Artificial Intelligence-Assisted Prediction of Late-Onset Cardiomyopathy Among Childhood Cancer Survivors

Presence and utility of electrocardiographic abnormalities in long-term childhood cancer survivors

BACKGROUND

We assessed the prevalence and diagnostic value of ECG abnormalities for cardiomyopathy surveillance in childhood cancer survivors.

METHODS

In this cross-sectional study, 1381 survivors (≥5 years) from the Dutch Childhood Cancer Survivor Study part 2 and 272 siblings underwent a long-term follow-up ECG and echocardiography. We compared ECG abnormality prevalences using the Minnesota Code between survivors and siblings, and within biplane left ventricular ejection fraction (LVEF) categories. Among 880 survivors who received anthracycline, mitoxantrone or heart radiotherapy, logistic regression models using least absolute shrinkage and selection operator identified ECG abnormalities associated with three abnormal LVEF categories (<52% in male/<54% in female, <50% and <45%). We assessed the overall contribution of these ECG abnormalities to clinical regression models predicting abnormal LVEF, assuming an absence of systolic dysfunction with a <1% threshold probability.

RESULTS

16% of survivors (52% female, mean age 34.7 years) and 14% of siblings had major ECG abnormalities. ECG abnormalities increased with decreasing LVEF. Integrating selected ECG data into the baseline model significantly improved prediction of sex-specific abnormal LVEF (c-statistic 0.66 vs 0.71), LVEF <50% (0.66 vs 0.76) and LVEF <45% (0.80 vs 0.86). While no survivor met the preset probability threshold in the first two models, the third model used five ECG variables to predict LVEF <45% and was applicable for ruling out (sensitivity 93%, specificity 56%, negative predictive value 99.6%). Calibration and internal validation tests performed well.

CONCLUSION

A clinical prediction model with ECG data (left bundle branch block, left atrial enlargement, left heart axis, Cornell's criteria for left ventricular hypertrophy and heart rate) may aid in ruling out LVEF <45%.

Development and validation of an electrocardiographic artificial intelligence model for detection of peripartum cardiomyopathy

BACKGROUND

This study used electrocardiogram data in conjunction with artificial intelligence methods as a noninvasive tool for detecting peripartum cardiomyopathy.

OBJECTIVE

This study aimed to assess the efficacy of an artificial intelligence-based heart failure detection model for peripartum cardiomyopathy detection.

STUDY DESIGN

We first built a deep-learning model for heart failure detection using retrospective data at the University of Tennessee Health Science Center. Cases were adult and nonpregnant female patients with a heart failure diagnosis; controls were adult nonpregnant female patients without heart failure. The model was then tested on an independent cohort of pregnant women at the University of Tennessee Health Science Center with or without peripartum cardiomyopathy. We also tested the model in an external cohort of pregnant women at Atrium Health Wake Forest Baptist. Key outcomes were assessed using the area under the receiver operating characteristic curve. We also repeated our analysis using only lead I electrocardiogram as an input to assess the feasibility of remote monitoring via wearables that can capture single-lead electrocardiogram data.

RESULTS

The University of Tennessee Health Science Center heart failure cohort comprised 346,339 electrocardiograms from 142,601 patients. In this cohort, 60% of participants were Black and 37% were White, with an average age (standard deviation) of 53 (19) years. The heart failure detection model achieved an area under the curve of 0.92 on the holdout set. We then tested the ability of the heart failure model to detect peripartum cardiomyopathy in an independent University of Tennessee Health Science Center cohort of pregnant women and an external Atrium Health Wake Forest Baptist cohort of pregnant women. The independent University of Tennessee Health Science Center cohort included 158 electrocardiograms from 115 patients; our deep-learning model achieved an area under the curve of 0.83 (0.77-0.89) for this data set. The external Atrium Health Wake Forest Baptist cohort involved 80 electrocardiograms from 43 patients; our deep-learning model achieved an area under the curve of 0.94 (0.91-0.98) for this data set. For identifying peripartum cardiomyopathy diagnosed ≥10 days after delivery, the model achieved an area under the curve of 0.88 (0.81-0.94) for the University of Tennessee Health Science Center cohort and of 0.96 (0.93-0.99) for the Atrium Health Wake Forest Baptist cohort. When we repeated our analysis by building a heart failure detection model using only lead-I electrocardiograms, we obtained similarly high detection accuracies, with areas under the curve of 0.73 and 0.93 for the University of Tennessee Health Science Center and Atrium Health Wake Forest Baptist cohorts, respectively.

