Machine learning-based mortality prediction of patients undergoing cardiac resynchronization therapy: the SEMMELWEIS-CRT score

Left Bundle Branch Area Pacing versus Biventricular Pacing for Cardiac Resynchronization Therapy on Morbidity and Mortality

BACKGROUND

Left bundle branch area pacing (LBBAP) has emerged as an alternative to biventricular pacing (BVP) for cardiac resynchronization therapy (CRT). We aimed to compare the morbidity and mortality associated with LBBAP versus BVP in patients undergoing CRT implantation.

METHODS

Consecutive patients who received CRT from two high-volume implantation centers were retrospectively recruited. The primary endpoint was a composite of all-cause death and heart failure hospitalization, and the secondary endpoint was all-cause death.

RESULTS

A total of 491 patients receiving CRT (154 via LBBAP and 337 via BVP) were included, with a median follow-up of 31 months. The primary endpoint was reached by 21 (13.6%) patients in the LBBAP group, as compared with 74 (22.0%) patients in the BVP group [hazard ratio (HR) 0.70, 95% confidence interval (CI) 0.43-1.14, P = 0.15]. There were 10 (6.5%) deaths in the LBBAP group, as compared with 31 (9.2%) in the BVP group (HR 0.91, 95% CI 0.44-1.86, P = 0.79). No significant difference was observed in the risk of either the primary or secondary endpoint between LBBAP and BVP after multivariate Cox regression (HR 0.74, 95% CI 0.45-1.23, P = 0.24, and HR 0.77, 95% CI 0.36-1.67, P = 0.51, respectively) or propensity score matching (HR 0.72, 95% CI 0.41-1.29, P = 0.28, and HR 0.69, 95% CI 0.29-1.65, P = 0.40, respectively).

CONCLUSION

LBBAP was associated with a comparable effect on morbidity and mortality relative to BVP in patients with indications for CRT.

Role of Machine Learning and Artificial Intelligence in Arrhythmias and Electrophysiology

Machine learning (ML), a subset of artificial intelligence (AI) centered on machines learning from extensive datasets, stands at the forefront of a technological revolution shaping various facets of society. Cardiovascular medicine has emerged as a key domain for ML applications, with considerable efforts to integrate these innovations into routine clinical practice. Within cardiac electrophysiology, ML applications, especially in the automated interpretation of electrocardiograms, have garnered substantial attention in existing literature. However, less recognized are the diverse applications of ML in cardiac electrophysiology and arrhythmias, spanning basic science research on arrhythmia mechanisms, both experimental and computational, as well as contributions to enhanced techniques for mapping cardiac electrical function and translational research related to arrhythmia management. This comprehensive review delves into various ML applications within the scope of this journal, organized into 3 parts. The first section provides a fundamental understanding of general ML principles and methodologies, serving as a foundational resource for readers interested in exploring ML applications in arrhythmia research. The second part offers an in-depth review of studies in arrhythmia and electrophysiology that leverage ML methodologies, showcasing the broad potential of ML approaches. Each subject is thoroughly outlined, accompanied by a review of notable ML research advancements. Finally, the review delves into the primary challenges and future perspectives surrounding ML-driven cardiac electrophysiology and arrhythmias research.

A new method of modeling the multi-stage decision-making process of CRT using machine learning with uncertainty quantification

AIMS

The purpose of this study is to create a multi-stage machine learning model to predict cardiac resynchronization therapy (CRT) response for heart failure (HF) patients. This model exploits uncertainty quantification to recommend additional collection of single-photon emission computed tomography myocardial perfusion imaging (SPECT MPI) variables if baseline clinical variables and features from electrocardiogram (ECG) are not sufficient.

METHODS

218 patients who underwent rest-gated SPECT MPI were enrolled in this study. CRT response was defined as an increase in left ventricular ejection fraction (LVEF) > 5% at a 6+-1 month follow-up. A multi-stage ML model was created by combining two ensemble models: Ensemble 1 was trained with clinical variables and ECG; Ensemble 2 included Ensemble 1 plus SPECT MPI features. Uncertainty quantification from Ensemble 1 allowed for multi-stage decision-making to determine if the acquisition of SPECT data for a patient is necessary. The performance of the multi-stage model was compared with that of Ensemble models 1 and 2.

RESULTS

The response rate for CRT was 55.5% (n = 121) with overall male gender 61.0% (n = 133), an average age of 62.0+-11.8, and LVEF of 27.7+-11.0. The multi-stage model performed similarly to Ensemble 2 (which utilized the additional SPECT data) with AUC of 0.75 vs. 0.77, accuracy of 0.71 vs. 0.69, sensitivity of 0.70 vs. 0.72, and specificity 0.72 vs. 0.65, respectively. However, the multi-stage model only required SPECT MPI data for 52.7% of the patients across all folds.

CONCLUSIONS

By using rule-based logic stemming from uncertainty quantification, the multi-stage model was able to reduce the need for additional SPECT MPI data acquisition without sacrificing performance.

Application of machine learning methods for predicting under-five mortality: analysis of Nigerian demographic health survey 2018 dataset

BACKGROUND

Under-five mortality remains a significant public health issue in developing countries. This study aimed to assess the effectiveness of various machine learning algorithms in predicting under-five mortality in Nigeria and identify the most relevant predictors.

METHODS

The study used nationally representative data from the 2018 Nigeria Demographic and Health Survey. The study evaluated the performance of the machine learning models such as the artificial neural network, k-nearest neighbourhood, Support Vector Machine, Naïve Bayes, Random Forest, and Logistic Regression using the true positive rate, false positive rate, accuracy, precision, F-measure, Matthew's correlation coefficient, and the Area Under the Receiver Operating Characteristics.

RESULTS

The study found that machine learning models can accurately predict under-five mortality, with the Random Forest and Artificial Neural Network algorithms emerging as the best models, both achieving an accuracy of 89.47% and an AUROC of 96%. The results show that under-five mortality rates vary significantly across different characteristics, with wealth index, maternal education, antenatal visits, place of delivery, employment status of the woman, number of children ever born, and region found to be the top determinants of under-five mortality in Nigeria.

CONCLUSIONS

The findings suggest that machine learning models can be useful in predicting U5M in Nigeria with high accuracy. The study emphasizes the importance of addressing social, economic, and demographic disparities among the population in Nigeria. The study's findings can inform policymakers and health workers about developing targeted interventions to reduce under-five mortality in Nigeria.

Adopting artificial intelligence in cardiovascular medicine: a scoping review

Recent years have witnessed significant transformations in cardiovascular medicine, driven by the rapid evolution of artificial intelligence (AI). This scoping review was conducted to capture the breadth of AI applications within cardiovascular science. Employing a structured approach, we sourced relevant articles from PubMed, with an emphasis on journals encompassing general cardiology and digital medicine. We applied filters to highlight cardiovascular articles published in journals focusing on general internal medicine, cardiology and digital medicine, thereby identifying the prevailing trends in the field. Following a comprehensive full-text screening, a total of 140 studies were identified. Over the preceding 5 years, cardiovascular medicine's interplay with AI has seen an over tenfold augmentation. This expansive growth encompasses multiple cardiovascular subspecialties, including but not limited to, general cardiology, ischemic heart disease, heart failure, and arrhythmia. Deep learning emerged as the predominant methodology. The majority of AI endeavors in this domain have been channeled toward enhancing diagnostic and prognostic capabilities, utilizing resources such as hospital datasets, electrocardiograms, and echocardiography. A significant uptrend was observed in AI's application for omics data analysis. However, a clear gap persists in AI's full-scale integration into the clinical decision-making framework. AI, particularly deep learning, has demonstrated robust applications across cardiovascular subspecialties, indicating its transformative potential in this field. As we continue on this trajectory, ensuring the alignment of technological progress with medical ethics becomes crucial. The abundant digital health data today further accentuates the need for meticulous systematic reviews, tailoring them to each cardiovascular subspecialty.

Artificial intelligence models in prediction of response to cardiac resynchronization therapy: a systematic review

The aim of the presented review is to summarize the literature data on the accuracy and clinical applicability of artificial intelligence (AI) models as a valuable alternative to the current guidelines in predicting cardiac resynchronization therapy (CRT) response and phenotyping of patients eligible for CRT implantation. This systematic review was performed according to the PRISMA guidelines. After a search of Scopus, PubMed, Cochrane Library, and Embase databases, 675 records were identified. Twenty supervised (prediction of CRT response) and 9 unsupervised (clustering and phenotyping) AI models were analyzed qualitatively (22 studies, 14,258 patients). Fifty-five percent of AI models were based on retrospective studies. Unsupervised AI models were able to identify clusters of patients with significantly different rates of primary outcome events (death, heart failure event). In comparison to the guideline-based CRT response prediction accuracy of 70%, supervised AI models trained on cohorts with > 100 patients achieved up to 85% accuracy and an AUC of 0.86 in their prediction of response to CRT for echocardiographic and clinical outcomes, respectively. AI models seem to be an accurate and clinically applicable tool in phenotyping of patients eligible for CRT implantation and predicting potential responders. In the future, AI may help to increase CRT response rates to over 80% and improve clinical decision-making and prognosis of the patients, including reduction of mortality rates. However, these findings must be validated in randomized controlled trials.

