Cardiovascular Event Prediction by Machine Learning: The Multi-Ethnic Study of Atherosclerosis

Elevated N-terminal prohormone of brain natriuretic peptide among persons living with HIV in a South African peri-urban township

AIMS

Efforts to improve access to antiretroviral therapy (ART) have shifted morbidity and mortality among persons living with HIV (PLWH) from AIDS to non-communicable diseases, such as cardiovascular disease (CVD). However, contemporary data on CVD among PLWH in sub-Saharan Africa in the current ART era are lacking. The aim of this study was to assess the burden of cardiac stress among PLWH in South Africa via measurement of N-terminal prohormone of brain natriuretic peptide (NT-proBNP).

METHODS AND RESULTS

NT-proBNP was measured at baseline in 224 PLWH enrolled in a sub-study of a tuberculosis vaccine trial in Khayelitsha township near Cape Town, South Africa. Thresholds were applied at the assay's limit of detection (≥137 pg/mL) and a level indicative of symptomatic heart failure in the acute setting (>300 pg/mL). Mean (SD) age of participants was 39 (6) years, 86% were female, and 19% were hypertensive. Mean (SD) duration of HIV diagnosis was 8.3 (3.9) years and CD4 + count was 673 (267) with 79% prescribed ART for a duration of 5.6 (2.7) years. Thirty-one percent of participants had NT-proBNP > 300 pg/mL. Elevated vs. undetectable NT-proBNP level was associated with older age (P = 0.04), no ART (P = 0.03), and higher plasma tumour necrosis factor-α (P = 0.01).

CONCLUSIONS

Among South African PLWH largely free of known CVD and on ART with high CD4 + counts and few comorbidities, we observed a high proportion with elevated NT-proBNP levels, suggesting the burden of cardiac stress in this population may be high. This observation underscores the need for more in-depth research, including the current effect of HIV on heart failure risk among a growing ART-treated population in sub-Saharan Africa.

Collaborative AI and Laboratory Medicine integration in precision cardiovascular medicine

Artificial Intelligence (AI) is a broad term that combines computation with sophisticated mathematical models and in turn allows the development of complex algorithms which are capable to simulate human intelligence such as problem solving and learning. It is devised to promote a significant paradigm shift in the most diverse areas of medical knowledge. On the other hand, Cardiology is a vast field dealing with diseases relating to the heart, the circulatory system, and includes coronary heart disease, cerebrovascular disease, rheumatic heart disease and other conditions. AI has emerged as a promising tool in cardiovascular medicine which is aimed in augmenting the effectiveness of the cardiologist and to extend better quality to patients. It has the ability to support decision‑making and improve diagnostic and prognostic performance. Attempt has also been made to explore novel genotypes and phenotypes in existing cardiovascular diseases, improve the quality of patient care, to enablecost-effectiveness with reducereadmissionand mortality rates. Our review addresses the integration of AI and laboratory medicine as an accelerator of personalization care associated with the precision and the need of value creation services in cardiovascular medicine.

Artificial Intelligence and Texture Analysis in Cardiac Imaging

PURPOSE OF REVIEW

The aim of this structured review is to summarize the current research applications and opportunities arising from artificial intelligence (AI) and texture analysis with regard to cardiac imaging.

RECENT FINDINGS

Current research findings suggest tremendous potential for AI in cardiac imaging, especially with regard to objective image analyses, overcoming the limitations of an observer-dependent subjective image interpretation. Researchers have used this technique across multiple imaging modalities, for instance to detect myocardial scars in cardiac MR imaging, to predict contrast enhancement in non-contrast studies, and to improve image acquisition and reconstruction. AI in medical imaging has the potential to provide novel, much-needed applications for improving patient care pertaining to the cardiovascular system. While several shortcomings are still present in the current methodology, AI may serve as a resourceful assistant to radiologists and clinicians alike.

Cardiovascular risk assessment in patients with rheumatoid arthritis using carotid ultrasound B-mode imaging

Rheumatoid arthritis (RA) is a systemic chronic inflammatory disease that affects synovial joints and has various extra-articular manifestations, including atherosclerotic cardiovascular disease (CVD). Patients with RA experience a higher risk of CVD, leading to increased morbidity and mortality. Inflammation is a common phenomenon in RA and CVD. The pathophysiological association between these diseases is still not clear, and, thus, the risk assessment and detection of CVD in such patients is of clinical importance. Recently, artificial intelligence (AI) has gained prominence in advancing healthcare and, therefore, may further help to investigate the RA-CVD association. There are three aims of this review: (1) to summarize the three pathophysiological pathways that link RA to CVD; (2) to identify several traditional and carotid ultrasound image-based CVD risk calculators useful for RA patients, and (3) to understand the role of artificial intelligence in CVD risk assessment in RA patients. Our search strategy involves extensively searches in PubMed and Web of Science databases using search terms associated with CVD risk assessment in RA patients. A total of 120 peer-reviewed articles were screened for this review. We conclude that (a) two of the three pathways directly affect the atherosclerotic process, leading to heart injury, (b) carotid ultrasound image-based calculators have shown superior performance compared with conventional calculators, and (c) AI-based technologies in CVD risk assessment in RA patients are aggressively being adapted for routine practice of RA patients.

Machine Learning Prediction Models for In-Hospital Mortality After Transcatheter Aortic Valve Replacement

OBJECTIVES

This study sought to develop and compare an array of machine learning methods to predict in-hospital mortality after transcatheter aortic valve replacement (TAVR) in the United States.

BACKGROUND

Existing risk prediction tools for in-hospital complications in patients undergoing TAVR have been designed using statistical modeling approaches and have certain limitations.

METHODS

Patient data were obtained from the National Inpatient Sample database from 2012 to 2015. The data were randomly divided into a development cohort (n = 7,615) and a validation cohort (n = 3,268). Logistic regression, artificial neural network, naive Bayes, and random forest machine learning algorithms were applied to obtain in-hospital mortality prediction models.

RESULTS

A total of 10,883 TAVRs were analyzed in our study. The overall in-hospital mortality was 3.6%. Overall, prediction models' performance measured by area under the curve were good (>0.80). The best model was obtained by logistic regression (area under the curve: 0.92; 95% confidence interval: 0.89 to 0.95). Most obtained models plateaued after introducing 10 variables. Acute kidney injury was the main predictor of in-hospital mortality ranked with the highest mean importance in all the models. The National Inpatient Sample TAVR score showed the best discrimination among available TAVR prediction scores.

CONCLUSIONS

Machine learning methods can generate robust models to predict in-hospital mortality for TAVR. The National Inpatient Sample TAVR score should be considered for prognosis and shared decision making in TAVR patients.

Cardiovascular/stroke risk predictive calculators: a comparison between statistical and machine learning models

Background

Statistically derived cardiovascular risk calculators (CVRC) that use conventional risk factors, generally underestimate or overestimate the risk of cardiovascular disease (CVD) or stroke events primarily due to lack of integration of plaque burden. This study investigates the role of machine learning (ML)-based CVD/stroke risk calculators (CVRC_ML) and compares against statistically derived CVRC (CVRC_Stat) based on (I) conventional factors or (II) combined conventional with plaque burden (integrated factors).

Methods

The proposed study is divided into 3 parts: (I) statistical calculator: initially, the 10-year CVD/stroke risk was computed using 13 types of CVRC_Stat (without and with plaque burden) and binary risk stratification of the patients was performed using the predefined thresholds and risk classes; (II) ML calculator: using the same risk factors (without and with plaque burden), as adopted in 13 different CVRC_Stat, the patients were again risk-stratified using CVRC_ML based on support vector machine (SVM) and finally; (III) both types of calculators were evaluated using AUC based on ROC analysis, which was computed using combination of predicted class and endpoint equivalent to CVD/stroke events.

