Machine learning prediction in cardiovascular diseases: a meta-analysis

Interpretable machine learning models based on shear-wave elastography radiomics for predicting cardiovascular disease in diabetic kidney disease patients

BACKGROUND

The risk of cardiovascular complications is significantly elevated in patients with diabetic kidney disease (DKD). Recognizing the link between the progression of DKD and an increased risk of cardiovascular disease (CVD), it is crucial to focus on the early prediction and management of CVD risk factors among these patients to potentially enhance their health outcomes.

OBJECTIVE

This study sought to bridge the existing gap by developing and validating machine learning (ML) models that utilize clinical data and shear wave elastography (SWE) radiomics features to identify patients at risk of CVD, ultimately aiming to improve the management of DKD.

MATERIALS AND METHODS

This study conducted a retrospective analysis of 586 patients with DKD, dividing them into training and external validation cohorts. We categorized patients based on the presence or absence of CVD. Utilizing SWE imaging, we extracted and standardized radiomics features to develop multiple ML models. These models underwent internal validation using radiomics features alone, clinical data, or a combination thereof. The optimal model was then identified, and its feature importance was assessed through the Shapley Additive Explanations (SHAP) method, before proceeding to external validation.

RESULTS

Among the 586 patients analyzed, 30.7% (180/586) were identified as at risk for CVD. The study pinpointed six significant radiomics features related to CVD, alongside six critical pieces of clinical data. The Support Vector Machine (SVM) model outperformed others in both internal and external validations. Further, SHAP analysis highlighted five principal determinants of CVD risk, comprising three clinical indicators and two SWE radiomics features.

CONCLUSIONS

This study highlights the effectiveness of an SVM model that combines clinical and radiomics features in predicting CVD risk among DKD patients. It enables early prediction of CVD in this patient group, thereby supporting the implementation of timely and suitable interventions.

Machine Learning and Omics Analysis in Aortic Aneurysm

Aortic aneurysm is a life-threatening condition and mechanisms underlying its formation and progression are still incompletely understood. Omics approach has brought new insights to identify a broad spectrum of biomarkers and better understand cellular and molecular pathways involved. Omics generate a large amount of data and several studies have highlighted that artificial intelligence (AI) and techniques such as machine learning (ML)/deep learning (DL) can be of use in analyzing such complex datasets. However, only a few studies have so far reported the use of ML/DL for omics analysis in aortic aneurysms. The aim of this study is to summarize recent advances on the use of ML/DL for omics analysis to decipher aortic aneurysm pathophysiology and develop patient-tailored risk prediction models. In the light of current knowledge, we discuss current limits and highlight future directions in the field.

Artificial Intelligence Algorithms in Cardiovascular Medicine: An Attainable Promise to Improve Patient Outcomes or an Inaccessible Investment?

PURPOSE OF REVIEW

This opinion paper highlights the advancements in artificial intelligence (AI) technology for cardiovascular disease (CVD), presents best practices and transformative impacts, and addresses current concerns that must be resolved for broader adoption.

RECENT FINDINGS

With the evolution of digitization in data collection, large amounts of data have become available, surpassing the human capacity for processing and analysis, thus enabling the application of AI. These models can learn complex spatial and temporal patterns from large amounts of data, providing patient-specific outputs. These advantages have resulted, at the moment, in more than 900 AI-based devices being approved, today, by regulatory entities, for clinical use, with similar to improved performance and efficiency compared to traditional technologies. However, issues such as model generalization, bias, transparency, interpretability, accountability, and data privacy remain significant barriers for broad adoption of these technologies. AI shows great promise in enhancing CVD care through more accurate and efficient approaches. Yet, widespread adoption is hindered by unresolved concerns of interested stakeholders. Addressing these challenges is crucial for fully integrating AI into clinical practice and shaping the future of CVD prevention, diagnosis and treatment.

Development and validation of a cardiovascular risk prediction model for Sri Lankans using machine learning

INTRODUCTION AND OBJECTIVES

Sri Lankans do not have a specific cardiovascular (CV) risk prediction model and therefore, World Health Organization(WHO) risk charts developed for the Southeast Asia Region are being used. We aimed to develop a CV risk prediction model specific for Sri Lankans using machine learning (ML) of data of a population-based, randomly selected cohort of Sri Lankans followed up for 10 years and to validate it in an external cohort.

MATERIAL AND METHODS

The cohort consisted of 2596 individuals between 40-65 years of age in 2007, who were followed up for 10 years. Of them, 179 developed hard CV diseases (CVD) by 2017. We developed three CV risk prediction models named model 1, 2 and 3 using ML. We compared predictive performances between models and the WHO risk charts using receiver operating characteristic curves (ROC). The most predictive and practical model for use in primary care, model 3 was named "SLCVD score" which used age, sex, smoking status, systolic blood pressure, history of diabetes, and total cholesterol level in the calculation. We developed an online platform to calculate the SLCVD score. Predictions of SLCVD score were validated in an external hospital-based cohort.

RESULTS

Model 1, 2, SLCVD score and the WHO risk charts predicted 173, 162, 169 and 10 of 179 observed events and the area under the ROC (AUC) were 0.98, 0.98, 0.98 and 0.52 respectively. During external validation, the SLCVD score and WHO risk charts predicted 56 and 18 respectively of 119 total events and AUCs were 0.64 and 0.54 respectively.

CONCLUSIONS

SLCVD score is the first and only CV risk prediction model specific for Sri Lankans. It predicts the 10-year risk of developing a hard CVD in Sri Lankans. SLCVD score was more effective in predicting Sri Lankans at high CV risk than WHO risk charts.

Development and validation of a machine learning model to predict myocardial blood flow and clinical outcomes from patients' electrocardiograms

We develop a machine learning (ML) model using electrocardiography (ECG) to predict myocardial blood flow reserve (MFR) and assess its prognostic value for major adverse cardiovascular events (MACEs). Using 3,639 ECG-positron emission tomography (PET) and 17,649 ECG-single-photon emission computed tomography (SPECT) data pairs, the ML model is trained with a swarm intelligence approach and support vector regression (SVR). The model achieves a receiver-operator curve (ROC) area under the curve (AUC) of 0.83, with a sensitivity and specificity of 0.75. An ECG-MFR value below 2 is significantly associated with MACE, with hazard ratios (HRs) of 3.85 and 3.70 in the discovery and validation phases, respectively. The model's C-statistic is 0.76, with a net reclassification improvement (NRI) of 0.35. Validated in an independent cohort, the ML model using ECG data offers superior MACE prediction compared to baseline clinical models, highlighting its potential for risk stratification in patients with coronary artery disease (CAD) using the accessible 12-lead ECG.

Leveraging Artificial Intelligence to Optimize the Care of Peripheral Artery Disease Patients

Peripheral artery disease is a major atherosclerotic disease that is associated with poor outcomes such as limb loss, cardiovascular morbidity, and death. Artificial intelligence (AI) has seen increasing integration in medicine, and its various applications can optimize the care of peripheral artery disease (PAD) patients in diagnosis, predicting patient outcomes, and imaging interpretation. In this review, we introduce various AI applications such as natural language processing, supervised machine learning, and deep learning, and we analyze the current literature in which these algorithms have been applied to PAD.

Machine-learning clustering analysis identifies novel phenogroups in patients with ST-elevation acute myocardial infarction

BACKGROUND

Machine learning clustering of patients with ST-elevation acute myocardial infarction (STEMI) may provide important insights into their risk profile, management and prognosis.

METHODS

All adult discharges for STEMI in the National Inpatient Sample (October 2015 to December 2019) were included, excluding patients with prior myocardial infarction. Machine-learning clustering analysis was used to define clusters based on 21 clinical attributes of interest. Main outcomes of the study were cluster-based comparison of risk profile, in-hospital clinical outcomes and utilization of invasive management. Binomial hierarchical multivariable logistic regression with adjusted odds ratios (aOR) and 95% confidence intervals (95% CI) was used to detect the between-cluster differences.

RESULTS

Out of overall 470,960 STEMI cases, the machine-learning analysis revealed 4 different clusters with 205,640 (cluster 0: 'behavioural risk cluster'), 146,400 (cluster 1: 'least comorbidity cluster'), 45,100 (cluster 2: 'diabetes with end-organ damage cluster') and 73,820 (cluster 3: 'cardiometabolic cluster') cases. Attributes with the highest importance for clustering were hypertension and diabetes. After multivariable adjustment, patients from 'diabetes with end-organ damage cluster' exhibited the worst mortality, MACCE and ischemic stroke (p < 0.001 for all), as well as the lowest utilization of invasive management (p < 0.001 for all), in comparison to other clusters. Patients from 'behavioural risk cluster' exhibited the best in-hospital prognosis and the highest utilization of invasive management, compared to other clusters (p < 0.001 for all).

CONCLUSIONS

Machine learning driven clustering of inpatients with STEMI reveals important population subgroups with distinct prevalence, risk profile, prognosis and management. Data driven approaches may identify high risk phenogroups and warrants further study.

