Fig. 2

Radscore, calibration curves, and decision curves to predict non-response to neoadjuvant therapy. a Radscore for each patient in the training cohort (n = 362), internal testing cohort (n = 91), and external validation cohort (n = 110). b Calibration curves for Model_NoRes_NAT in the training cohort, internal testing cohort, and external validation cohort. The dotted reference line indicates perfect calibration. c Decision curves for Model_NoRes_NAT in the training cohort, internal testing cohort, and external validation cohort. The y-axis represents the net benefit. The x-axis represents the threshold probability. The red line represents Model_NoRes_NAT. The gray line represents the assumption that all patients showed non-response to neoadjuvant therapy. The black line represents the hypothesis that no patients showed non-response to neoadjuvant therapy. The threshold probability is where the expected benefit of treatment is equal to the expected benefit of avoiding treatment. Model_NoRes_NAT, pre-NAT multiparametric magnetic resonance imaging-based radiomics model