803).

• #1 Integrative proteomic analyses across common cardiac diseases yield mechanistic insights and enhanced prediction | Nature Cardiovascular Research

  https://www.nature.com/articles/s44161-024-00567-0

  Cardiac diseases represent common highly morbid conditions for which molecular mechanisms remain incompletely understood. […] Models incorporating proteomic data (versus clinical risk factors alone) improved prediction for coronary artery disease, heart failure and atrial fibrillation. […] The present study tested the associations of circulating proteins with common cardiac diseases (coronary artery disease, heart failure, atrial fibrillation and aortic stenosis) in the UKB-PPP. […] Primary analyses identified 820 proteindisease associations reflecting 441 unique proteins at Bonferroni-corrected P<0.05. [...] The strongest proteindisease associations (by P value) were observed for atrial fibrillation, with N-terminal pro-B-type natriuretic peptide (NT-proBNP) and B-type natriuretic peptide (NPPB, also known as BNP) yielding hazard ratios (HRs) of 1.74 and 1.62, respectively, for each s.d. increase in circulating protein levels.

• #1 An active learning machine technique based prediction of cardiovascular heart disease from UCI-repository database | Scientific Reports

  https://www.nature.com/articles/s41598-023-40717-1

  Heart disease is a significant global cause of mortality, and predicting it through clinical data analysis poses challenges. […] Machine learning (ML) has emerged as a valuable tool for diagnosing and predicting heart disease by analyzing healthcare data. […] The motivation behind predicting Cardiovascular Heart Disease lies in its potential to save lives, improves health outcomes, and allocates healthcare resources efficiently. […] The key contributions encompass early intervention, personalized medicine, technological advancements, the impact on public health, and ongoing research, all of which collectively work toward reducing the burden of CHD on both individual patients and society as a whole. […] Heart disease diagnostics aren’t as good as they could be, and there is a huge need for better big-data analysis in cardiovascular system redesign and patient outcomes.

• #2 An active learning machine technique based prediction of cardiovascular heart disease from UCI-repository database | Scientific Reports

  https://www.nature.com/articles/s41598-023-40717-1

  Heart disease is a significant global cause of mortality, and predicting it through clinical data analysis poses challenges. […] Machine learning (ML) has emerged as a valuable tool for diagnosing and predicting heart disease by analyzing healthcare data. […] The motivation behind predicting Cardiovascular Heart Disease lies in its potential to save lives, improves health outcomes, and allocates healthcare resources efficiently. […] The key contributions encompass early intervention, personalized medicine, technological advancements, the impact on public health, and ongoing research, all of which collectively work toward reducing the burden of CHD on both individual patients and society as a whole. […] Heart disease diagnostics aren’t as good as they could be, and there is a huge need for better big-data analysis in cardiovascular system redesign and patient outcomes.

• #2 Prediction of mortality from coronary heart disease among diverse populations: is there a common predictive function?

  https://pmc.ncbi.nlm.nih.gov/articles/PMC1767330/

  Objectives: To examine the generalisability of multivariate risk functions from diverse populations in three contexts: ordering risk, magnitude of relative risks, and estimation of absolute risk. […] Main outcome measure: Death from coronary heart disease (CHD). […] The prediction of absolute risk was not very accurate in most of the cases when a model derived from one study was applied to a different study. […] When considered qualitatively, the major risk factors are associated with CHD mortality in a diverse set of populations. However, when considered quantitatively, there was significant heterogeneity in all three aspects: ordering risk, magnitude of relative risks, and estimation of absolute risk. […] The absolute risk was overestimated when applying the northern European model to southern European populations and vice versa.

• #2 Prediction of mortality from coronary heart disease among diverse populations: is there a common predictive function?

  https://pmc.ncbi.nlm.nih.gov/articles/PMC1767330/

  The Framingham predictive function for white middle aged men overestimated absolute CHD risks in Japanese American and Hispanic men and Native American women, in France, in Sweden, and in Italy. […] We showed that there was significant variation between the studies in all aspects of multivariate risk.

• #2 Risk prediction for coronary heart disease by a genetic risk score – results from the Heinz Nixdorf Recall study | BMC Medical Genetics | Full Text

  https://bmcmedgenet.biomedcentral.com/articles/10.1186/s12881-020-01113-y

  A Genetic risk score for coronary artery disease (CAD) improves the ability of predicting coronary heart disease (CHD). It is unclear whether i) the use of a CAD genetic risk score is superior to the measurement of coronary artery calcification (CAC) for CHD risk assessment and ii) the CHD risk assessment using a CAD genetic risk score differs between men and women. […] Our results suggest that the CAD genetic risk score could be useful for CHD risk prediction, at least in men belonging to the higher genetic risk group, but it does not outbalance the value of CT-based quantification of CAC which works independently on both men and women and allows better risk stratification in both the genders. […] The important findings of our study are i) the CAD genetic risk score was associated with incident CHD, showing stronger effect in men with no effect observed in women, ii) the effect of genetic risk score did not alter even after adjusting for CAC, iii) the CAD genetic risk score was associated with incident CHD only in the group with presence of CAC, showing stronger effect in men and no effect in women and iv) CAD genetic risk score was associated with CAC with stronger effect observed in men.