Results

An Institutional Review Board approved 202 patients (156 males and 46 females) of Japanese ethnicity were recruited for this study with a mean age of 69±11 years. The AUC for 13 different types of CVRC_Stat calculators were: AECRS2.0 (AUC 0.83, P<0.001), QRISK3 (AUC 0.72, P<0.001), WHO (AUC 0.70, P<0.001), ASCVD (AUC 0.67, P<0.001), FRS_cardio (AUC 0.67, P<0.01), FRS_stroke (AUC 0.64, P<0.001), MSRC (AUC 0.63, P=0.03), UKPDS56 (AUC 0.63, P<0.001), NIPPON (AUC 0.63, P<0.001), PROCAM (AUC 0.59, P<0.001), RRS (AUC 0.57, P<0.001), UKPDS60 (AUC 0.53, P<0.001), and SCORE (AUC 0.45, P<0.001), while the AUC for the CVRC_ML with integrated risk factors (AUC 0.88, P<0.001), a 42% increase in performance. The overall risk-stratification accuracy for the CVRC_ML with integrated risk factors was 92.52% which was higher compared all the other CVRC_Stat.

Conclusions

ML-based CVD/stroke risk calculator provided a higher predictive ability of 10-year CVD/stroke compared to the 13 different types of statistically derived risk calculators including integrated model AECRS 2.0.

Detection and Severity Assessment of Peripheral Occlusive Artery Disease via Deep Learning Analysis of Arterial Pulse Waveforms: Proof-of-Concept and Potential Challenges

Toward the ultimate goal of affordable and non-invasive screening of peripheral occlusive artery disease (PAD), the objective of this work is to investigate the potential of deep learning-based arterial pulse waveform analysis in detecting and assessing the severity of PAD. Using an established transmission line model of arterial hemodynamics, a large number of virtual patients associated with PAD of a wide range of severity and the corresponding arterial pulse waveform data were created. A deep convolutional neural network capable of detecting and assessing the severity of PAD based on the analysis of brachial and ankle arterial pulse waveforms was constructed, evaluated for efficacy, and compared with the state-of-the-art ankle-brachial index (ABI) using the virtual patients. The results suggested that deep learning may diagnose PAD more accurately and robustly than ABI. In sum, this work demonstrates the initial proof-of-concept of deep learning-based arterial pulse waveform analysis for affordable and convenient PAD screening as well as presents challenges that must be addressed for real-world clinical applications.

Machine Learning Based Suicide Ideation Prediction for Military Personnel

Military personnel have greater psychological stress and are at higher suicide attempt risk compared with the general population. High mental stress may cause suicide ideations which are crucially driving suicide attempts. However, traditional statistical methods could only find a moderate degree of correlation between psychological stress and suicide ideation in non-psychiatric individuals. This article utilizes machine learning techniques including logistic regression, decision tree, random forest, gradient boosting regression tree, support vector machine and multilayer perceptron to predict the presence of suicide ideation by six important psychological stress domains of the military males and females. The accuracies of all the six machine learning methods are over 98%. Among them, the multilayer perceptron and support vector machine provide the best predictions of suicide ideation approximately to 100%. As compared with the BSRS-5 score ≥7, a conventional criterion, for the presence of suicide ideation ≥1, the proposed algorithms can improve the performances of accuracy, sensitivity, specificity, precision, the AUC of ROC curve and the AUC of PR curve up to 5.7%, 35.9%, 4.6%, 65.2%, 4.3% and 53.2%, respectively; and for the presence of more severely intense suicide ideation ≥2, the improvements are 6.1%, 26.2%, 5.8%, 83.5%, 2.8% and 64.7%, respectively.

An artificial neural network approach for predicting hypertension using NHANES data

This paper focus on a neural network classification model to estimate the association among gender, race, BMI, age, smoking, kidney disease and diabetes in hypertensive patients. It also shows that artificial neural network techniques applied to large clinical data sets may provide a meaningful data-driven approach to categorize patients for population health management, and support in the control and detection of hypertensive patients, which is part of the critical factors for diseases of the heart. Data was obtained from the National Health and Nutrition Examination Survey from 2007 to 2016. This paper utilized an imbalanced data set of 24,434 with (69.71%) non-hypertensive patients, and (30.29%) hypertensive patients. The results indicate a sensitivity of 40%, a specificity of 87%, precision of 57.8% and a measured AUC of 0.77 (95% CI [75.01-79.01]). This paper showed results that are to some degree more effectively than a previous study performed by the authors using a statistical model with similar input features that presents a calculated AUC of 0.73. This classification model can be used as an inference agent to assist the professionals in diseases of the heart field, and can be implemented in applications to assist population health management programs in identifying patients with high risk of developing hypertension.

Epidemiology of Atrial Fibrillation in the 21st Century: Novel Methods and New Insights

Accompanying the aging of populations worldwide, and increased survival with chronic diseases, the incidence and prevalence of atrial fibrillation (AF) are rising, justifying the term global epidemic. This multifactorial arrhythmia is intertwined with common concomitant cardiovascular diseases, which share classical cardiovascular risk factors. Targeted prevention programs are largely missing. Prevention needs to start at an early age with primordial interventions at the population level. The public health dimension of AF motivates research in modifiable AF risk factors and improved precision in AF prediction and management. In this review, we summarize current knowledge in an attempt to untangle these multifaceted associations from an epidemiological perspective. We discuss disease trends, preventive opportunities offered by underlying risk factors and concomitant disorders, current developments in diagnosis and risk prediction, and prognostic implications of AF and its complications. Finally, we review current technological (eg, eHealth) and methodological (artificial intelligence) advances and their relevance for future prevention and disease management.

Age and sex affect deep learning prediction of cardiometabolic risk factors from retinal images

Deep neural networks can extract clinical information, such as diabetic retinopathy status and individual characteristics (e.g. age and sex), from retinal images. Here, we report the first study to train deep learning models with retinal images from 3,000 Qatari citizens participating in the Qatar Biobank study. We investigated whether fundus images can predict cardiometabolic risk factors, such as age, sex, blood pressure, smoking status, glycaemic status, total lipid panel, sex steroid hormones and bioimpedance measurements. Additionally, the role of age and sex as mediating factors when predicting cardiometabolic risk factors from fundus images was studied. Predictions at person-level were made by combining information of an optic disc centred and a macula centred image of both eyes with deep learning models using the MobileNet-V2 architecture. An accurate prediction was obtained for age (mean absolute error (MAE): 2.78 years) and sex (area under the curve: 0.97), while an acceptable performance was achieved for systolic blood pressure (MAE: 8.96 mmHg), diastolic blood pressure (MAE: 6.84 mmHg), Haemoglobin A1c (MAE: 0.61%), relative fat mass (MAE: 5.68 units) and testosterone (MAE: 3.76 nmol/L). We discovered that age and sex were mediating factors when predicting cardiometabolic risk factors from fundus images. We have found that deep learning models indirectly predict sex when trained for testosterone. For blood pressure, Haemoglobin A1c and relative fat mass an influence of age and sex was observed. However, achieved performance cannot be fully explained by the influence of age and sex. In conclusion we confirm that age and sex can be predicted reliably from a fundus image and that unique information is stored in the retina that relates to blood pressure, Haemoglobin A1c and relative fat mass. Future research should focus on stratification when predicting person characteristics from a fundus image.