Phenogrouping and risk stratification of patients undergoing cardiac resynchronization therapy upgrade using topological data analysis

Choosing the optimal device during cardiac resynchronization therapy (CRT) upgrade can be challenging. Therefore, we sought to provide a solution for identifying patients in whom upgrading to a CRT-defibrillator (CRT-D) is associated with better long-term survival than upgrading to a CRT-pacemaker (CRT-P). To this end, we first applied topological data analysis to create a patient similarity network using 16 clinical features of 326 patients without prior ventricular arrhythmias who underwent CRT upgrade. Then, in the generated circular network, we delineated three phenogroups exhibiting significant differences in clinical characteristics and risk of all-cause mortality. Importantly, only in the high-risk phenogroup was upgrading to a CRT-D associated with better survival than upgrading to a CRT-P (hazard ratio: 0.454 (0.228-0.907), p = 0.025). Finally, we assigned each patient to one of the three phenogroups based on their location in the network and used this labeled data to train multi-class classifiers to enable the risk stratification of new patients. During internal validation, an ensemble of 5 multi-layer perceptrons exhibited the best performance with a balanced accuracy of 0.898 (0.854-0.942) and a micro-averaged area under the receiver operating characteristic curve of 0.983 (0.980-0.986). To allow further validation, we made the proposed model publicly available ( https://github.com/tokmarton/crt-upgrade-risk-stratification ).

Combination of personalized computational modeling and machine learning for optimization of left ventricular pacing site in cardiac resynchronization therapy

Introduction: The 30-50% non-response rate to cardiac resynchronization therapy (CRT) calls for improved patient selection and optimized pacing lead placement. The study aimed to develop a novel technique using patient-specific cardiac models and machine learning (ML) to predict an optimal left ventricular (LV) pacing site (ML-PS) that maximizes the likelihood of LV ejection fraction (LVEF) improvement in a given CRT candidate. To validate the approach, we evaluated whether the distance D_PS between the clinical LV pacing site (ref-PS) and ML-PS is associated with improved response rate and magnitude. Materials and methods: We reviewed retrospective data for 57 CRT recipients. A positive response was defined as a more than 10% LVEF improvement. Personalized models of ventricular activation and ECG were created from MRI and CT images. The characteristics of ventricular activation during intrinsic rhythm and biventricular (BiV) pacing with ref-PS were derived from the models and used in combination with clinical data to train supervised ML classifiers. The best logistic regression model classified CRT responders with a high accuracy of 0.77 (ROC AUC = 0.84). The LR classifier, model simulations and Bayesian optimization with Gaussian process regression were combined to identify an optimal ML-PS that maximizes the ML-score of CRT response over the LV surface in each patient. Results: The optimal ML-PS improved the ML-score by 17 ± 14% over the ref-PS. Twenty percent of the non-responders were reclassified as positive at ML-PS. Selection of positive patients with a max ML-score >0.5 demonstrated an improved clinical response rate. The distance D_PS was shorter in the responders. The max ML-score and D_PS were found to be strong predictors of CRT response (ROC AUC = 0.85). In the group with max ML-score > 0.5 and D_PS< 30 mm, the response rate was 83% compared to 14% in the rest of the cohort. LVEF improvement in this group was higher than in the other patients (16 ± 8% vs. 7 ± 8%). Conclusion: A new technique combining clinical data, personalized heart modelling and supervised ML demonstrates the potential for use in clinical practice to assist in optimizing patient selection and predicting optimal LV pacing lead position in HF candidates for CRT.

Machine learning and physical based modeling for cardiac hypertrophy

Background and objective

Predicting the long-term expansion and remodeling of the left ventricle in patients is challenging task but it has the potential to be clinically very useful.

Methods

In our study, we present machine learning models based on random forests, gradient boosting, and neural networks, used to track cardiac hypertrophy. We collected data from multiple patients, and then the model was trained using the patient's medical history and present level of cardiac health. We also demonstrate a physical-based model, using the finite element procedure to simulate the development of cardiac hypertrophy.

Results

Our models were used to forecast the evolution of hypertrophy over six years. The machine learning model and finite element model provided similar results.

Conclusions

The finite element model is much slower, but it's more accurate compared to the machine learning model since it's based on physical laws guiding the hypertrophy process. On the other hand, the machine learning model is fast but the results can be less trustworthy in some cases. Both of our models, enable us to monitor the development of the disease. Because of its speed machine learning model is more likely to be used in clinical practice. Further improvements to our machine learning model could be achieved by collecting data from finite element simulations, adding them to the dataset, and retraining the model. This can result in a fast and more accurate model combining the advantages of physical-based and machine learning modeling.

Feasibility and safety of left bundle branch area pacing-cardiac resynchronization therapy in elderly patients

BACKGROUND

Left bundle branch area pacing (LBBAP) is an emerging technique to achieve cardiac resynchronization therapy (CRT), but its feasibility and safety in elderly patients with heart failure with reduced ejection fraction and left bundle branch block is hardly investigated.

METHODS

We enrolled consecutive patients with an indication for CRT comparing pacing parameters and complication rates of LBBAP-CRT in elderly patients (≥ 75 years) versus younger patients (< 75 years) over a 6-month follow-up.

RESULTS

LBBAP was successful in 55/60 enrolled patients (92%), among which 25(45%) were elderly. In both groups, LBBAP significantly reduced the QRS duration (elderly group: 168 ± 15 ms to 136 ± 12 ms, p < 0.0001; younger group: 166 ± 14 ms to 134 ± 11 ms, p < 0.0001) and improved LVEF (elderly group: 28 ± 5% to 40 ± 7%, p < 0.0001; younger group: 29 ± 5% to 41 ± 8%, p < 0.0001). The pacing threshold was 0.9 ± 0.8 V in the elderly group vs. 0.7 ± 0.5 V in the younger group (p = 0.350). The R wave was 9.5 ± 3.9 mV in elderly patients vs. 10.7 ± 2.7 mV in younger patients (p = 0.341). The fluoroscopic (elderly: 13 ± 7 min vs. younger: 11 ± 7 min, p = 0.153) and procedural time (elderly: 80 ± 20 min vs. younger: 78 ± 16 min, p = 0.749) were comparable between groups. Lead dislodgement occurred in 2(4%) patients, 1 in each group (p = 1.000). Intraprocedural septal perforation occurred in three patients (5%), 2(8%) in the elderly group (p = 0.585). One patient (2%) in the elderly group had a pocket infection.

CONCLUSIONS

LBBAP is a feasible and safe technique for delivering physiological pacing in elderly patients who are candidates for CRT with suitable pacing parameters and low complication rates.

Risk scores in cardiac resynchronization therapy-A review of the literature

Cardiac resynchronization therapy (CRT) for selected heart failure (HF) patients improves symptoms and reduces morbidity and mortality; however, the prognosis of HF is still poor. There is an emerging need for tools that might help in optimal patient selection and provide prognostic information for patients and their families. Several risk scores have been created in recent years; although, no literature review is available that would list the possible scores for the clinicians. We identified forty-eight risk scores in CRT and provided the calculation methods and formulas in a ready-to-use format. The reviewed score systems can predict the prognosis of CRT patients; some of them have even provided an online calculation tool. Significant heterogeneity is present between the various risk scores in terms of the variables incorporated and some variables are not yet used in daily clinical practice. The lack of cross-validation of the risk scores limits their routine use and objective selection. As the number of prognostic markers of CRT is overwhelming, further studies might be required to analyze and cross-validate the data.

Answering Clinical Questions Using Machine Learning: Should We Look at Diastolic Blood Pressure When Tailoring Blood Pressure Control?

Background: The guidelines recommend intensive blood pressure control. Randomized trials have focused on the relevance of the systolic blood pressure (SBP) lowering, leaving the safety of the diastolic blood pressure (DBP) reduction unresolved. There are data available which show that low DBP should not stop clinicians from achieving SBP targets; however, registries and analyses of randomized trials present conflicting results. The purpose of the study was to apply machine learning (ML) algorithms to determine, whether DBP is an important risk factor to predict stroke, heart failure (HF), myocardial infarction (MI), and primary outcome in the SPRINT trial database. Methods: ML experiments were performed using decision tree, random forest, k-nearest neighbor, naive Bayesian, multi-layer perceptron, and logistic regression algorithms, including and excluding DBP as the risk factor in an unselected and selected (DBP < 70 mmHg) study population. Results: Including DBP as the risk factor did not change the performance of the machine learning models evaluated using accuracy, AUC, mean, and weighted F-measure, and was not required to make proper predictions of stroke, MI, HF, and primary outcome. Conclusions: Analyses of the SPRINT trial data using ML algorithms imply that DBP should not be treated as an independent risk factor when intensifying blood pressure control.

