BMC Medical Imaging

Table 2 The diagnostic performance of each model in training and validation sets

From: Development of a multi-phase CT-based radiomics model to differentiate heterotopic pancreas from gastrointestinal stromal tumor

	Training set					Validation set
	AUC	ACC	SEN	SPE		AUC	ACC	SEN	SPE
Plain phase
0.5 mm	0.846 (0.743–0.949)	0.833	0.868	0.786		0.906 (0.777-1)	0.893	0.813	1.000
1 mm	1 (1–1)	1.000	1.000	1.000		0.789 (0.617–0.963)	0.786	0.789	0.778
2 mm	0.895 (0.818–0.972)	0.864	0.816	0.929		0.906 (0.791-1)	0.857	0.875	0.833
Arterial phase
0.5 mm	0.920 (0.851–0.989)	0.864	0.868	0.857		0.939 (0.839-1)	0.893	0.875	0.917
1 mm	0.972 (0.931-1)	0.939	0.921	0.964		0.901 (0.775-1)	0.857	0.875	0.833
2 mm	0.946 (0.884-1)	0.909	0.895	0.929		0.865 (0.729-1)	0.821	0.938	0.667
Venous phase
0.5 mm	0.959 (0.907-1)	0.924	0.895	0.964		0.854 (0.702-1)	0.821	0.813	0.833
1 mm	0.864 (0.775–0.952)	0.833	0.895	0.750		0.823 (0.648–0.998)	0.821	0.750	0.917
2 mm	0.943 (0.889–0.996)	0.894	0.947	0.821		0.901 (0.754-1)	0.893	0.875	0.917
Delayed phase
0.5 mm	0.919 (0.838-1)	0.879	0.868	0.893		0.807 (0.635–0.980)	0.821	0.813	0.833
1 mm	0.962 (0.898-1)	0.955	0.974	0.929		0.766 (0.587–0.944)	0.714	0.563	0.917
2 mm	0.848 (0.750–0.946)	0.803	0.811	0.793		0.743 (0.537–0.949)	0.786	0.941	0.545
Combined phase
0.5 mm	0.997 (0.991-1)	0.985	1.000	0.964		0.953 (0.881-1)	0.929	0.938	0.917
1 mm	0.996 (0.989-1)	0.970	0.947	1.000		0.880 (0.748-1)	0.857	0.938	0.750
2 mm	0.937 (0.875–0.999)	0.879	0.895	0.857		0.885 (0.741-1)	0.857	0.812	0.917

AUC, the area under the receiver operating characteristic curve; ACC, accuracy; SEN, sensitivity; SPE, specificity

Back to article page

ISSN: 1471-2342

Contact us

General enquiries: journalsubmissions@springernature.com