Clinico-radiological characteristic-based machine learning in reducing unnecessary prostate biopsies of PI-RADS 3 lesions with dual validation

OBJECTIVE

To evaluate machine learning-based classifiers in detecting clinically significant prostate cancer (PCa) with Prostate Imaging Reporting and Data System (PI-RADS) score 3 lesions.

METHODS

We retrospectively enrolled 346 patients with PI-RADS 3 lesions at two institutions. All patients underwent prostate multiparameter MRI (mpMRI) and transperineal MRI-ultrasonography (MRI-US)-targeted biopsy. We collected data on age, pre-biopsy serum prostate-specific antigen (PSA) level, prostate volume (PV), PSA density (PSAD), the location of suspicious PI-RADS 3 lesions, and histopathology results. Four machine learning-based classifiers-logistic regression, support vector machine, eXtreme Gradient Boosting (XGBoost), and random forest-were trained using datasets from Nanjing Drum Tower Hospital. External validation was carried out using datasets from Molinette Hospital.

RESULTS

Among 287 PI-RADS 3 patients, prostate cancer was proven pathologically in 59 (20.6%), and 228 (79.4%) had benign lesions. For 380 PI-RADS 3 lesions, 81 (21.3%) were proven to be PCa and 299 (78.7%) benign. Among four classifiers, the random forest classifier had the best performance in both patient-based and lesion-based datasets, with overall accuracy of 0.713 and 0.860, sensitivity of 0.857 and 0.613, and area under curve (AUC) of 0.771 and 0.832, respectively. In external validation, our best classifiers had an AUC of 0.688 with the best sensitivity (0.870) and specificity (0.500) in the 59 PI-RADS 3 patients in Molinette Hospital dataset.

CONCLUSIONS

The machine learning-based random forest classifier provided a reliable probability if a PI-RADS 3 patient was benign.

KEY POINTS

• Machine learning-based classifiers could combine the clinical characteristics with accessible information on image report of PI-RADS 3 patient to generate a probability of malignancy. • This probability could assist surgeons to make diagnostic decisions with more confidence and higher efficiency.

Machine learning algorithms estimating prognosis and guiding therapy in adult congenital heart disease: data from a single tertiary centre including 10 019 patients

AIMS

To assess the utility of machine learning algorithms on estimating prognosis and guiding therapy in a large cohort of patients with adult congenital heart disease (ACHD) or pulmonary hypertension at a single, tertiary centre.

METHODS AND RESULTS

We included 10 019 adult patients (age 36.3 ± 17.3 years) under follow-up at our institution between 2000 and 2018. Clinical and demographic data, ECG parameters, cardiopulmonary exercise testing, and selected laboratory markers where collected and included in deep learning (DL) algorithms. Specific DL-models were built based on raw data to categorize diagnostic group, disease complexity, and New York Heart Association (NYHA) class. In addition, models were developed to estimate need for discussion at multidisciplinary team (MDT) meetings and to gauge prognosis of individual patients. Overall, the DL-algorithms-based on over 44 000 medical records-categorized diagnosis, disease complexity, and NYHA class with an accuracy of 91.1%, 97.0%, and 90.6%, respectively in the test sample. Similarly, patient presentation at MDT-meetings was predicted with a test sample accuracy of 90.2%. During a median follow-up time of 8 years, 785 patients died. The automatically derived disease severity-score derived from clinical information was related to survival on Cox analysis independently of demographic, exercise, laboratory, and ECG parameters.

CONCLUSION

We present herewith the utility of machine learning algorithms trained on large datasets to estimate prognosis and potentially to guide therapy in ACHD. Due to the largely automated process involved, these DL-algorithms can easily be scaled to multi-institutional datasets to further improve accuracy and ultimately serve as online based decision-making tools.

Artificial intelligence: improving the efficiency of cardiovascular imaging

INTRODUCTION

Artificial intelligence (AI) describes the use of computational techniques to mimic human intelligence. In healthcare, this typically involves large medical datasets being used to predict a diagnosis, identify new disease genotypes or phenotypes, or guide treatment strategies. Noninvasive imaging remains a cornerstone for the diagnosis, risk stratification, and management of patients with cardiovascular disease. AI can facilitate every stage of the imaging process, from acquisition and reconstruction, to segmentation, measurement, interpretation, and subsequent clinical pathways.

AREAS COVERED

In this paper, we review state-of-the-art AI techniques and their current applications in cardiac imaging, and discuss the future role of AI as a precision medicine tool.

EXPERT OPINION

Cardiovascular medicine is primed for scalable AI applications which can interpret vast amounts of clinical and imaging data in greater depth than ever before. AI-augmented medical systems have the potential to improve workflow and provide reproducible and objective quantitative results which can inform clinical decisions. In the foreseeable future, AI may work in the background of cardiac image analysis software and routine clinical reporting, automatically collecting data and enabling real-time diagnosis and risk stratification.

Calcium scoring: a personalized probability assessment predicts the need for additional or alternative testing to coronary CT angiography

Objective

To assess whether anthropometrics, clinical risk factors, and coronary artery calcium score (CACS) can predict the need of further testing after coronary CT angiography (CTA) due to non-diagnostic image quality and/or the presence of significant stenosis.

Methods

Consecutive patients who underwent coronary CTA due to suspected coronary artery disease (CAD) were included in our retrospective analysis. We used multivariate logistic regression and receiver operating characteristics analysis containing anthropometric factors: body mass index, heart rate, and rhythm irregularity (model 1); and parameters used for pre-test likelihood estimation: age, sex, and type of angina (model 2); and also added total calcium score (model 3) to predict downstream testing.

Results

We analyzed 4120 (45.7% female, 57.9 ± 12.1 years) patients. Model 3 significantly outperformed models 1 and 2 (area under the curve, 0.84 [95% CI 0.83–0.86] vs. 0.56 [95% CI 0.54–0.58] and 0.72 [95% CI 0.70–0.74], p < 0.001). For patients with sinus rhythm of 50 bpm, in case of non-specific angina, CACS above 435, 756, and 944; in atypical angina CACS above 381, 702, and 890; and in typical angina CACS above 316, 636, and 824 correspond to 50%, 80%, and 90% probability of further testing, respectively. However, higher heart rates and arrhythmias significantly decrease these cutoffs (p < 0.001).

Conclusion

CACS significantly increases the ability to identify patients in whom deferral from coronary CTA may be advised as CTA does not lead to a final decision regarding CAD management. Our results provide individualized cutoff values for given probabilities of the need of additional testing, which may facilitate personalized decision-making to perform or defer coronary CTA.

Key Points

• Anthropometric parameters on their own are insufficient predictors of downstream testing. Adding parameters of the Diamond and Forrester pre-test likelihood test significantly increases the power of prediction.

• Total CACS is the most important independent predictor to identify patients in whom coronary CTA may not be recommended as CTA does not lead to a final decision regarding CAD management.

• We determined specific CACS cutoff values based on the probability of downstream testing by angina-, arrhythmia-, and heart rate–based groups of patients to help individualize patient management.

Electronic supplementary material

The online version of this article (10.1007/s00330-020-06921-7) contains supplementary material, which is available to authorized users.

Artificial Intelligence in Cardiology: Present and Future

Artificial intelligence (AI) is a nontechnical, popular term that refers to machine learning of various types but most often to deep neural networks. Cardiology is at the forefront of AI in medicine. For this review, we searched PubMed and MEDLINE databases with no date restriction using search terms related to AI and cardiology. Articles were selected for inclusion on the basis of relevance. We highlight the major achievements in recent years in nearly all areas of cardiology and underscore the mounting evidence suggesting how AI will take center stage in the field. Artificial intelligence requires a close collaboration among computer scientists, clinical investigators, clinicians, and other users in order to identify the most relevant problems to be solved. Best practices in the generation and implementation of AI include the selection of ideal data sources, taking into account common challenges during the interpretation, validation, and generalizability of findings, and addressing safety and ethical concerns before final implementation. The future of AI in cardiology and in medicine in general is bright as the collaboration between investigators and clinicians continues to excel.

