order of scans randomized. The animals in each group of experiments were subsequently euthanized and the lesions confirmed by tissue histopathology. Results: Both lesion enhancement (SI post -SI pre /noise) and conspicuity post-contrast (SI lesion -SI brain /noise) approximately doubled when comparing 1.5 to 3 T using identical SE technique (8.0 AE 2.9 vs 16.2 AE 3.8 for lesion conspicuity, 1.5 vs 3 T, n ¼ 6, p ¼ 0.002). This improvement was substantially reduced when comparing lesion conspicuity for the short TE 2D GRE technique at 3 T to SE technique at 1.5 T, with both sequences optimized for the respective field strength (7.6 AE 2.5 vs 10.7 AE 3.0, 1.5 vs 3 T, p ¼ 0.06). However, when lesion enhancement was compared, a near-doubling was still observed (7.7 AE 1.9 vs 14.4 AE 2.2, 1.5 vs 3 T, p ¼ 0.001). These comparisons were all performed with acquisition time matched scans. 3D spoiled GRE provided an improvement of 20% over SE technique at 3 T for lesion conspicuity, in a subsequent trial with acquisition time not matched (5:10 vs 1:52 min:s, n ¼ 5). In a second similar trial, 3D MP-RAGE proved to be not significantly different when compared with SE technique at 3 T for lesion conspicuity, with acquisition time similarly not matched (6:03 vs 1:52 min:s, n ¼ 5). Conclusion: Using a standardized animal model with both matched and field strength optimized imaging techniques, this study shows a significant benefit of 3 T compared with 1.5 T for the visualization of contrast enhancement in brain tumor imaging. Normalized for scan time, in terms of lesion conspicuity post-contrast, the four evaluated scan techniques ranked at 3 T as follows: SE>3D spoiled GRE>MP-RAGE>2D spoiled GRE. Rationale: Microbeam radiation therapy is a form of radiosurgery first dedicated to the treatment of brain tumors. It uses arrays of synchrotrongenerated X-rays microbeams of very high doses (typically 625 Gy) [1,2]. Microbeams are typically few micrometers large (25 mm) and a few hundred micrometers spaced (200 mm). It has been shown that this particular irradiation geometry spares normal tissues surrounding the tumor, through a rapid repair of normal brain vasculature [3]. Previous experiments have shown that, despite a good tumor eradication rate (5/11), a 100 mm spacing unidirectional irradiation (skin dose 625 Gy, width 25 mm) was too invasive for normal tissue. On the contrary, a 200 mm spacing unidirectional irradiation preserved healthy tissue with a low tumor eradication rate (2/32) [4]. Objectives: The purpose of this study was to enhance the potential of the 200 mm spacing irradiation protocol by combining MRT with drug intratumoral injection. After diagnosis of the tumor by MRI, 9L tumorbearing rats were laterally irradiated with 51 microbeams (625 Gy, 25 mm, 200 mm) 14 days after implantation using the Tecomet multislit collimator installed at the ESRF biomedical beamline. Doses of 10 ml of gadolinium (Magnevist, Lab. Guerbet) were manually injected at the theoretical center of the tumor, 20 min before irradiation. Result...