CONCLUSION

Artificial intelligence can accurately detect peripartum cardiomyopathy from electrocardiograms alone. A simple electrocardiographic artificial intelligence-based peripartum screening could result in a timelier diagnosis. Given that results with 1-lead electrocardiogram data were similar to those obtained using all 12 leads, future studies will focus on remote screening for peripartum cardiomyopathy using smartwatches that can capture single-lead electrocardiogram data.

Advancing the care of individuals with cancer through innovation & technology: Proceedings from the cardiology oncology innovation summit 2020 and 2021

As cancer therapies increase in effectiveness and patients' life expectancies improve, balancing oncologic efficacy while reducing acute and long-term cardiovascular toxicities has become of paramount importance. To address this pressing need, the Cardiology Oncology Innovation Network (COIN) was formed to bring together domain experts with the overarching goal of collaboratively investigating, applying, and educating widely on various forms of innovation to improve the quality of life and cardiovascular healthcare of patients undergoing and surviving cancer therapies. The COIN mission pillars of innovation, collaboration, and education have been implemented with cross-collaboration among academic institutions, private and public establishments, and industry and technology companies. In this report, we summarize proceedings from the first two annual COIN summits (inaugural in 2020 and subsequent in 2021) including educational sessions on technological innovations for establishing best practices and aligning resources. Herein, we highlight emerging areas for innovation and defining unmet needs to further improve the outcome for cancer patients and survivors of all ages. Additionally, we provide actionable suggestions for advancing innovation, collaboration, and education in cardio-oncology in the digital era.

Machine Learning in Cardio-Oncology: New Insights from an Emerging Discipline

A growing body of evidence on a wide spectrum of adverse cardiac events following oncologic therapies has led to the emergence of cardio-oncology as an increasingly relevant interdisciplinary specialty. This also calls for better risk-stratification for patients undergoing cancer treatment. Machine learning (ML), a popular branch discipline of artificial intelligence that tackles complex big data problems by identifying interaction patterns among variables, has seen increasing usage in cardio-oncology studies for risk stratification. The objective of this comprehensive review is to outline the application of ML approaches in cardio-oncology, including deep learning, artificial neural networks, random forest and summarize the cardiotoxicity identified by ML. The current literature shows that ML has been applied for the prediction, diagnosis and treatment of cardiotoxicity in cancer patients. In addition, role of ML in gender and racial disparities for cardiac outcomes and potential future directions of cardio-oncology are discussed. It is essential to establish dedicated multidisciplinary teams in the hospital and educate medical professionals to become familiar and proficient in ML in the future.

Informatics tools to implement late cardiovascular risk prediction modeling for population management of high-risk childhood cancer survivors

BACKGROUND

Clinical informatics tools to integrate data from multiple sources have the potential to catalyze population health management of childhood cancer survivors at high risk for late heart failure through the implementation of previously validated risk calculators.

METHODS

The Oklahoma cohort (n = 365) harnessed data elements from Passport for Care (PFC), and the Duke cohort (n = 274) employed informatics methods to automatically extract chemotherapy exposures from electronic health record (EHR) data for survivors 18 years old and younger at diagnosis. The Childhood Cancer Survivor Study (CCSS) late cardiovascular risk calculator was implemented, and risk groups for heart failure were compared to the Children's Oncology Group (COG) and the International Guidelines Harmonization Group (IGHG) recommendations. Analysis within the Oklahoma cohort assessed disparities in guideline-adherent care.

RESULTS

The Oklahoma and Duke cohorts both observed good overall concordance between the CCSS and COG risk groups for late heart failure, with weighted kappa statistics of .70 and .75, respectively. Low-risk groups showed excellent concordance (kappa > .9). Moderate and high-risk groups showed moderate concordance (kappa .44-.60). In the Oklahoma cohort, adolescents at diagnosis were significantly less likely to receive guideline-adherent echocardiogram surveillance compared with survivors younger than 13 years old at diagnosis (odds ratio [OD] 0.22; 95% confidence interval [CI]: 0.10-0.49).