Incidental Coronary Artery Calcification and Stroke Risk in Patients With Atrial Fibrillation

OBJECTIVE. Atrial fibrillation (AF) is a major risk factor for stroke. The CHA₂DS_2-VASc score is used to risk stratify patients, and the score includes known coronary artery disease (CAD) as a variable. The aim of this study was to assess if the presence of incidental coronary artery calcification (CAC), without known CAD, is associated with stroke independent of CHA₂DS₂-VASc variables. MATERIALS AND METHODS. A retrospective review of health records was performed for patients who had AF, a chest CT scan performed within 1 year, and a subsequent visit for stroke. Patients with CAD and other vascular disease, a mechanical valve, or who were older than 74 years old were excluded. Included patients were one-to-one matched by age and CHA₂DS₂-VASc risk factors to patients who had had similar follow-up but who did not have a stroke. Nongated CT images were reviewed for CAC. Univariate and Cox regression analyses were performed. RESULTS. A total of 203 patients met the study criteria, and 203 matched patients without stroke were identified. Median age was 61 years old with stroke and 62 years old without stroke (p = 0.99). In both groups, 82 (39.0%) were women and the median CHA₂DS₂-VASc was 2 (interquartile range, 1-2). Anticoagulation medication was prescribed to 46 (22.7%) patients in the group who had had a stroke and 52 (25.6%) in the group without stroke (p = 0.49). On Cox regression analysis, CAC was associated with stroke (hazard ratio [HR], 1.47; 95% CI, 1.10-1.97; p < 0.01) and mortality (adjusted HR, 1.41; 95% CI, 1.02-1.95; p = 0.04). CONCLUSION. Patients with AF and incidental CAC depicted on chest CT have an increased risk of stroke and mortality beyond established risk factors.

An Electrocardiographic System With Anthropometrics via Machine Learning to Screen Left Ventricular Hypertrophy among Young Adults

The prevalence of physiological and pathological left ventricular hypertrophy (LVH) among young adults is about 5%. A use of electrocardiographic (ECG) voltage criteria and machine learning for the ECG parameters to identify the presence of LVH is estimated only 20-30% in the general population. The aim of this study is to develop an ECG system with anthropometric data using machine learning to increase the accuracy and sensitivity for a screen of LVH. In a large sample of 2,196 males, aged 17-45 years, the support vector machine (SVM) classifier is used as the machine learning method for 31 characteristics including age, body height and body weight in addition to 28 ECG parameters such as axes, intervals and voltages to link the output of LVH. The diagnosis of LVH is based on the echocardiographic criteria for young males to be 116 gram/meter2 (left ventricular mass (LVM)/body surface area) or 49 gram/meter2.7 (LVM/body height2.7). On the purpose of increasing sensitivity, the specificity is adjusted around 70-75% and all data tested in proposed model reveal high sensitivity to 86.7%. The area under curve (AUC) of the Precision-Recall (PR) curve is 0.308 in the proposed model which is better than 0.109 and 0.077 using Cornell and Sokolow-Lyon voltage criteria for LVH, respectively. Our system provides a novel screening tool using age, body height, body weight and ECG data to identify most of the LVH among young adults. It provides a fast, accurate and practical diagnosis tool to identify LVH.

Artificial Intelligence in Cardiology: Concepts, Tools and Challenges - "The Horse is the One Who Runs, You Must Be the Jockey"

The recent advances at hardware level and the increasing requirement of personalization of care associated with the urgent needs of value creation for the patients has helped Artificial Intelligence (AI) to promote a significant paradigm shift in the most diverse areas of medical knowledge, particularly in Cardiology, for its ability to support decision-making and improve diagnostic and prognostic performance. In this context, the present work does a non-systematic review of the main papers published on AI in Cardiology, focusing on its main applications, potential impacts and challenges.

Machine learning for characterizing risk of type 2 diabetes mellitus in a rural Chinese population: the Henan Rural Cohort Study

With the development of data mining, machine learning offers opportunities to improve discrimination by analyzing complex interactions among massive variables. To test the ability of machine learning algorithms for predicting risk of type 2 diabetes mellitus (T2DM) in a rural Chinese population, we focus on a total of 36,652 eligible participants from the Henan Rural Cohort Study. Risk assessment models for T2DM were developed using six machine learning algorithms, including logistic regression (LR), classification and regression tree (CART), artificial neural networks (ANN), support vector machine (SVM), random forest (RF) and gradient boosting machine (GBM). The model performance was measured in an area under the receiver operating characteristic curve, sensitivity, specificity, positive predictive value, negative predictive value and area under precision recall curve. The importance of variables was identified based on each classifier and the shapley additive explanations approach. Using all available variables, all models for predicting risk of T2DM demonstrated strong predictive performance, with AUCs ranging between 0.811 and 0.872 using laboratory data and from 0.767 to 0.817 without laboratory data. Among them, the GBM model performed best (AUC: 0.872 with laboratory data and 0.817 without laboratory data). Performance of models plateaued when introduced 30 variables to each model except CART model. Among the top-10 variables across all methods were sweet flavor, urine glucose, age, heart rate, creatinine, waist circumference, uric acid, pulse pressure, insulin, and hypertension. New important risk factors (urinary indicators, sweet flavor) were not found in previous risk prediction methods, but determined by machine learning in our study. Through the results, machine learning methods showed competence in predicting risk of T2DM, leading to greater insights on disease risk factors with no priori assumption of causality.

A Novel Fracture Prediction Model Using Machine Learning in a Community-Based Cohort

The prediction of fracture risk in osteoporotic patients has been a topic of interest for decades, and models have been developed for the accurate prediction of fracture, including the fracture risk assessment tool (FRAX). As machine-learning methodologies have recently emerged as a potential model for medical prediction tools, we aimed to develop a novel fracture prediction model using machine-learning methods in a prospective community-based cohort. In this study, 2227 participants (1257 females) with a baseline bone mineral density (BMD) and trabecular bone score were enrolled from the Ansung cohort. The primary endpoint was the fragility fractures reported by patients or confirmed by X-rays. We used 3 different models: CatBoost, support vector machine (SVM), and logistic regression. During a mean 7.5-year follow-up (range, 2.5 to 10 years), fragility fractures occurred in 537 (25.6%) of participants. In predicting total fragility fractures, the area under the curve (AUC) values of the CatBoost, SVM, and logistic regression models were 0.688, 0.500, and 0.614, respectively. The AUC value of CatBoost was significantly better than that of FRAX (0.663; p < 0.001), whereas the the SVM and logistic regression models were not. Compared with the conventional models such as SVM and logistic regression, the CatBoost model had the best performance in predicting total fragility fractures (p < 0.001). According to feature importance in the CatBoost model, the top predicting factors (listed in order) were total hip, lumbar spine, and femur neck BMD, subjective arthralgia score, serum creatinine, and homocysteine. The latter three factors were listed higher than conventional predictors such as age or previous fracture history. In summary, we hereby report the development of a prediction model for fragility fractures using a machine-learning method, CatBoost, which outperforms the FRAX model as well as two conventional machine-learning models. The model was also able to propose novel high-ranking predictors. © 2020 The Authors. JBMR Plus published by Wiley Periodicals, Inc. on behalf of American Society for Bone and Mineral Research.

