Better efficacy in differentiating WHO grade II from III oligodendrogliomas with machine-learning than radiologist’s reading from conventional T1 contrast-enhanced and fluid attenuated inversion recovery images

Table 2 Diagnostic performance of comparison of radiomics and human assessment

	Sensitivity	Specificity	AUC	ACC
Radiomics (T1 CE)	0.672	0.789	0.798 (95% CI: 0.699, 0.896)	0.735
Radiomics (FLAIR)	0.700	0.683	0.774 (95% CI: 0.671, 0.877)	0.689
Radiomics (T1 CE + FLAIR)	0.778	0.783	0.861 (95% CI: 0.783, 0.940)	0.781
Reader1	0.824	0.632	0.700 (95% CI: 0.519, 0.880)	0.722
Reader2	0.706	0.684	0.687 (95% CI: 0.507, 0.867)	0.694
Reader3	0.647	0.632	0.714 (95% CI 0.545–0.883)	0.667

ISSN: 1471-2377