. 2023 Nov 21;13:1219071. doi: 10.3389/fonc.2023.1219071

Table 2.

performance of the optimal Rss model and the optimal R_FF model for different molecular receptor statuses discrimination.

Classification tasks	Model	Training/Validation cohort					Test cohort
Classification tasks	Model	AUC	ACC	SEN	SPE	P value	AUC	ACC	SEN	SPE	P value
HR+ vs. HR-	Rss (DWI₆₀₀₎	0.767	0.736	0.770	0.653		0.768	0.693	0.689	0.706
HR+ vs. HR-	R_FF (DWI₆₀₀+DWI₈₀₀+DCE₅)	0.778	0.737	0.753	0.699	0.066	0.726	0.659	0.673	0.613	0.028
TNBC vs. HEBC	Rss (ADC)	0.769	0.656	0.667	0.650		0.718	0.692	0.684	0.700
TNBC vs. HEBC	R_FF(ADC+DCE₂+DCE₄)	0.787	0.692	0.655	0.720	0.043	0.773	0.645	0.636	0.650	0.017
TNBC vs. non-TNBC	Rss (ADC)	0.784	0.727	0.683	0.734		0.735	0.707	0.611	0.721
TNBC vs. non-TNBC	R_FF(ADC+DWI₆₀₀+T2WI+DCE₂)	0.818	0.718	0.705	0.721	0.042	0.773	0.767	0.636	0.780	0.025

P value: compared the performance between the optimal Rss model and the optimal R_FF model in the training/validation cohort and test cohort of each discriminative task. Significant values (P < 0.05) are presented in bold. HR, hormone receptor; HEBC: human epidermal growth factor receptor 2 enriched BC; TNBC: triple-negative breast cancer. AUC, area under the receiver-operating characteristic curve; SEN, sensitivity; SPE, specificity; ACC, accuracy.