The diffusion coefficient (D) and T2* reflect clinically significant tumor characteristics, including cellularity and hypoxia. Methods to non-invasively assess D and T2* at diagnosis and early follow up are desired. The aim of this work was to evaluate the sensitivity of D and T2* to radiotherapy effects in a neuroendocrine tumor model, and to evaluate clustering as an objective method to facilitate D and T2* analysis. 20 mice were examined using MRI before and repeatedly for 2 weeks after tumor irradiation (8Gy) . We show that D and T2* are potential response biomarkers, and that clustering improves response prediction for T2*.