The aim of this study is to quantify inter-observer variability in delineating pancreatic neuroendocrine neoplasm (PanNEN) on Computerized Tomography (CT) images and its impact on radiomic features (RF), subsequently to this determination, to use CT texture analysis to predict, histological characteristics of PanNEN on CT scans.

CT imaging is the most widely used modality for studying radiomic features due to its ability to assess tissue density, shape, texture and size before, during and after therapy. To the best of the investigator's knowledge, the impact of inter-observer delineation variability on the reliability of CT RF for PanNEN patients, including Hounsfield unit (HU) values-, shape-, and texture-based features, has not yet been assessed. One this has been determined, an additional evaluation will be conducted to correlate the morphologically observed images with their histopathological characteristics.

The ultimate potential objective of this research is to identify and predict characteristics of aggressiveness of PanNEN in CT scans.

Inclusion Criteria:

• > 18 years of age
• pancreatic neuroendocrine neoplasm with intervention and biopsy
• availability of pre-operatory CT scan with or without IV contrast agent- inInstitution from 2009-2017

Exclusion Criteria:

• pregnant women

Monocentric, retrospective, observational study. Subjects who fulfill the inclusion criteria will be randomly chosen from our Institutional data-base. Thirty patients will be used to evaluate inter-observer variability and forty patients will be used to evaluate histological characteristics on CT-scans with and without contrast agent).

Imaged based and clinical variables will be used to construct an overall status of the patient.