Table 2.

Results of response prediction on the training and independent test sets using different clinical and QUS feature sets. Acc: Accuracy; Sen: Sensitivity; Spec: Specificity.

Feature set	Training set				Test set
	Acc	Sen	Spec	AUC	Acc	Sen	Spec	AUC
Clinical Features: Tumor Size, ER/PR, HER2, Age	77.0%	51.9%	83.8%	0.678	69.1%	40.0%	80.0%	0.6
Unsegmented Core: $M_C^{\mathit{MBF}}$, $SN{R}_C^{\mathit{EAC}}$, $SN{R}_C^{\mathit{SI}}$, $M_C^{\mathit{SI}}$	81.0%	81.5%	80.8%	0.853	74.5%	66.6%	77.5%	0.79
Unsegmented Core and Margin: $M_C^{\mathit{MBF}}$, $SN{R}_m^{\mathit{SI}}$, $SN{R}_C^{\mathit{EAC}}$, $M_C^{\mathit{SI}}$	81.0%	77.8%	81.8%	0.86	76.4%	66.6%	80.0%	0.76
Intra-Tumor Regions and Margin: $SN{R}_1^{\mathit{SI}}$, $SN{R}_m^{\mathit{ESD}}$, $M_{3-1}^{\mathit{EAC}}$, $M_1^{\mathit{MBF}}$	90.5%	88.9%	91.0%	0.974	85.4%	86.6%	85.0%	0.89

The best value in each column is underlined