Clinician's guide to trustworthy and responsible artificial intelligence in cardiovascular imaging

A growing number of artificial intelligence (AI)-based systems are being proposed and developed in cardiology, driven by the increasing need to deal with the vast amount of clinical and imaging data with the ultimate aim of advancing patient care, diagnosis and prognostication. However, there is a critical gap between the development and clinical deployment of AI tools. A key consideration for implementing AI tools into real-life clinical practice is their "trustworthiness" by end-users. Namely, we must ensure that AI systems can be trusted and adopted by all parties involved, including clinicians and patients. Here we provide a summary of the concepts involved in developing a "trustworthy AI system." We describe the main risks of AI applications and potential mitigation techniques for the wider application of these promising techniques in the context of cardiovascular imaging. Finally, we show why trustworthy AI concepts are important governing forces of AI development.

Electrocardiogram-based deep learning improves outcome prediction following cardiac resynchronization therapy

AIMS

This study aims to identify and visualize electrocardiogram (ECG) features using an explainable deep learning-based algorithm to predict cardiac resynchronization therapy (CRT) outcome. Its performance is compared with current guideline ECG criteria and QRSAREA.

METHODS AND RESULTS

A deep learning algorithm, trained on 1.1 million ECGs from 251 473 patients, was used to compress the median beat ECG, thereby summarizing most ECG features into only 21 explainable factors (FactorECG). Pre-implantation ECGs of 1306 CRT patients from three academic centres were converted into their respective FactorECG. FactorECG predicted the combined clinical endpoint of death, left ventricular assist device, or heart transplantation [c-statistic 0.69, 95% confidence interval (CI) 0.66-0.72], significantly outperforming QRSAREA and guideline ECG criteria [c-statistic 0.61 (95% CI 0.58-0.64) and 0.57 (95% CI 0.54-0.60), P < 0.001 for both]. The addition of 13 clinical variables was of limited added value for the FactorECG model when compared with QRSAREA (Δ c-statistic 0.03 vs. 0.10). FactorECG identified inferolateral T-wave inversion, smaller right precordial S- and T-wave amplitude, ventricular rate, and increased PR interval and P-wave duration to be important predictors for poor outcome. An online visualization tool was created to provide interactive visualizations (https://crt.ecgx.ai).

CONCLUSION

Requiring only a standard 12-lead ECG, FactorECG held superior discriminative ability for the prediction of clinical outcome when compared with guideline criteria and QRSAREA, without requiring additional clinical variables. End-to-end automated visualization of ECG features allows for an explainable algorithm, which may facilitate rapid uptake of this personalized decision-making tool in CRT.

Artificial intelligence in cardiology: Hope for the future and power for the present

Cardiovascular disease (CVD) is the principal cause of mortality and morbidity globally. With the pressures for improved care and translation of the latest medical advances and knowledge to an actionable plan, clinical decision-making for cardiologists is challenging. Artificial Intelligence (AI) is a field in computer science that studies the design of intelligent agents which take the best feasible action in a situation. It incorporates the use of computational algorithms which simulate and perform tasks that traditionally require human intelligence such as problem solving and learning. Whilst medicine is arguably the last to apply AI in its everyday routine, cardiology is at the forefront of AI revolution in the medical field. The development of AI methods for accurate prediction of CVD outcomes, non-invasive diagnosis of coronary artery disease (CAD), detection of malignant arrythmias through wearables, and diagnosis, treatment strategies and prediction of outcomes for heart failure (HF) patients, demonstrates the potential of AI in future cardiology. With the advancements of AI, Internet of Things (IoT) and the promotion of precision medicine, the future of cardiology will be heavily based on these innovative digital technologies. Despite this, ethical dilemmas regarding the implementation of AI technologies in real-world are still unaddressed.

Contemporary Applications of Machine Learning for Device Therapy in Heart Failure

Despite a better understanding of the underlying pathogenesis of heart failure (HF), pharmacotherapy, surgical, and percutaneous interventions do not prevent disease progression in all patients, and a significant proportion of patients end up requiring advanced therapies. Machine learning (ML) is gaining wider acceptance in cardiovascular medicine because of its ability to incorporate large, complex, and multidimensional data and to potentially facilitate the creation of predictive models not constrained by many of the limitations of traditional statistical approaches. With the coexistence of "big data" and novel advanced analytic techniques using ML, there is ever-increasing research into applying ML in the context of HF with the goal of improving patient outcomes. Through this review, the authors describe the basics of ML and summarize the existing published reports regarding contemporary applications of ML in device therapy for HF while highlighting the limitations to widespread implementation and its future promises.

Critical appraisal of artificial intelligence-based prediction models for cardiovascular disease

The medical field has seen a rapid increase in the development of artificial intelligence (AI)-based prediction models. With the introduction of such AI-based prediction model tools and software in cardiovascular patient care, the cardiovascular researcher and healthcare professional are challenged to understand the opportunities as well as the limitations of the AI-based predictions. In this article, we present 12 critical questions for cardiovascular health professionals to ask when confronted with an AI-based prediction model. We aim to support medical professionals to distinguish the AI-based prediction models that can add value to patient care from the AI that does not.

Severity of Hospitalized Children with Anti-NMDAR Autoimmune Encephalitis

Background: Information on the clinical characteristics and severity of autoimmune encephalitis with antibodies against the N-methyl-d-aspartate receptor (NMDAR) in children is attracting more and more attention in the field of pediatric research. Methods: In this retrospective cohort study, all cases (n = 67) were enrolled from a tertiary children's hospital, from 2017 to 2020. We compared severe cases that received intensive care unit (ICU) care with nonsevere cases that did not receive ICU care and used machine learning algorithm to predict the severity of children, as well as using immunologic and viral nucleic acid tests to identify possible pathogenic triggers. Results: Mean age of children was 8.29 (standard deviation 4.09) years, and 41 (61.19%) were girls. Eleven (16.42%) were admitted to the ICU, and 56 (83.58%) were admitted to neurology ward. Ten individual parameters were statistically significant differences between severe cases and nonsevere cases (P < .05), including headache, abnormal mental behavior or cognitive impairment, seizures, concomitant tumors, sputum/blood pathogens, blood globulin, blood urea nitrogen, blood immunoglobulin G, blood immunoglobulin M, and number of polynucleated cells in cerebrospinal fluid. Random forest regression model presented that the overall prediction power of severity reached 0.806, among which the number of polynucleated cells in cerebrospinal fluid contributed the most. Potential pathogenic causes exhibited that the proportion of mycoplasma was the highest, followed by Epstein-Barr virus. Conclusion: Our findings provided evidence for early identification of autoimmune encephalitis in children, especially in severe cases.

Using machine learning for mortality prediction and risk stratification in atezolizumab-treated cancer patients: Integrative analysis of eight clinical trials

BACKGROUND

Few models exist to predict mortality in cancer patients receiving immunotherapy. Our aim was to build a machine learning-based risk stratification model for predicting mortality in atezolizumab-treated cancer patients.

METHODS

Data from 2538 patients in eight atezolizumab-treated cancer clinical trials across three cancer types (non-small-cell lung cancer, bladder transitional cell carcinoma, and renal cell carcinoma) were included. The whole cohort was randomly split into development and validation cohorts in a 7:3 ratio. Machine-learning algorithms (extreme gradient boosting, random forest, logistic regression with lasso regularization, support vector machine, and K-nearest neighbor) were applied to develop prediction models. Model performance was mainly assessed by area under the receiver operating characteristic curve (AUC) value, calibration plot, and decision curve analysis. The probability of death risk was then stratified.

RESULTS

One thousand and three hundred and seventy-nine (54.33%) patients died. The random forest (RF) model was overall the best in terms of predictive performance, with the AUC of 0.844 (95% confidence interval [CI]: 0.826-0.862) in the development cohort and 0.786 (95% CI: 0.754-0.818) in the validation cohort for predicting mortality. Twelve baseline variables contributing to mortality prediction in the RF model were C-reactive protein, PD-L1 level, cancer type, prior liver metastasis, derived neutrophil-to-lymphocyte ratio, alkaline phosphatase, albumin, hemoglobin, white blood cell count, number of metastatic sites, pulse rate, and Eastern Cooperative Oncology Group (ECOG) performance status. A total of 1782 (70.2%) patients were separated into the high-risk and 756 (29.8%) low-risk groups. Patients in the high-risk group were significantly more likely to die, experience disease progression, discontinue study, and discontinue treatment than patients in the low-risk group (all p values < 0.001). Risk groups were not associated with immune-related adverse events and grades 3-5 treatment-related adverse events (all p values > 0.05).

