BMC Medical Imaging

Table 2 ROC analysis by the four classifiers in testing set

From: Radiomics-based classification of hepatocellular carcinoma and hepatic haemangioma on precontrast magnetic resonance images

Decision Tree				Random Forest
Images	AUC	Sensitivity	Specificity	Images	AUC	Sensitivity	Specificity
In-phase	0.63	0.67	0.64	In-phase	0.74	0.83	0.62
Out-phase	0.73	0.83	0.68	Out-phase	0.78	0.79	0.77
T2WI	0.68	0.72	0.68	T2WI	0.76	0.90	0.68
DWI	0.63	0.69	0.62	DWI	0.73	0.81	0.66
Combined	0.76	0.80	0.69	Combined	0.86	0.82	0.786
K Nearest Neighbours				Logistic Regression
Images	AUC	Sensitivity	Specificity	Images	AUC	Sensitivity	Specificity
In-phase	0.66	0.73	0.57	In-phase	0.72	0.81	0.64
Out-phase	0.58	0.57	0.60	Out-phase	0.77	0.74	0.79
T2WI	0.63	0.69	0.56	T2WI	0.80	0.82	0.78
DWI	0.65	0.69	0.62	DWI	0.65	0.83	0.47
Combined	0.76	0.778	0.653	Combined	0.89	0.822	0.714

AUC, area under receiver-operating characteristic curve; T2WI, T2-weighted imaging; DWI, diffusion-weighted imaging

Back to article page

ISSN: 1471-2342

Contact us

General enquiries: journalsubmissions@springernature.com