Preoperative Cervical Lymph Node Metastasis Prediction in Papillary Thyroid Carcinoma: A Noninvasive Clinical Multimodal Radiomics (CMR) Nomogram Analysis

Table 3

Quantitative indices pertaining to the six models for data from two cohorts.

Model	Cohorts	AUC (95% CI)	ACC	SEN	SPE	PPV	NPV
MOGONET	Training	0.84 (0.76–0.93)	0.82	0.89	0.71	0.80	0.83
	Validation	0.84 (0.71–0.96)	0.77	0.76	0.94	0.88	0.74
Combined-radiomic	Training	0.82 (0.74–0.90)	0.80	0.85	0.74	0.77	0.83
	Validation	0.77 (0.61–0.94)	0.77	0.93	0.67	0.64	0.94
DWI-radiomic	Training	0.74 (0.56–0.91)	0.72	0.89	0.51	0.70	0.78
	Validation	0.74 (0.56–0.91)	0.74	0.91	0.53	0.71	0.82
CE-T1WC-radiomic	Training	0.79 (0.70–0.89)	0.77	0.74	0.80	0.83	0.70
	Validation	0.68 (0.50–0.85)	0.64	0.73	0.53	0.67	0.60
T2WI-radiomic	Training	0.71 (0.61–0.82)	0.69	0.57	0.85	0.83	0.60
	Validation	0.52 (0.33–0.71)	0.51	0.45	0.59	0.59	0.45
US-radiomic	Training	0.69 (0.59–0.80)	0.69	0.92	0.39	0.66	0.80
	Validation	0.66 (0.48–0.85)	0.59	0.91	0.18	0.59	0.60

AUC, area under the receiver operating characteristic curve; ACC, accuracy; SEN, sensitivity; SPE, specificity; PPV, positive predictive value; NPV, negative predictive value; CI, confidence interval.