CONCLUSION

RF model has good performance in mortality prediction and risk stratification for cancer patients receiving atezolizumab monotherapy.

Lifestyle and chronic kidney disease: A machine learning modeling study

Background

Individual lifestyle varies in the real world, and the comparative efficacy of lifestyles to preserve renal function remains indeterminate. We aimed to systematically compare the effects of lifestyles on chronic kidney disease (CKD) incidence, and establish a lifestyle scoring system for CKD risk identification.

Methods

Using the data of the UK Biobank cohort, we included 470,778 participants who were free of CKD at the baseline. We harnessed the light gradient boosting machine algorithm to rank the importance of 37 lifestyle factors (such as dietary patterns, physical activity (PA), sleep, psychological health, smoking, and alcohol) on the risk of CKD. The lifestyle score was calculated by a combination of machine learning and the Cox proportional-hazards model. A CKD event was defined as an estimated glomerular filtration rate <60 ml/min/1.73 m², mortality and hospitalization due to chronic renal failure, and self-reported chronic renal failure, initiated renal replacement therapy.

Results

During a median of the 11-year follow-up, 13,555 participants developed the CKD event. Bread, walking time, moderate activity, and vigorous activity ranked as the top four risk factors of CKD. A healthy lifestyle mainly consisted of whole grain bread, walking, moderate physical activity, oat cereal, and muesli, which have scored 12, 12, 10, 7, and 7, respectively. An unhealthy lifestyle mainly included white bread, tea >4 cups/day, biscuit cereal, low drink temperature, and processed meat, which have scored −12, −9, −7, −4, and −3, respectively. In restricted cubic spline regression analysis, a higher lifestyle score was associated with a lower risk of CKD event (p for linear relation < 0.001). Compared to participants with the lifestyle score < 0, participants scoring 0–20, 20–40, 40–60, and >60 exhibited 25, 42, 55, and 70% lower risk of CKD event, respectively. The C-statistic of the age-adjusted lifestyle score for predicting CKD events was 0.710 (0.703–0.718).

Conclusion

A lifestyle scoring system for CKD prevention was established. Based on the system, individuals could flexibly choose healthy lifestyles and avoid unhealthy lifestyles to prevent CKD.

Introduction to supervised machine learning in clinical epidemiology

Machine learning refers to a series of processes in which a computer finds rules from a vast amount of data. With recent advances in computer technology and the availability of a wide variety of health data, machine learning has rapidly developed and been applied in medical research. Currently, there are three types of machine learning: supervised, unsupervised, and reinforcement learning. In medical research, supervised learning is commonly used for diagnoses and prognoses, while unsupervised learning is used for phenotyping a disease, and reinforcement learning for maximizing favorable results, such as optimization of total patients' waiting time in the emergency department. The present article focuses on the concept and application of supervised learning in medicine, the most commonly used machine learning approach in medicine, and provides a brief explanation of four algorithms widely used for prediction (random forests, gradient-boosted decision tree, support vector machine, and neural network). Among these algorithms, the neural network has further developed into deep learning algorithms to solve more complex tasks. Along with simple classification problems, deep learning is commonly used to process medical imaging, such as retinal fundus photographs for diabetic retinopathy diagnosis. Although machine learning can bring new insights into medicine by processing a vast amount of data that are often beyond human capacity, algorithms can also fail when domain knowledge is neglected. The combination of algorithms and human cognitive ability is a key to the successful application of machine learning in medicine.

Prediction of All-Cause Mortality Following Percutaneous Coronary Intervention in Bifurcation Lesions Using Machine Learning Algorithms

Stratifying prognosis following coronary bifurcation percutaneous coronary intervention (PCI) is an unmet clinical need that may be fulfilled through the adoption of machine learning (ML) algorithms to refine outcome predictions. We sought to develop an ML-based risk stratification model built on clinical, anatomical, and procedural features to predict all-cause mortality following contemporary bifurcation PCI. Multiple ML models to predict all-cause mortality were tested on a cohort of 2393 patients (training, n = 1795; internal validation, n = 598) undergoing bifurcation PCI with contemporary stents from the real-world RAIN registry. Twenty-five commonly available patient-/lesion-related features were selected to train ML models. The best model was validated in an external cohort of 1701 patients undergoing bifurcation PCI from the DUTCH PEERS and BIO-RESORT trial cohorts. At ROC curves, the AUC for the prediction of 2-year mortality was 0.79 (0.74–0.83) in the overall population, 0.74 (0.62–0.85) at internal validation and 0.71 (0.62–0.79) at external validation. Performance at risk ranking analysis, k-center cross-validation, and continual learning confirmed the generalizability of the models, also available as an online interface. The RAIN-ML prediction model represents the first tool combining clinical, anatomical, and procedural features to predict all-cause mortality among patients undergoing contemporary bifurcation PCI with reliable performance.

Utilizing longitudinal data in assessing all-cause mortality in patients hospitalized with heart failure

AIMS

Risk stratification in patients with a new onset or worsened heart failure (HF) is essential for clinical decision making. We have utilized a novel approach to enrich patient level prognostication using longitudinally gathered data to develop ML-based algorithms predicting all-cause 30, 90, 180, 360, and 720 day mortality.

METHODS AND RESULTS

In a cohort of 2449 HF patients hospitalized between 1 January 2011 and 31 December 2017, we utilized 422 parameters derived from 151 451 patient exams. They included clinical phenotyping, ECG, laboratory, echocardiography, catheterization data or percutaneous and surgical interventions reflecting the standard of care as captured in individual electronic records. The development of predictive models consisted of 101 iterations of repeated random subsampling splits into balanced training and validation sets. ML models yielded area under the receiver operating characteristic curve (AUC-ROC) performance ranging from 0.83 to 0.89 on the outcome-balanced validation set in predicting all-cause mortality at aforementioned time-limits. The 1 year mortality prediction model recorded an AUC of 0.85. We observed stable model performance across all HF phenotypes: HFpEF 0.83 AUC, HFmrEF 0.85 AUC, and HFrEF 0.86 AUC, respectively. Model performance improved when utilizing data from more hospital contacts compared with only data collected at baseline.

CONCLUSIONS

Our findings present a novel, patient-level, comprehensive ML-based algorithm for predicting all-cause mortality in new or worsened heart failure. Its robust performance across phenotypes throughout the longitudinal patient follow-up suggests its potential in point-of-care clinical risk stratification.

Utilizing Artificial Intelligence to Enhance Health Equity Among Patients with Heart Failure

Patients with heart failure (HF) are heterogeneous with various intrapersonal and interpersonal characteristics contributing to clinical outcomes. Bias, structural racism, and social determinants of health have been implicated in unequal treatment of patients with HF. Through several methodologies, artificial intelligence (AI) can provide models in HF prediction, prognostication, and provision of care, which may help prevent unequal outcomes. This review highlights AI as a strategy to address racial inequalities in HF; discusses key AI definitions within a health equity context; describes the current uses of AI in HF, strengths and harms in using AI; and offers recommendations for future directions.

Machine Learning-Based Personalized Risk Prediction Model for Mortality of Patients Undergoing Mitral Valve Surgery: The PRIME Score

Background

Mitral valve surgery (MVS) is an effective treatment for mitral valve diseases. There is a lack of reliable personalized risk prediction models for mortality in patients undergoing mitral valve surgery. Our aim was to develop a risk stratification system to predict all-cause mortality in patients after mitral valve surgery.

Methods

Different machine learning models for the prediction of all-cause mortality were trained on a derivation cohort of 1,883 patients undergoing mitral valve surgery [split into a training cohort (70%) and internal validation cohort (30%)] to predict all-cause mortality. Forty-five clinical variables routinely evaluated at discharge were used to train the models. The best performance model (PRIME score) was tested in an externally validated cohort of 220 patients undergoing mitral valve surgery. The model performance was evaluated according to the area under the curve (AUC). Net reclassification improvement (NRI) and integrated discrimination improvement (IDI) were compared with existing risk strategies.

Results

After a median follow-up of 2 years, there were 133 (7.063%) deaths in the derivation cohort and 17 (7.727%) deaths in the validation cohort. The PRIME score showed an AUC of 0.902 (95% confidence interval [CI], 0.849–0.956) in the internal validation cohort and 0.873 (95% CI: 0.769–0.977) in the external validation cohort. In the external validation cohort, the performance of the PRIME score was significantly improved compared with that of the existing EuroSCORE II (NRI = 0.550, [95% CI 0.001–1.099], P = 0.049, IDI = 0.485, [95% CI 0.230–0.741], P < 0.001).

