Disease Prediction Research Articles

Generalization Investigation in Heart Disease Prediction: Comparative Analysis of MLP, Random Forest, and SVM with Cross-Domain Datasets

The traditional way of predicting heart disease is usually through the subjective judgment of doctors, which is high subjectivity. Current machine learning methods do not pay enough attention to its universality. The two datasets used in this study are from Kaggle and preprocessed in advance, including standardization by using Z-Score and dimensionality reduction by KPCA. The two datasets were designated as the source domain and the target domain. The source domain dataset was further divided into two subsets with an 80:20 split, where 80% was used for training the model and the remaining 20% served as the test set. The trained model was then applied to the target domain to compare the differences in prediction results. Through the exploration of Multilayer Perceptron (MLP), Random Forest and Support Vector Machine (SVM), for the data set used in this study, when MLP uses a simple structure, the difference in prediction accuracy dropped from 13.8% to 5.07. For Random Forest, by increasing the number of decision trees and decreasing the minimum number of samples required for splitting, the difference is reduced from 15.6% to 9.32%. By modifying the penalty parameter C value of SVM, the difference on different datasets is reduced from 13.39% to 4.46%. This study is one of the few to explore the generalizations of heart disease prediction. The results demonstrate that the generalization performance of the model for heart disease prediction can be significantly enhanced through appropriate modifications to its structure and hyperparameters.

Enhancing Heart Disease Prediction through Machine Learning: A Comparative Analysis of Algorithm Generalization across Datasets with Various Distributions

Heart disease, due to its high prevalence and mortality, remains a key area of global research. Although traditional diagnostic methods are effective, they are often invasive and time-consuming, highlighting the need for non-invasive, AI-based approaches. A significant challenge in real-world applications is ensuring model generalization across different datasets, particularly when the datasets are small. In this study, the performance of machine learning models, including Decision Tree, Random Forest, Support Vector Machine (SVM), and Multi-Layer Perceptron (MLP), was evaluated on two distinct heart disease datasets with different distributions and relatively small sizes. Two datasets with varying distributions were used for training and testing, with the primary focus on assessing model generalization in crossdataset applications. It is shown in the results that, while the Decision Tree model performed best after hyperparameter tuning, the improvements in Random Forest and MLP were limited, and SVM exhibited a decline in performance after tuning in the cross-dataset task. It was found that grid search tuning has limitations in cross-dataset scenarios, especially with small datasets, where complex models are prone to overfitting. The study demonstrates that, with smaller datasets, simpler models like Decision Trees often adapt better to different datasets. Furthermore, transfer learning and domain adaptation techniques are suggested as crucial for improving model generalization. Future research should focus on employing these techniques to enhance the robustness and accuracy of heart disease prediction models across diverse datasets.

Artificial intelligence – NPHIs’ role in utilizing its opportunities and addressing the risks

Abstract Issue & Problem National Public Health Institutes (NPHIs) in many countries are at the forefront of action to mitigate the impact of current and future threats to health as well as improving population health and wellbeing. Through research, collaboration, and support, NPHIs can approach these new realities together. One of the many emerging realities that NPHIs will need to engage in is how NPHIs best can utilise the powers of artificial intelligence and deal with the ethical and legal issues that come with it. Methods In preparation of IANPHI’s (International Association of National Public Health Institutes) Europe Regional network meeting in April 2024 a survey was conducted among the 42 members. 16 countries replied to questions, around e.g., 1) current applications of (generative) artificial intelligence (AI) in the work of institutes, 2) which experiences were made and 3) which measures NPHIs take to best utilize the possibilities and overcome challenges of AI technologies. Within four weeks 16 replies were shared. Lessons & Conclusions Some institutes have not yet adopted AI, while others report multiple applications but the primary areas of application vary considerably among Europe. Applications include summarizing and reformatting text; translation; transcribing and documenting meetings; search information in documents or databases; coding; creating visuals; literature review; surveillance of infectious diseases; disease forecasting; analysis of transmission dynamics; vaccine efficacy assessment and development; image recognition; diagnosis and prediction of diseases from medical datasets; text mining; symptom extraction; social listening; prediction of nitrate concentration in groundwater; drug analysis and measuring psychological resilience in social crisis. Key messages • AI poses opportunities and risks for public health and work of national public health institutes and capacity building initiatives are essential. • IANPHI decided to develop a joint framework for the use of AI in its member organisations with a focus on research and responsible use.

