Machine Learning – ABC CARDIOL

ABCHF

Original Article

Global Mapping of Research on Artificial Intelligence Applied to Myocarditis

Guilherme Cesar Fernandes de Oliveira da Costa et al. Jose Gregorio Valero Rodriguez Giulia Fernandes de Oliveira da Costa Adriana Munford Lima Pimentel Evandro Tinoco Mesquita

Abstract

Background

Myocarditis remains a diagnostic challenge, and interest in using artificial intelligence to support noninvasive diagnostic evaluation has grown rapidly. Although the literature at the intersection of artificial intelligence and myocarditis has expanded substantially, it remains fragmented and lacks comprehensive synthesis.

Objectives

To map the global scientific output on the application of artificial intelligence to myocarditis by identifying leading contributors, collaboration patterns, and emerging thematic trends.

Methods

We conducted a bibliometric analysis of articles on artificial intelligence applied to myocarditis indexed in the Web of Science Core Collection. The R package bibliometrix and VOSviewer were used to perform bibliometric network and thematic analyses.

Results

A total of 47 articles were included. Scientific output increased markedly after 2021, coinciding with the COVID-19 pandemic, and peaked in 2025. China led in publication volume, whereas the United Kingdom and Germany achieved the greatest citation impact. No publications originated from Latin America, the Caribbean, or Africa. International collaboration networks remain fragmented. Keyword and thematic analyses identified the integration of deep learning with cardiovascular magnetic resonance imaging as the principal motor theme, reflecting a shift away from traditional endomyocardial biopsy. The most highly cited authors were primarily computer and data scientists.

Conclusions

Research at the intersection of artificial intelligence and myocarditis is an emerging field undergoing methodological consolidation, with growth accelerated by the COVID-19 pandemic. Cardiovascular magnetic resonance imaging has become the primary data source for artificial intelligence-based applications. The principal challenge appears to be translational, underscoring the need for stronger collaboration between algorithm developers and clinicians. The successful integration of these models into routine clinical practice will likely depend on multidisciplinary collaboration and validation across diverse, multicenter populations.

Read Full Article

ABC

Original Article

Artificial Intelligence-Derived ECG-Age as a Predictor of Mortality and Cardiovascular Events: A Systematic Review and Meta-Analysis

Isadora Cristine Reis Sguizzato Bozzi et al. Maria Clara de Araujo Gontijo Lima Antonio Luiz Pinho Ribeiro Gabriela Miana de Mattos Paixão

Abstract

Background

Artificial intelligence (AI)-derived electrocardiographic age (ECG-age) and the difference between ECG-age and chronological age (delta-age) are emerging biomarkers of cardiovascular aging and adverse outcomes, but their prognostic value remains unclear.

Objectives

We performed a systematic review and meta-analysis to evaluate associations between AI-derived ECG-age or delta-age and clinical outcomes.

Methods

Nine databases were searched through May 1, 2025. Eligible studies evaluated mortality or cardiovascular events and reported measures of association. Pooled hazard ratios (pHRs) with 95% confidence intervals (CIs) were calculated via fixed- or random-effects models, with statistical significance set at p < 0.05. The protocol was registered in PROSPERO (CRD420251042467).

Results

Ten studies (2021–2025) with over 550,000 participants from East Asia, the Americas, and the UK were included. Most used convolutional neural networks to estimate ECG-age; delta-age was calculated as ECG-age minus chronological age. Elevated delta-age was associated with increased all-cause mortality (pHR = 1.83, 95% CI: 1.45–2.32) and cardiovascular mortality (pHR = 2.63, 95% CI: 1.93–3.58). Three studies reported increased risk of atrial fibrillation (pHR = 1.96, 95% CI: 1.43–2.69), although data were limited and heterogeneous. Descriptive analyses suggested that greater delta-age predicts incident heart failure and stroke.

Conclusions

AI-derived ECG-age, particularly delta-age, is associated with all-cause and cardiovascular mortality, supporting its role as a noninvasive biomarker of cardiovascular aging. Evidence for atrial fibrillation is suggestive but limited. Standardized algorithms, robust external validation, and prospective multicenter studies are needed to confirm clinical utility and integrate ECG-age into risk stratification, even in asymptomatic individuals.