Machine learning can predict survival of patients with heart failure from serum creatinine and ejection fraction alone

BACKGROUND

Cardiovascular diseases kill approximately 17 million people globally every year, and they mainly exhibit as myocardial infarctions and heart failures. Heart failure (HF) occurs when the heart cannot pump enough blood to meet the needs of the body.Available electronic medical records of patients quantify symptoms, body features, and clinical laboratory test values, which can be used to perform biostatistics analysis aimed at highlighting patterns and correlations otherwise undetectable by medical doctors. Machine learning, in particular, can predict patients' survival from their data and can individuate the most important features among those included in their medical records.

METHODS

In this paper, we analyze a dataset of 299 patients with heart failure collected in 2015. We apply several machine learning classifiers to both predict the patients survival, and rank the features corresponding to the most important risk factors. We also perform an alternative feature ranking analysis by employing traditional biostatistics tests, and compare these results with those provided by the machine learning algorithms. Since both feature ranking approaches clearly identify serum creatinine and ejection fraction as the two most relevant features, we then build the machine learning survival prediction models on these two factors alone.

RESULTS

Our results of these two-feature models show not only that serum creatinine and ejection fraction are sufficient to predict survival of heart failure patients from medical records, but also that using these two features alone can lead to more accurate predictions than using the original dataset features in its entirety. We also carry out an analysis including the follow-up month of each patient: even in this case, serum creatinine and ejection fraction are the most predictive clinical features of the dataset, and are sufficient to predict patients' survival.

CONCLUSIONS

This discovery has the potential to impact on clinical practice, becoming a new supporting tool for physicians when predicting if a heart failure patient will survive or not. Indeed, medical doctors aiming at understanding if a patient will survive after heart failure may focus mainly on serum creatinine and ejection fraction.

Heart Disease and Stroke Statistics-2020 Update: A Report From the American Heart Association

BACKGROUND

The American Heart Association, in conjunction with the National Institutes of Health, annually reports on the most up-to-date statistics related to heart disease, stroke, and cardiovascular risk factors, including core health behaviors (smoking, physical activity, diet, and weight) and health factors (cholesterol, blood pressure, and glucose control) that contribute to cardiovascular health. The Statistical Update presents the latest data on a range of major clinical heart and circulatory disease conditions (including stroke, congenital heart disease, rhythm disorders, subclinical atherosclerosis, coronary heart disease, heart failure, valvular disease, venous disease, and peripheral artery disease) and the associated outcomes (including quality of care, procedures, and economic costs).

METHODS

The American Heart Association, through its Statistics Committee, continuously monitors and evaluates sources of data on heart disease and stroke in the United States to provide the most current information available in the annual Statistical Update. The 2020 Statistical Update is the product of a full year's worth of effort by dedicated volunteer clinicians and scientists, committed government professionals, and American Heart Association staff members. This year's edition includes data on the monitoring and benefits of cardiovascular health in the population, metrics to assess and monitor healthy diets, an enhanced focus on social determinants of health, a focus on the global burden of cardiovascular disease, and further evidence-based approaches to changing behaviors, implementation strategies, and implications of the American Heart Association's 2020 Impact Goals.

RESULTS

Each of the 26 chapters in the Statistical Update focuses on a different topic related to heart disease and stroke statistics.

CONCLUSIONS

The Statistical Update represents a critical resource for the lay public, policy makers, media professionals, clinicians, healthcare administrators, researchers, health advocates, and others seeking the best available data on these factors and conditions.

High Altitude as a Cause of Congenital Heart Defects: A Medical Hypothesis Rediscovered in Ecuador

Background: There are ∼83 million people living at high altitude (>2500 m) worldwide who endure chronic hypoxia conditions. This article aims to analyze the relationship between high altitude, identified in several cities in Ecuador, and the prevalence of congenital heart disease (CHD). Methods: Set in Ecuador, this epidemiological observational cross-sectional study analyzes data over a range of 18 years (from 2000 to 2017), including 34,904 reported cases of CHD, with a mean of 1939 cases per year. Results: The mean prevalence rate of CHD found is 70.6 per 10,000 live newborns. A K-means analysis resulted in three clusters. Cluster 1 shows the lowest altitude and prevalence of CHD, with an average of 2619 m and 63.02 cases per 10,000 live newborns. Cluster 2 presents the second highest altitude and prevalence of CHD, with an average of 2909 m and 72.04 cases per 10,000 live newborns. Cluster 3 shows the highest values of altitude and prevalence of CHD, with an average of 3176 m and 86.62 cases per 10,000 live newborns. Pearson's coefficient is 0.979, so the correlation between the variables is positive. An altitude ranging from 2500 to 2750 m relates to a prevalence of CHD of ≤71 cases per 10,000 live newborns. An altitude ranging from 2751 to 3000 m relates to a prevalence of CHD of >71 and <89 cases per 10,000 live newborns. An altitude ranging between 3001 and 3264 m relates to a prevalence of CHD of ≥89 cases per 10,000 live newborns. Conclusions: The findings show that high altitude (>2500 m), ethnicity (Native American), rural locations, and limited access to health care are factors that influence and increase the prevalence rate of CHD. A correlation coefficient of 0.914 shows the direct relationship between high altitude and prevalence rates of CHD. For each year elapsed, the prevalence of CHD increased by 3.33 cases per 10,000 live newborns.

Hypertension Drug Therapy

Hypertension is still the number one global killer. No matter what causes are, lowering blood pressure can significantly reduce cardiovascular complications, cardiovascular death, and total death. Unfortunately, some hypertensive individuals simply do not know having hypertension. Some knew it but either not being treated or treated but blood pressure does not achieve goal. The reasons for inadequate control of blood pressure are many. One important reason is that we are not very familiar with antihypertensive agents and less attention has been paid to comorbidities, complications as well as the hypertension-modified target organ damage in patients with hypertension. The right antihypertensive drug was not given to the right hypertensive patients at right time. This reviewer studied comprehensively the literature, hopefully that the review will help improve antihypertensive drug selection and antihypertensive therapy.

Evaluation of Risk Prediction Models of Atrial Fibrillation (from the Multi-Ethnic Study of Atherosclerosis [MESA])

Atrial fibrillation (AF) is prevalent and strongly associated with higher cardiovascular disease (CVD) risk. Machine learning is increasingly used to identify novel predictors of CVD risk, but prediction improvements beyond established risk scores are uncertain. We evaluated improvements in predicting 5-year AF risk when adding novel candidate variables identified by machine learning to the CHARGE-AF Enriched score, which includes age, race/ethnicity, height, weight, systolic and diastolic blood pressure, current smoking, use of antihypertensive medication, diabetes, and NT-proBNP. We included 3,534 participants (mean age, 61.3 years; 52.0% female) with complete data from the prospective Multi-Ethnic Study of Atherosclerosis. Incident AF was defined based on study electrocardiograms and hospital discharge diagnosis ICD-9 codes, supplemented by Medicare claims. Prediction performance was evaluated using Cox regression and a parsimonious model was selected using LASSO. Within 5 years of baseline, 124 participants had incident AF. Compared with the CHARGE-AF Enriched model (c-statistic, 0.804), variables identified by machine learning, including biomarkers, cardiac magnetic resonance imaging variables, electrocardiogram variables, and subclinical CVD variables, did not significantly improve prediction. A 23-item score derived by machine learning achieved a c-statistic of 0.806, whereas a parsimonious model including the clinical risk factors age, weight, current smoking, NT-proBNP, coronary artery calcium score, and cardiac troponin-T achieved a c-statistic of 0.802. This analysis confirms that the CHARGE-AF Enriched model and a parsimonious 6-item model performed similarly to a more extensive model derived by machine learning. In conclusion, these simple models remain the gold standard for risk prediction of AF, although addition of the coronary artery calcium score should be considered.

