Adoption of multi-slice spiral computed tomography and multiparametric magnetic resonance imaging in the differential diagnosis of pancreatic cancer and chronic pancreatitis

This study aimed to assess the diagnostic value of combining computed tomography (CT) and magnetic resonance imaging (MRI) for differentiating pancreatic cancer (PCa) from chronic pancreatitis (CP). A total of 340 patients with pancreatic diseases were retrospectively included, comprising 173 cases of PCa and 167 cases of CP. The patients were randomly divided into a training set of 238 cases and a validation set of 102 cases, with the validation set comprising 52 cases of PCa (Group A) and 50 cases of CP (Group B). All patients underwent multi-slice spiral CT (MSCT) and multi-parametric MRI examinations. Imaging features and CT attenuation values were analyzed. Diagnostic sensitivity (Sen), specificity (Spe), and accuracy (Acc), and the area under the receiver operating characteristic curve (AUC) were compared among CT, MRI, and their combined application. The inter-group comparison indicated that pancreatic volume enlargement, invasion of surrounding tissues and major blood vessels, peripheral lymph node enlargement, inconspicuous arterial phase (AP) enhancement, and smooth dilation of the pancreatic duct (PD) and bile duct (BD) in Group A were markedly higher than those in Group B. Conversely, lesion calcification, fibrosis and pseudocyst formation, and irregular dilation of the PD and BD were less common in Group A than in Group B (<i>P</i><0.001 for pancreatic enlargement, surrounding tissue/vessel invasion, peripheral lymph node enlargement, inconspicuous AP enhancement, smooth ductal dilation, and lesion calcification; <i>P</i>=0.001 for fibrosis and pseudocyst formation; <i>P</i>=0.012 for irregular ductal dilation). The CT attenuation values of Group A were markedly lower than those of Group B during the AP, portal venous phase (PVP), and delayed phase (DP) (all <i>P</i><0.001), whereas the apparent diffusion coefficient (ADC) values were significantly higher (1.463 ± 0.098 vs. 1.158 ± 0.134, <i>P</i>=0.000). In the training set data, the efficacy of the CT+MRI diagnosis of PCa and CP (Sen=96.2%, Spe=92%, Acc=96%, AUC=0.972) was significantly better than that of CT alone (Sen=86.5%, etc.) and MRI alone (Sen=88.5%, etc.), with statistically significant differences in all compared indicators (<i>P</i><0.05 or <i>P</i><0.001). In the validation set, the efficacy of the CT+MRI diagnosis of PCa and CP (Sen=93.4%, Spe=89.7%, Acc=91.6%, AUC=0.951) was significantly better than that of CT alone (Sen=84.3%, etc.) and MRI alone (Sen=86%, etc.), with statistically significant differences in all compared indicators (all <i>P</i><0.05), and the efficacy trend was completely consistent with that of the training set. In summary, the combined application of CT and MRI provides superior Sen, Spe, and Acc for differentiating pancreatic cancer from chronic pancreatitis compared with either modality alone.