Antibodies against phosphorylcholine in prediction of cardiovascular disease among women: a population-based prospective cohort study

Abstract Background and aims Antibodies against phosphorylcholine (anti-PC) have been reported as associated with protection against atherosclerosis, cardiovascular disease (CVD) and other chronic inflammatory diseases. Underlying potential mechanisms have been demonstrated and include anti-inflammatory, clearance of dead cells and inhibition of oxidized low-density lipoprotein-effects. The aim is to study the role of IgM anti-PC and incident CVD among women, where less is known than among men in the general population. Methods In a total of 932 women, 66 ± 6 years old at baseline, from the population-based Swedish Mammography Cohort IgM anti-PC levels of sera were measured using ELISA. Prospective associations with any first CVD, ischemic heart disease (IHD), myocardial infarction (MI) and ischemic stroke were assessed using Cox proportional hazard regression, generating hazard ratios (HR) and 95% confidence intervals (CI). The model was adjusted for potential confounding factors. Results Over the course of 16 years (14,900 person-years), we identified 113 cases of composite CVD, - 69 cases of ischemic heart disease (IHD), 44 cases of myocardial infarction (MI) and 50 cases of ischemic stroke. IgM antiPC was statistically significantly inversely associated with risk of CVD, IHD and MI, but not with ischemic stroke. Comparing the highest tertile with lowest, we observed multivariable-adjusted HR of 0.26 (95% CI: 0.10–0.66; p-trend <0.01) for MI. Conclusions IgM anti-PC may play an active role in inhibition of CVD development in women, especially MI. Elevating anti-PC concentration through immunization, may be a future development in CVD prevention.

Prediction of mortality, future arrhythmia and cardiovascular disease: an artificial intelligence-enhanced electrocardiography platform

Abstract Background Artificial intelligence-enhanced electrocardiography (AI-ECG) can be used to identify existing disease, but could additionally be used to predict occurrence of future disease and death. Purpose We developed an AI-ECG platform that predicts mortality and a wide spectrum of future arrhythmia and cardiovascular disease. Methods The AI-ECG risk estimation platform (AIRE) was developed in a dataset of 1,163,401 ECGs from 189,539 patients from a secondary care setting in the USA. Uniquely, AIRE uses deep learning with a residual convolutional neural network with a discrete-time survival loss function to output subject-specific survival curves (Fig 1A). AIRE was fine-tuned to additionally predict the endpoints described below. Results AIRE accurately predicts risk of all-cause mortality (C-index 0.775 (0.773-0.776) and can differentiate prognosis in high- and low-risk subjects, even when only cardiologist-defined normal ECGs are considered (Fig 2B). Additionally, in Cox regression analyses, AIRE was superior to conventional risk factors (Fig 1B). AIRE accurately predicted future ventricular arrhythmia (c-index 0.760 (0.756-0.763)), future complete heart block (CHB) (0.809 (0.805-0.814), future atrial fibrillation (0.753 (0.751-0.756), future atherosclerotic cardiovascular disease (0.696 (0.694-0.698)) and future heart failure (0.787 (0.785-0.789)) . AIRE was superior to left ventricular ejection fraction (LVEF) in predicting risk of future ventricular arrhythmias in subjects with LVEF <50% (C-index 0.649 (0.638 – 0.660) vs 0.615 (0.603 – 0.626) p < 0.0001 ). For CHB prediction, AIRE was able to further risk stratify subjects with bifascicular block and syncope into low (1.9% 5 year CHB), intermediate (8.8%), and high risk (20.7%) groups, and could guide treatment decisions. We externally validated our findings in four diverse, transnational cohorts from Brazil and the UK, including volunteers, primary care subjects and patients with Chagas disease, and found AIRE accurately predicted all-cause mortality in all four external cohorts (Fig 2). Non-mortality endpoints were successful externally validated in the UKB. In explainability analyses, we found features of QRS morphology, low QRS voltage, poor precordial R wave progression, inverted and biphasic T waves and ST segment changes were the most significant morphological features associated with high predicted mortality (Fig 1D). Through phenome- and genome-wide association studies (GWAS), we identified candidate biological pathways including changes in cardiac structure (LV mass, LV trabeculation and left atrial size). GWAS identified variants adjacent to VGLL2, KCNQ1, CCDC91 and TBX3 which have been described in association with QT interval, biological aging and metabolic syndrome (Fig 1E). Conclusion AIRE could be used worldwide across a range of clinical contexts for risk estimation of not only mortality but a wide spectrum of future arrhythmia and cardiovascular diseases.Figure 1Figure 2

The added value of device measured physical activity to the prediction of incident cardiovascular disease.

