Value of computed tomography imaging features and texture analysis in the differential diagnosis of acinar cell carcinoma of the pancreas

Abstract

Objective To explore the value of computed tomography (CT) imaging features and texture analysis in the differential diagnosis of acinar cell carcinoma of the pancreas (ACCP). Methods The clinical data of patients who were patholo-gically diagnosed with ACCP, pancreatic ductal adenocarcinoma (PDAC), or solid pseudopapillary neoplasm (SPN) of the pancreas in our hospital from January 2017 to October 2018 were retrospectively analyzed. The clinical manifestations, tumor markers, CT imaging features, and texture parameters were compared and analyzed. Results The contrast-enhanced CT images of 5 patients with ACCP showed that most of the lesions were solid-cystic masses with a large size and more solid components. The CT images of 12 patients with PDAC showed that most of the lesions were solid masses with a small size. The CT images of 12 patients with SPN showed that most of the lesions were solid-cystic masses with a size greater than that of ACCP lesions and fewer solid components, showing the “floating cloud sign”. Eight texture parameters were valuable for the differential diagnosis between ACCP and PDAC (Z=2.015-2.530,P<0.05). Four texture parameters were valuable for the differential diagnosis between ACCP and SPN (Z=2.108-2.864,P<0.05). Conclusion The diagnosis of ACCP is relatively difficult. The CT imaging features and texture analysis contribute to the diagnosis of ACCP and its differentiation with PDAC and SPN.

Keywords

• pancreas|carcinoma, acinar cell|tomography, x-ray computed|diagnosis, differential|image proces-sing, computer-assisted