Neutrophils remain detrimentally active in hydroxyurea-treated patients with sickle cell disease

Neutrophilia is a feature of sickle cell disease (SCD) that has been consistently correlated with clinical severity and has been shown to remain highly activated even at steady state. In addition to induction of fetal hemoglobin (HbF), hydroxyurea (HU) leads to reduction in neutrophil count and their adhesion properties, which contributes to the clinical efficacy of HU in SCD. Although HU reduces the frequency and severity of acute vaso-occlusive crises (VOCs) and chest syndrome, HU therapy does not abolish these acute clinical events. In this study we investigated whether neutrophils in SCD patients whilst on HU therapy retained features of detrimental pro-inflammatory activity. Freshly isolated neutrophils from SCD patients on HU therapy at steady state and from ethnic-matched healthy controls were evaluated ex vivo for their degranulation response and production of neutrophil extracellular traps (NETs). Unstimulated SCD patient neutrophils already produced NETs within 30 minutes, compared to none for healthy neutrophils, and by 4 hours, these neutrophils produced significantly more NETs than the control neutrophils (P = 0.0079**). Higher numbers of neutrophils from SCD patients also showed higher degree of degranulation-related intracellular features compared to healthy neutrophils, including rough-textured cellular membranes (P = 0.03*), double-positivity for F-Actin and CD63 (P = 0.02*) and re-located CD63 within cytoplasm more efficiently than their healthy counterparts (P = 0.02*). The neutrophils from SCD donors released more myeloperoxidase (P = 0.02*) in the absence of any trigger. Our data showed that neutrophils from patients with SCD at steady state remained active during hydroxyurea treatment and are likely to be able to contribute to the SCD pro-inflammatory environment.

Radiomics for Distinguishing Myocardial Infarction from Myocarditis at Late Gadolinium Enhancement at MRI: Comparison with Subjective Visual Analysis

Purpose

To evaluate whether radiomics features of late gadolinium enhancement (LGE) regions at cardiac MRI enable distinction between myocardial infarction (MI) and myocarditis and to compare radiomics with subjective visual analyses by readers with different experience levels.

Materials and Methods

In this retrospective, institutional review board-approved study, consecutive MRI examinations of 111 patients with MI and 62 patients with myocarditis showing LGE were included. By using open-source software, classification performances attained from two-dimensional (2D) and three-dimensional (3D) texture analysis, shape, and first-order descriptors were compared, applying five different machine learning algorithms. A nested, stratified 10-fold cross-validation was performed. Classification performances were compared through Wilcoxon signed-rank tests. Supervised and unsupervised feature selection techniques were tested; the effect of resampling MR images was analyzed. Subjective image analysis was performed on 2D and 3D image sets by two independent, blinded readers with different experience levels.

Results

When trained with recursive feature elimination (RFE), a support vector machine achieved the best results (accuracy: 88%) for 2D features, whereas linear discriminant analysis (LDA) showed the highest accuracy (85%) for 3D features (P <.05). When trained with principal component analysis (PCA), LDA attained the highest accuracy with both 2D (86%) and 3D (89%; P =.4) features. Results found for classifiers trained with spline resampling were less accurate than those achieved with one-dimensional (1D) nearest-neighbor interpolation (P <.05), whereas results for classifiers trained with 1D nearest-neighbor interpolation and without resampling were similar (P =.1). As compared with the radiomics approach, subjective visual analysis performance was lower for the less experienced and higher for the experienced reader for both 2D and 3D data.

Conclusion

Radiomics features of LGE permit the distinction between MI and myocarditis with high accuracy by using either 2D features and RFE or 3D features and PCA.© RSNA, 2019Supplemental material is available for this article.

Application of machine learning to determine top predictors of noncalcified coronary burden in psoriasis: An observational cohort study

BACKGROUND

Psoriasis is associated with elevated risk of heart attack and increased accumulation of subclinical noncalcified coronary burden by coronary computed tomography angiography (CCTA). Machine learning algorithms have been shown to effectively analyze well-characterized data sets.

OBJECTIVE

In this study, we used machine learning algorithms to determine the top predictors of noncalcified coronary burden by CCTA in psoriasis.

METHODS

The analysis included 263 consecutive patients with 63 available variables from the Psoriasis Atherosclerosis Cardiometabolic Initiative. The random forest algorithm was used to determine the top predictors of noncalcified coronary burden by CCTA. We evaluated our results using linear regression models.

RESULTS

Using the random forest algorithm, we found that the top 10 predictors of noncalcified coronary burden were body mass index, visceral adiposity, total adiposity, apolipoprotein A1, high-density lipoprotein, erythrocyte sedimentation rate, subcutaneous adiposity, small low-density lipoprotein particle, cholesterol efflux capacity and the absolute granulocyte count. Linear regression of noncalcified coronary burden yielded results consistent with our machine learning output.

LIMITATION

We were unable to provide external validation and did not study cardiovascular events.

CONCLUSION

Machine learning methods identified the top predictors of noncalcified coronary burden in psoriasis. These factors were related to obesity, dyslipidemia, and inflammation, showing that these are important targets when treating comorbidities in psoriasis.

Extreme Gradient Boosting Model Has a Better Performance in Predicting the Risk of 90-Day Readmissions in Patients with Ischaemic Stroke

OBJECT

Ischemic stroke readmission within 90 days of hospital discharge is an important quality of care metric. The readmission rates of ischemic stroke patients are usually higher than those of patients with other chronic diseases. Our aim was to identify the ischemic stroke readmission risk factors and establish a 90-day readmission prediction model for first-time ischemic stroke patients.

METHODS

The readmission prediction model was developed using the extreme gradient boosting (XGboost) model, which can generate an ensemble of classification trees and assign a predictive risk score to each feature. The patient data were split into a training set (5159) and a validation set (911). The prediction results were evaluated with the receiver operating characteristic (ROC) curve and time-dependent ROC curve, which were compared with the outputs from the logistic regression (LR) model.

RESULTS

A total of 6070 adult patients (39.6% female, median age 67 years) without any ischemic attack (IS) history were included, and 520 (8.6%) were readmitted within 90 days. The XGboost-based prediction model achieved a standard area under the curve (AUC) value of .782 (.729-.834), and the best time-dependent AUC value was .808 in 54 days for the validation set. In contrast, the LR model yielded a standard AUC value of .771 (.714-.828) and best time-dependent AUC value of .797.

CONCLUSIONS

The XGboost model obtained a better risk prediction for 90-day readmission for first-time ischemic stroke patients than the LR model. This model can also reveal the high risk factors for stroke readmission in first-time ischemic stroke patients.

Machine learning in cardiovascular magnetic resonance: basic concepts and applications

Machine learning (ML) is making a dramatic impact on cardiovascular magnetic resonance (CMR) in many ways. This review seeks to highlight the major areas in CMR where ML, and deep learning in particular, can assist clinicians and engineers in improving imaging efficiency, quality, image analysis and interpretation, as well as patient evaluation. We discuss recent developments in the field of ML relevant to CMR in the areas of image acquisition & reconstruction, image analysis, diagnostic evaluation and derivation of prognostic information. To date, the main impact of ML in CMR has been to significantly reduce the time required for image segmentation and analysis. Accurate and reproducible fully automated quantification of left and right ventricular mass and volume is now available in commercial products. Active research areas include reduction of image acquisition and reconstruction time, improving spatial and temporal resolution, and analysis of perfusion and myocardial mapping. Although large cohort studies are providing valuable data sets for ML training, care must be taken in extending applications to specific patient groups. Since ML algorithms can fail in unpredictable ways, it is important to mitigate this by open source publication of computational processes and datasets. Furthermore, controlled trials are needed to evaluate methods across multiple centers and patient groups.