Abstract Background Conventional risk prediction models for Major Adverse Cardio/Cerebrovascular Events (MACCE) primarily rely on non-modifiable factors such as age, sex, and physiological measurements, including LDL/HDL cholesterol and blood pressure readings, that are prone to variations across medical practices and are often unavailable in many parts of the world. In this study, we investigate the potential of inexpensive wearable measurements of physical activity to enhance MACCE prediction when physiological measurements are unavailable. Purpose To assess the added value of physical activity measurements, collected via wrist-worn accelerometers, to established cardiovascular risk prediction models. Additionally, we seek to investigate whether physical activity measurements can substitute for common physiological measurements in established cardiovascular risk models. Methods We obtained accelerometer-measured physical activity over 7 days between 2013 and 2015 from 69,898 UK Biobank (UKB) participants without a prior history of MACCE, defined as death or hospitalisation due to (1) myocardial infarction, (2) heart failure, or (3) Stroke/transient ischaemic attack. We assessed the added value of daily step counts, sleep duration, and a deep learning-derived activity health score (Fig 1), calculated from complete 7-day accelerometer recordings, to established clinical risk models (SCORE2, QRISK3 and AHA) before and after excluding measurements of cholesterol and systolic blood pressure (SBP). The primary outcome was the first recorded MACCE within 6 years. Risk scores were developed in men and women using Cox proportional hazards models. We assessed predictive performance using Harrel’s C-index, net reclassification index (NRI), and net benefit at the recommended treatment threshold for each model. Results Of the 69,898 UKB participants in the study, 3,386 (4.8%) experienced a MACCE over a 6-year follow-up period. In place of HDL ratio and SBP, the addition of deep-learned activity health scores modestly improved performance in the best clinical baseline, QRISK3 for Female participants ΔC-index women:0.003 (95% CI 0.002-0.004); Δnet benefit 0.02(0.01-0.03); NRI 0.04, (0.02-0.07), and outperformed manually extracted steps and sleep measurements. We report no significant improvements for men. We find a greater increase in model performance when both HDL ratio and SBP, plus activity health scores, are included in the model. ΔC-index women:0.008, (0.007-0.009); men: 0.003 (0.002-0.004) (Fig 2). Conclusion Our findings indicate that in the absence of cholesterol and systolic blood pressure, the addition of device-measured physical activity modestly improves the performance of clinical risk scores among individuals without prior cardiovascular disease (CVD). These findings could further refine intervention strategies for targeted prevention of CVD, particularly in settings where physiological measurements may be unavailable.

Machine-learning based prediction of obstructive coronary artery disease using integrated submodules of clinical information, chest x-ray, and electrocardiography

Abstract Background The early diagnosis of obstructive coronary artery disease (CAD) is critical. However, delays in diagnosis may occur due to subsequent high-cost tests, contributing to substantial medical expenses. A more insightful interpretation of fundamental diagnostic tools, such as chest x-ray (CXR) and electrocardiography (ECG), could mitigate medical costs and facilitate a timely diagnosis. Purpose This study explores the application of machine learning to interpret each modality and create a consolidated prediction model for significant CAD, aiming to construct a diagnostic model based on primary tests, including CXR, ECG, and clinical information. Methods Clinical information-ECG-CXR paired data were generated from 19,140 patients, with CAD confirmed through coronary angiography (CAG). Machine learning (ML) was employed to develop submodules for each modality, and various integration methods were explored. The area under the curve (AUC) served as the outcome metric for predicting obstructive CAD. Results The development set comprised data from 17,976 patients, with CAG-confirmed obstructive CAD observed in approximately 60% of patients. The obstructive CAD group exhibited characteristics such as advanced age, male predominance, a higher prevalence of chest pain, and more risk factors. The ML model, incorporating clinical information, CXR, and ECG with submodules, demonstrated the optimal prediction of obstructive CAD (AUC 0.722). This model significantly outperformed the model without submodules (0.722 vs. 0.638, p<0.0001) in obstructive CAD prediction. Conclusions Through the integration of submodules encompassing clinical information and basic tests, leveraging a high-quality database, the integrated ML algorithm offers improved prediction capabilities for CAG-confirmed obstructive CAD. This underscores the potential role of machine learning in clinical decision-making based on diverse modalities with distinct characteristics.

Polygenic prediction of coronary artery disease among 130,000 Mexican adults

Abstract Background Coronary artery disease (CAD) is a leading cause of premature mortality globally. Polygenic risk scores (PRSs) have been reported to enhance CAD prediction in populations of European ancestry but evidence from other populations is limited. We assessed the transferability and relevance of eight different external PRSs to the odds of premature CAD in a Mexican population with high levels of relatedness and ancestry admixture. Methods 128,388 genotyped participants aged 35-74 years who were recruited between 1998 and 2004 into the Mexico City Prospective Study (MCPS) were included. We selected and recreated eight previously-reported CAD PRSs. The number of single nucleotide polymorphism variants included in the PRSs varied from 44 to 6,472,620. Premature CAD was defined as either self-reported history of heart attack or angina at recruitment or death before age 75 with CAD recorded on the death certificate. Age and sex-adjusted logistic regression was used to estimate the ability of each PRS to predict CAD, both overall and separately by age, sex and estimated proportion of Indigenous American ancestry. Analyses were repeated after further adjustment for conventional risk factors (blood pressure, adiposity, smoking, diabetes and highest education achieved). Risk discrimination was assessed using C-statistic. Results Of the 128,388 participants, 68% were women, mean (SD) age was 50.2±10.9 years and the average proportion of Indigenous American ancestry was 67%. CAD was identified for 3,871 (3.0%) participants; prior CAD was reported by 1,669 (1.3%) participants while 2,360 (1.8%) participants died before age 75 years with CAD mentioned on the death certificate. All eight PRSs were positively and approximately log-linearly associated with the odds of CAD, but the strength of association varied between the PRSs, ranging from a 7% increase per 1SD higher PRS (odds ratio [OR] 1.07, 95% CI 1.04-1.11) to a 33% increase (OR 1.33, 95% CI 1.29-1.37). Collectively, age, sex and (any) PRS resulted in a C-statistic of 0.70-0.71. C-statistics were slightly lower among those with a higher proportion of indigenous American ancestry. For four of the eight PRSs (including three of the four multi-ancestry PRSs) the increase in odds of CAD per 1SD higher PRS was significantly greater for men than women (eg, for the PRS with the strongest overall association with CAD the OR per 1SD higher PRS was 1.43 [95% CI 1.36-1.49] in men and 1.25 [1.20-1.31] in women). Further inclusion of conventional vascular risk factors reduced the strengths of associations of each PRS with CAD only slightly but increased the model C statistic to 0.74 in all cases. Conclusions In this population of admixed Mexican adults, existing PRSs predicted CAD reasonably well and independently of conventional vascular risk factors. Ancestry-specific instruments that more closely represent genetic variation in Mexico may further enhance polygenic prediction of CAD risk.Table 1.Baseline characteristicsFigure 1.Main analysis results

