Development of heart failure risk prediction models based on a multi-marker approach using random forest algorithms

Artificial intelligence and heart failure: A state-of-the-art review

Heart failure (HF) is a heterogeneous syndrome affecting more than 60 million individuals globally. Despite recent advancements in understanding of the pathophysiology of HF, many issues remain including residual risk despite therapy, understanding the pathophysiology and phenotypes of patients with HF and preserved ejection fraction, and the challenges related to integrating a large amount of disparate information available for risk stratification and management of these patients. Risk prediction algorithms based on artificial intelligence (AI) may have superior predictive ability compared to traditional methods in certain instances. AI algorithms can play a pivotal role in the evolution of HF care by facilitating clinical decision making to overcome various challenges such as allocation of treatment to patients who are at highest risk or are more likely to benefit from therapies, prediction of adverse outcomes, and early identification of patients with subclinical disease or worsening HF. With the ability to integrate and synthesize large amounts of data with multidimensional interactions, AI algorithms can supply information with which physicians can improve their ability to make timely and better decisions. In this review, we provide an overview of the AI algorithms that have been developed for establishing early diagnosis of HF, phenotyping HF with preserved ejection fraction, and stratifying HF disease severity. This review also discusses the challenges in clinical deployment of AI algorithms in HF, and the potential path forward for developing future novel learning-based algorithms to improve HF care.

Development and comparison of machine learning-based models for predicting heart failure after acute myocardial infarction

AIMS

Heart failure (HF) is one of the common adverse cardiovascular events after acute myocardial infarction (AMI), but the predictive efficacy of numerous machine learning (ML) built models is unclear. This study aimed to build an optimal model to predict the occurrence of HF in AMI patients by comparing seven ML algorithms.

METHODS

Cohort 1 included AMI patients from 2018 to 2019 divided into HF and control groups. All first routine test data of the study subjects were collected as the features to be selected for the model, and seven ML algorithms with screenable features were evaluated. Cohort 2 contains AMI patients from 2020 to 2021 to establish an early warning model with external validation. ROC curve and DCA curve to analyze the diagnostic efficacy and clinical benefit of the model respectively.

RESULTS

The best performer among the seven ML algorithms was XgBoost, and the features of XgBoost algorithm for troponin I, triglycerides, urine red blood cell count, γ-glutamyl transpeptidase, glucose, urine specific gravity, prothrombin time, prealbumin, and urea were ranked high in importance. The AUC of the HF-Lab9 prediction model built by the XgBoost algorithm was 0.966 and had good clinical benefits.

CONCLUSIONS

This study screened the optimal ML algorithm as XgBoost and developed the model HF-Lab9 will improve the accuracy of clinicians in assessing the occurrence of HF after AMI and provide a reference for the selection of subsequent model-building algorithms.

Novel Biomarkers and the Multiple-Marker Approach in Early Detection, Prognosis, and Risk Stratification of Cardiac Diseases: A Narrative Review

Cardiac diseases are a primary cause of mortality worldwide, underscoring the importance of early identification and risk stratification to enhance patient outcomes. Biomarkers have become important tools for the risk assessment of cardiovascular disease and monitoring disease progression. This narrative review focuses on the multiple-marker approach, which involves simultaneously evaluating several biomarkers for the early detection and risk stratification of heart diseases. The review covers the clinical applications of novel biomarkers, such as high-sensitivity troponin, galectin-3, source of tumorigenicity 2, B-type natriuretic peptide and N-terminal pro-B-type natriuretic peptide, growth differentiation factor 15, myeloperoxidase, fatty acid-binding protein, C-reactive protein, lipoprotein-associated phospholipase A2, microRNAs, circulating endothelial cells, and ischemia-modified albumin. These biomarkers have demonstrated potential in identifying people who are at high risk for developing heart disease and in providing prognostic data. Given the complexity of cardiac illnesses, the multiple-marker approach to risk assessment is extremely beneficial. Implementing the multiple-marker strategy can improve risk stratification, diagnostic accuracy, and patient care in heart disease patients.

Research on Establishing Corneal Edema after Phacoemulsification Prediction Model Based on Variable Selection with Copula Entropy

BACKGROUND

Corneal edema (CE) affects the outcome of phacoemulsification. Effective ways to predict the CE after phacoemulsification are needed.