Machine-Learning for Phenotyping and Prognostication of Myocardial Infarction and Injury in Suspected Acute Coronary Syndrome

Background

Clinical work-up for suspected acute coronary syndrome (ACS) is resource intensive.

Objectives

This study aimed to develop a machine learning model for digitally phenotyping myocardial injury and infarction and predict 30-day events in suspected ACS patients.

Methods

Training and testing data sets, predominantly derived from electronic health records, included suspected ACS patients presenting to 6 and 26 South Australian hospitals, respectively. All index presentations and 30-day death and myocardial infarction (MI) were adjudicated using the Fourth Universal Definition of MI. We developed 2 diagnostic prediction models which phenotype myocardial injury and infarction according to the Fourth UDMI (chronic myocardial injury vs acute myocardial injury patterns, the latter further differentiated into acute non-ischaemic myocardial injury, Types 1 and 2 MI) using eXtreme Gradient Boosting (XGB) and deep-learning (DL). We also developed an event prediction model for risk prediction of 30-day death or MI using XGB. Analyses were performed in Python 3.6.

Results

The training and testing data sets had 6,722 and 8,869 participants, respectively. The diagnostic prediction XGB and deep learning models achieved an area under the curve of 99.2% ± 0.1% and 98.8% ± 0.2%, respectively, for differentiating an acute myocardial injury pattern from no injury or chronic myocardial injury pattern and achieved 95.5% ± 0.2% and 94.6% ± 0.9%, respectively, for differentiating type 1 MI from type 2 MI or acute nonischemic myocardial injury. The 30-day death/MI event prediction model achieved an area under the curve of 88.5% ± 0.5%.

Conclusions

Machine learning models can digitally phenotype suspected ACS patients at index presentation and predict subsequent events within 30 days. These models require external validation in a randomized clinical trial to evaluate their impact in clinical practice.

Using machine learning-based algorithms to construct cardiovascular risk prediction models for Taiwanese adults based on traditional and novel risk factors

OBJECTIVE

To develop and validate machine learning models for predicting coronary artery disease (CAD) within a Taiwanese cohort, with an emphasis on identifying significant predictors and comparing the performance of various models.

METHODS

This study involved a comprehensive analysis of clinical, demographic, and laboratory data from 8,495 subjects in Taiwan Biobank (TWB) after propensity score matching to address potential confounding factors. Key variables included age, gender, lipid profiles (T-CHO, HDL_C, LDL_C, TG), smoking and alcohol consumption habits, and renal and liver function markers. The performance of multiple machine learning models was evaluated.

RESULTS

The cohort comprised 1,699 individuals with CAD identified through self-reported questionnaires. Significant differences were observed between CAD and non-CAD individuals regarding demographics and clinical features. Notably, the Gradient Boosting model emerged as the most accurate, achieving an AUC of 0.846 (95% confidence interval [CI] 0.819-0.873), sensitivity of 0.776 (95% CI, 0.732-0.820), and specificity of 0.759 (95% CI, 0.736-0.782), respectively. The accuracy was 0.762 (95% CI, 0.742-0.782). Age was identified as the most influential predictor of CAD risk within the studied dataset.

CONCLUSION

The Gradient Boosting machine learning model demonstrated superior performance in predicting CAD within the Taiwanese cohort, with age being a critical predictor. These findings underscore the potential of machine learning models in enhancing the prediction accuracy of CAD, thereby supporting early detection and targeted intervention strategies.

TRIAL REGISTRATION

Not applicable.

Risk prediction of heart diseases in patients with breast cancer: A deep learning approach with longitudinal electronic health records data

Accurately predicting heart disease risks in patients with breast cancer is crucial for clinical decision support and patient safety. This study developed and evaluated predictive models for six heart diseases using real-world electronic health records (EHRs) data. We incorporated a trainable decay mechanism to handle missing values in the long short-term memory (LSTM) model, creating LSTM-D models to predict heart disease risk based on longitudinal EHRs data. Additionally, we deployed NLP methods to extract breast cancer phenotypes from clinical texts, integrating unstructured and structured data to enhance predictions. Our LSTM-D models outperformed baseline models in predicting congestive heart failure, coronary artery disease, cardiomyopathy, myocardial infarction, transient ischemic attack, and aortic regurgitation, with AUC scores ranging from 0.7189 to 0.9548. Observation windows of 12-24 months were found optimal for model performance. This research advances precise, personalized care strategies, enabling early intervention and improved management of cardiovascular risks in breast cancer survivors.

A Systematic Review of Artificial Intelligence Models for Time-to-Event Outcome Applied in Cardiovascular Disease Risk Prediction

Artificial intelligence (AI) based predictive models for early detection of cardiovascular disease (CVD) risk are increasingly being utilised. However, AI based risk prediction models that account for right-censored data have been overlooked. This systematic review (PROSPERO protocol CRD42023492655) includes 33 studies that utilised machine learning (ML) and deep learning (DL) models for survival outcome in CVD prediction. We provided details on the employed ML and DL models, eXplainable AI (XAI) techniques, and type of included variables, with a focus on social determinants of health (SDoH) and gender-stratification. Approximately half of the studies were published in 2023 with the majority from the United States. Random Survival Forest (RSF), Survival Gradient Boosting models, and Penalised Cox models were the most frequently employed ML models. DeepSurv was the most frequently employed DL model. DL models were better at predicting CVD outcomes than ML models. Permutation-based feature importance and Shapley values were the most utilised XAI methods for explaining AI models. Moreover, only one in five studies performed gender-stratification analysis and very few incorporate the wide range of SDoH factors in their prediction model. In conclusion, the evidence indicates that RSF and DeepSurv models are currently the optimal models for predicting CVD outcomes. This study also highlights the better predictive ability of DL survival models, compared to ML models. Future research should ensure the appropriate interpretation of AI models, accounting for SDoH, and gender stratification, as gender plays a significant role in CVD occurrence.

Random Forest and Feature Importance Measures for Discriminating the Most Influential Environmental Factors in Predicting Cardiovascular and Respiratory Diseases

BACKGROUND

Several studies suggest that environmental and climatic factors are linked to the risk of mortality due to cardiovascular and respiratory diseases; however, it is still unclear which are the most influential ones. This study sheds light on the potentiality of a data-driven statistical approach by providing a case study analysis.

METHODS

Daily admissions to the emergency room for cardiovascular and respiratory diseases are jointly analyzed with daily environmental and climatic parameter values (temperature, atmospheric pressure, relative humidity, carbon monoxide, ozone, particulate matter, and nitrogen dioxide). The Random Forest (RF) model and feature importance measure (FMI) techniques (permutation feature importance (PFI), Shapley Additive exPlanations (SHAP) feature importance, and the derivative-based importance measure (κALE)) are applied for discriminating the role of each environmental and climatic parameter. Data are pre-processed to remove trend and seasonal behavior using the Seasonal Trend Decomposition (STL) method and preliminary analyzed to avoid redundancy of information.

RESULTS

The RF performance is encouraging, being able to predict cardiovascular and respiratory disease admissions with a mean absolute relative error of 0.04 and 0.05 cases per day, respectively. Feature importance measures discriminate parameter behaviors providing importance rankings. Indeed, only three parameters (temperature, atmospheric pressure, and carbon monoxide) were responsible for most of the total prediction accuracy.

CONCLUSIONS

Data-driven and statistical tools, like the feature importance measure, are promising for discriminating the role of environmental and climatic factors in predicting the risk related to cardiovascular and respiratory diseases. Our results reveal the potential of employing these tools in public health policy applications for the development of early warning systems that address health risks associated with climate change, and improving disease prevention strategies.

Machine learning models based on fluid immunoproteins that predict non-AIDS adverse events in people with HIV

Despite the success of antiretroviral therapy (ART), individuals with HIV remain at risk for experiencing non-AIDS adverse events (NAEs), including cardiovascular complications and malignancy. Several surrogate immune biomarkers in blood have shown predictive value in predicting NAEs; however, composite panels generated using machine learning may provide a more accurate advancement for monitoring and discriminating NAEs. In a nested case-control study, we aimed to develop machine learning models to discriminate cases (experienced an event) and matched controls using demographic and clinical characteristics alongside 49 plasma immunoproteins measured prior to and post-ART initiation. We generated support vector machine (SVM) classifier models for high-accuracy discrimination of individuals aged 30-50 years who experienced non-fatal NAEs at pre-ART and one-year post-ART. Extreme gradient boosting generated a high-accuracy model at pre-ART, while K-nearest neighbors performed poorly all around. SVM modeling may offer guidance to improve disease monitoring and elucidate potential therapeutic interventions.

Predicting severity of acute appendicitis with machine learning methods: a simple and promising approach for clinicians

BACKGROUNDS

Acute Appendicitis (AA) is one of the most common surgical emergencies worldwide. This study aims to investigate the predictive performances of 6 different Machine Learning (ML) algorithms for simple and complicated AA.