Multi-modality deep learning model for prediction of chronic obstructive coronary artery disease

Abstract Background/Introduction The incidence of coronary artery disease (CAD) is on the rise with an increasing burden of stable ischemic heart disease (SIHD) and high prevalence of obesity, diabetes, hyperlipidemia, and other cardiovascular risk factors. Pre-test probability estimation for CAD based on demographic features and characterization of chest pain is insufficient to accurately assess whether further evaluation for CAD is warranted. Purpose To develop and validate a deep learning (DL) algorithm utilizing electrocardiogram (ECG) waveforms and clinical features to predict obstructive CAD. Methods Clinical data from subjects undergoing invasive angiography over a 4-year period at a quaternary care center exclusive of subjects who had concern for acute coronary syndromes were used for model training. ECGs obtained up to 5 years prior to angiography were included; ECGs were partitioned 8:1:1 for training, validation, and test datasets. Obstructive CAD was defined as receiving percutaneous coronary intervention during elective angiography. Convolutional neural network models were developed for ECG waveforms alone (DL-ECG), clinical features from standard risk scores (DL-Clinical), and the combination of ECG waveforms and clinical features (DL-ECG+Clinical); a commonly used pre-test probability estimation tool derived from the CAD Consortium study was used as a comparison (PTP). The models were evaluated based on area under the curve receiver operating characteristics (AUC). Shapley additive explanations (SHAP) analysis was performed to understand feature importance in the DL models. Results A total of 40,602 ECGs and 8,361 coronary angiograms from 7,201 patients were included in the analysis. The median age was 66 (IQR 56-75), 71.9% were men, 34.9% had hypertension, 30.6% had diabetes mellitus, and 39.0% had dyslipidemia. The PTP model (AUC 0.733 [0.717-0.750]) had similar performance as the DL-Clinical model (AUC 0.762 [0.746-0.778]). The DL-ECG model (AUC 0.741 [0.726-0.758]) had similar performance as both the clinical feature models. The DL-ECG+Clinical model (AUC 0.807 [0.793-0.822]) trained on ECG waveforms and clinical features had a superior performance (Figure 1). SHAP analysis for the DL-ECG+Clinical model revealed that ECG waveforms had the largest influence on the model’s output followed by age and chest pain type (Figure 2). Conclusions Multi-modality DL model utilizing ECG and commonly used features can improve prediction of obstructive CAD in SIHD compared to traditional pre-test probability estimates. Prospective research is warranted to determine whether incorporation of DL methods to ECG analysis effectively improves diagnosis and outcomes in obstructive CAD.Figure 1.Figure 2.

Lipidomics-based deep learning algorithms can enhance prediction of obstructive coronary artery disease

Abstract Background Lipidomics emerge as a promising research field with the potential to help in personalized risk stratification and also improve our understanding on the functional role of individual lipid species in the metabolic perturbations occurring in coronary artery disease (CAD). The existing technological challenges to detect diverse yet structurally similar lipids and their isomers, as well as the need for novel statistical and computational tools to handle the high-dimensional lipidome may be responsible for the reduced clinical translation of the very first lipidomics research outcomes. Methods To that end, this post-hoc analysis of the prospective CorLipid trial aimed to investigate the predictive capability of a lipidomics panel for obstructive CAD risk through an extreme gradient boosting (XGBoost) machine learning (ML) approach. The lipid profiles of 146 individuals with suspected CAD were investigated through liquid chromatography-mass spectrometry. Fasting blood samples were drawn prior to invasive coronary angiography execution. Obstructive CAD was defined as SYNTAX Score (SS) >0 compared to non-obstructive CAD (SS=0). Results Study participants (75.3% male, mean age: 61 ±10.5 years old) were divided into two categories based on the existence of obstructive CAD (54.8% with SS>0 and 45.2% with SS=0). In total, 517 lipid species were identified, from which 290 lipid species were finally quantified in participants’ serum [glycerophospholipids (52.1%), glycerolipids (28.6%) and sphingolipids (19.3%)]. The levels of glycerophospholipid, sphingolipid and glycerolipid classes were significantly different in patients with obstructive CAD. Finally, a ML XGBoost algorithm identified a panel of 17 serum biomarkers (5 sphingolipids, 7 glycerophospholipids, triacylglycerols, galectin-3, glucose, LDL and LDH) as totally sensitive (100% sensitivity, 62.1% specificity and 100% negative predictive value) for the prediction of obstructive CAD. Conclusions These findings provide molecular insights into the role of dysregulated lipid metabolism in the development and progression of CAD while validating the existing body of evidence from similar research studies. Further (ML-based) investigation of lipid metabolism could hold promises for novel therapeutic strategies and improvement of the existing risk stratification schemes.Lipidomics panel importance

