Skip to main content
. 2022 Jul 11;12:937050. doi: 10.3389/fonc.2022.937050

Table 4.

Comparison of the performance of the models.

Comparison DeLong’s test* (p-value) in the testing cohort DeLong’s test* (p-value) training cohorts
BPGT vs. MPGT
Clinical model vs. radiomics model 0.404 0.103
Clinical model vs. radiomics nomogram 0.041 0.010
Radiomics model vs. radiomics nomogram 0.406 0.172
PA vs. MPGT
Clinical model vs. radiomics model 0.163 0.009
Clinical model vs. radiomics nomogram 0.048 0.013
Radiomics model vs. radiomics nomogram 0.434 0.100
WT vs. MPGT
Clinical model vs. radiomics model 0.359 0.046
Clinical model vs. radiomics nomogram 0.016 0.013
Radiomics model vs. radiomics nomogram 0.603 0.616
PA vs. WT
Clinical model vs. radiomics model 0.034 0.001
Clinical model vs. radiomics nomogram 0.006 0.000
Radiomics model vs. radiomics nomogram 0.395 0.519

p-value < 0.05 indicated a statistically significant difference. *Test for the comparison of the difference of AUC; Vs, versus; PA, pleomorphic adenomas; WT, Warthin tumor; BPGT, benign parotid gland tumor, MPGT, malignant parotid gland tumor.