METHODS

Data regarding operated AA patients between 2012 and 2022 were analyzed retrospectively. Based on operative findings, patients were evaluated under two groups: perforated AA and none-perforated AA. The features that showed statistical significance (p < 0.05) in both univariate and multivariate analysis were included in the prediction models as input features. Five different error metrics and the area under the receiver operating characteristic curve (AUC) were used for model comparison.

RESULTS

A total number of 1132 patients were included in the study. Patients were divided into training (932 samples), testing (100 samples), and validation (100 samples) sets. Age, gender, neutrophil count, lymphocyte count, Neutrophil to Lymphocyte ratio, total bilirubin, C-Reactive Protein (CRP), Appendix Diameter, and PeriAppendicular Liquid Collection (PALC) were significantly different between the two groups. In the multivariate analysis, age, CRP, and PALC continued to show a significant difference in the perforated AA group. According to univariate and multivariate analysis, two data sets were used in the prediction model. K-Nearest Neighbors and Logistic Regression algorithms achieved the best prediction performance in the validation group with an accuracy of 96%.

CONCLUSION

The results showed that using only three input features (age, CRP, and PALC), the severity of AA can be predicted with high accuracy. The developed prediction model can be useful in clinical practice.

Phenotype clustering of hospitalized high-risk patients with COVID-19 - a machine learning approach within the multicentre, multinational PCHF-COVICAV registry

IMTRODUCTION

The high-risk population of patients with cardiovascular (CV) disease or risk factors (RF) suffering from COVID-19 is heterogeneous. Several predictors for impaired prognosis have been identified. However, with machine learning (ML) approaches, certain phenotypes may be confined to classify the affected population and to predict outcome. This study aimed to phenotype patients using unsupervised ML technique within the International Postgraduate Course Heart Failure Registry for patients hospitalized with COVID-19 and Cardiovascular disease and/or RF (PCHF-COVICAV).

MATERIAL AND METHODS

Patients from the eight centres with follow-up data available from the PCHF-COVICAV registry were included in this ML analysis (K-medoids algorithm).

RESULTS

Out of 617 patients included into the prospective part of the registry, 458 [median age: 76 (IQR:65-84) years, 55% male] were analyzed and 46 baseline variables, including demographics, clinical status, comorbidities and biochemical characteristics were incorporated into the ML. Three clusters were extracted by this ML method. Cluster 1 (n = 181) represents mainly women with the least number of overall comorbidities and cardiovascular RF. Cluster 2 (n = 227) is characterized mainly by men with non-CV conditions and less severe symptoms of infection. Cluster 3 (n=50) mainly represents men with the highest prevalence of cardiac comorbidities and RF, more extensive inflammation and organ dysfunction with the highest 6-month all-cause mortality risk.

CONCLUSIONS

The ML process has identified three important clinical clusters from hospitalized COVID-19 CV and/or RF patients. The cluster of males with severe CV disease, particularly HF, and multiple RF presenting with increased inflammation had a particularly poor outcome.

Machine learning methods in predicting the risk of malignant transformation of oral potentially malignant disorders: A systematic review

BACKGROUND

Oral Potentially Malignant Disorders (OPMDs) refer to a heterogenous group of clinical presentations with heightened rate of malignant transformation. Identification of risk levels in OPMDs is crucial to determine the need for active intervention in high-risk patients and routine follow-up in low-risk ones. Machine learning models has shown tremendous potential in several areas of dentistry that strongly suggest its application to estimate rate of malignant transformation of precancerous lesions.

METHODS

A comprehensive literature search was performed on Pubmed/MEDLINE, Web of Science, Scopus, Embase, Cochrane Library database to identify articles including machine learning models and algorithms to predict malignant transformation in OPMDs. Relevant bibliographic data, study characteristics, and outcomes were extracted for eligible studies. Quality of the included studies was assessed through the IJMEDI checklist.

RESULTS

Fifteen articles were found suitable for the review as per the PECOS criteria. Amongst all studies, highest sensitivity (100%) was recorded for U-net architecture, Peaks Random forest model, and Partial least squares discriminant analysis (PLSDA). Highest specificity (100%) was noted for PLSDA. Range of overall accuracy in risk prediction was between 95.4% and 74%.

CONCLUSION

Machine learning proved to be a viable tool in risk prediction, demonstrating heightened sensitivity, automation, and improved accuracy for predicting transformation of OPMDs. It presents an effective approach for incorporating multiple variables to monitor the progression of OPMDs and predict their malignant potential. However, its sensitivity to dataset characteristics necessitates the optimization of input parameters to maximize the efficiency of the classifiers.

Machine learning prediction models for diabetic kidney disease: systematic review and meta-analysis

BACKGROUND

Machine learning is increasingly recognized as a viable approach for identifying risk factors associated with diabetic kidney disease (DKD). However, the current state of real-world research lacks a comprehensive systematic analysis of the predictive performance of machine learning (ML) models for DKD.

OBJECTIVES

The objectives of this study were to systematically summarize the predictive capabilities of various ML methods in forecasting the onset and the advancement of DKD, and to provide a basic outline for ML methods in DKD.

METHODS

We have searched mainstream databases, including PubMed, Web of Science, Embase, and MEDLINE databases to obtain the eligible studies. Subsequently, we categorized various ML techniques and analyzed the differences in their performance in predicting DKD.

RESULTS

Logistic regression (LR) was the prevailing ML method, yielding an overall pooled area under the receiver operating characteristic curve (AUROC) of 0.83. On the other hand, the non-LR models also performed well with an overall pooled AUROC of 0.80. Our t-tests showed no statistically significant difference in predicting ability between LR and non-LR models (t = 1.6767, p > 0.05).

CONCLUSION

All ML predicting models yielded relatively satisfied DKD predicting ability with their AUROCs greater than 0.7. However, we found no evidence that non-LR models outperformed the LR model. LR exhibits high performance or accuracy in practice, while it is known for algorithmic simplicity and computational efficiency compared to others. Thus, LR may be considered a cost-effective ML model in practice.

Bibliometric Analysis and Visualization of Scientific Literature on Heart Disease Classification Using a Logistic Regression Model

With the advancement in artificial intelligence, the use of machine learning algorithms for clinical prediction has increased tremendously. Logistic regression is one of the powerful machine learning algorithms that can be used to predict the probability of a variable. Logistic regression is very popular among medical researchers owing to its simplicity, interpretability, and solid statistical foundation. This study aims to investigate the research productivity of heart disease classification using a logistic regression model to analyze the current patterns and potential future trends through bibliometric analysis. Additionally, it aims to highlight the impact and quality of research in the area, identify prominent research groups, the countries actively contributing to the field, which will help the researchers and healthcare professionals to pinpoint research gaps, influential authors, and make informed decisions and invest resources accordingly. The data is collected from a database of Scopus spanning from 2019 to 2023. We have used two bibliometric software, Biblioshiny (Aria and Cuccurullo, 2017) and VOSviewer (Centre for Science and Technology Studies (CWTS), Leiden University, the Netherlands), to analyze the bibliographic data regarding the citation count, contribution of authors, publication count, the contribution of institutions, etc. There are 2331 documents under study which were fed into both software to analyze the data. With 700 documents, China topped the list of most productive countries indicating the vast contribution of the country followed by India and the United States. Contributions of the Harvard Medical School, Boston, MA, United States are found to be the greatest with six papers. The most productive author is Wang Y with 73 documents. Analysis of trending topics reveals that the field progressing towards using support vector machines (SVM), k-nearest neighbours (KNN), and naïve Bayes algorithms. The article has only considered data from Scopus excluding literature indexed in other databases which limits the potential coverage of the data. Also, the work focuses on recent developments excluding older literature from 2019 which could be a limitation. Furthermore, since the study is a bibliometric analysis targeting the use of logistic regression for heart disease prediction, powerful techniques such as SVM, decision trees, random forests, neural networks and deep learning have not been included, which could be another limitation.

Optimizing prediction accuracy for early recurrent lumbar disc herniation with a directional mutation-guided SVM model

Percutaneous endoscopic lumbar discectomy (PELD) is one of the main means of minimally invasive spinal surgery, and is an effective means of treating lumbar disc herniation, but its early recurrence is still difficult to predict. With the development of machine learning technology, the auxiliary model based on the prediction of early recurrent lumbar disc herniation (rLDH) and the identification of causative risk factors have become urgent problems in current research. However, the screening ability of current models for key factors affecting the prediction of rLDH, as well as their predictive ability, needs to be improved. Therefore, this paper presents a classification model that utilizes wrapper feature selection, developed through the integration of an enhanced bat algorithm (BDGBA) and support vector machine (SVM). Among them, BDGBA increases the population diversity and improves the population quality by introducing directional mutation strategy and guidance-based strategy, which in turn allows the model to secure better subsets of features. Furthermore, SVM serves as the classifier for the wrapper feature selection method, with its classification prediction results acting as a fitness function for the feature subset. In the proposed prediction method, BDGBA is used as an optimizer for feature subset filtering and as an objective function for feature subset evaluation based on the classification results of the support vector machine, which improves the interpretability and prediction accuracy of the model. In order to verify the performance of the proposed method, this paper proves the performance of the model through global optimization experiments and prediction experiments on real data sets. The accuracy of the proposed rLDH prediction model is 93.49% and sensitivity is 88.33%. The experimental results show that Level of herniated disk, Modic change, Disk height, Disk length, and Disk width are the key factors for predicting rLDH, and the proposed method is an effective auxiliary diagnosis method.