Read Full Article

ABC

Original Article

A New Risk Score Based on Machine Learning in Patients with Acute Heart Failure: The ML-HF score

Matheus Bissa Duarte Ferreira et al. Jorge Tadashi Daikubara Neto Gustavo S. Pereira da Cunha Rafael Moretti … Jessica Tamires Reichert Lucas Müller Prado Raphael Henrique Déa Cirino Sidney C Smith Jr Fábio Papa Taniguchi Andrei C. Sposito Odilson M. Silvestre Wilson Nadruz Miguel Morita Fernandes-Silva

Abstract

Background

Conventional prognostic assessment scores often fall short in performance to predict mortality in patients with acute heart failure (AHF).

Objective

To develop and validate a machine learning-based prognostic score to predict in-hospital death in patients with AHF and compare its performance with the main traditional scores.

Methods

Patients admitted for AHF in Brazilian hospitals from the “Best Practices in Cardiology Program” from 2016 – 2022 were included. Clinical data, laboratory results, and the World Health Organization Quality of Life (WHOQOL-Bref) at hospital admission were collected. The outcome was in-hospital death. The model was trained using 70% of admissions (training set) and validated with the remaining 30% (test set). The ML-HF score area under the ROC curve (AUC) was compared with the AUC of the traditional scores Acute Decompensated Heart Failure National Registry (ADHERE) and Get With the Guidelines–Heart Failure (GWTG-HF). The level of significance was p<0.05.

Results

One thousand and one hundred fifty-seven patients hospitalized for AHF were included. The five most important variables of the ML-HF score were: Physical Health Domain Quality (WHOQOL-BREF), serum sodium, serum urea, serum creatinine, and systolic blood pressure at hospital admission. In the test set, the ML-HF score showed an adequate model calibration (Hosmer-Lemeshow test p value=0.056) and discrimination (AUC=0.722 [CI95%, 0.661-0.783]), which was superior to GWTG-HF (AUC=0.616 [CI95%, 0.529-0.702; p=0.014]) and the ADHERE (AUC=0.601 [CI95%, 0.511-0.691; p=0.006]) scores.

Conclusion

We developed and validated a score using machine learning to predict in-hospital death in patients with AHF, which outperformed the traditional scores.

Read Full Article

ABC

Editorial

Applicability of Machine Learning Algorithms in the Diagnosis of Arrhythmias – How Long Until The Machine Starts Teaching Us?

Marcelo Martins Pinto Filho

The electrocardiogram (ECG), in clinical use since the early 1900s, remains a beloved tool for cardiologists, clinicians, and medical students. It is arguably the most important bedside instrument in cardiology, essential for managing arrhythmias and acute ischemic syndromes, as well as guiding electrophysiologic care. Still, advances in electrocardiography interpretation have been only modest in the last hundred years. Devices have become more portable, digital, and capable of longer recordings, but core interpretive principles remain rooted in early 20th-century methods. In recent years, this has been reshaped by the rise of computational technologies, especially artificial intelligence (AI). Tools such as deep neural networks (DNNs) and machine learning models (MLMs) are enabling a new era in ECG analysis, the AI-enhanced electrocardiography (AI-enhanced ECG).

In this issue of the ABC Cardiol, Guimarães do Nascimento et al. report on a systematic review of studies evaluating the applicability of AI-enhanced ECG in the diagnosis of cardiac arrhythmias. They included studies of different methodologies that addressed the role of AI-enhanced ECG in the detection of long-QT syndrome (LQTS) (one study), corrected QT interval (cQTi) (one study), and atrial fibrillation (AF) (eleven studies).

[…]

Read Full Article

ABC

Original Article

Applicability of Machine Learning Algorithms in Diagnosis of Atrial Fibrillation and LQTS by Electrocardiogram Interpretation: A Systematic Review

Paulo Cainan Guimarães do Nascimento et al. Matthews Silva Martins Alex Cleber Improta-Caria Roque Aras Junior

Abstract

Background

Machine Learning (ML) is a type of algorithm that autonomously learns to recognize complex patterns. In the diagnostic context of cardiac arrhythmias, these algorithms have shown significant advancements due to their ability to provide automated interpretation and pattern recognition in electrocardiograms (ECGs).