ClinicAl verSus algorIthmic predictioN of Obstructive Coronary Artery Disease - the CASINO study

Abstract Background The estimation of pretest probability (PTP) is a key step when evaluating patients with suspected coronary artery disease (CAD). Current European guidelines recommend using an algorithm based on age, sex, and symptom typicality. However, many cardiologists do not regularly use this method, relying instead on overall clinical impression. The aim of this study was to compare clinical and algorithmic prediction of obstructive CAD in a set of symptomatic patients with suspected chronic coronary syndrome (CCS). Methods In this survey study, 10 cardiologists from were asked to estimate the probability of obstructive CAD (on a scale of 1-99% and without the aid of any score or algorithm) for 120 anonymized clinical vignettes of outpatients who underwent diagnostic workup for suspected CCS. The provided information included age, sex, symptom description, cardiovascular risk factors, and know comorbidities. Whenever non-invasive tests had been performed, physicians were also asked to estimate the probability of obstructive CAD after knowing that results. Obstructive CAD was defined as diameter stenosis ≥50% confirmed by invasive coronary angiography. Results Among the 120 included patients (62 women, mean age 61±10 years), symptoms consisted of chest pain in 104 and dyspnea in the remainder. The median PTP of obstructive CAD was 15% (IQR 10-25%) according to the ESC algorithm, and 30% (IQR 15-50%) according to cardiologists’ estimates (p<0.001). The observed prevalence of obstructive CAD was 20% (n=24), yielding suitable calibration for the ESC model (predicted vs. observed p=0.631), and significant overestimation for clinicians’ assessments (predicted vs. observed p=0.014) – Figure 1A. Despite using the same data, cardiologists’ PTP estimates varied widely within the same patients (mean inter-observer coefficient of variation 56%±22%). In a subset of 29 patients whose PTP was appraised twice by the same physicians, the mean absolute difference between assessments (intra-observer variability) was 22%±8%. The ESC algorithm showed fair discriminative power to identify patients with obstructive CAD (C-statistic 0.68, 95% CI 0.58-0.79, p=0.005), whereas the C-statistic for clinicians ranged from 0.53 to 0.72 (average 0.62±0.05, p values for comparisons 0.001-0.677) – Fig 1B. In the subset of 70 patients with prior non-invasive tests, clinicians significantly changed their predictions after knowing test results (mean absolute difference 22%±12%, p<0.001), improving their discriminative power significantly (mean change in C-statistic of 0.08±0.09, p=0.020). Conclusion Clinicians’ assessments of PTP in symptomatic patients display high inter- and intra-observer variability, and tend to overestimate the likelihood of obstructive CAD. Despite using more clinical data, cardiologists fail to outperform the ESC algorithm. The systematic use of this tool should be promoted in order to improve disease prediction and guide non-invasive testing.

Cardiovascular risk prediction using metabolomic biomarkers and polygenic risk scores: a cohort study and modelling analyses

Abstract Background/Introduction Metabolomic biomarker scores and polygenic risk scores (PRS) have shown promise for improving cardiovascular disease (CVD) prediction, but have not yet been evaluated in the context of current prediction models (SCORE2) and ESC recommendations for 10-year prediction of fatal and non-fatal CVD. Purpose To assess whether metabolomics biomarkers and PRSs can improve 10-year CVD risk prediction when added to SCORE2, and whether improvements are meaningful at ESC 2021 recommended risk thresholds for treatment consideration. Methods Metabolomics biomarker scores were constructed and compared to PRS and SCORE2 in 170,000 UK Biobank participants (5,096 incident CVD cases) aged 40–69 without previous CVD, diabetes, or lipid-lowering treatment. Improvement in risk discrimination when added to SCORE2 was assessed using Harrel’s C-index. Improvement in risk stratification following ESC guideline risk thresholds was assessed using categorical net reclassification. Population modelling was subsequently applied to estimate the impact on CVD prevention if applied at scale. Results Risk discrimination provided by SCORE2 (C-index: 0.718) was similarly improved by addition of metabolomic biomarker scores (ΔC-index: 0.011 [0.009–0.014]) and PRSs (ΔC-index 0.009; [0.007–0.012]). Addition of both metabolomic biomarker scores and PRSs to SCORE2 yielded the largest improvement risk discrimination, with ΔC-index 0.019 (0.016–0.022). Concomitant improvements in risk stratification were observed in categorical net reclassification index, with net case reclassification of 13.04% (11.67–14.41%). Modelling metabolic biomarker scores and PRSs for targeted risk-reclassification increased the number of CVD events prevented per 100,000 screened from 201 to 370 (ΔCVDprevented: 170 [158–182]) while essentially maintaining the number of statins prescribed per CVD event prevented. Conclusions Combining metabolomic biomarker scores and PRSs with SCORE2 moderately enhances prediction of first-onset CVD, and could have substantial population health benefit if applied at scale.