Pulmonary Hypertension Detection Non-Invasively at Point-of-Care Using a Machine-Learned Algorithm

Artificial intelligence, particularly machine learning, has gained prominence in medical research due to its potential to develop non-invasive diagnostics. Pulmonary hypertension presents a diagnostic challenge due to its heterogeneous nature and similarity in symptoms to other cardiovascular conditions. Here, we describe the development of a supervised machine learning model using non-invasive signals (orthogonal voltage gradient and photoplethysmographic) and a hand-crafted library of 3298 features. The developed model achieved a sensitivity of 87% and a specificity of 83%, with an overall Area Under the Receiver Operator Characteristic Curve (AUC-ROC) of 0.93. Subgroup analysis showed consistent performance across genders, age groups and classes of PH. Feature importance analysis revealed changes in metrics that measure conduction, repolarization and respiration as significant contributors to the model. The model demonstrates promising performance in identifying pulmonary hypertension, offering potential for early detection and intervention when embedded in a point-of-care diagnostic system.

Predictive performance of machine learning compared to statistical methods in time-to-event analysis of cardiovascular disease: a systematic review protocol

BACKGROUND

Globally, cardiovascular disease (CVD) remains the leading cause of death, warranting effective management and prevention measures. Risk prediction tools are indispensable for directing primary and secondary prevention strategies for CVD and are critical for estimating CVD risk. Machine learning (ML) methodologies have experienced significant advancements across numerous practical domains in recent years. Several ML and statistical models predicting CVD time-to-event outcomes have been developed. However, it is not known as to which of the two model types-ML and statistical models-have higher discrimination and calibration in this regard. Hence, this planned work aims to systematically review studies that compare ML with statistical methods in terms of their predictive abilities in the case of time-to-event data with censoring.

METHODS

Original research articles published as prognostic prediction studies, which involved the development and/or validation of a prognostic model, within a peer-reviewed journal, using cohort or experimental design with at least a 12-month follow-up period will be systematically reviewed. The review process will adhere to the Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies checklist.

ETHICS AND DISSEMINATION

Ethical approval is not required for this review, as it will exclusively use data from published studies. The findings of this study will be published in an open-access journal and disseminated at scientific conferences.

PROSPERO REGISTRATION NUMBER

CRD42023484178.

A comparative evaluation of low-density lipoprotein cholesterol estimation: Machine learning algorithms versus various equations

BACKGROUND

Given the critical importance of Low-density lipoprotein cholesterol (LDL-C) levels in determining cardiovascular risk, it is essential to measure LDL-C accurately. Since the Friedewald formula generates incorrect predictions in many circumstances, new equations have been developed to overcome the Friedewald equations' shortcomings. This study aimed to compare estimated LDL-C with directly measured LDL-C (dLDL-C), as well as their performance in predicting LDL-C, utilizing Friedewald, extended Martin-Hopkins, Sampson, de Cordova, and Vujovic formulas and five machine learning (ML) algorithms.

METHODS

A total of 29,504 samples from the ISLAB-2 Core Laboratory were included in the study. All statistical analysis was performed using R version 4.1.2. Statistical Language.

RESULTS

Bayesian-Regularized Neural Network (BRNN) (r = 0.957) and Random Forest (RF) (r = 0.957) algorithms showed a higher correlation with dLDL-C than the other equations in all-testing dataset. All ML algorithms demonstrated less bias than pre-existing LDL-C equations with dLDL-C and outperformed the LDL-C estimation equations in terms of concordance in all-testing dataset.

CONCLUSIONS

The results of our research indicate that when compared to conventional equations, ML algorithms are much more effective in predicting LDL-C. ML algorithms, aided by a vast dataset, could have the capability to predict LDL-C levels even in cases where triglyceride levels are high, unlike the limited usage of Friedewald formula.

A novel method to guide biomarker combinations to optimize the sensitivity

Logistic regression has demonstrated its utility in classifying binary labeled datasets through the maximum likelihood approach. However, in numerous biological and clinical contexts, the aim is often to determine coefficients that yield the highest sensitivity at the pre-specified specificity or vice versa. Therefore, the application of logistic regression is limited in such settings. To this end, we have developed an improved regression framework, SMAGS, for binary classification that, for a given specificity, finds the linear decision rule that yields the maximum sensitivity. Furthermore, we employed the method for feature selection to find the features that are satisfying the sensitivity maximization goal. We compared our method with normal logistic regression by applying it to real clinical data as well as synthetic data. In the real application data (colorectal cancer dataset), we found 14% improvement of sensitivity at 98.5% specificity.

Availability and implementation

Software is made available in Python ( https://github.com/smahmoodghasemi/SMAGS ).

Adverse Effects of Meditation: Autonomic Nervous System Activation and Individual Nauseous Responses During Samadhi Meditation in the Czech Republic

Buddhist meditation practices, including Samadhi meditation, which forms the basis for mindfulness practice, are broadly promoted as pathways to wellbeing, but evidence of their adverse effects is emerging. In a single-group observational study with assessments of autonomic system before, during, and after Samadhi meditation, we explore the relationship between post-meditation nausea symptoms and the degree of change in autonomic system activity during meditation as compared to before and after in 57 university students (42 women; mean age = 22.6) without any previous experience in meditation or yoga practices. We hypothesize that nauseous feelings in meditation are connected to a rapid increase of activity in the sympathetic nervous system, as indicated by decreased heart-rate variability (HRV). We additionally explore links between meditation-induced nausea and two markers of parasympathetic activity: increased HRV and vasovagal syncope. Engaging in meditation and increased nausea during meditation were both associated with increased markers of HRV parasympathetic activity, but 12 individuals with markedly higher nausea demonstrated increased HRV markers of sympathetic activity during meditation. Vasovagal syncope was observed but found to be unrelated to nausea levels. Drivers of adverse effects of meditation in some individuals require further investigation.

Causal machine learning for predicting treatment outcomes

Causal machine learning (ML) offers flexible, data-driven methods for predicting treatment outcomes including efficacy and toxicity, thereby supporting the assessment and safety of drugs. A key benefit of causal ML is that it allows for estimating individualized treatment effects, so that clinical decision-making can be personalized to individual patient profiles. Causal ML can be used in combination with both clinical trial data and real-world data, such as clinical registries and electronic health records, but caution is needed to avoid biased or incorrect predictions. In this Perspective, we discuss the benefits of causal ML (relative to traditional statistical or ML approaches) and outline the key components and steps. Finally, we provide recommendations for the reliable use of causal ML and effective translation into the clinic.

Development of a Non-Invasive Machine-Learned Point-of-Care Rule-Out Test for Coronary Artery Disease

The current standard of care for coronary artery disease (CAD) requires an intake of radioactive or contrast enhancement dyes, radiation exposure, and stress and may take days to weeks for referral to gold-standard cardiac catheterization. The CAD diagnostic pathway would greatly benefit from a test to assess for CAD that enables the physician to rule it out at the point of care, thereby enabling the exploration of other diagnoses more rapidly. We sought to develop a test using machine learning to assess for CAD with a rule-out profile, using an easy-to-acquire signal (without stress/radiation) at the point of care. Given the historic disparate outcomes between sexes and urban/rural geographies in cardiology, we targeted equal performance across sexes in a geographically accessible test. Noninvasive photoplethysmogram and orthogonal voltage gradient signals were simultaneously acquired in a representative clinical population of subjects before invasive catheterization for those with CAD (gold-standard for the confirmation of CAD) and coronary computed tomographic angiography for those without CAD (excellent negative predictive value). Features were measured from the signal and used in machine learning to predict CAD status. The machine-learned algorithm achieved a sensitivity of 90% and specificity of 59%. The rule-out profile was maintained across both sexes, as well as all other relevant subgroups. A test to assess for CAD using machine learning on a noninvasive signal has been successfully developed, showing high performance and rule-out ability. Confirmation of the performance on a large clinical, blinded, enrollment-gated dataset is required before implementation of the test in clinical practice.

Artificial intelligence significantly facilitates development in the mental health of college students: a bibliometric analysis

Objective

College students are currently grappling with severe mental health challenges, and research on artificial intelligence (AI) related to college students mental health, as a crucial catalyst for promoting psychological well-being, is rapidly advancing. Employing bibliometric methods, this study aim to analyze and discuss the research on AI in college student mental health.

