In order to investigate the value of preoperative X-ray computed tomography (CT) in predicting the pathological grade of pancreatic neuroendocrine tumors. This paper retrospectively analyzed the CT image examination of pancreatic neuroendocrine tumors, the image characteristics of G-NEC detected by CT image, and the detection of GST by spiral CT. In order to clearly diagnose and evaluate the size and scope of the focus, whether there is adjacent tissue invasion, metastasis, and treatment effect, CT, MR, PET-CT, nuclide specific imaging, and other imaging methods are widely used in the medical treatment of pNEN patients. These imaging methods have the advantages of noninvasive, rapid imaging, objective image medium, and strong repeatability. If the pathological grade of pNEN patients can be obtained by imaging examination before operation, it will be of great benefit to the formulation of treatment strategies and the prediction of clinical outcomes. Combining CT image performance with imaging omics characteristics to establish a prediction model that can develop a better auxiliary decision-making tool for clinical practice. Different pathological grades prompt clinicians to provide personalized and accurate medical treatment for patients, and reduce excessive medical treatment or wrong judgment caused by unclear preoperative diagnostic information.