1 AN ARTIFICIAL NEURAL NETWORK TO PREDICT TIME OF REPLANNING FOR TOMOTHERAPY TREATMENTS N.MAFFEI, G.GUIDI, C.VECCHI, G.BALDAZZI Physics Department, University of Bologna, via Irnerio 40 40138 Bologna, Italy maffei.nicola87@gmail.com G. GUIDI, A.CIARMATORI, T. COSTI Medical Physics DepartŵeŶt,UŶiversity Hospital”PolicliŶico”, ModeŶa Italy , via del Pozzo 71 40121 Modena, Italy guidi.gabriele@policlinico.mo.it Abstract The study has analyzed and validated methods for Adaptive Radiation Therapy (RT), whose goal is the optimization of the daily radiation treatment based on the anatomical variations and patient dosimetry. Using IGRT (Image-Guided Radiation Therapy) techniques, was conducted an analysis of 51 patients treated by Tomotherapy, subdivided by three pathology: Head and Neck (H & N), Prostate Adenocarcinoma (ADK) and Lung Stereotactic Body Radiation Therapy (SBRT). Neural networks, developed and implemented in this work, allow to identify, within the statistical sample, the cases that reveal criticality not in line with the average trend of the patients. Getting information about the complex anatomy affected by deviations from the initial constraints and setting them from the time point of view, it becomes possible to plan the clinical activity and make the methods of warping usable for correction of the daily treatment delivery. Among the various anatomical regions analyzed, the use of Adaptive RT techniques proved to be especially useful for parts subject to temporal variations in the course of the therapy. In particular, parotid glands, rectum, bladder, and lungs were sensitive organs for the study. The work lays the foundation for research studies regarding deformation of organ through biomechanical approach to validate a method intended, presumably in the near future, the real clinical practice. Keywords: Adaptive Radiation Therapy, Neural Network, IGRT, Tomotherapy, Biomechanics.