METHODS

On the basis of data from patients conforming to the protocol of the AGSPC trial, 17 variables were selected to predict CE after phacoemulsification by constructing a CE nomogram through multivariate logistic regression, which was improved via variable selection with copula entropy. The prediction models were evaluated using predictive accuracy, the area under the receiver operating characteristic curve (AUC), and decision curve analysis (DCA).

RESULTS

Data from 178 patients were used to construct prediction models. After copula entropy variable selection, which shifted the variables used for prediction in the CE nomogram from diabetes, best corrected visual acuity (BCVA), lens thickness and cumulative dissipated energy (CDE) to CDE and BCVA in the Copula nomogram, there was no significant change in predictive accuracy (0.9039 vs. 0.9098). There was also no significant difference in AUCs between the CE nomogram and the Copula nomogram (0.9637, 95% CI 0.9329-0.9946 vs. 0.9512, 95% CI 0.9075-0.9949; p = 0.2221). DCA suggested that the Copula nomogram has clinical application.

CONCLUSIONS

This study obtained a nomogram with good performance to predict CE after phacoemulsification, and showed the improvement of copula entropy for nomogram models.

Advances in congestive heart failure biomarkers

Congestive heart failure (CHF) is the leading cause of morbidity and mortality in the elderly worldwide. Although many biomarkers associated with in heart failure, these are generally prognostic and identify patients with moderate and severe disease. Unfortunately, the role of biomarkers in decision making for early and advanced heart failure remains largely unexplored. Previous studies suggest the natriuretic peptides have the potential to improve the diagnosis of heart failure, but they still have significant limitations related to cut-off values. Although some promising cardiac biomarkers have emerged, comprehensive data from large cohort studies is lacking. The utility of multiple biomarkers that reflect various pathophysiologic pathways are increasingly being explored in heart failure risk stratification and to diagnose disease conditions promptly and accurately. MicroRNAs serve as mediators and/or regulators of renin-angiotensin-induced cardiac remodeling by directly targeting enzymes, receptors and signaling molecules. The role of miRNA in HF diagnosis is a promising area of research and further exploration may offer both diagnostic and prognostic applications and phenotype-specific targets. In this review, we provide insight into the classification of different biochemical and molecular markers associated with CHF, examine clinical usefulness in CHF and highlight the most clinically relevant.

Identification of energy metabolism-related biomarkers for risk prediction of heart failure patients using random forest algorithm

Objective

Energy metabolism plays a crucial role in the improvement of heart dysfunction as well as the development of heart failure (HF). The current study is designed to identify energy metabolism-related diagnostic biomarkers for predicting the risk of HF due to myocardial infarction.

Methods

Transcriptome sequencing data of HF patients and non-heart failure (NF) people (GSE66360 and GSE59867) were obtained from gene expression omnibus (GEO) database. Energy metabolism-related differentially expressed genes (DEGs) were screened between HF and NF samples. The subtyping consistency analysis was performed to enable the samples to be grouped. The immune infiltration level among subtypes was assessed by single sample gene set enrichment analysis (ssGSEA). Random forest algorithm (RF) and support vector machine (SVM) were applied to identify diagnostic biomarkers, and the receiver operating characteristic curves (ROC) was plotted to validate the accuracy. Predictive nomogram was constructed and validated based on the result of the RF. Drug screening and gene-miRNA network were analyzed to predict the energy metabolism-related drugs and potential molecular mechanism.

Results

A total of 22 energy metabolism-related DEGs were identified between HF and NF patients. The clustering analysis showed that HF patients could be classified into two subtypes based on the energy metabolism-related genes, and functional analyses demonstrated that the identified DEGs among two clusters were mainly involved in immune response regulating signaling pathway and lipid and atherosclerosis. ssGSEA analysis revealed that there were significant differences in the infiltration levels of immune cells between two subtypes of HF patients. Random-forest and support vector machine algorithm eventually identified ten diagnostic markers (MEF2D, RXRA, PPARA, FOXO1, PPARD, PPP3CB, MAPK14, CREB1, MEF2A, PRMT1) for risk prediction of HF patients, and the proposed nomogram resulted in good predictive performance (GSE66360, AUC = 0.91; GSE59867, AUC = 0.84) and the clinical usefulness in HF patients. More importantly, 10 drugs and 15 miRNA were predicted as drug target and hub miRNA that associated with energy metabolism-related genes, providing further information on clinical HF treatment.