Objective

To analyze and identify the applicability, validity, and feasibility of ML algorithm models in the diagnostic process of cardiac arrhythmias through automated electrocardiogram interpretation.

Methods

This systematic literature review was reported according to the PRISMA guidelines. The searches were conducted in the Cochrane Library, EMBASE, LILACS, and PubMed between February 2022 and November 2022. The study period encompasses articles published between 2017 and 2022.

Results

The database search yielded 119 results, covering three subthemes: Long QT Syndrome (LQTS), corrected QT interval (QTc), and atrial fibrillation (AF). AF was the most prevalent theme. The sample sizes were quite variable. The outcomes were mostly satisfactory. In the diagnosis of LQTS using Artificial Intelligence (AI), the algorithm outperformed conventional methods in diagnostic distinction. In the evaluation of QTc, there was no difference between the AI-integrated ECG and the conventional ECG. In the diagnosis of AF, the algorithms, models, and devices demonstrated high sensitivity and specificity, along with greater accuracy.

Conclusion

ML models in the diagnostic process of cardiac arrhythmias are feasible and rapidly developing. They demonstrate accuracy values between 96.4% and 98.2%, sensitivity between 92.8% and 99.4%, and specificity between 95% and 98.1%, particularly in the diagnosis of atrial fibrillation.

Read Full Article

IJCS

Review Article

Artificial intelligence for nuclear cardiology: Perspectives and challenges

Caitlyn Gilbert et al. Alec Chunta Robert Miller

Abstract

Recent advancements in artificial intelligence (AI) have significantly enhanced the efficiency and quality of nuclear cardiology imaging. AI algorithms, such as generative adversarial networks (GAN) and deep learning (DL) models, are improving image acquisition by reducing noise and correcting motion to minimize radiation exposure while maintaining diagnostic accuracy. AI has also demonstrated the potential to create synthetic attenuation correction (AC) imaging, accounting for soft-tissue attenuation without added computed tomography (CT) imaging. In patients with CT attenuation imaging, AI can automate the segmentation of coronary artery calcification (CAC), epicardial adipose tissue (EAT), and other structures to maximize the clinical information obtained from these added studies. Furthermore, machine learning (ML) models that combine clinical, imaging, and stress test data are advancing risk prediction for coronary artery disease (CAD) and ischemia, assisting clinicians in decision-making and improving patient outcomes. While many of these applications have been developed using myocardial perfusion imaging (MPI), they are increasingly applied to other nuclear cardiology studies. This review explores these AI applications and their impact on diagnostic accuracy, workflow efficiency, and risk stratification while also highlighting potential future directions and challenges.

Read Full Article

IJCS

Review Article

Computer-Interpreted Electrocardiograms: Impact on Cardiology Practice

Shyla Gupta et al. Anthony H. Kashou Robert Herman Stephen Smith … Adam May Ana G. Múnera Echeverri Mildren Del Sueldo Ana C. Berni Juan Farina Sebastian Garcia-Zamora Adrian Baranchuk

Abstract

In the realm of modern cardiology, the integration of computer-interpreted electrocardiograms (CI-ECGs) has marked the beginning of a new era of diagnostic precision and efficiency. Contemporary electrocardiogram (ECG) integration systems, applying algorithms and artificial intelligence, have modernized the interpretation of heart rhythms and cardiac morphology. Due to their ability to rapidly analyze and interpret ECG recordings CI-ECGs have already profoundly impacted clinical practice. This review explores the evolution of computer interpreted ECG technology, evaluates the pros and cons of current automatic reporting systems, analyzes the growing role of artificial intelligence on ECG interpretation technologies, and discusses emerging applications that may have transformative effects on patient outcomes. Emphasis is placed on the role of ECGs in the automatic diagnosis of occlusion myocardial infarctions (OMI). AI models enhance accuracy and efficiency in ECG interpretation, offering insights into cardiac function and aiding timely detection of concerning patterns for accurate clinical diagnoses. The shift to AI-driven diagnostics has emphasized the importance of data in the realm of cardiology by improving patient care. The integration of novel AI models in ECG analysis has created a promising future for ECG diagnostics through a synergistic fusion of feature-based machine learning models, deep learning approaches, and clinical acumen. Overall, CI-ECGs have transformed cardiology practice, offering rapid, accurate, and standardized analyses. These systems reduce interpretation time significantly, allowing for quick identification of abnormalities. However, sole reliance on automated interpretations may overlook nuanced findings, risking diagnostic errors. Therefore, a balanced approach in integrating automated analysis with clinical judgment is necessary.

