Conventional treatment of rectal cancer is chemoradation followed by total mesorectal excision (TME) . However, 15-20% of the patients show no residual tumor cells in histological examinations. Local reoccurrence rate for these pathology complete responders (pCR) is 6-10%, meaning 10 to 16 patients need extensive mutilating surgery to prevent one local reoccurrence.¹ This questions the present clinical practice of surgical removal of the whole mesorectum in all patients, whether or not a pCR is suspected. The crucial step for safe omission of surgery is the selection of patients with a complete response. Currently, MRI is the most important diagnostic tool to define the regression status of the patient.² However, the prediction of pCR is not accurate enough to safely omit surgery in patients, who clinically show a complete response. Therefore, ultra-high field MRI will be developed for response assessment, which will need validation of the imaging with pathology. As a first step, high resolution MR images of ex-vivo recta are compared and validated using pathological examinations. The contrast in ex-vivo material decreases due to removal of blood flow and to tissue preservation using formaldehyde, which creates a bias in image validation with pathology when comparing in- and ex-vivo MR images. To translate MR contrast from ex vivo to in vivo, care must be taken to incorporate contrast alterations as transverse magnetization relaxation times (T2 and T2*) change due to altered tissue properties.[3-4] [A3] In this study we investigate the transverse relaxation effects at 7T during the fixation process in rectal specimens as a preparation for 7T MRI validation of rectal tumor regression with pathology.Pig recta (Vion Food Group NV, Netherlands, Boxtel) were cleaned from feces and positioned in a closeable PVC tube. To ensure full fixation of the tissue, the tube and rectal lumen are filled with a 4% formaldehyde solution using a syringe. The tube was positioned in a 32 channel high density coil array (MR Coils, the Netherlands, Drunen) in the center of a 7T MRI system (Philips Healthcare, Best). B0 shimming was applied and B1 mapping was obtained followed by quantitative T2 and T2* mapping at a spatial isotropic resolution of 0.5 mm. Other imaging parameters for T2 mapping were: TR, 2500ms; TE, 10ms; echo train, 6; field of view, 100x100 mm2, slice thickness; and for T2* mapping were: TR, 119ms; cTE, 2ms; echotrain, 16; flip-angle, 10˚; field of view, 100x100x20mm3; These sequences were repeated every two hours including recording of time since fixation. To obtain data points closest to surgery, also MRI scans were obtained prior to fixation. Here, Galden D05 (Solvey Plastisc, Netherlands, Klundert), a perfluoropolyether fluorinated fluid without hydrogen protons, was used to minimize B0 distortions. Processing of the images was done using Matlab 2014b (Mathworks, United States, Natick, MA). For both T2 and T2* calculations, regions of interest (ROI) were drawn on different tissue types, namely, the circular muscle (tissue 1), submucosa (tissue 2), and the mucosa (tissue 3). Signal intensity within a ROI was averaged in every echo before solving the exponential fit to decrease noise contribution.Using the high dense receiver arrays at 7T, images with good SNR could be obtained that allowed mapping of relaxation time at high spatial resolution (see fig 1b and 2b). ROI analyses of the relaxation parameters showed a T2* reduction in the circular muscle of 25% over 8 hours. As a consequence, the clear contrast between circular muscle and mucosa at the start of fixation is absent at the end. For T2, only a rapid change was observed in mucosa from pre-fixation to fixation while for the remaining tissues these remained constant over fixation time as is shown by the trend line during fixation.Quantification of T2 and T2* times over time allowed us to visualize the changes of these tissue properties due to fixation. In addition, the high spatial resolution enables delineation of multiple anatomical structures in the tissue-dense rectal specimen. Differentiation of different tissues will provide a more accurate comparison between pathology and in vivo images. Though high resolution imaging is inherent to very long scan times, in vivo images will benefit from the higher contrast with respect to the ex vivo images, favoring the future validation process.