Pre-existing and machine learning-based models for cardiovascular risk prediction

Abstract

Predicting the risk of cardiovascular disease is the key to primary prevention. Machine learning has attracted attention in analyzing increasingly large, complex healthcare data. We assessed discrimination and calibration of pre-existing cardiovascular risk prediction models and developed machine learning-based prediction algorithms. This study included 222,998 Korean adults aged 40–79 years, naïve to lipid-lowering therapy, had no history of cardiovascular disease. Pre-existing models showed moderate to good discrimination in predicting future cardiovascular events (C-statistics 0.70–0.80). Pooled cohort equation (PCE) specifically showed C-statistics of 0.738. Among other machine learning models such as logistic regression, treebag, random forest, and adaboost, the neural network model showed the greatest C-statistic (0.751), which was significantly higher than that for PCE. It also showed improved agreement between the predicted risk and observed outcomes (Hosmer–Lemeshow χ2 = 86.1, P < 0.001) than PCE for whites did (Hosmer–Lemeshow χ2 = 171.1, P < 0.001). Similar improvements were observed for Framingham risk score, systematic coronary risk evaluation, and QRISK3. This study demonstrated that machine learning-based algorithms could improve performance in cardiovascular risk prediction over contemporary cardiovascular risk models in statin-naïve healthy Korean adults without cardiovascular disease. The model can be easily adopted for risk assessment and clinical decision making.

Striving towards excellence in research on biomarkers

“…Machine learning–based models including logistic regression, treebag, support vector machine, random forest, adaboost and the neural network model show high C-statistic values and have been used to produce prediction algorithms, especially in cardiovascular risk prediction. [ 24 ]…”

A Multimodal Approach to Discover Biomarkers for Taxane-Induced Peripheral Neuropathy (TIPN): A Study Protocol

“…We anticipate that not all biomarkers from each domain will be needed, but that a parsimonious set of biomarkers from multiple modalities may improve predictive accuracy. The overall result will be a predictive risk score for TIPN that can be used to classify patients as low, moderate, or high risk to guide clinical management, 104,106 based on extent of impacts on physical functioning and likelihood of treatment discontinuation.…”

History of Curriculums of the Department of Korean Traditional Music in Seoul National University

“…SNGAN+DGflow (Ansari et al, 2020) 9.62 AutoGAN (Gong et al, 2019) 12.4 TransGAN (Jiang et al, 2021) 9.26 StyleGAN2 w/o ADA (Karras et al, 2020) 8.32 StyleGAN2 w/ ADA (Karras et al, 2020) 2.92 DDGAN (T=1) (Xiao et al, 2021) 16.68 DDGAN (Xiao et al, 2021) 3.75 RGM (Choi et al, 2023b) 2.47 Diffusion NCSN (Song & Ermon, 2019) 25.3 DDPM (Ho et al, 2020) 3.21 Score SDE (VE) (Song et al, 2021b) 2.20 Score SDE (VP) (Song et al, 2021b) 2.41 DDIM (50 steps) (Song et al, 2021a) 4.67 CLD (Dockhorn et al, 2022) 2.25 Subspace Diffusion (Jing et al, 2022) 2.17 LSGM (Vahdat et al, 2021) 2.10 VAE&EBM NVAE 23.5 Glow (Kingma & Dhariwal, 2018) 48.9 PixelCNN (Van Oord et al, 2016) 65.9 VAEBM (Xiao et al, 2020) 12.2 Recovery EBM (Gao et al, 2021) 9.58…”