Table 4:

Phase classification performance of models on the C4KC-KiTS dataset: XGBoost, ResNet3D 18-layer (r3d_18), Mixed Convolution Network 18-layer (mc3_18), R(2+1)D 18-layer (r2plus1d_18), and TotalSegmentator (ts_phase). Models are evaluated using AUC, Sensitivity, Specificity, PPV, F1 Score, and Accuracy.

	AUC	Sensitivity	Specificity	PPV	F1-score	Accuracy	p-value
Non-contrast
XGBoost	0.994	0.981	0.996	0.990	0.985	0.981	—
r3d_18	0.992	0.981	0.973	0.929	0.954	0.981	NaN
mc3_18	0.989	0.971	0.940	0.852	0.908	0.971	1.000
r2plus1d_18	0.992	0.981	0.986	0.963	0.972	0.981	NaN
ts_phase	0.984	0.971	0.996	0.990	0.981	0.990	1.000
Arterial/Venous
XGBoost	0.994	0.961	0.974	0.975	0.968	0.961	—
r3d_18	0.961	0.876	0.878	0.884	0.880	0.876	<0.001
mc3_18	0.925	0.838	0.772	0.796	0.816	0.838	<0.001
r2plus1d_18	0.917	0.800	0.777	0.792	0.796	0.800	<0.001
ts_phase	0.620	0.614	0.626	0.635	0.624	0.620	<0.001
Delayed
XGBoost	0.991	0.956	0.974	0.915	0.935	0.956	—
r3d_18	0.926	0.670	0.917	0.701	0.685	0.670	<0.001
mc3_18	0.862	0.406	0.911	0.569	0.474	0.406	<0.001
r2plus1d_18	0.877	0.494	0.867	0.517	0.505	0.494	<0.001
ts_phase	0.469	0.197	0.741	0.180	0.188	0.620	<0.001

P-values indicate the significance of accuracy differences compared to XGBoost (p<0.001 considered significant).