Conclusion

Machine learning-based model (the PRIME score) that integrate clinical, demographic, imaging, and laboratory features demonstrated superior performance for the prediction of mortality patients after mitral valve surgery compared with the traditional risk model EuroSCORE II.

Clinical Trial Registration

[http://www.clinicaltrials.gov], identifier [NCT05141292].

Nonlaboratory-based risk assessment model for coronary heart disease screening: Model development and validation

BACKGROUND

Identifying groups at high risk of coronary heart disease (CHD) is important to reduce mortality due to CHD. Although machine learning methods have been introduced, many require laboratory or imaging parameters, which are not always readily available; thus, their wide applications are limited.

OBJECTIVE

The aim of this study was to develop and validate a simple, efficient, and joint machine learning model for identifying individuals at high risk of CHD using easily obtainable nonlaboratory parameters.

METHODS

This prospective study used data from the Henan Rural Cohort Study, which was conducted in rural areas of Henan Province, China, between July 2015 and September 2017. A joint machine learning model was developed by selecting and combining four base machine learning algorithms, including logistic regression (LR), artificial neural network (ANN), random forest (RF), and gradient boosting machine (GBM). We used readily accessible variables, including demographics, medical and family history, lifestyle and dietary factors, and anthropometric data, to inform the model. The model was also externally validated by a cohort of individuals from the Dongfeng-Tongji cohort study. Model discrimination was assessed by using the area under the receiver operating characteristic curve (AUC), and calibration was measured by using the Brier score (BS).

RESULTS

A total of 38 716 participants (mean [SD] age, 55.64[12.19] years; 23449[60.6%] female) from the Henan Rural Cohort Study and 17 958 subjects (mean [SD] age, 62.74 [7.59] years; 10,076 [56.1%] female) from the Dongfeng-Tongji cohort study were included in the analysis. Age, waist circumference, pulse pressure, heart rate, family history of CHD, education level, family history of type 2 diabetes mellitus (T2DM), and family history of dyslipidaemia were strongly associated with the development of CHD. In regard to internal validation, the model we built demonstrated good discrimination (AUC, 0.844 (95% CI 0.828-0.860)) and had acceptable calibration (BS, 0. 066). In regard to external validation, the model performed well with clearly useful discrimination (AUC, 0.792 (95% CI 0.774-0.810)) and robust calibration (BS, 0.069).

CONCLUSIONS

In this study, the novel and simple, machine learning-based model comprising readily accessible variables accurately identified individuals at high risk of CHD. This model has the potential to be widely applied for large-scale screening of CHD populations, especially in medical resource-constrained settings.

TRIAL REGISTRATION

The Henan Rural Cohort Study has been registered at the Chinese Clinical Trial Register. (Trial registration: ChiCTR-OOC-15006699. Registered 6 July 2015 - Retrospectively registered) http://www.chictr.org.cn/showproj.aspx?proj=11375.

Big Data in electrophysiology

The quantity of data produced and captured in medicine today is unprecedented. Technological improvements and automation have expanded the traditional statistical methods and enabled the analysis of Big Data. This has permitted the discovery of new associations with a granularity that was previously hidden to human eyes. In the first part of this review, the authors would like to provide an overview of basic Machine Learning (ML) principles and techniques in order to better understand their application in recent publications about cardiac arrhythmias. In the second part, ML-enabled advances in disease detection and diagnosis, outcome prediction, and novel disease characterization in topics like electrocardiography, atrial fibrillation, ventricular arrhythmias, and cardiac devices are presented. Finally, the limitations and challenges of applying ML in clinical practice, such as validation, replication, generalizability, and regulatory issues, are discussed. More carefully designed studies and collaborations are needed for ML to become feasible, trustworthy, accurate, and reproducible and to reach its full potential for patient-oriented precision medicine.

Artificial Intelligence and Machine Learning in Cardiac Electrophysiology

Cardiac electrophysiology requires the processing of several patient-specific data points in real time to provide an accurate diagnosis and determine an optimal therapy. Expanding beyond the traditional tools that have been used to extract information from patient-specific data, machine learning offers a new set of advanced tools capable of revealing previously unknown data patterns and features. This new tool set can substantially improve the speed and level of confidence with which electrophysiologists can determine patient-specific diagnoses and therapies. The ability to process substantial amounts of data in real time also paves the way to novel techniques for data collection and visualization. Extended realities such as virtual and augmented reality can now enable the real-time visualization of 3-dimensional images in space. This enables improved preprocedural planning and intraprocedural interventions. Machine learning supplemented with novel visualization technologies could substantially improve patient care and outcomes by helping physicians to make more informed patient-specific decisions. This article presents current applications of machine learning and their use in cardiac electrophysiology.

Identifying patients with heart failure in susceptible to de novo acute kidney injury: a machine learning approach (Preprint)

Background

Studies have shown that more than half of patients with heart failure (HF) with acute kidney injury (AKI) have newonset AKI, and renal function evaluation markers such as estimated glomerular filtration rate are usually not repeatedly tested during the hospitalization. As an independent risk factor, delayed AKI recognition has been shown to be associated with the adverse events of patients with HF, such as chronic kidney disease and death.

Objective

The aim of this study is to develop and assess of an unsupervised machine learning model that identifies patients with HF and normal renal function but who are susceptible to de novo AKI.

Methods

We analyzed an electronic health record data set that included 5075 patients admitted for HF with normal renal function, from which 2 phenogroups were categorized using an unsupervised machine learning algorithm called K-means clustering. We then determined whether the inferred phenogroup index had the potential to be an essential risk indicator by conducting survival analysis, AKI prediction, and the hazard ratio test.

Results

The AKI incidence rate in the generated phenogroup 2 was significantly higher than that in phenogroup 1 (group 1: 106/2823, 3.75%; group 2: 259/2252, 11.50%; P<.001). The survival rate of phenogroup 2 was consistently lower than that of phenogroup 1 (P<.005). According to logistic regression, the univariate model using the phenogroup index achieved promising performance in AKI prediction (sensitivity 0.710). The generated phenogroup index was also significant in serving as a risk indicator for AKI (hazard ratio 3.20, 95% CI 2.55-4.01). Consistent results were yielded by applying the proposed model on an external validation data set extracted from Medical Information Mart for Intensive Care (MIMIC) III pertaining to 1006 patients with HF and normal renal function.

Conclusions

According to a machine learning analysis on electronic health record data, patients with HF who had normal renal function were clustered into separate phenogroups associated with different risk levels of de novo AKI. Our investigation suggests that using machine learning can facilitate patient phengrouping and stratification in clinical settings where the identification of high-risk patients has been challenging.

The Prognostic Value of Anemia in Patients with Preserved, Mildly Reduced and Recovered Ejection Fraction

Data on the relevance of anemia in heart failure (HF) patients with an ejection fraction (EF) > 40% by subgroup—preserved (HFpEF), mildly reduced (HFmrEF) and the newly defined recovered EF (HFrecEF)—are scarce. Patients with HF symptoms, elevated NT-proBNP, EF ≥ 40% and structural abnormalities were registered in the HFpEF-HFmrEF database. We described the outcome of our HFpEF-HFmrEF cohort by the presence of anemia. Additionally, HFrecEF patients were also selected from HFrEF patients who underwent resynchronization and, as responders, reached 40% EF. Using propensity score matching (PSM), 75 pairs from the HFpEF-HFmrEF and HFrecEF groups were matched by their clinical features. After PMS, we compared the survival of the HFpEF-HFmrEF and HFrecEF groups. Log-rank, uni-and multivariate regression analyses were performed. From 375 HFpEF-HFmrEF patients, 42 (11%) died during the median follow-up time of 1.4 years. Anemia (HR 2.77; 95%CI 1.47–5.23; p < 0.01) was one of the strongest mortality predictors, which was also confirmed by the multivariate analysis (aHR 2.33; 95%CI 1.21–4.52; p = 0.01). Through PSM, the outcomes for HFpEF-HFmrEF and HFrecEF patients with anemia were poor, exhibiting no significant difference. In HFpEF-HFmrEF, anemia was an independent mortality predictor. Its presence multiplied the mortality risk in those with EF ≥ 40%, regardless of HF etiology.

Machine learning-derived major adverse event prediction of patients undergoing transvenous lead extraction: Using the ESC EHRA EORP European lead extraction ConTRolled ELECTRa registry

BACKGROUND

Transvenous lead extraction (TLE) remains a high-risk procedure.