Methods

Publications pertaining to AI and college student mental health were retrieved from the Web of Science core database. The distribution of publications were analyzed to gage the predominant productivity. Data on countries, authors, journal, and keywords were analyzed using VOSViewer, exploring collaboration patterns, disciplinary composition, research hotspots and trends.

Results

Spanning 2003 to 2023, the study encompassed 1722 publications, revealing notable insights: (1) a gradual rise in annual publications, reaching its zenith in 2022; (2) Journal of Affective Disorders and Psychiatry Research emerged were the most productive and influential sources in this field, with significant contributions from China, the United States, and their affiliated higher education institutions; (3) the primary mental health issues were depression and anxiety, with machine learning and AI having the widest range of applications; (4) an imperative for enhanced international and interdisciplinary collaboration; (5) research hotspots exploring factors influencing college student mental health and AI applications.

Conclusion

This study provides a succinct yet comprehensive overview of this field, facilitating a nuanced understanding of prospective applications of AI in college student mental health. Professionals can leverage this research to discern the advantages, risks, and potential impacts of AI in this critical field.

Precision medicine: Hype or hope?

In recent years, precision medicine has steadily risen to the forefront of many aspects of medicine, including cardiology [1]. While this field has exponentially expanded and advanced in the last few years, a lot of questions remain regarding exact definition, usage, and clinical applications [2,3]. This review will provide a brief synopsis of the current state of precision medicine, its limitations, future directions, as well as analyze emerging clinical applications in cardiology.

Cardiovascular Diseases Diagnosis Using an ECG Multi-Band Non-Linear Machine Learning Framework Analysis

BACKGROUND

cardiovascular diseases (CVDs), which encompass heart and blood vessel issues, stand as the leading cause of global mortality for many people.

METHODS

the present study intends to perform discrimination between seven well-known CVDs (bundle branch block, cardiomyopathy, myocarditis, myocardial hypertrophy, myocardial infarction, valvular heart disease, and dysrhythmia) and one healthy control group, respectively, by feeding a set of machine learning (ML) models with 10 non-linear features extracted every 1 s from electrocardiography (ECG) lead signals of a well-known ECG database (PTB diagnostic ECG database) using multi-band analysis performed by discrete wavelet transform (DWT). The ML models were trained and tested using a leave-one-out cross-validation approach, assessing the individual and combined capabilities of features, per each lead or combined, to distinguish between pairs of study groups and for conducting a comprehensive all vs. all analysis.

RESULTS

the Accuracy discrimination results ranged between 73% and 100%, the Recall between 68% and 100%, and the AUC between 0.42 and 1.

CONCLUSIONS

the results suggest that our method is a good tool for distinguishing CVDs, offering significant advantages over other studies that used the same dataset, including a multi-class comparison group (all vs. all), a wider range of binary comparisons, and the use of classical non-linear analysis under ECG multi-band analysis performed by DWT.

Navigating the AI frontiers in cardiovascular research: a bibliometric exploration and topic modeling

Artificial intelligence (AI) has emerged as a promising field in cardiovascular disease (CVD) research, offering innovative approaches to enhance diagnosis, treatment, and patient outcomes. In this study, we conducted bibliometric analysis combined with topic modeling to provide a comprehensive overview of the AI research landscape in CVD. Our analysis included 23,846 studies from Web of Science and PubMed, capturing the latest advancements and trends in this rapidly evolving field. By employing LDA (Latent Dirichlet Allocation) we identified key research themes, trends, and collaborations within the AI-CVD domain. The findings revealed the exponential growth of AI-related research in CVD, underscoring its immense potential to revolutionize cardiovascular healthcare. The annual scientific publication of machine learning papers in CVD increases continuously and significantly since 2016, with an overall annual growth rate of 22.8%. Almost half (46.2%) of the growth happened in the last 5 years. USA, China, India, UK and Korea were the top five productive countries in number of publications. UK, Germany and Australia were the most collaborative countries with a multiple country publication (MCP) value of 42.8%, 40.3% and 40.0% respectively. We observed the emergence of twenty-two distinct research topics, including "stroke and robotic rehabilitation therapy," "robotic-assisted cardiac surgery," and "cardiac image analysis," which persisted as major topics throughout the years. Other topics, such as "retinal image analysis and CVD" and "biomarker and wearable signal analyses," have recently emerged as dominant areas of research in cardiovascular medicine. Convolutional neural network appears to be the most mentioned algorithm followed by LSTM (Long Short-Term Memory) and KNN (K-Nearest Neighbours). This indicates that the future direction of AI cardiovascular research is predominantly directing toward neural networks and image analysis. As AI continues to shape the landscape of CVD research, our study serves as a comprehensive guide for researchers, practitioners, and policymakers, providing valuable insights into the current state of AI in CVD research. This study offers a deep understanding of research trends and paves the way for future directions to maximiz the potential of AI to effectively combat cardiovascular diseases.

Machine learning-driven predictions and interventions for cardiovascular occlusions

BACKGROUND

Cardiovascular diseases remain a leading cause of global morbidity and mortality, with heart attacks and strokes representing significant health challenges. The accurate, early diagnosis and management of these conditions are paramount in improving patient outcomes. The specific disease, cardiovascular occlusions, has been chosen for the study due to the significant impact it has on public health. Cardiovascular diseases are a leading cause of mortality globally, and occlusions, which are blockages in the blood vessels, are a critical factor contributing to these conditions.

OBJECTIVE

By focusing on cardiovascular occlusions, the study aims to leverage machine learning to improve the prediction and management of these events, potentially helping to reduce the incidence of heart attacks, strokes, and other related health issues. The use of machine learning in this context offers the promise of developing more accurate and timely interventions, thus improving patient outcomes.

METHODS

We analyze diverse datasets to assess the efficacy of various machine learning algorithms in predicting heart attacks and strokes, comparing their performance to pinpoint the most accurate and reliable models. Additionally, we classify individuals by their predicted risk levels and examine key features that correlate with the incidence of cardiovascular events. The PyCaret machine learning library's Classification Module was key in developing predictive models which were evaluated with stratified cross-validation for reliable performance estimates.

RESULTS

Our findings suggest that machine learning can significantly improve the prediction accuracy for heart attacks and strokes, facilitating earlier and more precise interventions. We also discuss the integration of machine learning models into clinical practice, addressing potential challenges and the need for healthcare professionals to interpret and apply these predictions effectively.

CONCLUSIONS

The use of machine learning for risk stratification and the identification of modifiable factors may empower preemptive approaches to cardiovascular care, ultimately aiming to reduce the occurrence of life-threatening events and improve long-term patient health trajectories.

Cluster Analysis reveals Socioeconomic Disparities among Elective Spine Surgery Patients

This work demonstrates the use of cluster analysis in detecting fair and unbiased novel discoveries. Given a sample population of elective spinal fusion patients, we identify two overarching subgroups driven by insurance type. The Medicare group, associated with lower socioeconomic status, exhibited an over-representation of negative risk factors. The findings provide a compelling depiction of the interwoven socioeconomic and racial disparities present within the healthcare system, highlighting their consequential effects on health inequalities. The results are intended to guide design of fair and precise machine learning models based on intentional integration of population stratification.

Predicting Survival of End-Stage Heart Failure Patients Receiving HeartMate-3: Comparing Machine Learning Methods

HeartMate 3 is the only durable left ventricular assist devices (LVAD) currently implanted in the United States. The purpose of this study was to develop a predictive model for 1 year mortality of HeartMate 3 implanted patients, comparing standard statistical techniques and machine learning algorithms. Adult patients registered in the Society of Thoracic Surgeons, Interagency Registry for Mechanically Assisted Circulatory Support (STS-INTERMACS) database, who received primary implant with a HeartMate 3 between January 1, 2017, and December 31, 2019, were included. Epidemiological, clinical, hemodynamic, and echocardiographic characteristics were analyzed. Standard logistic regression and machine learning (elastic net and neural network) were used to predict 1 year survival. A total of 3,853 patients were included. Of these, 493 (12.8%) died within 1 year after implantation. Standard logistic regression identified age, Model End Stage Liver Disease (MELD)-XI score, right arterial (RA) pressure, INTERMACS profile, heart rate, and etiology of heart failure (HF), as important predictor factors for 1 year mortality with an area under the curve (AUC): 0.72 (0.66-0.77). This predictive model was noninferior to the ones developed using the elastic net or neural network. Standard statistical techniques were noninferior to neural networks and elastic net in predicting 1 year survival after HeartMate 3 implantation. The benefit of using machine-learning algorithms in the prediction of outcomes may depend on the type of dataset used for analysis.

