A comparison of 2D and 3D magnetic resonance imaging-based intratumoral and peritumoral radiomics models for the prognostic prediction of endometrial cancer: a pilot study

Yan, Ruixin; Qin, Siyuan; Xu, Jiajia; Zhao, Weili; Xin, Peijin; Xing, Xiaoying; Lang, Ning

doi:10.1186/s40644-024-00743-2

Table 2 Logistic regression with five-fold cross-validation for predictive performance of the models

From: A comparison of 2D and 3D magnetic resonance imaging-based intratumoral and peritumoral radiomics models for the prognostic prediction of endometrial cancer: a pilot study

Model		Group	Mean AUC (95%CI)	ACC	SEN	SPE
LVSI	2D_intra	Train	0.810 (0.718–0.903)	0.730	0.737	0.728
	2D_intra	Val	0.725 (0.505–0.937)	0.666	0.639	0.674
	2D_peri	Train	0.784 (0.695–0.874)	0.682	0.731	0.669
	2D_peri	Val	0.706 (0.490–0.920)	0.644	0.668	0.637
	2D_{intra + peri}	Train	0.853 (0.773–0.933)	0.730	0.789	0.713
	2D_{intra + peri}	Val	0.720 (0.493–0.944)	0.667	0.639	0.674
	3D_intra	Train	0.738 (0.634–0.844)	0.675	0.647	0.683
	3D_intra	Val	0.738 (0.534–0.933)	0.684	0.614	0.704
	3D_peri	Train	0.813 (0.782–0.898)	0.726	0.769	0.713
	3D_peri	Val	0.805 (0.625–0.977	0.695	0.771	0.674
	3D_{intra + peri}	Train	0.816 (0.730–0.901)	0.737	0.750	0.733
	3D_{intra + peri}	Val	0.812 (0.633–0.979)	0.737	0.743	0.726
DMI	2D_intra	Train	0.784 (0.699–0.871)	0.762	0.636	0.819
	2D_intra	Val	0.760 (0.582–0.938)	0.736	0.600	0.800
	2D_peri	Train	0.928 (0.889–0.972)	0.826	0.936	0.775
	2D_peri	Val	0.773 (0.614–0.932)	0.684	0.727	0.663
	2D_{intra + peri}	Train	0.839 (0.768–0.913)	0.783	0.741	0.803
	2D_{intra + peri}	Val	0.799 (0.638–0.963)	0.753	0.691	0.782
	3D_intra	Train	0.761 (0.679–0.847)	0.700	0.641	0.727
	3D_intra	Val	0.719 (0.538–0.904)	0.679	0.618	0.707
	3D_peri	Train	0.833 (0.761–0.909)	0.757	0.791	0.742
	3D_peri	Val	0.804 (0.654–0.953)	0.724	0.745	0.715
	3D_{intra + peri}	Train	0.845 (0.776–0.917)	0.779	0.782	0.777
	3D_{intra + peri}	Val	0.807 (0.653–0.959)	0.770	0.745	0.782
Stage	2D_intra	Train	0.773 (0.673–0.872)	0.703	0.703	0.703
	2D_intra	Val	0.704 (0.478–0.928)	0.625	0.701	0.683
	2D_peri	Train	0.760 (0.658–0.860)	0.716	0.699	0.703
	2D_peri	Val	0.706 (0.491–0.919)	0.704	0.701	0.701
	2D_{intra + peri}	Train	0.814 (0.719–0.909)	0.757	0.725	0.731
	2D_{intra + peri}	Val	0.687 (0.472–0.896)	0.575	0.693	0.667
	3D_intra	Train	0.782 (0.688–0.877)	0.663	0.732	0.717
	3D_intra	Val	0.738 (0.540–0.932)	0.621	0.694	0.678
	3D_peri	Train	0.824 (0.747–0.908)	0.730	0.748	0.744
	3D_peri	Val	0.763 (0.581–0.942)	0.704	0.715	0.713
	3D_{intra + peri}	Train	0.864 (0.795–0.933)	0.717	0.854	0.825
	3D_{intra + peri}	Val	0.804 (0.640–0.967)	0.700	0.833	0.805

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

Back to article page

ISSN: 1470-7330

Contact us

General enquiries: journalsubmissions@springernature.com

Cancer Imaging

Contact us