Multimodality MRI-based radiomics for aggressiveness prediction in papillary thyroid cancer

BMC Medical Imaging

Table 3 The results of independent test set data

Models	AUC	ACC	SEN	SPE	PPV	NPV
CE-T1WI	0.789 [0.622, 0.907]	0.778 [0.642,0.914]	0.824 [0.566,0.962]	0.737 [0.488,0.909]	0.737 [0.480,0.912]	0.824 [0.566,0.962]
T2WI	0.830 [0.668,0.934]	0.778 [0.642,0.914]	0.647 [0.383,0.858]	0.895 [0.669,0.987]	0.846 [0.546,0.981]	0.739 [0.516,0.898]
DWI	0.944 [0.813,0.993]	0.861 [0.748,0.974]	0.824 [0.566,0.962]	0.895 [0.669,0.987]	0.875 [0.605,0.986]	0.850 [0.621,0.968]
Combined	0.960 [0.836,0.997]	0.917 [0.827,1.000]	0.882 [0.636,0.985]	0.947 [0.740,0.999]	0.937 [0.686,0.999]	0.900 [0.683,0.988]

95% confidence intervals are included in brackets. AUC, ACC, SEN, SPE, PPV, NPV are abbreviations of area under curve, accuracy, sensitivity, specificity, accuracy, negative predictive value and positive predictive value, respectively

ISSN: 1471-2342