SCORE2-ASIA risk prediction algorithms: revised models to estimate 10-year risk of cardiovascular disease in the Asia Pacific Region

Abstract Background Most current cardiovascular disease (CVD) prediction tools have been developed and validated solely in Western countries, while there are significant differences between Asian and Western populations in risk factor distributions and CVD incidence patterns. This study outlines the recalibration and validation of the SCORE2-ASIA model to estimate 10-year fatal and non-fatal CVD risk among adults aged 40-69 years with no previous CVD or diabetes in the Asia Pacific Region. Methods The sex-specific and competing risk-adjusted SCORE2 models were systematically recalibrated to reflect CVD incidence in the Asian clinical practice. We defined four risk regions in Asia according to country-specific CVD mortality rates (low risk (<100 CVD deaths per 100,000), moderate risk (100 to <150 CVD deaths per 100,000), high risk (150 to <300 CVD deaths per 100,000), and very high risk (≥300 CVD deaths per 100,000)). Models were recalibrated using region representative data on expected incidences and risk factor distributions, utilising the methodological approaches developed in recalibration of the SCORE2 models for Europe. The SCORE-ASIA model was externally validated in all risk regions using population-based cohort data from independents cohort data. To ensure the European coefficients well applied to Asian populations, the SCORE2 model derivation process was repeated in external validation cohorts. Results Region-specific incidence was estimated using country-specific data on CVD mortality and data on the ratio of fatal to total events on 9,496,212 individuals from 5 Asian countries (496,212 CVD events). For external validation, we analysed data from 12 additional cohorts in 13 Asian countries (4,775,976 individuals, 176,011 CVD events). After applying the derived risk prediction models to external validation cohorts, the pooled external validation C-index was 0.709 (95%CI 0.676-0.743). In the nationally representative CDMD cohort from Singapore (population completely separate from recalibration data) there was adequate agreement between predicted and observed risks. Risk factor coefficients derived in Asian populations were largely similar to the original SCORE2 coefficients, indicating that the original SCORE2 coefficients apply well to the Asian population. Predicted CVD risk varied several-fold across Asian risk regions. For example, the estimated 10-year CVD risk for a 50-year-old non-smoker, with a systolic blood pressure of 140mmHg, total cholesterol of 5.5mmol/L, and HDL-cholesterol of 1.3mmol/L, ranged from 12% for men in low-risk countries to 26% for men in very-high-risk countries, and from 7% for women in low-risk countries to 26% for women in very-high-risk countries. Conclusions The SCORE2-ASIA model accurately predicts 10-year risk of CVD for apparently healthy people in Asian countries after systematic recalibration, thereby enhancing the identification of individuals at higher risk of developing CVD across the Asia Pacific Region.

Machine learning methods for diagnostic disease prediction and treatment decisions in parvovirus B19 positive viral cardiomyopathy

Abstract Background Parvovirus B19 (B19V) is the most common type of virus found in endomyocardial biopsies (EMB) from patients with inflammatory heart disease. However, the detection of genomic B19V-DNA appears to have little clinical significance and is often considered an incidental finding. Recent analyses of a large cohort of B19V-positive patients revealed that the detection of RNA transcripts rather than DNA is of clinical importance, as B19V-RNA-positive patients have a significantly worse cardiovascular outcome. Here, we applied machine learning (ML) methods to assess the impact of B19V transcriptional activity on cardiac inflammation and patient clinical outcome. Methods and results A total of 305 patients with a known clinical course and a positive endomyocardial B19V-DNA test were included in this study. Follow-up measurements of the B19V-specific transcripts VP1/2 and NS1 revealed significantly worse clinical courses than in patients without B19V-RNA detection (P=0.019, hazard ratio (95% CI)=1.62 (1.10-2.57)). LVEF at follow-up was significantly lower in patients with detected B19V-RNA (P=0.007), and adverse events such as worsening LV function (P=0.003) or stable systolic dysfunction (P=0.013) were more frequent in this group. Importantly, analysis of inflammatory markers in EMB showed no difference in patients with and without RNA detection (P=0.432), suggesting that the deterioration in clinical outcome was related to B19V transcriptional activity rather than an increase in inflammation. The analysis of area under the receiver operating characteristic curve (AUROC) of B19V-RNA-negative patients revealed that the macrophage marker MAC-1 (AUROC=0.66, P=0.002) was the single most prognostically relevant parameter, while the T cell marker CD3 (AUROC=0.73, P<0.001) was identified as the most prognostically relevant marker in B19V-RNA-positive patients. The linear combination of several inflammatory markers increased the AUROC value to 0.67 in B19V-RNA-negative patients and to 0.74 in B19V-RNA-positive patients. To further increase prognostic accuracy, we trained ML models on features derived from EMB inflammatory markers. For the prediction of major adverse cardiac events within 60 months, the best ML-models achieved an AUROC of 0.87 in B19V-RNA-positive and of 0.86 in B19V-RNA-negative patients, while models that neglect the differentiation of patients with regard to the presence of B19V RNA transcripts only achieved an AUROC of up to 0.76. Conclusions These results show that B19V transcriptional activity is an independent risk marker for adverse cardiac events in patients with viral cardiomyopathy and thus confirm the clinical-therapeutic importance of myocardial biopsy. Furthermore, we present an ML-based algorithm that accurately indicates disease progression in B19V-positive cardiomyopathy patients and is thus helpful for appropriate therapy recommendation.

