Table 2.

Performance metrics of the deep learning model on multi-modal MRI for glioma grading and molecular subtyping

Tasks	Modality	Accuracy	Sensitivity	Specificity	PPV	NPV	F1-score
OGG/GBM	A + B	0.80 ± 0.07	0.76 ± 0.16	0.87 ± 0.16	0.91 ± 0.11	0.67 ± 0.05	0.78 ± 0.09
	A + C	0.83 ± 0.11	0.90 ± 0.12	0.73 ± 0.25	0.82 ± 0.15	0.82 ± 0.15	0.77 ± 0.17
	B + C	0.77 ± 0.07	0.80 ± 0.19	0.73 ± 0.25	0.84 ± 0.13	0.74 ± 0.14	0.71 ± 0.12
	A + B + C	0.89 ± 0.11	0.90 ± 0.12	0.87 ± 0.16	0.91 ± 0.11	0.82 ± 0.15	0.87 ± 0.11
LGG/HGG	A + B	0.77 ± 0.11	0.71 ± 0.19	0.87 ± 0.16	0.89 ± 0.14	0.64 ± 0.08	0.76 ± 0.12
	A + C	0.74 ± 0.14	0.76 ± 0.16	0.73 ± 0.25	0.81 ± 0.17	0.62 ± 0.10	0.72 ± 0.17
	B + C	0.83 ± 0.11	0.86 ± 0.20	0.80 ± 0.27	0.89 ± 0.14	0.79 ± 0.18	0.77 ± 0.17
	A + B + C	0.86 ± 0.09	0.91 ± 0.11	0.80 ± 0.27	0.89 ± 0.14	0.82 ± 0.15	0.81 ± 0.17
IDH1(+)/(−)	A + B	0.69 ± 0.11	0.75 ± 0.22	0.60 ± 0.25	0.75 ± 0.14	0.65 ± 0.18	0.60 ± 0.12
	A + C	0.80 ± 0.07	0.86 ± 0.12	0.73 ± 0.25	0.84 ± 0.14	0.75 ± 0.13	0.75 ± 0.14
	B + C	0.80 ± 0.07	0.81 ± 0.19	0.80 ± 0.27	0.84 ± 0.13	0.74 ± 0.14	0.74 ± 0.14
	A + B + C	0.89 ± 0.11	0.81 ± 0.19	1.00 ± 0.00	1.00 ± 0.00	0.80 ± 0.17	0.85 ± 0.14
ATRX(−)/(+)	A + B	0.51 ± 0.15	0.55 ± 0.19	0.47 ± 0.16	0.57 ± 0.14	0.45 ± 0.15	0.48 ± 0.14
	A + C	0.66 ± 0.15	0.65 ± 0.20	0.67 ± 0.30	0.72 ± 0.16	0.62 ± 0.22	0.60 ± 0.16
	B + C	0.63 ± 0.15	0.60 ± 0.12	0.67 ± 0.30	0.75 ± 0.21	0.54 ± 0.14	0.59 ± 0.16
	A + B + C	0.71 ± 0.13	0.70 ± 0.19	0.73 ± 0.25	0.81 ± 0.17	0.60 ± 0.10	0.67 ± 0.15

A T2 FLAIR, B T1WI + C, C QSM, OGG other grade glioma (grade II + III), GBM glioblastoma multiforme (grade IV), LGG low-grade glioma (grade II), HGG high-grade glioma (grade III + IV), IDH1(+) IDH1 mutation, IDH1(−) IDH1 wildtype, ATRX(−) ATRX expression loss, ATRX(+) ATRX retention, PPV positive predictive value, NPV negative predictive value

The best results are marked in bold