Emerging Strategies and Effective Prevention Measures for Investigating the Association Between Stroke and Sudden Cardiac Fatality

Stroke-related cardiac death is a significant concern for patients with stroke and their healthcare providers. It is a complex and multifaceted condition that requires careful management of both modifiable and non-modifiable risk factors. This review provides an overview of the pathophysiology, risk factors, and prevention strategies for stroke-related cardiac death. The review highlights the importance of identifying and managing modifiable risk factors such as hypertension, diabetes, and lifestyle factors, as well as non-modifiable risk factors such as age and genetics. Additionally, the review explores emerging strategies for prevention, including the use of wearable devices and genetic testing to identify patients at risk, stem cell therapy and gene therapy for cardiac dysfunction, and precision medicine for personalized treatment plans. Despite some limitations to this review, it provides valuable insights into the current understanding of stroke-related cardiac death and identifies important areas for future research. Ultimately, the implementation of evidence-based prevention strategies and personalized treatment plans has the potential to improve outcomes for patients with stroke and reduce the burden of stroke-related cardiac death in the population.

Antenatal prediction models for outcomes of extremely and very preterm infants based on machine learning

PURPOSE

Predicting individual risks for adverse outcomes in preterm infants is necessary for perinatal management and antenatal counseling for their parents. To evaluate whether a machine learning approach can improve the prediction of severe infant outcomes beyond the performance of conventional logistic models, and to identify maternal and fetal factors that largely contribute to these outcomes.

METHODS

A population-based retrospective study was performed using clinical data of 31,157 infants born at < 32 weeks of gestation and weighing ≤ 1500 g, registered in the Neonatal Research Network of Japan between 2006 and 2015. We developed a conventional logistic model and 6 types of machine learning models based on 12 maternal and fetal factors. Discriminative ability was evaluated using the area under the receiver operating characteristic curves (AUROCs), and the importance of each factor in terms of its contribution to outcomes was evaluated using the SHAP (SHapley Additive exPlanations) value.

RESULTS

The AUROCs of the most discriminative machine learning models were better than those of the conventional models for all outcomes. The AUROCs for in-hospital death and short-term adverse outcomes in the gradient boosting decision tree were significantly higher than those in the conventional model (p = 0.015 and p = 0.002, respectively). The SHAP value analyses showed that gestational age, birth weight, and antenatal corticosteroid treatment were the three most important factors associated with severe infant outcomes.

CONCLUSION

Machine learning models improve the prediction of severe infant outcomes. Moreover, the machine learning approach provides insight into the potential risk factors for severe infant outcomes.

Influence of cardiovascular risk factors and treatment exposure on cardiovascular event incidence: Assessment using machine learning algorithms

Assessment of the influence of cardiovascular risk factors (CVRF) on cardiovascular event (CVE) using machine learning algorithms offers some advantages over preexisting scoring systems, and better enables personalized medicine approaches to cardiovascular prevention. Using data from four different sources, we evaluated the outcomes of three machine learning algorithms for CVE prediction using different combinations of predictive variables and analysed the influence of different CVRF-related variables on CVE prediction when included in these algorithms. A cohort study based on a male cohort of workers applying populational data was conducted. The population of the study consisted of 3746 males. For descriptive analyses, mean and standard deviation were used for quantitative variables, and percentages for categorical ones. Machine learning algorithms used were XGBoost, Random Forest and Naïve Bayes (NB). They were applied to two groups of variables: i) age, physical status, Hypercholesterolemia (HC), Hypertension, and Diabetes Mellitus (DM) and ii) these variables plus treatment exposure, based on the adherence to the treatment for DM, hypertension and HC. All methods point out to the age as the most influential variable in the incidence of a CVE. When considering treatment exposure, it was more influential than any other CVRF, which changed its influence depending on the model and algorithm applied. According to the performance of the algorithms, the most accurate was Random Forest when treatment exposure was considered (F1 score 0.84), followed by XGBoost. Adherence to treatment showed to be an important variable in the risk of having a CVE. These algorithms could be applied to create models for every population, and they can be used in primary care to manage interventions personalized for every subject.

Machine learning for prediction of viral hepatitis: A systematic review and meta-analysis

BACKGROUND

Lack of accurate and timely diagnosis of hepatitis poses obstacles to effective treatment, disease progression prevention, complication reduction, and life-saving interventions of patients. Utilizing machine learning can greatly enhance the achievement of timely and precise disease diagnosis. Therefore, we carried out this systematic review and meta-analysis to explore the performance of machine learning algorithms in predicting viral hepatitis.

METHODS

Using an extensive literature search in PubMed, Scopus, and Web of Science databases until June 15, 2023, English publications on hepatitis prediction using machine learning algorithms were included. Two authors independently extracted pertinent information from the selected studies. The PRISMA 2020 checklist was followed for study selection and result reporting. The risk of bias was checked using the International Journal of Medical Informatics (IJMEDI) checklist. Data were analyzed using the 'metandi' command in Stata 17.

RESULTS

Twenty-one original studies were included, covering 82 algorithms. Sixteen studies utilized five algorithms to predict hepatitis B. Ten studies used five algorithms for hepatitis C prediction. For hepatitis B prediction, the SVM algorithms demonstrated the highest sensitivity (90.0%; 95% confidence interval (CI): 77.0%-96.0%), specificity (94%; 95% CI: 90.0%-97.0%), and a diagnostic odds ratio (DOR) of 145 (95% CI: 37.0-559.0). In the case of hepatitis C, the KNN algorithms exhibited the highest sensitivity (80%; 95% CI:30.0%-97.0%), specificity (95%; 95% CI: 58.0%-99.0%), and DOR (72; 95% CI: 3.0-1644.0) for prediction.

CONCLUSION

SVM and KNN demonstrated superior performance in predicting hepatitis. The proper algorithm along with clinical practice could improve hepatitis prediction and management.

Machine Learning Models for ASCVD Risk Prediction in an Asian Population - How to Validate the Model is Important

Introduction

Atherosclerotic cardiovascular disease (ASCVD) is prevalent worldwide including Taiwan, however widely accepted tools to assess the risk of ASCVD are lacking in Taiwan. Machine learning models are potentially useful for risk evaluation. In this study we used two cohorts to test the feasibility of machine learning with transfer learning for developing an ASCVD risk prediction model in Taiwan.

Methods

Two multi-center observational registry cohorts, T-SPARCLE and T-PPARCLE were used in this study. The variables selected were based on European, U.S. and Asian guidelines. Both registries recorded the ASCVD outcomes of the patients. Ten-fold validation and temporal validation methods were used to evaluate the performance of the binary classification analysis [prediction of major adverse cardiovascular (CV) events in one year]. Time-to-event analyses were also performed.

Results

In the binary classification analysis, eXtreme Gradient Boosting (XGBoost) and random forest had the best performance, with areas under the receiver operating characteristic curve (AUC-ROC) of 0.72 (0.68-0.76) and 0.73 (0.69-0.77), respectively, although it was not significantly better than other models. Temporal validation was also performed, and the data showed significant differences in the distribution of various features and event rate. The AUC-ROC of XGBoost dropped to 0.66 (0.59-0.73), while that of random forest dropped to 0.69 (0.62-0.76) in the temporal validation method, and the performance also became numerically worse than that of the logistic regression model. In the time-to-event analysis, most models had a concordance index of around 0.70.

Conclusions

Machine learning models with appropriate transfer learning may be a useful tool for the development of CV risk prediction models and may help improve patient care in the future.

Machine learning prediction models for clinical management of blood-borne viral infections: a systematic review of current applications and future impact

BACKGROUND

Machine learning (ML) prediction models to support clinical management of blood-borne viral infections are becoming increasingly abundant in medical literature, with a number of competing models being developed for the same outcome or target population. However, evidence on the quality of these ML prediction models are limited.

OBJECTIVE

This study aimed to evaluate the development and quality of reporting of ML prediction models that could facilitate timely clinical management of blood-borne viral infections.

METHODS

We conducted narrative evidence synthesis following the synthesis without meta-analysis guidelines. We searched PubMed and Cochrane Central Register of Controlled Trials for all studies applying ML models for predicting clinical outcomes associated with hepatitis B virus (HBV), human immunodeficiency virus (HIV), or hepatitis C virus (HCV).

RESULTS

We found 33 unique ML prediction models aiming to support clinical decision making. Overall, 12 (36.4%) focused on HBV, 10 (30.3%) on HCV, 10 on HIV (30.3%) and two (6.1%) on co-infection. Among these, six (18.2%) addressed the diagnosis of infection, 16 (48.5%) the prognosis of infection, eight (24.2%) the prediction of treatment response, two (6.1%) progression through a cascade of care, and one (3.03%) focused on the choice of antiretroviral therapy (ART). Nineteen prediction models (57.6%) were developed using data from high-income countries. Evaluation of prediction models was limited to measures of performance. Detailed information on software code accessibility was often missing. Independent validation on new datasets and/or in other institutions was rarely done.

CONCLUSION

Promising approaches for ML prediction models in blood-borne viral infections were identified, but the lack of robust validation, interpretability/explainability, and poor quality of reporting hampered their clinical relevance. Our findings highlight important considerations that can inform standard reporting guidelines for ML prediction models in the future (e.g., TRIPOD-AI), and provides critical data to inform robust evaluation of the models.

