Table 4.

AUCs of Nomogram, radiomics Signature, ER status, and PR status for pCR prediction in Trarining and Test sets.

	Nomogram(95% CI)	Signature (95% CI)	ER status(95% CI)	PR status(95% CI)
Training set	0.888 (0.814–0.94)	0.835 (0.753–0.899)	0.658 (0.562–0.746)	0.693 (0.598–0.777)
Nomogram vs. other metrics		0.0956	<0.0001*	0.0015*
Signature vs. other metrics			0.0133#	0.075
ER status vs. PR status				0.06392
Test set	0.879 (0.752–0.955)	0.834 (0.699–0.926)	0.716 (0.567–0.837)	0.642 (0.491–0.775)
Nomogram vs. other metrics		0.1446	0.0384*	0.0066*
Signature vs. other metrics			0.2239	0.0366#
ER status vs. PR status				0.5265

p-value refers to Delong test for the differences of AUCs between different metrics in different cohorts,

^*p < 0.05, with significant differences for AUCs of Signature, ER status, and PR status compared with that of nomogram.

^#p < 0.05, with significant differences for AUCs of ER status and PR status compared with that of Signature.