CONCLUSIONS

Clinical informatics tools represent a feasible approach to leverage discrete treatment-related data elements from PFC or the EHR to successfully implement previously validated late cardiovascular risk prediction models on a population health level. Concordance of CCSS, COG, and IGHG risk groups using real-world data informs current guidelines and identifies inequities in guideline-adherent care.

Risk Prediction Models for Myocardial Dysfunction and Heart Failure in Patients with Current or Prior Cancer

PURPOSE OF REVIEW

Cancer patients are at risk for treatment-related myocardial dysfunction and heart failure during or after treatment. Risk prediction models have the potential to play an important role in identifying patients at high or low risk in order to take appropriate measures. Here, we review their current role.

RECENT FINDINGS

More and more risk prediction models are currently being developed. Unfortunately, they vary widely in their ability to identify patients and survivors at risk for myocardial dysfunction or heart failure, from very poor to strong. Part of this variation might be explained by methodological limitations of the models, but due to a lack of reporting it is not possible to completely assess this. There lies great potential in the improvement of the quality and the use of risk prediction models to inform patients and clinicians on the absolute risk of cardiac events in order to guide care.

Multimodality Advanced Cardiovascular and Molecular Imaging for Early Detection and Monitoring of Cancer Therapy-Associated Cardiotoxicity and the Role of Artificial Intelligence and Big Data

Cancer mortality has improved due to earlier detection via screening, as well as due to novel cancer therapies such as tyrosine kinase inhibitors and immune checkpoint inhibitions. However, similarly to older cancer therapies such as anthracyclines, these therapies have also been documented to cause cardiotoxic events including cardiomyopathy, myocardial infarction, myocarditis, arrhythmia, hypertension, and thrombosis. Imaging modalities such as echocardiography and magnetic resonance imaging (MRI) are critical in monitoring and evaluating for cardiotoxicity from these treatments, as well as in providing information for the assessment of function and wall motion abnormalities. MRI also allows for additional tissue characterization using T1, T2, extracellular volume (ECV), and delayed gadolinium enhancement (DGE) assessment. Furthermore, emerging technologies may be able to assist with these efforts. Nuclear imaging using targeted radiotracers, some of which are already clinically used, may have more specificity and help provide information on the mechanisms of cardiotoxicity, including in anthracycline mediated cardiomyopathy and checkpoint inhibitor myocarditis. Hyperpolarized MRI may be used to evaluate the effects of oncologic therapy on cardiac metabolism. Lastly, artificial intelligence and big data of imaging modalities may help predict and detect early signs of cardiotoxicity and response to cardioprotective medications as well as provide insights on the added value of molecular imaging and correlations with cardiovascular outcomes. In this review, the current imaging modalities used to assess for cardiotoxicity from cancer treatments are discussed, in addition to ongoing research on targeted molecular radiotracers, hyperpolarized MRI, as well as the role of artificial intelligence (AI) and big data in imaging that would help improve the detection and prognostication of cancer-treatment cardiotoxicity.

Artificial intelligence opportunities in cardio-oncology: Overview with spotlight on electrocardiography

Cardiovascular disease is a leading cause of death among cancer survivors, second only to cancer recurrence or development of new tumors. Cardio-oncology has therefore emerged as a relatively new specialty focused on prevention and management of cardiovascular consequences of cancer therapies. Yet challenges remain regarding precision and accuracy with predicting individuals at highest risk for cardiotoxicity. Barriers such as access to care also limit screening and early diagnosis to improve prognosis. Thus, developing innovative approaches for prediction and early detection of cardiovascular illness in this population is critical. In this review, we provide an overview of the present state of machine learning applications in cardio-oncology. We begin by outlining some factors that should be considered while utilizing machine learning algorithms. We then examine research in which machine learning has been applied to improve prediction of cardiac dysfunction in cancer survivors. We also highlight the use of artificial intelligence (AI) in conjunction with electrocardiogram (ECG) to predict cardiac malfunction and also atrial fibrillation (AF), and we discuss the potential role of wearables. Additionally, the article summarizes future prospects and critical takeaways for the application of machine learning in cardio-oncology. This study is the first in a series on artificial intelligence in cardio-oncology, and complements our manuscript on echocardiography and other forms of imaging relevant to cancer survivors cared for in cardiology clinical practice.

