Deep learning vs. CFD-based CT-FFR: similar diagnostic accuracy for coronary stenosis

DL and CFD-based CT-FFR show similar high diagnostic performance for detecting hemodynamically significant stenosis, with AUCs of 0.90 (95% CI 0.88-0.93) and 0.89 (95% CI 0.86-0.92).

• Both methods were strongly correlated with invasive FFR (DL: rho=0.71; CFD: rho=0.68; both p<0.05).
• In gray-zone lesions, correct classification rates were similar (DL 86.4% vs CFD 84.6%, p=0.690).
• Retrospective study of 220 patients (277 vessels) from two centers.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.