Table 2.

Sensitivity, specificity, and AUC along with the 95% CI of AUC for each classifier trained on the full set. Sensitivity and specificity presented are for the optimal operating point determined using a metric for cut-off value that minimizes $m={(1-\text{sensitivity})}^2+{(1-\text{specificity})}^2$. Because all lesions were referred for biopsy, the sensitivity and specificity of the data set were not calculated for clinical assessment.

Classifier	DCE	T2w	DWI	Feature fusion	Classifier fusion
AUC [95% CI]	0.84 [0.82, 0.87]	0.83 [0.80, 0.86]	0.69 [0.62, 0.75]	0.87 [0.84, 0.89]	0.87 [0.84, 0.89]
Sensitivity (%)	75.7	76.3	61.4	79.1	79.0
Specificity (%)	76.3	74.5	62.9	77.2	78.4