Detecting transthyretin amyloid cardiomyopathy (ATTR-CM) using machine learning: an evaluation of the performance of an algorithm in a UK setting

OBJECTIVE

The aim of this study was to evaluate the potential real-world application of a machine learning (ML) algorithm, developed and trained on heart failure (HF) cohorts in the USA, to detect patients with undiagnosed wild type cardiac amyloidosis (ATTRwt) in the UK.

DESIGN

In this retrospective observational study, anonymised, linked primary and secondary care data (Clinical Practice Research Datalink GOLD and Hospital Episode Statistics, respectively, were used to identify patients diagnosed with HF between 2009 and 2018 in the UK. International Classification of Diseases (ICD)-10 clinical modification codes were matched to equivalent Read (primary care) and ICD-10 WHO (secondary care) diagnosis codes used in the UK. In the absence of specific Read or ICD-10 WHO codes for ATTRwt, two proxy case definitions (definitive and possible cases) based on the degree of confidence that the contributing codes defined true ATTRwt cases were created using ML.

PRIMARY OUTCOME MEASURE

Algorithm performance was evaluated primarily using the area under the receiver operating curve (AUROC) by comparing the actual versus algorithm predicted case definitions at varying sensitivities and specificities.

RESULTS

The algorithm demonstrated strongest predictive ability when a combination of primary care and secondary care data were used (AUROC: 0.84 in definitive cohort and 0.86 in possible cohort). For primary care or secondary care data alone, performance ranged from 0.68 to 0.78.

CONCLUSION

The ML algorithm, despite being developed in a US population, was effective at identifying patients that may have ATTRwt in a UK setting. Its potential use in research and clinical care to aid identification of patients with undiagnosed ATTRwt, possibly enabling earlier diagnosis in the disease pathway, should be investigated.

Scoping Meta-Review of Methods Used to Assess Artificial Intelligence-Based Medical Devices for Heart Failure

Artificial intelligence and machine learning (AI/ML) are playing increasingly important roles, permeating the field of medical devices (MDs). This rapid progress has not yet been matched by the Health Technology Assessment (HTA) process, which still needs to define a common methodology for assessing AI/ML-based MDs. To collect existing evidence from the literature about the methods used to assess AI-based MDs, with a specific focus on those used for the management of heart failure (HF), the International Federation of Medical and Biological Engineering (IFMBE) conducted a scoping meta-review. This manuscript presents the results of this search, which covered the period from January 1974 to October 2022. After careful independent screening, 21 reviews, mainly conducted in North America and Europe, were retained and included. Among the findings were that deep learning is the most commonly utilised method and that electronic health records and registries are among the most prevalent sources of data for AI/ML algorithms. Out of the 21 included reviews, 19 focused on risk prediction and/or the early diagnosis of HF. Furthermore, 10 reviews provided evidence of the impact on the incidence/progression of HF, and 13 on the length of stay. From an HTA perspective, the main areas requiring improvement are the quality assessment of studies on AI/ML (included in 11 out of 21 reviews) and their data sources, as well as the definition of the criteria used to assess the selection of the most appropriate AI/ML algorithm.

Social Determinants of Health Data Improve the Prediction of Cardiac Outcomes in Females with Breast Cancer

Cardiovascular disease is the leading cause of mortality among breast cancer (BC) patients aged 50 and above. Machine Learning (ML) models are increasingly utilized as prediction tools, and recent evidence suggests that incorporating social determinants of health (SDOH) data can enhance its performance. This study included females ≥ 18 years diagnosed with BC at any stage. The outcomes were the diagnosis and time-to-event of major adverse cardiovascular events (MACEs) within two years following a cancer diagnosis. Covariates encompassed demographics, risk factors, individual and neighborhood-level SDOH, tumor characteristics, and BC treatment. Race-specific and race-agnostic Extreme Gradient Boosting ML models with and without SDOH data were developed and compared based on their C-index. Among 4309 patients, 11.4% experienced a 2-year MACE. The race-agnostic models exhibited a C-index of 0.78 (95% CI 0.76-0.79) and 0.81 (95% CI 0.80-0.82) without and with SDOH data, respectively. In non-Hispanic Black women (NHB; n = 765), models without and with SDOH data achieved a C-index of 0.74 (95% CI 0.72-0.76) and 0.75 (95% CI 0.73-0.78), respectively. Among non-Hispanic White women (n = 3321), models without and with SDOH data yielded a C-index of 0.79 (95% CI 0.77-0.80) and 0.79 (95% CI 0.77-0.80), respectively. In summary, including SDOH data improves the predictive performance of ML models in forecasting 2-year MACE among BC females, particularly within NHB.

Coronary Stent Infections - A Systematic Review and Meta-Analysis

BACKGROUND

Coronary stent infection (CSI) represents a rare but potentially fatal complication of percutaneous coronary interventions (PCI). A systematic review and meta-analysis of published reports was performed to profile CSI and its management strategies.

METHODS

Online database searches were performed using MeSH and keywords. The primary outcome of the study was in-hospital mortality. A unique Artificial Intelligence-based predictive model was developed for need for delayed surgery and probability of survival on medical therapy alone.

RESULTS

A total of 79 subjects were included in the study. Twenty eight (35.0 %) patients had type 2 diabetes mellitus. Subjects most commonly reported symptoms within the first week of the procedure (43 %). Fever was the most common initial symptom (72 %). Thirty eight percent of patients presented with acute coronary syndrome. The presence of mycotic aneurysms was described in 62 % of patients. Staphylococcus species were the most common (65 %) isolated organism. The primary outcome of in-hospital mortality was seen in a total of 24 patients out of 79 (30.3 %). A comparative univariate analysis comparing those encountering in-hospital mortality versus otherwise revealed the presence of structural heart disease (83 % mortality vs 17 % survival, p = 0.009), and the presence of non ST elevation acute coronary syndrome (11 % mortality vs 88 % survival, p = 0.03), to be a statistically significant factor predicting in-hospital mortality. In an analysis between patients with successful versus failed initial medical therapy, patients from private teaching hospitals (80.0 % vs 20.0 %; p = 0.01, n = 10) had a higher survival with medical therapy alone.

CONCLUSION

CSI is a highly under-studied disease entity with largely unknown risk factors and clinical outcomes. Larger studies are needed to further define the characteristics of CSI. (PROSPERO ID CRD42021216031).

Machine Learning Prediction of Objective Hearing Loss With Demographics, Clinical Factors, and Subjective Hearing Status

OBJECTIVE

Hearing loss (HL) is highly prevalent, yet underrecognized and underdiagnosed. Lack of standardized screening, awareness, cost, and access to hearing testing present barriers to HL identification. To facilitate prescreening and selection of patients who warrant audiometric evaluation, we developed a machine learning (ML) model to predict speech-frequency pure-tone average (PTA).

STUDY DESIGN

Cross-sectional study.

SETTING

National Health and Nutrition Examination Survey (NHANES).

METHODS

The cohort included 8918 adults (≥20 years) who completed audiometric testing with NHANES (2012-2018). The primary outcome measure was the prediction of better hearing ear speech-frequency PTA. Relevant predictors included demographics, medical conditions, and subjective assessment of hearing. Supervised ML with a tree-based architecture was used. Regression performance was determined by the mean absolute error (MAE) with binary classification assessed with area under the receiver operating characteristic curve (AUC).

RESULTS

Using the full set of predictors, the test set MAE between the ML-predicted and actual PTA was 5.29 dB HL (95% confidence interval [CI]: 4.97-5.61). The 5 most influential predictors of higher PTA were increased age, worse subjective hearing, male gender, increased body mass index, and history of smoking. The 5-factor abbreviated model performed comparably to the extended feature set with MAE 5.36 (95% CI: 5.03-5.69) and AUC for PTA > 25 dB HL of 0.92 (95% CI: 0.90-0.94).

CONCLUSION

The ML model was able to predict PTA with patient demographics, clinical factors, and subjective hearing status. ML-based prediction may be used to identify individuals who could benefit most from audiometric evaluation.

A machine-learning based bio-psycho-social model for the prediction of non-obstructive and obstructive coronary artery disease

BACKGROUND

Mechanisms of myocardial ischemia in obstructive and non-obstructive coronary artery disease (CAD), and the interplay between clinical, functional, biological and psycho-social features, are still far to be fully elucidated.

OBJECTIVES

To develop a machine-learning (ML) model for the supervised prediction of obstructive versus non-obstructive CAD.

METHODS

From the EVA study, we analysed adults hospitalized for IHD undergoing conventional coronary angiography (CCA). Non-obstructive CAD was defined by a stenosis < 50% in one or more vessels. Baseline clinical and psycho-socio-cultural characteristics were used for computing a Rockwood and Mitnitski frailty index, and a gender score according to GENESIS-PRAXY methodology. Serum concentration of inflammatory cytokines was measured with a multiplex flow cytometry assay. Through an XGBoost classifier combined with an explainable artificial intelligence tool (SHAP), we identified the most influential features in discriminating obstructive versus non-obstructive CAD.

