Table 3. Diagnostic performance analysis of the four statistically significant features for the differentiation between PTs and FAs.

Features	AUC	Cutoff value	Sensitivity	Specificity	PPV	NPV
HWR	0.733	0.525	0.875	0.534	0.969	0.206
ML	0.610	6.000	0.750	0.445	0.928	0.157
E	0.758	0.035	0.725	0.662	0.946	0.228
GE	0.759	2.295	0.925	0.459	0.978	0.190

PTs, phyllodes tumors; FAs, fibroadenomas; HWR, height-to-width ratio; ML, margin lobulation; E, energy; GE, grey entropy; AUC, area under the curve; PPV, positive predictive value; NPV, negative predictive value.