Fig. 4.
Graphs show performance evaluation of models created using combinations of baseline participants characteristics for discriminating the prediction of revascularization. Receiver operating characteristics performed best for discriminating when adding FFRCT to risk factors (diabetes, hypertension, dyslipidemia, and smoking), symptoms (dyspnoea and chest pain) and stenosis (+/-70%), AUC 0.95 [0.91–0.98], risk factors alone with AUC 0.53 [0.48–0.59], Risk factors and symptoms, AUC 0.72 [0.64–0.79] and risk factors, symptoms and stenosis AUC 0.72[0.64–0.79], p < 0.001