Machine learning to predict cardiovascular risk

AIMS

To analyse the predictive capacity of 15 machine learning methods for estimating cardiovascular risk in a cohort and to compare them with other risk scales.

METHODS

We calculated cardiovascular risk by means of 15 machine-learning methods and using the SCORE and REGICOR scales and in 38 527 patients in the Spanish ESCARVAL RISK cohort, with 5-year follow-up. We considered patients to be at high risk when the risk of a cardiovascular event was over 5% (according to SCORE and machine learning methods) or over 10% (using REGICOR). The area under the receiver operating curve (AUC) and the C-index were calculated, as well as the diagnostic accuracy rate, error rate, sensitivity, specificity, positive and negative predictive values, positive likelihood ratio, and number needed to treat to prevent a harmful outcome.

RESULTS

The method with the greatest predictive capacity was quadratic discriminant analysis, with an AUC of 0.7086, followed by Naive Bayes and neural networks, with AUCs of 0.7084 and 0.7042, respectively. REGICOR and SCORE ranked 11th and 12th, respectively, in predictive capacity, with AUCs of 0.63. Seven machine learning methods showed a 7% higher predictive capacity (AUC) as well as higher sensitivity and specificity than the REGICOR and SCORE scales.

CONCLUSIONS

Ten of the 15 machine learning methods tested have a better predictive capacity for cardiovascular events and better classification indicators than the SCORE and REGICOR risk assessment scales commonly used in clinical practice in Spain. Machine learning methods should be considered in the development of future cardiovascular risk scales.

Development and verification of prediction models for preventing cardiovascular diseases

Objectives

Cardiovascular disease (CVD) is one of the major causes of death worldwide. For improved accuracy of CVD prediction, risk classification was performed using national time-series health examination data. The data offers an opportunity to access deep learning (RNN-LSTM), which is widely known as an outstanding algorithm for analyzing time-series datasets. The objective of this study was to show the improved accuracy of deep learning by comparing the performance of a Cox hazard regression and RNN-LSTM based on survival analysis.

Methods and findings

We selected 361,239 subjects (age 40 to 79 years) with more than two health examination records from 2002–2006 using the National Health Insurance System-National Health Screening Cohort (NHIS-HEALS). The average number of health screenings (from 2002–2013) used in the analysis was 2.9 ± 1.0. Two CVD prediction models were developed from the NHIS-HEALS data: a Cox hazard regression model and a deep learning model. In an internal validation of the NHIS-HEALS dataset, the Cox regression model showed a highest time-dependent area under the curve (AUC) of 0.79 (95% CI 0.70 to 0.87) for in females and 0.75 (95% CI 0.70 to 0.80) in males at 2 years. The deep learning model showed a highest time-dependent AUC of 0.94 (95% CI 0.91 to 0.97) for in females and 0.96 (95% CI 0.95 to 0.97) in males at 2 years. Layer-wise Relevance Propagation (LRP) revealed that age was the variable that had the greatest effect on CVD, followed by systolic blood pressure (SBP) and diastolic blood pressure (DBP), in that order.

Conclusion

The performance of the deep learning model for predicting CVD occurrences was better than that of the Cox regression model. In addition, it was confirmed that the known risk factors shown to be important by previous clinical studies were extracted from the study results using LRP.

Artificial Intelligence Will Transform Cardiac Imaging-Opportunities and Challenges

Artificial intelligence (AI) using machine learning techniques will change healthcare as we know it. While healthcare AI applications are currently trailing behind popular AI applications, such as personalized web-based advertising, the pace of research and deployment is picking up and about to become disruptive. Overcoming challenges such as patient and public support, transparency over the legal basis for healthcare data use, privacy preservation, technical challenges related to accessing large-scale data from healthcare systems not designed for Big Data analysis, and deployment of AI in routine clinical practice will be crucial. Cardiac imaging and imaging of other body parts is likely to be at the frontier for the development of applications as pattern recognition and machine learning are a significant strength of AI with practical links to image processing. Many opportunities in cardiac imaging exist where AI will impact patients, medical staff, hospitals, commissioners and thus, the entire healthcare system. This perspective article will outline our vision for AI in cardiac imaging with examples of potential applications, challenges and some lessons learnt in recent years.

Value of machine learning in predicting TAVI outcomes

BACKGROUND

Transcatheter aortic valve implantation (TAVI) has become a commonly applied procedure for high-risk aortic valve stenosis patients. However, for some patients, this procedure does not result in the expected benefits. Previous studies indicated that it is difficult to predict the beneficial effects for specific patients. We aim to study the accuracy of various traditional machine learning (ML) algorithms in the prediction of TAVI outcomes.

METHODS AND RESULTS

Clinical and laboratory data from 1,478 TAVI patients from a single centre were collected. The outcome measures were improvement of dyspnoea and mortality. Three experiments were performed using (1) screening data, (2) laboratory data, and (3) the combination of both. Five well-established ML techniques were implemented, and the models were evaluated based on the area under the curve (AUC). Random forest classifier achieved the highest AUC (0.70) for predicting mortality. Logistic regression had the highest AUC (0.56) in predicting improvement of dyspnoea.

CONCLUSIONS

In our single-centre TAVI population, the tree-based models were slightly more accurate than others in predicting mortality. However, ML models performed poorly in predicting improvement of dyspnoea.

Identification of clinically relevant features in hypertensive patients using penalized regression: a case study of cardiovascular events

Appropriate management of hypertensive patients relies on the accurate identification of clinically relevant features. However, traditional statistical methods may ignore important information in datasets or overlook possible interactions among features. Machine learning may improve the prediction accuracy and interpretability of regression models by identifying the most relevant features in hypertensive patients. We sought the most relevant features for prediction of cardiovascular (CV) events in a hypertensive population. We used the penalized regression models least absolute shrinkage and selection operator (LASSO) and elastic net (EN) to obtain the most parsimonious and accurate models. The clinical parameters and laboratory biomarkers were collected from the clinical records of 1,471 patients receiving care at Mostoles University Hospital. The outcome was the development of major adverse CV events. Cox proportional hazards regression was performed alone and with penalized regression analyses (LASSO and EN), producing three models. The modeling was performed using 10-fold cross-validation to fit the penalized models. The three predictive models were compared and statistically analyzed to assess their classification accuracy, sensitivity, specificity, discriminative power, and calibration accuracy. The standard Cox model identified five relevant features, while LASSO and EN identified only three (age, LDL cholesterol, and kidney function). The accuracies of the models (prediction vs. observation) were 0.767 (Cox model), 0.754 (LASSO), and 0.764 (EN), and the areas under the curve were 0.694, 0.670, and 0.673, respectively. However, pairwise comparison of performance yielded no statistically significant differences. All three calibration curves showed close agreement between the predicted and observed probabilities of the development of a CV event. Although the performance was similar for all three models, both penalized regression analyses produced models with good fit and fewer features than the Cox regression predictive model but with the same accuracy. This case study of predictive models using penalized regression analyses shows that penalized regularization techniques can provide predictive models for CV risk assessment that are parsimonious, highly interpretable, and generalizable and that have good fit. For clinicians, a parsimonious model can be useful where available data are limited, as such a model can offer a simple but efficient way to model the impact of the different features on the prediction of CV events. Management of these features may lower the risk for a CV event. Graphical Abstract In a clinical setting, with numerous biological and laboratory features and incomplete datasets, traditional statistical methods may ignore important information and overlook possible interactions among features. Our aim was to identify the most relevant features to predict cardiovascular events in a hypertensive population, using three different regression approaches for feature selection, to improve the prediction accuracy and interpretability of regression models by identifying the relevant features in these patients.