Conclusion

This study identified ten energy metabolism-related diagnostic markers using random forest algorithm, which may help optimize risk stratification and clinical treatment in HF patients.

The amputation and mortality of inpatients with diabetic foot ulceration in the COVID-19 pandemic and postpandemic era: A machine learning study

This study aimed to explore the clinical characteristic and outcomes of inpatients with diabetic foot ulceration (DFU) in 2019 (prelockdown) and 2020 (postlockdown) due to the COVID-19 pandemic, at an emergency medical service unit. Prediction models for mortality and amputation were developed to describe the risk factors using a machine learning-based approach. Hospitalized DFU patients (N = 23) were recruited after the lockdown in 2020 and matched with corresponding inpatients (N = 23) before lockdown in 2019. Six widely used machine learning models were built and internally validated using 3-fold cross-validation to predict the risk of amputation and death in DFU inpatients under the COVID-19 pandemic. Previous DF ulcers, prehospital delay, and mortality were significantly higher in 2020 compared to 2019. Diabetic foot patients in 2020 had higher hs-CRP levels (P = .037) but lower hemoglobin levels (P = .017). The extreme gradient boosting (XGBoost) performed best in all models for predicting amputation and mortality with the highest area under the curve (0.86 and 0.94), accuracy (0.80 and 0.90), sensitivity (0.67 and 1.00), and negative predictive value (0.86 and 1.00). A long delay in admission and a higher risk of mortality was observed in patients with DFU who attended the emergency center during the COVID-19 post lockdown. The XGBoost model can provide evidence-based risk information for patients with DFU regarding their amputation and mortality. The prediction models would benefit DFU patients during the COVID-19 pandemic.

Application of metagenomic next-generation sequencing in the diagnosis of pulmonary invasive fungal disease

Background

Metagenomic next-generation sequencing (mNGS) is increasingly being used to detect pathogens directly from clinical specimens. However, the optimal application of mNGS and subsequent result interpretation can be challenging. In addition, studies reporting the use of mNGS for the diagnosis of invasive fungal infections (IFIs) are rare.

Objective

We critically evaluated the performance of mNGS in the diagnosis of pulmonary IFIs, by conducting a multicenter retrospective analysis. The methodological strengths of mNGS were recognized, and diagnostic cutoffs were determined.

Methods

A total of 310 patients with suspected pulmonary IFIs were included in this study. Conventional microbiological tests (CMTs) and mNGS were performed in parallel on the same set of samples. Receiver operating characteristic (ROC) curves were used to evaluate the performance of the logarithm of reads per kilobase per million mapped reads [lg(RPKM)], and read counts were used to predict true-positive pathogens.

Result

The majority of the selected patients (86.5%) were immunocompromised. Twenty species of fungi were detected by mNGS, which was more than was achieved with standard culture methods. Peripheral blood lymphocyte and monocyte counts, as well as serum albumin levels, were significantly negatively correlated with fungal infection. In contrast, C-reactive protein and procalcitonin levels showed a significant positive correlation with fungal infection. ROC curves showed that mNGS [and especially lg(RPKM)] was superior to CMTs in its diagnostic performance. The area under the ROC curve value obtained for lg(RPKM) in the bronchoalveolar lavage fluid of patients with suspected pulmonary IFIs, used to predict true-positive pathogens, was 0.967, and the cutoff value calculated from the Youden index was −5.44.

Conclusions

In this study, we have evaluated the performance of mNGS-specific indicators that can identify pathogens in patients with IFIs more accurately and rapidly than CMTs, which will have important clinical implications.

Early prediction of clinical scores for left ventricular reverse remodeling using extreme gradient random forest, boosting, and logistic regression algorithm representations

Objective

At present, there is no early prediction model of left ventricular reverse remodeling (LVRR) for people who are in cardiac arrest with an ejection fraction (EF) of ≤35% at first diagnosis; thus, the purpose of this article is to provide a supplement to existing research.

Materials and methods

A total of 109 patients suffering from heart attack with an EF of ≤35% at first diagnosis were involved in this single-center research study. LVRR was defined as an absolute increase in left ventricular ejection fraction (LVEF) from ≥10% to a final value of >35%, with analysis features including demographic characteristics, diseases, biochemical data, echocardiography, and drug therapy. Extreme gradient boosting (XGBoost), random forest, and logistic regression algorithm models were used to distinguish between LVRR and non-LVRR cases and to obtain the most important features.