Application of artificial intelligence to study the causality in public health: a systematic review

Abstract Introduction In epidemiology, establishing causation requires meeting rigorous criteria beyond mere association. Traditional methodologies often struggle to provide definitive causal conclusions due to confounding variables. However, leveraging artificial Intelligence (AI) and machine Learning (ML) offers a promising solution by modeling complex interactions among relevant factors. This systematic review explores the AI techniques utilized to uncover causal relationships in public health applications. Methods A literature search was conducted in Pubmed, Web of Science, and Scopus, employing the following search strategy: (Causalit*[Title/abstract] OR Causation*[Title/abstract] OR Causal[Title/abstract]) AND (‘Machine Learning’[Title/abstract] OR “Deep Learning” [Title/abstract] OR ‘Artificial Intelligence’ [Title/abstract] OR Algorithm* [Title/abstract]). The search aimed to identify studies encompassing causal inference techniques utilizing ML and AI methodologies in public health. Results From a total of 28,230 articles published up to August 2023, after removing duplicates (11,158) and articles that did not meet the inclusion criteria (16,168 post title/abstract screening, and 856 post full-text screening), the systematic review included 48 articles. Bayesian additive regression trees, Bayesian network models, causal forests analysis, and graphical causal models emerged as the prevailing causal methodologies. AI-driven causal methodologies have been investigated across various domains within public health, encompassing disease prediction, evaluation of treatment efficacy, identification of risk factors, analysis of health behaviors, precision Public Health initiatives, and environmental health assessment. Discussion While our systematic review affirms the potential of AI-driven causal approaches across several public health domains, further research is warranted before these emerging techniques can be effectively integrated into real settings. Key messages • In epidemiology, establishing causation requires meeting rigorous criteria beyond mere association and traditional methodologies often struggle to provide definitive causal conclusions. • This systematic review explores the AI techniques utilized to uncover causal relationships in public health applications.

Proteomic and genomic evaluation of plasma protein biomarkers for atrial fibrillation

Abstract Background Atrial fibrillation (AF) is a common arrhythmia and an important cause of ischemic stroke. Several plasma proteins have been suggested as biomarkers for the risk of AF and subsequent stroke in previous studies, but validation in independent cohorts is needed. Large-scale integration of proteomics with genetic and clinical data holds potential for refinement of disease prediction and risk stratification. Purpose This study aimed to evaluate previously suggested plasma protein biomarkers of AF and ischemic stroke using plasma protein measurements in a large cohort. Methods Ten tentative protein biomarkers previously associated with AF or ischemic stroke in AF patients, were evaluated in a nested case-control study comprising 49,757 individuals from the UK biobank (UKB). Plasma proteins were measured by normalized protein expression on the Olink platform. Outcomes were incident AF and ischemic stroke. The proteins were evaluated in Cox regression models adjusted for sex, age, and clinical risk factors. To assess whether the proteins were independent predictors, they were evaluated in sensitivity analyses with additional adjustment for NT-proBNP levels. Polygenic risk scores were calculated for plasma protein concentrations using the PRS-CS-auto algorithm and tested for association with AF or ischemic stroke in an independent subset of the UKB cohort without plasma protein measurements (n=379,392). Results All ten plasma proteins were associated with incident AF in the primary model, and eight were associated with incident ischemic stroke (P-values < 0.0025 for all). Adjusted for NT-proBNP levels, the proteins interleukin-6, GDF15, angiopoietin 2, and Troponin-I remained independently associated with AF, and the proteins GDF15, Cystatin-C, interleukin-6, ADAMTS13, and FGF23 remained independently associated with ischemic stroke. Polygenic risk scores for NT-proBNP and angiopoietin 2 were associated with increased rates of AF. No polygenic scores were significantly associated with ischemic stroke. Conclusion Large-scale proteomic evaluation validated ten plasma proteins associated AF and found eight to be associated with ischemic stroke. Seven proteins were associated with AF or ischemic stroke independently of NT-proBNP levels. Additional genomic evaluation provided evidence of associations between incident AF and genetically predicted plasma levels of angiopoietin 2 and NT-proBNP. As proteomic analyses of the plasma proteome become cheaper and more readily available, these findings could contribute to future risk stratification and selection of individuals for AF screening programs.Proteins associated with AFProteins associated with ischemic stroke

