EJNMMI Physics

Table 4 AUC statistics for the leave-1-out training for model selection

From: A novel intelligent radiomic analysis of perfusion SPECT/CT images to optimize pulmonary embolism diagnosis in COVID-19 patients

	Pulmonary Embolism			Pneumonia			Healthy Lung
Radiomic model	AUC	95% CI	P-value*	AUC	95% CI	P-value*	AUC	95% CI	P-value
ModelPrct3	0.919	0.909–0.929	*	0.922	0.914–0.931	*	0.971	0.966–0.977	0.09
ModelPrct5	0.902	0.892–0.913	0.028	0.905	0.896–0.914	< 0.001	0.978	0.973–0.983	*
ModelMx3	0.835	0.822–0.848	< 0.001	0.825	0.814–0.837	< 0.001	0.928	0.920–0.937	< 0.001
ModelMx5	0.823	0.810–0.836	< 0.001	0.818	0.806–0.830	< 0.001	0.935	0.928–0.943	< 0.001

Back to article page