To propose and validate an automated pipeline combining deep learning and radiomics for detecting non-calcified plaques in coronary arteries using non-contrast calcium score CT, highlighting the clinical significance of early detection.
Key Findings:
Radiomics models for predicting non-calcified plaques showed moderate to good vessel-level diagnostic performance, with AUCs ranging from 0.700 to 0.855, indicating varying effectiveness across different models.
Interpretation:
The automated pipeline enables efficient detection of non-calcified coronary plaques in CACS, with combined-region models showing promise for future use.
Limitations:
The study's generalizability may be limited due to its retrospective nature, reliance on data from two medical sites, and potential biases in patient selection.
Conclusion:
The approach may facilitate further research and support the clinical translation of chest CT for large-scale CAD screening, emphasizing its potential to improve patient outcomes.