ECG-AI: electrocardiographic artificial intelligence model for prediction of heart failure

AIMS

Heart failure (HF) is a leading cause of death. Early intervention is the key to reduce HF-related morbidity and mortality. This study assesses the utility of electrocardiograms (ECGs) in HF risk prediction.

METHODS AND RESULTS

Data from the baseline visits (1987-89) of the Atherosclerosis Risk in Communities (ARIC) study was used. Incident hospitalized HF events were ascertained by ICD codes. Participants with good quality baseline ECGs were included. Participants with prevalent HF were excluded. ECG-artificial intelligence (AI) model to predict HF was created as a deep residual convolutional neural network (CNN) utilizing standard 12-lead ECG. The area under the receiver operating characteristic curve (AUC) was used to evaluate prediction models including (CNN), light gradient boosting machines (LGBM), and Cox proportional hazards regression. A total of 14 613 (45% male, 73% of white, mean age ± standard deviation of 54 ± 5) participants were eligible. A total of 803 (5.5%) participants developed HF within 10 years from baseline. Convolutional neural network utilizing solely ECG achieved an AUC of 0.756 (0.717-0.795) on the hold-out test data. ARIC and Framingham Heart Study (FHS) HF risk calculators yielded AUC of 0.802 (0.750-0.850) and 0.780 (0.740-0.830). The highest AUC of 0.818 (0.778-0.859) was obtained when ECG-AI model output, age, gender, race, body mass index, smoking status, prevalent coronary heart disease, diabetes mellitus, systolic blood pressure, and heart rate were used as predictors of HF within LGBM. The ECG-AI model output was the most important predictor of HF.

CONCLUSIONS

ECG-AI model based solely on information extracted from ECG independently predicts HF with accuracy comparable to existing FHS and ARIC risk calculators.

Development and multicenter validation of chest X-ray radiography interpretations based on natural language processing

Background

Artificial intelligence can assist in interpreting chest X-ray radiography (CXR) data, but large datasets require efficient image annotation. The purpose of this study is to extract CXR labels from diagnostic reports based on natural language processing, train convolutional neural networks (CNNs), and evaluate the classification performance of CNN using CXR data from multiple centers

Methods

We collected the CXR images and corresponding radiology reports of 74,082 subjects as the training dataset. The linguistic entities and relationships from unstructured radiology reports were extracted by the bidirectional encoder representations from transformers (BERT) model, and a knowledge graph was constructed to represent the association between image labels of abnormal signs and the report text of CXR. Then, a 25-label classification system were built to train and test the CNN models with weakly supervised labeling.

Results

In three external test cohorts of 5,996 symptomatic patients, 2,130 screening examinees, and 1,804 community clinic patients, the mean AUC of identifying 25 abnormal signs by CNN reaches 0.866 ± 0.110, 0.891 ± 0.147, and 0.796 ± 0.157, respectively. In symptomatic patients, CNN shows no significant difference with local radiologists in identifying 21 signs (p > 0.05), but is poorer for 4 signs (p < 0.05). In screening examinees, CNN shows no significant difference for 17 signs (p > 0.05), but is poorer at classifying nodules (p = 0.013). In community clinic patients, CNN shows no significant difference for 12 signs (p > 0.05), but performs better for 6 signs (p < 0.001).

Conclusion

We construct and validate an effective CXR interpretation system based on natural language processing.

Chest X-rays are accompanied by a report from the radiologist, which contains valuable diagnostic information in text format. Extracting and interpreting information from these reports, such as keywords, is time-consuming, but artificial intelligence (AI) can help with this. Here, we use a type of AI known as natural language processing to extract information about abnormal signs seen on chest X-rays from the corresponding report. We develop and test natural language processing models using data from multiple hospitals and clinics, and show that our models achieve similar performance to interpretation from the radiologists themselves. Our findings suggest that AI might help radiologists to speed up interpretation of chest X-ray reports, which could be useful not only in patient triage and diagnosis but also cataloguing and searching of radiology datasets.

Zhang et al. develop a natural language processing approach, based on the BERT model, to extract linguistic information from chest X-ray radiography reports. The authors establish a 25-label classification system for abnormal findings described in the reports and validate their model using data from multiple sites.