Differential Diagnosis of Solitary Fibrous Tumor/Hemangiopericytoma and Angiomatous Meningioma Using Three-Dimensional Magnetic Resonance Imaging Texture Feature Model

Table 1

AUC, ACC, cut-off, sensitivity, and specificity of the four texture feature model.

Task	ACC	AUC	Cut-off	Sensitivity	Specificity
Train (T1WI)	0.820	0.885	0.598	76.1%	87.9%
Test (T1WI)	0.712	0.830	0.598	60.0%	82.8%
Train (T2WI)	0.842	0.918	0.678	73.1%	95.5%
Test (T2WI)	0.814	0.864	0.678	73.3%	89.7%
Train (contrasted-T1WI)	0.744	0.815	0.676	55.2%	93.9%
Test (contrasted-T1WI)	0.695	0.772	0.676	60.0%	79.3%
Train (combined)	0.887	0.959	0.318	92.5%	84.8%
Test (combined)	0.831	0.939	0.318	90.0%	75.9%