OBJECTIVE

The purpose of this study was to develop a machine learning (ML)-based risk stratification system to predict the risk of major adverse events (MAEs) after TLE. A MAE was defined as procedure-related major complication and procedure-related death.

METHODS

We designed and evaluated an ML-based risk stratification system trained using the European Lead Extraction ConTRolled (ELECTRa) registry to predict the risk of MAEs in 3555 patients undergoing TLE and tested this on an independent registry of 1171 patients. ML models were developed, including a self-normalizing neural network (SNN), stepwise logistic regression model ("stepwise model"), support vector machines, and random forest model. These were compared with the ELECTRa Registry Outcome Score (EROS) for MAEs.

RESULTS

There were 53 MAEs (1.7%) in the training cohort and 24 (2.4%) in the test cohort. Thirty-two clinically important features were used to train the models. ML techniques were similar to EROS by balanced accuracy (stepwise model: 0.74 vs EROS: 0.70) and superior by area under the curve (support vector machines: 0.764 vs EROS: 0.677). The SNN provided a finite risk for MAE and accurately identified MAE in 14 of 169 "high (>80%) risk" patients (8.3%) and no MAEs in all 198 "low (<20%) risk" patients (100%).

CONCLUSION

ML models incrementally improved risk prediction for identifying those at risk of MAEs. The SNN has the additional advantage of providing a personalized finite risk assessment for patients. This may aid patient decision making and allow better preoperative risk assessment and resource allocation.

Prognostic Value of Machine Learning in Patients with Acute Myocardial Infarction

(1) Background: Patients with acute myocardial infarction (AMI) still experience many major adverse cardiovascular events (MACEs), including myocardial infarction, heart failure, kidney failure, coronary events, cerebrovascular events, and death. This retrospective study aims to assess the prognostic value of machine learning (ML) for the prediction of MACEs. (2) Methods: Five-hundred patients diagnosed with AMI and who had undergone successful percutaneous coronary intervention were included in the study. Logistic regression (LR) analysis was used to assess the relevance of MACEs and 24 selected clinical variables. Six ML models were developed with five-fold cross-validation in the training dataset and their ability to predict MACEs was compared to LR with the testing dataset. (3) Results: The MACE rate was calculated as 30.6% after a mean follow-up of 1.42 years. Killip classification (Killip IV vs. I class, odds ratio 4.386, 95% confidence interval 1.943–9.904), drug compliance (irregular vs. regular compliance, 3.06, 1.721–5.438), age (per year, 1.025, 1.006–1.044), and creatinine (1 µmol/L, 1.007, 1.002–1.012) and cholesterol levels (1 mmol/L, 0.708, 0.556–0.903) were independent predictors of MACEs. In the training dataset, the best performing model was the random forest (RDF) model with an area under the curve of (0.749, 0.644–0.853) and accuracy of (0.734, 0.647–0.820). In the testing dataset, the RDF showed the most significant survival difference (log-rank p = 0.017) in distinguishing patients with and without MACEs. (4) Conclusions: The RDF model has been identified as superior to other models for MACE prediction in this study. ML methods can be promising for improving optimal predictor selection and clinical outcomes in patients with AMI.

Machine Learning Prediction of Cardiac Resynchronisation Therapy Response From Combination of Clinical and Model-Driven Data

Background: Up to 30–50% of chronic heart failure patients who underwent cardiac resynchronization therapy (CRT) do not respond to the treatment. Therefore, patient stratification for CRT and optimization of CRT device settings remain a challenge.

Objective: The main goal of our study is to develop a predictive model of CRT outcome using a combination of clinical data recorded in patients before CRT and simulations of the response to biventricular (BiV) pacing in personalized computational models of the cardiac electrophysiology.

Materials and Methods: Retrospective data from 57 patients who underwent CRT device implantation was utilized. Positive response to CRT was defined by a 10% increase in the left ventricular ejection fraction in a year after implantation. For each patient, an anatomical model of the heart and torso was reconstructed from MRI and CT images and tailored to ECG recorded in the participant. The models were used to compute ventricular activation time, ECG duration and electrical dyssynchrony indices during intrinsic rhythm and BiV pacing from the sites of implanted leads. For building a predictive model of CRT response, we used clinical data recorded before CRT device implantation together with model-derived biomarkers of ventricular excitation in the left bundle branch block mode of activation and under BiV stimulation. Several Machine Learning (ML) classifiers and feature selection algorithms were tested on the hybrid dataset, and the quality of predictors was assessed using the area under receiver operating curve (ROC AUC). The classifiers on the hybrid data were compared with ML models built on clinical data only.

Results: The best ML classifier utilizing a hybrid set of clinical and model-driven data demonstrated ROC AUC of 0.82, an accuracy of 0.82, sensitivity of 0.85, and specificity of 0.78, improving quality over that of ML predictors built on clinical data from much larger datasets by more than 0.1. Distance from the LV pacing site to the post-infarction zone and ventricular activation characteristics under BiV pacing were shown as the most relevant model-driven features for CRT response classification.

Conclusion: Our results suggest that combination of clinical and model-driven data increases the accuracy of classification models for CRT outcomes.

Machine Learning Algorithms to Distinguish Myocardial Perfusion SPECT Polar Maps

Myocardial perfusion imaging (MPI) plays an important role in patients with suspected and documented coronary artery disease (CAD). Machine Learning (ML) algorithms have been developed for many medical applications with excellent performance. This study used ML algorithms to discern normal and abnormal gated Single Photon Emission Computed Tomography (SPECT) images. We analyzed one thousand and seven polar maps from a database of patients referred to a university hospital for clinically indicated MPI between January 2016 and December 2018. These studies were reported and evaluated by two different expert readers. The image features were extracted from a specific type of polar map segmentation based on horizontal and vertical slices. A senior expert reading was the comparator (gold standard). We used cross-validation to divide the dataset into training and testing subsets, using data augmentation in the training set, and evaluated 04 ML models. All models had accuracy >90% and area under the receiver operating characteristics curve (AUC) >0.80 except for Adaptive Boosting (AUC = 0.77), while all precision and sensitivity obtained were >96 and 92%, respectively. Random Forest had the best performance (AUC: 0.853; accuracy: 0,938; precision: 0.968; sensitivity: 0.963). ML algorithms performed very well in image classification. These models were capable of distinguishing polar maps remarkably into normal and abnormal.

Artificial intelligence and cardiology: Current status and perspective: Artificial Intelligence and Cardiology

The development of artificial intelligence (AI) began in the mid-20th century but has been rapidly accelerating in the past decade. Reflecting the development of digital health over the past few years, this trend is also seen in medicine. The field of cardiovascular medicine uses a wide variety and a large amount of biosignals, so there are many situations where AI can contribute. The development of AI is in progress for all aspects of the healthcare system, including the prevention, screening, and treatment of diseases and the prediction of the prognosis. AI is expected to be used to provide specialist-level medical care, even in a situation where medical resources are scarce. However, like other medical devices, the concept and mechanism of AI must be fully understood when used; otherwise, it may be used inappropriately, resulting in detriment to the patient. Therefore, it is important to understand what we need to know as a cardiologist handling AI. This review introduces the basics and principles of AI, then shows how far the current development of AI has come, and finally gives a brief introduction of how to start the AI development for those who want to develop their own AI.

Machine learning-based risk prediction of malignant arrhythmia in hospitalized patients with heart failure

AIMS

Predicting the risk of malignant arrhythmias (MA) in hospitalized patients with heart failure (HF) is challenging. Machine learning (ML) can handle a large volume of complex data more effectively than traditional statistical methods. This study explored the feasibility of ML methods for predicting the risk of MA in hospitalized HF patients.

METHODS AND RESULTS

We evaluated the baseline data and MA events of 2794 hospitalized HF patients in the HF cohort in Anhui Province and randomly divided the study population into training and validation sets in a 7:3 ratio. The Lasso-logistic regression, multivariate adaptive regression splines (MARS), classification and regression tree (CART), random forest (RF), and eXtreme gradient boosting (XGBoost) algorithms were used to construct risk prediction models in the training set, and model performance was verified in the validation set. The area under the receiver operating characteristic curve (AUC) and Brier score were employed to evaluate the discrimination and calibration of the model, respectively. Clinical utility of the Lasso-logistic regression model was analysed using decision curve analysis (DCA). The median (Q1, Q3) age of the study population was 70 (61, 77) years, and 39.5% were female. MA events occurred in 117 patients (4.2%) during hospitalization. In the training set (n = 1964), the AUC of the XGBoost model was 0.998 [95% confidence interval (CI) 0.997-1.000], which was higher than the other models (all P < 0.001). In the validation set (n = 830), there was no significant difference in AUC of Lasso-logistic model 1 [AUC: 0.867 (95% CI 0.819-0.915)], Lasso-logistic model 2 [AUC: 0.828 (95% CI 0.764-0.892)], MARS model [AUC: 0.852 (95% CI 0.793-0.910)], RF model [AUC: 0.804 (95% CI 0.726-0.881)], and XGBoost model [AUC: 0.864 (95% CI 0.810-0.918); all P > 0.05], which were higher than that of CART model [AUC: 0.743 (95% CI 0.661-0.824); all P < 0.05]. Brier scores for all prediction models were less than 0.05. DCA results showed that the Lasso-logistic model had a net clinical benefit. Oral antiarrhythmic drug, left bundle branch block, serum magnesium, d-dimer, and random blood glucose were significant predictors in half or more of the models.