Read Full Article

ABCIC

Original Article

Use of Artificial Intelligence to Assess Cardiac Function by Echocardiography: Systematic Review of the State of the Art

Henrique Alexsander Ferreira Neves et al. Bruna Sadae Yuasa Thamires Hadassa Leite Pereira Costa Isabela Ertes Santos … Yannirê Milagros Roman Benavides Marco Stephan Lofrano-Alves

Abstract

Background:

Echocardiography is part of the routine investigation of most heart diseases. However, its operator-dependent nature and the quality of the image obtained can cause significant variability in measurements and impair their interpretation. The method has advanced on its path of standardization and incorporation of new technologies in an attempt to reduce its deficiencies. Recently, part of this advancement was due to the incorporation of artificial intelligence (AI).

Objectives:

The objective of this work was to investigate the use of AI in the assessment of cardiac function through echocardiography, discussing its potential to improve the method.

Methods:

We conducted a systematic review in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA), searching for articles within the period of 2000 to 2023.

Results:

A total of 31 articles were included, classified into four categories of studies that used AI: 1) for image analysis and echocardiographic parameters (N = 14); 2) in the image acquisition process (N = 11); 3) for integration with clinical data (N = 3); and 4) for application to the exam flow (N = 3).

Conclusion:

AI provides automation and standardization of images and respective measurements that result in more accurate diagnoses, overcoming the limitations encountered by the conventional method. However, the use of AI is still an evolving reality within echocardiography.

Read Full Article

ABCIC

Case Report

New Application of FFRCT in Clinical Practice: Evaluation of Interarterial Anomalous Coronary Course

Fábio Bordin Trindade et al. Thamara Carvalho Morais Roberto Nery Dantas Junior Roberto Vitor Almeida Torres … Clarice Santos Parreira Soares José Rodrigues Parga Filho

Discussion

Anomalous origin and course of coronary arteries are rare congenital heart diseases, affecting less than 1% of the general population.^, Anomalous coronary origin with an interarterial (suprapulmonary) course is characterized by the course of the coronary artery between the ascending aorta and the pulmonary artery trunk, most commonly involving the right coronary artery.^, Most individuals are asymptomatic, but, among symptomatic patients, chest pain and dyspnea on exertion are the most prevalent complaints. There is also an increase in rates of arrhythmia, sudden death, and acute myocardial infarction.^, Sudden death is the main complication of this anatomical variant, occurring in approximately 30% of patients.^,, In these cases, the narrowing and stretching of the anomalous ostium, mainly during physical exercise and in stressful situations, with consequent reduction in coronary flow, is the substrate for potentially fatal ischemic alterations.^,, Since electrocardiogram, in most cases, does not reveal ischemic alterations, diagnosis is generally made through an incidental finding on imaging exams. These exams are of great importance, especially coronary CTA, which, in addition to being a non-invasive exam with high negative predictive value, allows detailed anatomical visualization (angle of the origin, presence of intramural trajectory in the aorta, degree of ostial/proximal luminal reduction) and the correct classification in relation to the pulmonary valve plane (suprapulmonary versus subpulmonary).^,,

