Skip to main content
. 2021 Mar 9;21:26. doi: 10.1186/s40644-021-00395-6

Table 2.

Univariate and multivariate logistic regression analysis of the radiomic signature and preoperative clinical parameters

Characteristics Univariate analysis Multivariate analysis
OR 95% CI P OR 95% CI P
Radiomics signature 3.616 1.977 ~ 6.612 <  0.001 * 3.321 1.646 ~ 6.701 0.001 *
Age 0.963 0.910 ~ 1.018 0.185
Symptom 2.023 0.821 ~ 4.985 0.122
Sex 0.814 0.302 ~ 2.193 0.683
MPD size 3.355 1.912 ~ 5.887 <  0.001 * 2.878 1.330 ~ 6.227 0.007*
Largest cyst size 2.525 1.019 ~ 6.261 0.042 * 0.413 0.088 ~ 1.929 0.261
Enhancing mural nodule 2.810 1.104 ~ 7.152 0.027 * 0.943 0.174 ~ 5.103 0.945
CA19–9 > 37kU/L 8.667 3.183 ~ 23.599 <  0.001 * 8.799 1.793 ~ 43.190 0.007*
CEA > 5 ng/mL 0.984 0.319 ~ 3.036 0.978

MPD Main pancreatic duct, CEA Carcinoembryonic antigen, CI Confidence internal

Significant parameters with P < 0.05 in the univariate analysis were included in the multivariate logistic regression analysis