CONCLUSIONS

The current study findings suggest that ML models based on the Lasso-logistic regression, MARS, RF, and XGBoost algorithms can effectively predict the risk of MA in hospitalized HF patients. The Lasso-logistic model had better clinical interpretability and ease of use than the other models.

What will we ask to artificial intelligence for cardiovascular medicine in the next decade?

Artificial intelligence (AI) comprises a wide range of technologies and methods with heterogeneous degrees of complexity, applications and abilities. In the cardiovascular field, AI holds the potential to fulfil many unsolved challenges, eventually translating into improved patient care. In particular, AI appears as the most promising tool to overcome the gap between ever-increasing data-rich technologies and their practical implementation in cardiovascular research, in the cardiologist routine, in the patient daily life and at the healthcare-policy level. A multiplicity of AI technologies is progressively pervading several aspects of precision cardiovascular medicine including early diagnosis, automated imaging processing and interpretation, disease sub-phenotyping, risk prediction and remote monitoring systems. Several methodological, logistical, educational and ethical challenges are emerging by integrating AI systems at any stage of cardiovascular medicine. This review will discuss the basics of AI methods, the growing body of evidence supporting the role of AI in the cardiovascular field and the challenges to overcome for an effective AI-integrated cardiovascular medicine.

One-year mortality after implantable defibrillator implantation: do risk stratification models help improving clinical practice?

PURPOSE

The purpose of this study was to assess the available mortality risk stratification models for implantable cardioverter defibrillator (ICD) and cardiac resynchronization therapy defibrillator (CRT-D) patients.

METHODS

We conducted a review of mortality risk stratification models and tested their ability to improve prediction of 1-year survival after implant in a database of patients who received a remotely controlled ICD/CRT-D device during routine care and included in the independent Home Monitoring Expert Alliance registry.

RESULTS

We identified ten predicting models published in peer-reviewed journals between 2000 and 2021 (Parkash, PACE, MADIT, aCCI, CHA2DS2-VASc quartiles, CIDS, FADES, Sjoblom, AAACC, and MADIT-ICD non-arrhythmic mortality score) that could be tested in our database as based on common demographic, clinical, echocardiographic, electrocardiographic, and laboratory variables. Our cohort included 1,911 patients with left ventricular dysfunction (median age 71, 18.3% female) from sites not using any risk stratification score for systematic patient screening. Patients received an ICD (53.8%) or CRT-D (46.2%) between 2011 and 2017, after standard physician evaluation. There were 56 deaths within 1-year post-implant, with an all-cause mortality rate of 2.9% (95% confidence interval [CI], 2.3-3.8%). Four predicting models (Parkash, MADIT, AAACC, and MADIT-ICD non-arrhythmic mortality score) were significantly associated with increased risk of 1-year mortality with hazard ratios ranging from 3.75 (CI, 1.31-10.7) to 6.53 (CI 1.52-28.0, p ≤ 0.014 for all four). Positive predictive values of 1-year mortality were below 25% for all models.

CONCLUSION

In our analysis, the models we tested conferred modest incremental predicting power to ordinary screening methods.

Machine Learning Identifies Clinical Parameters to Predict Mortality in Patients Undergoing Transcatheter Mitral Valve Repair

OBJECTIVES

The aim of this study was to develop a machine learning (ML)-based risk stratification tool for 1-year mortality in transcatheter mitral valve repair (TMVR) patients incorporating metabolic and hemodynamic parameters.

BACKGROUND

The lack of appropriate, well-validated, and specific means to risk-stratify patients with mitral regurgitation complicates the evaluation of prognostic benefits of TMVR in clinical trials and practice.

METHODS

A total of 1,009 TMVR patients from 3 university hospitals within the Heart Failure Network Rhineland were included; 1 hospital (n = 317) served as external validation. The primary endpoint was all-cause 1-year mortality. Model performance was assessed using receiver-operating characteristic curve analysis. In the derivation cohort, different ML algorithms were tested using 5-fold cross-validation. The final model, called MITRALITY (transcatheter mitral valve repair mortality prediction system) was tested in the validation cohort with respect to existing clinical scores.

RESULTS

Extreme gradient boosting was selected for the MITRALITY score, using only 6 baseline clinical features for prediction (in order of predictive importance): urea, hemoglobin, N-terminal pro-brain natriuretic peptide, mean arterial pressure, body mass index, and creatinine. In the external validation cohort, the MITRALITY score's area under the curve was 0.783 (95% CI: 0.716-0.849), while existing scores yielded areas under the curve of 0.721 (95% CI: 0.63-0.811) and 0.657 (95% CI: 0.536-0.778) at best.

CONCLUSIONS

The MITRALITY score is a novel, internally and externally validated ML-based tool for risk stratification of patients prior to TMVR, potentially serving future clinical trials and daily clinical practice.

Interpretable prediction of 3-year all-cause mortality in patients with heart failure caused by coronary heart disease based on machine learning and SHAP

BACKGROUND

This study sought to evaluate the performance of machine learning (ML) models and establish an explainable ML model with good prediction of 3-year all-cause mortality in patients with heart failure (HF) caused by coronary heart disease (CHD).

METHODS

We established six ML models using follow-up data to predict 3-year all-cause mortality. Through comprehensive evaluation, the best performing model was used to predict and stratify patients. The log-rank test was used to assess the difference between Kaplan-Meier curves. The association between ML risk and 3-year all-cause mortality was also assessed using multivariable Cox regression. Finally, an explainable approach based on ML and the SHapley Additive exPlanations (SHAP) method was deployed to calculate 3-year all-cause mortality risk and to generate individual explanations of the model's decisions.

RESULTS

The best performing extreme gradient boosting (XGBoost) model was selected to predict and stratify patients. Subjects with a higher ML score had a high hazard of suffering events (hazard ratio [HR]: 10.351; P < 0.001), and this relationship persisted with a multivariable analysis (adjusted HR: 5.343; P < 0.001). Age, N-terminal pro-B-type natriuretic peptide, occupation, New York Heart Association classification, and nitrate drug use were important factors for both genders.

CONCLUSIONS

The ML-based risk stratification tool was able to accurately assess and stratify the risk of 3-year all-cause mortality in patients with HF caused by CHD. ML combined with SHAP could provide an explicit explanation of individualized risk prediction and give physicians an intuitive understanding of the influence of key features in the model.

Machine learning identifies ICU outcome predictors in a multicenter COVID-19 cohort

BACKGROUND

Intensive Care Resources are heavily utilized during the COVID-19 pandemic. However, risk stratification and prediction of SARS-CoV-2 patient clinical outcomes upon ICU admission remain inadequate. This study aimed to develop a machine learning model, based on retrospective & prospective clinical data, to stratify patient risk and predict ICU survival and outcomes.

METHODS

A Germany-wide electronic registry was established to pseudonymously collect admission, therapeutic and discharge information of SARS-CoV-2 ICU patients retrospectively and prospectively. Machine learning approaches were evaluated for the accuracy and interpretability of predictions. The Explainable Boosting Machine approach was selected as the most suitable method. Individual, non-linear shape functions for predictive parameters and parameter interactions are reported.

RESULTS

1039 patients were included in the Explainable Boosting Machine model, 596 patients retrospectively collected, and 443 patients prospectively collected. The model for prediction of general ICU outcome was shown to be more reliable to predict "survival". Age, inflammatory and thrombotic activity, and severity of ARDS at ICU admission were shown to be predictive of ICU survival. Patients' age, pulmonary dysfunction and transfer from an external institution were predictors for ECMO therapy. The interaction of patient age with D-dimer levels on admission and creatinine levels with SOFA score without GCS were predictors for renal replacement therapy.

CONCLUSIONS

Using Explainable Boosting Machine analysis, we confirmed and weighed previously reported and identified novel predictors for outcome in critically ill COVID-19 patients. Using this strategy, predictive modeling of COVID-19 ICU patient outcomes can be performed overcoming the limitations of linear regression models. Trial registration "ClinicalTrials" (clinicaltrials.gov) under NCT04455451.