Recent studies have demonstrated that coronary CTA is an accurate test for identifying myocardial ischemia through FFRCT, when compared to invasive measurement (iFFR) by coronary angiography.^,,, The detection of ischemia in coronary CTA is of great importance in decision-making, mainly in plaques considered moderate (50% to 69% luminal reduction) or when there is diagnostic doubt, reducing the number of unnecessary referrals to coronary angiography in cases without ischemia on FFRct.^,, Due to the excellent accuracy between the methods in the analysis of a coronary tree without anomaly, the applicability of the method in the context of coronary anomaly has been extrapolated.^, Our group applied the most current version of a tool for calculating FFRCT currently available only on a research platform developed by Siemens Healthineers (cFRR, version 3.0.0) in the clinical case described. This research tool is available for installation on standard configuration computers, and it uses artificial intelligence tools, with reduced processing time. Currently, commercially available options require the step of sending images in DICOM format for processing in specific centers, with delivery of results at least 24 hours after the images have been sent.^, The research tool used in this clinical case has advantages, such as fast processing time on standard configuration computers in the analysis room and use of tomographic images from the standard routine protocol, without requiring the addition of a specific protocol or a higher dose of radiation, without the use of stressors.^, This tool, in general, has some limitations, such as difficulty in defining the coronary borders in the presence of excessive calcification, and the need for high-quality images, without movement artifacts, for adequate automatic detection of the central luminal line and the lines that delimit the coronary borders, allowing adequate calculation of FFRct. It is important to emphasize that this tool distinguishes the interarterial pathways associated with ischemia, considering only the flow obstacle during rest, without estimating the risk of ischemic events associated with dynamic changes secondary to intense exercise. Use of this tool is scarce in the literature, and it still does not have a robust body of evidence. Nevertheless, in the clinical case described, the FFRCT tool was applied in a context different than what is usual, where luminal reduction was not determined by coronary atheromatosis, but by ostial angulation and compression of the proximal segment of the anomalous interarterial pathway. The ischemic response was considered a parameter of poor prognosis; subsequently, the confirmatory invasive functional test (iFFR) was indicated to assist in the therapeutic decision.

[…]

Read Full Article

ABC

Review Article

Machine Learning in Medicine: Review and Applicability

Gabriela Miana de Mattos Paixão et al. Bruno Campos Santos Rodrigo Martins de Araujo Manoel Horta Ribeiro … Jermana Lopes de Moraes Antonio L. Ribeiro

Introduction

Machine learning (ML) is a branch of artificial intelligence (AI) that explores the study and construction of computational algorithms based on data learning,^, rather than preprogrammed instructions. The main objective of an ML model is to construct a computer system that learns from a predefined database and, in the end, generates a model for prediction, classification, or detection.

The application of ML in practice is mainly aimed at handling consolidated databases with heterogeneous information where there is a limitation to the use of conventional statistical techniques.^, ML algorithms are already widespread in different areas, such as banking systems for fraud detection, internet search engines, video surveillance systems, data security, business logistics, robotics, and, in medicine, diagnosis and prognosis. With the digitalization of medical records, laboratory tests, and imaging, there has been a growth in database are sources for the application of ML techniques, with the aims of prevention, early diagnosis, and treatment of diseases.

[…]

Read Full Article

ABCIC

Editorial

Artificial Intelligence and Cardiac Imaging: We need to talk about this

José de Arimateia Batista Araujo-Filho et al. Antonildes Nascimento Assunção Júnior Marco Antonio Gutierrez Cesar Higa Nomura

[…]

Read Full Article

ABC

Editorial

Artificial Algorithms Outperform Traditional Models in Predicting Coronary Artery Disease

Lutfu Askin et al. Okan Tanrıverdi Mustafa Cetin

Recent clinical recommendations indicate that additional tests for assessing anatomical (extent, severity, morphology) or functional (ventricular function, presence/extent of ischemia) aspects of chronic and symptomatic coronary artery disease (CAD) may be helpful in certain cases. Emergency physicians must determine whether to release the patient, do further non-invasive testing, or perform invasive angiography on patients with acute chest discomfort. Accepting anybody with chest pain may have unintended effects if discharged with unstable coronary disease. The likelihood of obstructive CAD should guide medical decisions. Machine learning (ML) algorithms can supplement the diagnostic and prognostic capabilities of conventional regression methods. The disparity between the applicability of such methods and the outcomes achieved with them was due to the data analysis software platforms used.

ML may use thoracic phase signal features to build final mathematical models that evaluate the existence of severe CAD. Cardiac phase space analysis seems similar to the most widely used functional stress tests and needs little patient time. The 2-year results showed that deep learning fractional flow reserve derived from CT (DL-FFRCT) may be used to guide revascularization, with high cancellation rate and low event rate. A positive DL-FFRCT for tandem lesions was linked with reduced major adverse cardiac events (MACEs) after 2 years. The ML-ischemia risk score (ML-IRS) obtained from quantitative coronary CT angiography enhanced the prediction of future revascularization and may be used to identify individuals who are likely to need revascularization if referred for cardiac catheterization. This machine learning score is linked with invasive fractional flow reserve (FFR) measures, providing external validation across two centers and augmenting clinical risk prediction models.