• #2 Cardiovascular Event Prediction by Machine Learning: The Multi-Ethnic Study of Atherosclerosis

  https://pmc.ncbi.nlm.nih.gov/articles/PMC5640485/

  A telephone interviewer contacted each participant (or representative) every six-nine months to inquire about all interim hospital admissions, outpatient diagnoses, and deaths. […] For incident HF as the endpoint, cardiac chamber stress (increased LV volume, and increased NT-proBNP levels), and decreased LV function from MRI were the most important markers. […] Increased fasting glucose levels were the most important risk factor for stroke, while high blood pressure, a known stroke risk, and age also featured in the top-20 list. […] Expectedly, a composite of atherosclerosis measures (low and high ABI, increased carotid IMT, decreased aortic distensibility) were among the most important predictors of CHD which represents a subset of CVD events, with CAC being by far the most important, reflecting the specific influence of coronary atherosclerosis. […] Inflammation, subclinical atherosclerosis, myocardial damage, and cardiac chamber stress were among the most important predictors across all outcomes.

• #2 Machine learning-based long-term outcome prediction in patients undergoing percutaneous coronary intervention – Liu – Cardiovascular Diagnosis and Therapy

  https://cdt.amegroups.org/article/view/69917/html

  Traditional prognostic risk assessment in patients with coronary artery disease undergoing percutaneous coronary intervention (PCI) is based on a limited selection of clinical and imaging findings. […] Machine learning (ML) models improved the prediction of long-term all-cause mortality in patients with coronary artery disease before PCI. […] The performance of the RF model was better than that of the other models, providing a meaningful stratification of long-term outcomes in coronary patients following PCI.

• #2 An active learning machine technique based prediction of cardiovascular heart disease from UCI-repository database | Scientific Reports

  https://www.nature.com/articles/s41598-023-40717-1

  The goal of predicting Cardiovascular Heart Disease is to develop accurate and reliable models that can assess an individual’s risk of developing various cardiovascular conditions, enabling early intervention, personalized treatment, and ultimately reducing the burden of heart disease on public health. […] The proposed method uses supervised ML classifiers to show how different models can predict the presence of cardiovascular disease and evaluate the performance of these classifiers, such as the random forest, decision tree, support vector machine, XGBoost, radial basis function, k-nearest neighbour, nave bayes and learning vector quantization. […] The final findings demonstrate that when the learning machine classifiers were put to use, the Naive Bayes and RBF neural networks achieved an accuracy of 94.78% when attempting to forecast the presence of coronary cardiovascular disease. However, the Learning Vector Quantization method achieved the highest categorization accuracy of 98.78%, with a specificity of 97.1% and sensitivity of 97.91%, a precision of 98.07% and 95.31%, and 97.89% F1score and F-measure values, respectively.

• #2 Use of machine learning to identify risk factors for coronary artery disease | PLOS One

  https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0284103

  In this retrospective, cross sectional cohort of United States adults, a machine learning model utilizing demographic, laboratory, physical examination, and lifestyle questionnaire data had strong predictive accuracy (AUROC = 0.89). The visualizations completed for the top four covariates were concordant with current literature around the relationship between these covariates and coronary artery disease: there is strong epidemiological and physiological evidence for the link between increased age and cholesterol as major risk factors for coronary artery disease.

• #2 Development of a multidimensional prediction model for long-term prognostic risk in patients with acute coronary syndromes after percutaneous coronary intervention: A retrospective observational cohort study | PLOS One

  https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0318445

  This study found that the inflammatory response after onset, physical performance and exercise tolerance before discharge, and daily activity after discharge were the independent risk factors for predicting the long-term prognosis of patients with ACS. The multidimensional prognostic model to risk-stratify for the patients with ACS, was better than the single factor model. This study also provides a theoretical basis that the prognosis of potentially high-risk patients can be improved by precise and rational exercise prescription.

• #2 Predictive model and risk analysis for coronary heart disease in people living with HIV using machine learning | BMC Medical Informatics and Decision Making | Full Text

  https://bmcmedinformdecismak.biomedcentral.com/articles/10.1186/s12911-024-02511-5

  The risk of CVD among PLHIV is 2.16 times than that in the general population. […] The mortality rate among people living with HIV (PLHIV) is 1.6 per 1000 people and has been observed to increase annually. […] Given that CVD has become the first cause of non-AIDS death among PLHIV, the management of CVD in PLHIV should be given focus to control the death rate. […] In PLHIV, the risk factors are similar but more severe than traditional CVD. […] Even after controlling for traditional CVD risk factors, people with HIV are twice as likely to develop CVD as those without HIV. […] Current risk factors fail to assist predicting potential CVD risk in PLHIV, so a more accurate predictive indicator is well positioned to be discovered. […] The LightGBM model demonstrated superior performance, achieving the highest AUC of 0.849 (95% CI, 0.8140.883), whereas the decision tree exhibited the lowest AUC one (0.753; 95% CI, 0.7040.803).

Zobacz więcej