Robustness of CT radiomic features against image discretization and interpolation in characterizing pancreatic neuroendocrine neoplasms

Pancreatic neuroendocrine neoplasms (panNEN) are relatively indolent in most cases and, despite still rare, their incidence [1] is rapidly growing. Surgical resection is the only potentially curative therapy [2] and, due to the lack of prediction tools [3,4], is widely applied with a likely overuse. Consequently, the assessment of image-based non-invasive biomarkers is becoming a relevant and topical field, including their potential in selecting patients that could avoid surgery [5]. PanNEN are usually well visible with contrast enhanced CT and show very high contouring agreement among observers [6], making radiomics a suitable approach to assess CT predictors.

Source: Physica Medica: European Journal of Medical Physics - July 12, 2020 Category: General Medicine Authors: Sara Loi, Martina Mori, Giulia Benedetti, Stefano Partelli, Sara Broggi, Giovanni Mauro Cattaneo, Diego Palumbo, Francesca Muffatti, Massimo Falconi, Francesco De Cobelli, Claudio Fiorino Tags: Original paper Source Type: research