[…]

Read Full Article

ABC

Original Article

Diagnostic Performance of a Machine Learning-Based CT-Derived FFR in Detecting Flow-Limiting Stenosis

Thamara Carvalho Morais et al. Antonildes Nascimento Assunção-Jr Roberto Nery Dantas Júnior Carla Franco Grego da Silva … Caroline Bastida de Paula Roberto Almeida Torres Tiago Augusto Magalhães César Higa Nomura Luiz Francisco Rodrigues de Ávila José Rodrigues Parga Filho

Abstract

Background

The non-invasive quantification of the fractional flow reserve (FFRCT) using a more recent version of an artificial intelligence-based software and latest generation CT scanner (384 slices) may show high performance to detect coronary ischemia.

Objectives

To evaluate the diagnostic performance of FFRCT for the detection of significant coronary artery disease (CAD) in contrast to invasive FFR (iFFR) using previous generation CT scanners (128 and 256- detector rows).

Methods

Retrospective study with patients referred to coronary artery CT angiography (CTA) and catheterization (iFFR) procedures. Siemens Somatom Definition Flash (256-detector rows) and AS+ (128-detector rows) CT scanners were used to acquire the images. The FFRCT and the minimal lumen area (MLA) were evaluated using a dedicated software (cFFR version 3.0.0, Siemens Healthineers, Forchheim, Germany). Obstructive CAD was defined as CTA lumen reduction ≥ 50%, and flow-limiting stenosis as iFFR ≤0.8. All reported P values are two-tailed, and when <0.05, they were considered statistically significant.

Results

Ninety-three consecutive patients (152 vessels) were included. There was good agreement between FFRCT and iFFR, with minimal FFRCT overestimation (bias: -0.02; limits of agreement:0.14-0.09). Different CT scanners did not modify the association between FFRCT and FFRi (p for interaction=0.73). The performance of FFRCT was significantly superior compared to the visual classification of coronary stenosis (AUC 0.93vs.0.61, p<0.001) and to MLA (AUC 0.93vs.0.75, p<0.001), reducing the number of false-positive cases. The optimal cut-off point for FFRCT using a Youden index was 0.85 (87% Sensitivity, 86% Specificity, 73% PPV, 94% NPV), with a reduction of false-positives.

Conclusion

Machine learning-based FFRCT using previous generation CT scanners (128 and 256-detector rows) shows good diagnostic performance for the detection of CAD, and can be used to reduce the number of invasive procedures.

Read Full Article

IJCS

Review Article

Phenotype mapping of heart failure with preserved ejection fraction

Evandro Tinoco Mesquita et al. Debora Carvalho Grion Miguel Camargo Kubrusly Bernardo Barcelos Fernandes Fumagalli Silva Érico Araújo Reis Santos

Abstract

Heart failure with preserved ejection fraction (HFPEF) has become the main phenotypic model of heart failure (HF) in community and referral patients in Brazil and in the world. Despite advances in the development of new drugs for HF treatment, there has been no significant improvement in mortality of this condition.

According to many studies, this can be explained by the heterogeneous nature of HF physiopathology, whose basic mechanisms may result in different clinical presentations, culminating in the emerging of different phenogroups in this syndrome. In this context, phenotype mapping of HFPEF has emerged as a possible solution, since it enables the development of clinical trials that establish specific therapeutic strategies for each phenotypic profile.

New technologies in the field of artificial intelligence have enabled the assessment of a large volume of data and infer intrinsic patterns and different outcomes. Thereby, it is possible to obtain mutually exclusive categories of HFPEF, with a phenotype mapping of the syndrome and grouping of patients according to their phenotypic features. Besides, other diseases can have the same clinical phenotype but different pathophysiological basis, the so called “phenocopies”.

These tools enable the analysis and categorization of the wide spectrum of heart failure, contributing to solve the dilemmas of the treatment of this syndrome.

Read Full Article