Using machine learning to predict one-year cardiovascular events in patients with severe dilated cardiomyopathy

PURPOSE

Dilated cardiomyopathy (DCM) is a common form of cardiomyopathy and it is associated with poor outcomes. A poor prognosis of DCM patients with low ejection fraction has been noted in the short-term follow-up. Machine learning (ML) could aid clinicians in risk stratification and patient management after considering the correlation between numerous features and the outcomes. The present study aimed to predict the 1-year cardiovascular events in patients with severe DCM using ML, and aid clinicians in risk stratification and patient management.

MATERIALS AND METHODS

The dataset used to establish the ML model was obtained from 98 patients with severe DCM (LVEF < 35%) from two centres. Totally 32 features from clinical data were input to the ML algorithm, and the significant features highly relevant to the cardiovascular events were selected by Information gain (IG). A naive Bayes classifier was built, and its predictive performance was evaluated using the area under the curve (AUC) of the receiver operating characteristics by 10-fold cross-validation.

RESULTS

During the 1-year follow-up, a total of 22 patients met the criterion of the study end-point. The top features with IG > 0.01 were selected for ML model, including left atrial size (IG = 0.240), QRS duration (IG = 0.200), and systolic blood pressure (IG = 0.151). ML performed well in predicting cardiovascular events in patients with severe DCM (AUC, 0.887 [95% confidence interval, 0.813-0.961]).

CONCLUSIONS

ML effectively predicted risk in patients with severe DCM in 1-year follow-up, and this may direct risk stratification and patient management in the future.

Cardiovascular disease risk prediction using automated machine learning: A prospective study of 423,604 UK Biobank participants

BACKGROUND

Identifying people at risk of cardiovascular diseases (CVD) is a cornerstone of preventative cardiology. Risk prediction models currently recommended by clinical guidelines are typically based on a limited number of predictors with sub-optimal performance across all patient groups. Data-driven techniques based on machine learning (ML) might improve the performance of risk predictions by agnostically discovering novel risk predictors and learning the complex interactions between them. We tested (1) whether ML techniques based on a state-of-the-art automated ML framework (AutoPrognosis) could improve CVD risk prediction compared to traditional approaches, and (2) whether considering non-traditional variables could increase the accuracy of CVD risk predictions.

METHODS AND FINDINGS

Using data on 423,604 participants without CVD at baseline in UK Biobank, we developed a ML-based model for predicting CVD risk based on 473 available variables. Our ML-based model was derived using AutoPrognosis, an algorithmic tool that automatically selects and tunes ensembles of ML modeling pipelines (comprising data imputation, feature processing, classification and calibration algorithms). We compared our model with a well-established risk prediction algorithm based on conventional CVD risk factors (Framingham score), a Cox proportional hazards (PH) model based on familiar risk factors (i.e, age, gender, smoking status, systolic blood pressure, history of diabetes, reception of treatments for hypertension and body mass index), and a Cox PH model based on all of the 473 available variables. Predictive performances were assessed using area under the receiver operating characteristic curve (AUC-ROC). Overall, our AutoPrognosis model improved risk prediction (AUC-ROC: 0.774, 95% CI: 0.768-0.780) compared to Framingham score (AUC-ROC: 0.724, 95% CI: 0.720-0.728, p < 0.001), Cox PH model with conventional risk factors (AUC-ROC: 0.734, 95% CI: 0.729-0.739, p < 0.001), and Cox PH model with all UK Biobank variables (AUC-ROC: 0.758, 95% CI: 0.753-0.763, p < 0.001). Out of 4,801 CVD cases recorded within 5 years of baseline, AutoPrognosis was able to correctly predict 368 more cases compared to the Framingham score. Our AutoPrognosis model included predictors that are not usually considered in existing risk prediction models, such as the individuals' usual walking pace and their self-reported overall health rating. Furthermore, our model improved risk prediction in potentially relevant sub-populations, such as in individuals with history of diabetes. We also highlight the relative benefits accrued from including more information into a predictive model (information gain) as compared to the benefits of using more complex models (modeling gain).

CONCLUSIONS

Our AutoPrognosis model improves the accuracy of CVD risk prediction in the UK Biobank population. This approach performs well in traditionally poorly served patient subgroups. Additionally, AutoPrognosis uncovered novel predictors for CVD disease that may now be tested in prospective studies. We found that the "information gain" achieved by considering more risk factors in the predictive model was significantly higher than the "modeling gain" achieved by adopting complex predictive models.

A Special Report on Changing Trends in Preventive Stroke/Cardiovascular Risk Assessment Via B-Mode Ultrasonography

PURPOSE OF REVIEW

Cardiovascular disease (CVD) and stroke risk assessment have been largely based on the success of traditional statistically derived risk calculators such as Pooled Cohort Risk Score or Framingham Risk Score. However, over the last decade, automated computational paradigms such as machine learning (ML) and deep learning (DL) techniques have penetrated into a variety of medical domains including CVD/stroke risk assessment. This review is mainly focused on the changing trends in CVD/stroke risk assessment and its stratification from statistical-based models to ML-based paradigms using non-invasive carotid ultrasonography.

RECENT FINDINGS

In this review, ML-based strategies are categorized into two types: non-image (or conventional ML-based) and image-based (or integrated ML-based). The success of conventional (non-image-based) ML-based algorithms lies in the different data-driven patterns or features which are used to train the ML systems. Typically these features are the patients' demographics, serum biomarkers, and multiple clinical parameters. The integrated (image-based) ML-based algorithms integrate the features derived from the ultrasound scans of the arterial walls (such as morphological measurements) with conventional risk factors in ML frameworks. Even though the review covers ML-based system designs for carotid and coronary ultrasonography, the main focus of the review is on CVD/stroke risk scores based on carotid ultrasound. There are two key conclusions from this review: (i) fusion of image-based features with conventional cardiovascular risk factors can lead to more accurate CVD/stroke risk stratification; (ii) the ability to handle multiple sources of information in big data framework using artificial intelligence-based paradigms (such as ML and DL) is likely to be the future in preventive CVD/stroke risk assessment.

Carotid artery imaging: The study of intra-plaque vascularization and hemorrhage in the era of the "vulnerable" plaque

Intraplaque hemorrhage (IPH) is one of the main factors involved in atherosclerotic plaque (AP) instability. Its recognition is crucial for the correct staging and management of patients with carotid artery plaques to limit ischemic stroke. Imaging plays a crucial role in identifying IPH, even if the great variability of intraplaque vascularization and the limitations of our current imaging technologies make it difficult. The intent of this review is to give a general overview of the main features of intraplaque vascularization and IPH on Ultrasound (US), Computed Tomography (CT), Magnetic Resonance (MR) and Nuclear Medicine, and a brief description on the future prospectives.

Automation and artificial intelligence in the clinical laboratory

The daily operation of clinical laboratories will be drastically impacted by two disruptive technologies: automation and artificial intelligence (the development and use of computer systems able to perform tasks that normally require human intelligence). These technologies will also expand the scope of laboratory medicine. Automation will result in increased efficiency but will require changes to laboratory infrastructure and a shift in workforce training requirements. The application of artificial intelligence to large clinical datasets generated through increased automation will lead to the development of new diagnostic and prognostic models. Together, automation and artificial intelligence will support the move to personalized medicine. Changes in pathology and clinical doctoral scientist training will be necessary to fully participate in these changes. KEYWORDS: Automation; artificial intelligence; deep learning; laboratory medicine.