Results

There were 47 cases (42%) of LVRR in patients suffering from heart failure with an EF of ≤35% at first diagnosis after optimal drug therapy. General statistical analysis and machine learning methods were combined to exclude a number of significant feature groups. The median duration of disease in the LVRR group was significantly lower than that in the non-LVRR group (7 vs. 48 months); the mean values of creatine kinase (CK) and MB isoenzyme of creatine kinase (CK-MB) in the LVRR group were lower than those in the non-LVRR group (80.11 vs. 94.23 U/L; 2.61 vs. 2.99 ng/ml; 27.19 vs. 28.54 mm). Moreover, AUC values for our feature combinations ranged from 97 to 94% and to 87% when using the XGBoost, random forest, and logistic regression techniques, respectively. The ablation test revealed that beats per minute (BPM) and disease duration had a greater impact on the model’s ability to accurately forecast outcomes.

Conclusion

Shorter disease duration, slightly lower CK and CK-MB levels, slightly smaller right and left ventricular and left atrial dimensions, and lower mean heart rates were found to be most strongly predictive of LVRR development (BPM).

Machine learning for subtype definition and risk prediction in heart failure, acute coronary syndromes and atrial fibrillation: systematic review of validity and clinical utility

BACKGROUND

Machine learning (ML) is increasingly used in research for subtype definition and risk prediction, particularly in cardiovascular diseases. No existing ML models are routinely used for cardiovascular disease management, and their phase of clinical utility is unknown, partly due to a lack of clear criteria. We evaluated ML for subtype definition and risk prediction in heart failure (HF), acute coronary syndromes (ACS) and atrial fibrillation (AF).

METHODS

For ML studies of subtype definition and risk prediction, we conducted a systematic review in HF, ACS and AF, using PubMed, MEDLINE and Web of Science from January 2000 until December 2019. By adapting published criteria for diagnostic and prognostic studies, we developed a seven-domain, ML-specific checklist.

RESULTS

Of 5918 studies identified, 97 were included. Across studies for subtype definition (n = 40) and risk prediction (n = 57), there was variation in data source, population size (median 606 and median 6769), clinical setting (outpatient, inpatient, different departments), number of covariates (median 19 and median 48) and ML methods. All studies were single disease, most were North American (n = 61/97) and only 14 studies combined definition and risk prediction. Subtype definition and risk prediction studies respectively had limitations in development (e.g. 15.0% and 78.9% of studies related to patient benefit; 15.0% and 15.8% had low patient selection bias), validation (12.5% and 5.3% externally validated) and impact (32.5% and 91.2% improved outcome prediction; no effectiveness or cost-effectiveness evaluations).

CONCLUSIONS

Studies of ML in HF, ACS and AF are limited by number and type of included covariates, ML methods, population size, country, clinical setting and focus on single diseases, not overlap or multimorbidity. Clinical utility and implementation rely on improvements in development, validation and impact, facilitated by simple checklists. We provide clear steps prior to safe implementation of machine learning in clinical practice for cardiovascular diseases and other disease areas.

RETRACTED: Down-regulated microRNA-195-5p and up-regulated CXCR4 attenuates the heart function injury of heart failure mice via inactivating JAK/STAT pathway

This article has been retracted: please see Elsevier Policy on Article Withdrawal (http://www.elsevier.com/locate/withdrawalpolicy). This article has been retracted at the request of the Editor-in-Chief. Concern was raised about the integrity of the images in Figure 6, which appear to contain suspected image duplications, as detailed here: https://pubpeer.com/publications/A31DE9EEF13ED6B88BCC86A9CAC8D9 and here: https://docs.google.com/spreadsheets/d/1r0MyIYpagBc58BRF9c3luWNlCX8VUvUuPyYYXzxWvgY/edit#gid=262337249. Most of these image duplications involve either pasting portions of one image into another, or rotating/flipping the image. Numerous additional suspected image duplications were detected within Figures 2A and 7A. The journal requested the corresponding author comment on these concerns and provide the raw data. The authors did not respond to this request and therefore the Editor-in-Chief decided to retract the article.