RESULTS

Among the overall EVA cohort (n = 509), 311 individuals (mean age 67 ± 11 years, 38% females; 67% obstructive CAD) with complete data were analysed. The ML-based model (83% accuracy and 87% precision) showed that while obstructive CAD was associated with higher frailty index, older age and a cytokine signature characterized by IL-1β, IL-12p70 and IL-33, non-obstructive CAD was associated with a higher gender score (i.e., social characteristics traditionally ascribed to women) and with a cytokine signature characterized by IL-18, IL-8, IL-23.

CONCLUSIONS

Integrating clinical, biological, and psycho-social features, we have optimized a sex- and gender-unbiased model that discriminates obstructive and non-obstructive CAD. Further mechanistic studies will shed light on the biological plausibility of these associations.

CLINICAL TRIAL REGISTRATION

NCT02737982.

Hyperparameter optimization for cardiovascular disease data-driven prognostic system

Prediction and diagnosis of cardiovascular diseases (CVDs) based, among other things, on medical examinations and patient symptoms are the biggest challenges in medicine. About 17.9 million people die from CVDs annually, accounting for 31% of all deaths worldwide. With a timely prognosis and thorough consideration of the patient's medical history and lifestyle, it is possible to predict CVDs and take preventive measures to eliminate or control this life-threatening disease. In this study, we used various patient datasets from a major hospital in the United States as prognostic factors for CVD. The data was obtained by monitoring a total of 918 patients whose criteria for adults were 28-77 years old. In this study, we present a data mining modeling approach to analyze the performance, classification accuracy and number of clusters on Cardiovascular Disease Prognostic datasets in unsupervised machine learning (ML) using the Orange data mining software. Various techniques are then used to classify the model parameters, such as k-nearest neighbors, support vector machine, random forest, artificial neural network (ANN), naïve bayes, logistic regression, stochastic gradient descent (SGD), and AdaBoost. To determine the number of clusters, various unsupervised ML clustering methods were used, such as k-means, hierarchical, and density-based spatial clustering of applications with noise clustering. The results showed that the best model performance analysis and classification accuracy were SGD and ANN, both of which had a high score of 0.900 on Cardiovascular Disease Prognostic datasets. Based on the results of most clustering methods, such as k-means and hierarchical clustering, Cardiovascular Disease Prognostic datasets can be divided into two clusters. The prognostic accuracy of CVD depends on the accuracy of the proposed model in determining the diagnostic model. The more accurate the model, the better it can predict which patients are at risk for CVD.

A Systematic Review: Do the Use of Machine Learning, Deep Learning, and Artificial Intelligence Improve Patient Outcomes in Acute Myocardial Ischemia Compared to Clinician-Only Approaches?

Cardiovascular diseases (CVDs) present a significant global health challenge and remain a primary cause of death. Early detection and intervention are crucial for improved outcomes in acute coronary syndrome (ACS), particularly acute myocardial infarction (AMI) cases. Artificial intelligence (AI) can detect heart disease early by analyzing patient information and electrocardiogram (ECG) data, providing invaluable insights into this critical health issue. However, the imbalanced nature of ECG and patient data presents challenges for traditional machine learning (ML) algorithms in performing unbiasedly. Investigators have proposed various data-level and algorithm-level solutions to overcome these challenges. In this study, we used a systematic literature review (SLR) approach to give an overview of the current literature and to highlight the difficulties of utilizing ML, deep learning (DL), and AI algorithms in predicting, diagnosing, and prognosis of heart diseases. We reviewed 181 articles from reputable journals published between 2013 and June 15, 2023, focusing on eight selected papers for in-depth analysis. The analysis considered factors such as heart disease type, algorithms used, applications, and proposed solutions and compared the benefits of algorithms combined with clinicians versus clinicians alone. This systematic review revealed that the current ML-based diagnostic approaches face several open problems and issues when implementing ML, DL, and AI in real-life settings. Although these algorithms show higher sensitivities, specificities, and accuracies in detecting heart disease, we must address the ethical concerns while implementing these models into clinical practice. The transparency of how these algorithms operate remains a challenge. Nevertheless, further exploration and research in ML, DL, and AI are necessary to overcome these challenges and fully harness their potential to improve health outcomes for patients with AMI.

The role of artificial intelligence in hypertensive disorders of pregnancy: towards personalized healthcare

INTRODUCTION

Guidelines advise ongoing follow-up of patients after hypertensive disorders of pregnancy (HDP) to assess cardiovascular risk and manage future patient-specific pregnancy conditions. However, there are limited tools available to monitor patients, with those available tending to be simple risk assessments that lack personalization. A promising approach could be the emerging artificial intelligence (AI)-based techniques, developed from big patient datasets to provide personalized recommendations for preventive advice.

AREAS COVERED

In this narrative review, we discuss the impact of integrating AI and big data analysis for personalized cardiovascular care, focusing on the management of HDP.

EXPERT OPINION

The pathophysiological response of women to pregnancy varies, and deeper insight into each response can be gained through a deeper analysis of the medical history of pregnant women based on clinical records and imaging data. Further research is required to be able to implement AI for clinical cases using multi-modality and multi-organ assessment, and this could expand both knowledge on pregnancy-related disorders and personalized treatment planning.

Bibliometric analysis of 100 top cited articles of heart failure-associated diseases in combination with machine learning

Objective

The aim of this paper is to analyze the application of machine learning in heart failure-associated diseases using bibliometric methods and to provide a dynamic and longitudinal bibliometric analysis of heart failure-related machine learning publications.

Materials and methods

Web of Science was screened to gather the articles for the study. Based on bibliometric indicators, a search strategy was developed to screen the title for eligibility. Intuitive data analysis was employed to analyze the top-100 cited articles and VOSViewer was used to analyze the relevance and impact of all articles. The two analysis methods were then compared to get conclusions.

Results

The search identified 3,312 articles. In the end, 2,392 papers were included in the study, which were published between 1985 and 2023. All articles were analyzed using VOSViewer. Key points of the analysis included the co-authorship map of authors, countries and organizations, the citation map of journal and documents and a visualization of keyword co-occurrence analysis. Among these 100 top-cited papers, with a mean of 122.9 citations, the most-cited article had 1,189, and the least cited article had 47. Harvard University and the University of California topped the list among all institutes with 10 papers each. More than one-ninth of the authors of these 100 top-cited papers wrote three or more articles. The 100 articles came from 49 journals. The articles were divided into seven areas according to the type of machine learning approach employed: Support Vector Machines, Convolutional Neural Networks, Logistic Regression, Recurrent Neural Networks, Random Forest, Naive Bayes, and Decision Tree. Support Vector Machines were the most popular method.

Conclusions

This analysis provides a comprehensive overview of the artificial intelligence (AI)-related research conducted in the field of heart failure, which helps healthcare institutions and researchers better understand the prospects of AI in heart failure and formulate more scientific and effective research plans. In addition, our bibliometric evaluation can assist healthcare institutions and researchers in determining the advantages, sustainability, risks, and potential impacts of AI technology in heart failure.

Risk prediction of heart failure in patients with ischemic heart disease using network analytics and stacking ensemble learning

BACKGROUND

Heart failure (HF) is a major complication following ischemic heart disease (IHD) and it adversely affects the outcome. Early prediction of HF risk in patients with IHD is beneficial for timely intervention and for reducing disease burden.

METHODS

Two cohorts, cases for patients first diagnosed with IHD and then with HF (N = 11,862) and control IHD patients without HF (N = 25,652), were established from the hospital discharge records in Sichuan, China during 2015-2019. Directed personal disease network (PDN) was constructed for each patient, and then these PDNs were merged to generate the baseline disease network (BDN) for the two cohorts, respectively, which identifies the health trajectories of patients and the complex progression patterns. The differences between the BDNs of the two cohort was represented as disease-specific network (DSN). Three novel network features were exacted from PDN and DSN to represent the similarity of disease patterns and specificity trends from IHD to HF. A stacking-based ensemble model DXLR was proposed to predict HF risk in IHD patients using the novel network features and basic demographic features (i.e., age and sex). The Shapley Addictive exPlanations method was applied to analyze the feature importance of the DXLR model.

RESULTS

Compared with the six traditional machine learning models, our DXLR model exhibited the highest AUC (0.934 ± 0.004), accuracy (0.857 ± 0.007), precision (0.723 ± 0.014), recall (0.892 ± 0.012) and F₁ score (0.798 ± 0.010). The feature importance showed that the novel network features ranked as the top three features, playing a notable role in predicting HF risk of IHD patient. The feature comparison experiment also indicated that our novel network features were superior to those proposed by the state-of-the-art study in improving the performance of the prediction model, with an increase in AUC by 19.9%, in accuracy by 18.7%, in precision by 30.7%, in recall by 37.4%, and in F₁ score by 33.7%.

CONCLUSIONS

Our proposed approach that combines network analytics and ensemble learning effectively predicts HF risk in patients with IHD. This highlights the potential value of network-based machine learning in disease risk prediction field using administrative data.