Forecasting the trend of tuberculosis incidence in Anhui Province based on machine learning optimization algorithm, 2013–2023

Tuberculosis has been one of the most common communicable diseases raising global concerns. Accurately predicting the incidence of Tuberculosis remains challenging. Here we constructed a time-series analysis and fusion tool using multi-source data, and aimed to more accurately predict the incidence trend of tuberculosis of Anhui Province from 2013 to 2023. Random forest algorithm (RF), Feature Recursive Elimination (RFE) and Least absolute shrinkage and selection operator (LASSO) were implemented to improve the derivation of features related to infectious diseases and feature work. Based on the characteristics of infectious disease data, a model of RF-RFE-LASSO integrated particle swarm optimization multiple inputs long short term memory recurrent neural network (RRL-PSO-MiLSTM) was created to perform more accurate prediction. Results showed that the PSO-MiLSTM achieved excellent prediction results compared with common single-input and multi-input time-series models (test set MSE:42.3555, MAE: 59.3333, RMSE: 146.7237, MAPE: 2.1133, R2: 0.8634). PSO-MiLSTM enriches and complements the methodological research content of calibrating the time-series predictive analysis of infectious diseases using multi-source data, and can be used as a brand-new benchmark for the analysis of influencing factors and trend prediction of infectious diseases at the public health level in the future, as well as providing a reference for incidence rate prediction of infectious diseases.

Bibliometric and Content Analysis of Articles on Artificial Intelligence in Healthcare

The use of artificial intelligence in the healthcare sector is becoming widespread for reasons such as analyzing digital patient data, including it in decision-making processes, improving the quality of healthcare services, and providing cost, time, and access advantages. This study aims to evaluate published articles on bibliometric indicators and the use of artificial intelligence in the healthcare sector and examine the content of the most cited articles. Articles about artificial intelligence in the health sector in the Web of Science database were included in the study using the criteria of “keyword, publication year, and publication language”. The research covers 2,680 articles published in English by 14,195 authors from 106 countries in 1084 journals between 2020-2024. 4,671 different keywords were used in the published articles. The country that published the most was “USA”, the journal was “Journal of Medical Internet Research”, the author was “Meng Ji”, and the most cited author was “Weihua Li”. The 55 most cited (≥50) articles focused on themes related to “diagnosis of COVID-19 disease”, “diagnosis of diseases”, “detection and classification of cancerous cells”, “identification of disease risk factors and disease prediction”, “prediction of treatment outcomes”, “prediction of disease course”, “personalized treatment recommendations”, “decision-making processes”, “ethical considerations, risks, and responsibilities”. With the COVID-19 pandemic, it is seen that the number of articles on artificial intelligence in the healthcare sector has increased exponentially. In the research, articles related to artificial intelligence in the health sector were examined, and a framework was created for researchers by revealing the most publishing countries, journals, authors, most cited authors, and keywords that were used the most.

Functional investigation and two-sample Mendelian randomization study of primary biliary cholangitis hub genes.

The identification of specific gene expression patterns is crucial for understanding the mechanisms underlying primary biliary cholangitis (PBC) and finding relevant biomarkers for diagnosis and therapeutic evaluation. To determine PBC-associated hub genes and assess their clinical utility for disease prediction. PBC expression data were obtained from the Gene Expression Omnibus database. Overlapping genes from differential expression analysis and weighted gene co-expression network analysis (WGCNA) were identified as key genes for PBC. Kyoto Encyclopedia of Genes and Genomes and Gene Ontology analyses were performed to explore the potential roles of key genes. Hub genes were identified in protein-protein interaction (PPI) networks using the Degree algorithm in Cytoscape software. The relationship between hub genes and immune cells was investigated. Finally, a Mendelian randomization study was conducted to determine the causal effects of hub genes on PBC. We identified 71 overlapping key genes using differential expression analysis and WGCNA. These genes were primarily enriched in pathways related to cytokine-cytokine receptor interaction, and Th1, Th2, and Th17 cell differentiation. We utilized Cytoscape software and identified five hub genes (CD247, IL10, CCL5, CCL3, and STAT3) in PPI networks. These hub genes showed a strong correlation with immune cell infiltration in PBC. However, inverse variance weighting analysis did not indicate the causal effects of hub genes on PBC risk. Hub genes can potentially serve as valuable biomarkers for PBC prediction and treatment, thereby offering significant clinical utility.

Advancement in disease prediction using machine learning

Advancement in disease prediction using machine learning