Using Machine-Learning for Prediction of the Response to Cardiac Resynchronization Therapy: The SMART-AV Study

OBJECTIVES

This study aimed to apply machine learning (ML) to develop a prediction model for short-term cardiac resynchronization therapy (CRT) response to identifying CRT candidates for early multidisciplinary CRT heart failure (HF) care.

BACKGROUND

Multidisciplinary optimization of cardiac resynchronization therapy (CRT) delivery can improve long-term CRT outcomes but requires substantial staff resources.

METHODS

Participants from the SMART-AV (SmartDelay-Determined AV Optimization: Comparison of AV Optimization Methods Used in Cardiac Resynchronization Therapy [CRT]) trial (n = 741; age: 66 ± 11 years; 33% female; 100% New York Heart Association HF class III-IV; 100% ejection fraction ≤35%) were randomly split into training/testing (80%; n = 593) and validation (20%; n = 148) samples. Baseline clinical, electrocardiographic, echocardiographic, and biomarker characteristics, and left ventricular (LV) lead position (43 variables) were included in 8 ML models (random forests, convolutional neural network, lasso, adaptive lasso, plugin lasso, elastic net, ridge, and logistic regression). A composite of freedom from death and HF hospitalization and a >15% reduction in LV end-systolic volume index at 6 months after CRT was the end point.

RESULTS

The primary end point was met by 337 patients (45.5%). The adaptive lasso model was the most more accurate (area under the receiver operating characteristic curve: 0.759; 95% confidence interval [CI]: 0.678-0.840), well calibrated, and parsimonious (19 predictors; nearly half potentially modifiable). Participants in the 5th quintile compared with those in the 1st quintile of the prediction model had 14-fold higher odds of composite CRT response (odds ratio: 14.0; 95% CI: 8.0-14.4). The model predicted CRT response with 70% accuracy, 70% sensitivity, and 70% specificity, and should be further validated in prospective studies.

CONCLUSIONS

ML predicts short-term CRT response and thus may help with CRT procedure and early post-CRT care planning. (SmartDelay-Determined AV Optimization: A Comparison of AV Optimization Methods Used in Cardiac Resynchronization Therapy [CRT] [SMART-AV]; NCT00677014).

Artificial intelligence in the diagnosis and management of arrhythmias

The field of cardiac electrophysiology (EP) had adopted simple artificial intelligence (AI) methodologies for decades. Recent renewed interest in deep learning techniques has opened new frontiers in electrocardiography analysis including signature identification of diseased states. Artificial intelligence advances coupled with simultaneous rapid growth in computational power, sensor technology, and availability of web-based platforms have seen the rapid growth of AI-aided applications and big data research. Changing lifestyles with an expansion of the concept of internet of things and advancements in telecommunication technology have opened doors to population-based detection of atrial fibrillation in ways, which were previously unimaginable. Artificial intelligence-aided advances in 3D cardiac imaging heralded the concept of virtual hearts and the simulation of cardiac arrhythmias. Robotics, completely non-invasive ablation therapy, and the concept of extended realities show promise to revolutionize the future of EP. In this review, we discuss the impact of AI and recent technological advances in all aspects of arrhythmia care.

A review of machine learning for cardiology

This paper reviews recent cardiology literature and reports how Artificial Intelligence Tools (specifically, Machine Learning techniques) are being used by physicians in the field. Each technique is introduced with enough details to allow the understanding of how it works and its intent, but without delving into details that do not add immediate benefits and require expertise in the field. We specifically focus on the principal Machine Learning based risk scores used in cardiovascular research. After introducing them and summarizing their assumptions and biases, we discuss their merits and shortcomings. We report on how frequently they are adopted in the field and suggest why this is the case based on our expertise in Machine Learning. We complete the analysis by reviewing how corresponding statistical approaches compare with them. Finally, we discuss the main open issues in applying Machine Learning tools to cardiology tasks, also drafting possible future directions. Despite the growing interest in these tools, we argue that there are many still underutilized techniques: while Neural Networks are slowly being incorporated in cardiovascular research, other important techniques such as Semi-Supervised Learning and Federated Learning are still underutilized. The former would allow practitioners to harness the information contained in large datasets that are only partially labeled, while the latter would foster collaboration between institutions allowing building larger and better models.

The Role of Artificial Intelligence and Machine Learning in Clinical Cardiac Electrophysiology

In recent years, artificial intelligence (AI) has found numerous applications in cardiology due in part to large digitized datasets and the evolution of high performance computing. In the discipline of cardiac electrophysiology (EP), a number of clinical, imaging, and electrical waveform data are considered in the diagnosis, prognostication and management of arrhythmias, which lend themselves well to automation through AI. But equally relevant, AI offers a unique opportunity to discover novel EP concepts and improve clinical care through its inherent, hierarchical tenets of self-learning. This review will focus on the application of AI in clinical EP and summarize state-of-the art, large, clinical studies in the following key domains: (1) ECG-based arrhythmia and disease classification, (2) atrial fibrillation source detection, (3) substrate and risk assessment for atrial fibrillation and ventricular tachyarrhythmias, and (4) predicting outcomes after cardiac resynchronization therapy. Many are small, single-center, proof-of-concept investigations, but they still demonstrate groundbreaking performance of deep learning, a subdomain of AI, which surpasses traditional statistical analysis. Larger studies, for instance classifying arrhythmias from ECG recordings, have further provided external validation of their high accuracy. Ultimately, the performance of AI is dependent on the quality of the input data and the rigor of algorithm development. The field is still nascent and several barriers will need to be overcome, including prospective validation in large, well-labelled datasets and more seamless information technology-based data collection/integration, before AI can be adopted into broader clinical EP practice. This review will conclude with a discussion of these challenges and future work.

A machine learning-based risk stratification model for ventricular tachycardia and heart failure in hypertrophic cardiomyopathy

BACKGROUND

Machine learning (ML) and artificial intelligence are emerging as important components of precision medicine that enhance diagnosis and risk stratification. Risk stratification tools for hypertrophic cardiomyopathy (HCM) exist, but they are based on traditional statistical methods. The aim was to develop a novel machine learning risk stratification tool for the prediction of 5-year risk in HCM. The goal was to determine if its predictive accuracy is higher than the accuracy of the state-of-the-art tools.

METHOD

Data from a total of 2302 patients were used. The data were comprised of demographic characteristics, genetic data, clinical investigations, medications, and disease-related events. Four classification models were applied to model the risk level, and their decisions were explained using the SHAP (SHapley Additive exPlanations) method. Unwanted cardiac events were defined as sustained ventricular tachycardia occurrence (VT), heart failure (HF), ICD activation, sudden cardiac death (SCD), cardiac death, and all-cause death.

RESULTS

The proposed machine learning approach outperformed the similar existing risk-stratification models for SCD, cardiac death, and all-cause death risk-stratification: it achieved higher AUC by 17%, 9%, and 1%, respectively. The boosted trees achieved the best performing AUC of 0.82. The resulting model most accurately predicts VT, HF, and ICD with AUCs of 0.90, 0.88, and 0.87, respectively.

CONCLUSIONS

The proposed risk-stratification model demonstrates high accuracy in predicting events in patients with hypertrophic cardiomyopathy. The use of a machine-learning risk stratification model may improve patient management, clinical practice, and outcomes in general.

Machine learning-based long-term outcome prediction in patients undergoing percutaneous coronary intervention

Background

Traditional prognostic risk assessment in patients with coronary artery disease undergoing percutaneous coronary intervention (PCI) is based on a limited selection of clinical and imaging findings. Machine learning (ML) can consider a higher number and complexity of variables and may be useful for characterising cardiovascular risk, predicting outcomes, and identifying biomarkers in large population studies.

Methods

We prospectively enrolled 9,680 consecutive patients with coronary artery disease who underwent PCI at our institution between January 2013 and December 2013. Clinical features were selected and used to train 6 different ML models (support vector machine, decision tree, random forest, gradient boosting decision tree, neural network, and logistic regression) to predict cardiovascular outcomes, 10-fold cross-validation to evaluate the performance of models.

Results

During the 5-year follow-up, 467 (4.82%) patients died. Eighty-seven risk baseline measurements were used to train ML models. Compared with the other models, the random forest model (RF-PCI) exhibited the best performance on predicting all-cause mortality (area under the receiver operating characteristic curve: 0.71±0.04). Calibration plots demonstrated a slight overprediction for patients using the RF-PCI model (Hosmer-Lemeshow test: P>0.05). The top 15 features related to PCI candidates' long-term prognosis, among which 11 were laboratory measures.

Conclusions

ML models improved the prediction of long-term all-cause mortality in patients with coronary artery disease before PCI. The performance of the RF model was better than that of the other models, providing a meaningful stratification.