Key result
An 8-feature machine learning model predicts midterm coronary plaque progression with an AUC of ~0.88.
Why the study?
Subclinical nonobstructive coronary atherosclerosis is often under-recognized but carries substantial risk, necessitating improved risk prediction methods.
Does a machine learning model integrating clinical and CCTA features predict midterm plaque progression in adults with subclinical nonobstructive coronary atherosclerosis?
Observational (n=664)
Does a machine learning model integrating clinical and CCTA features predict midterm plaque progression in adults with subclinical nonobstructive coronary atherosclerosis?
An explainable machine learning model integrating clinical and CCTA features accurately predicts midterm plaque progression in patients with subclinical nonobstructive coronary atherosclerosis.
Answers are grounded in this paper and cite their sources.
May aid progression risk stratification in subclinical CAS; leaves open clinical utility pending prospective validation.
BACKGROUND: Coronary atherosclerosis (CAS) is a leading cause of cardiovascular morbidity and mortality. Subclinical nonobstructive CAS, often under-recognized in practice, carries substantial risk. We developed a clinically interpretable machine learning model that integrated traditional risk factors with quantitative plaque features derived from coronary computed tomography angiography to predict midterm plaque progression. METHODS: We retrospectively studied 664 adults with subclinical nonobstructive CAS from a community screening program who underwent serial coronary computed tomography angiography (median follow-up: 56 months). Candidate variables were filtered using Spearman correlation and a bootstrap-enhanced least absolute shrinkage and selection operator. Seven algorithms were compared; the optimal model was explained using SHapley Additive exPlanations (SHAP; https://github.com/shap/shap). RESULTS: Participants had a mean age of 62 years; 53.6% were men. The random forest model performed best. The final 8-feature random forest achieved an area under the receiver operating characteristic curve of 0.875 (95% confidence interval, 0.827-0.920), with precision 0.863, accuracy 0.779, recall 0.726, F1 score 0.788, specificity 0.849, and negative predictive value 0.702. SHAP analysis identified fibrofatty plaque volume and coronary artery calcium category as the principal drivers of plaque progression. CONCLUSIONS: The proposed machine learning SHAP framework integrates imaging and clinical features to enable individualized risk stratification for subclinical nonobstructive CAS. Incorporating quantitative imaging biomarkers into longitudinal follow-up might facilitate the early identification of high-risk individuals and support the timely implementation of intensive lipid-lowering, anti-inflammatory therapy, and lifestyle modification, thereby maximizing plaque stabilization or regression and ultimately reducing the risk of future cardiovascular events.
No takes yet. Share an insight, caveat, or question.
Su et al. (2025) conducted an observational in Subclinical nonobstructive coronary atherosclerosis (n=664). 8-feature random forest machine learning model was evaluated on Prediction of midterm plaque progression (AUC 0.875, 95% CI 0.827-0.920). An 8-feature random forest machine learning model predicted midterm coronary plaque progression with an area under the receiver operating characteristic curve of 0.875.
Synapse has enriched 10 related papers. Consider them for comparative context: