Magnetic resonance (MR) evaluation of intracranial hemorrhage is often challenging due to the variable appearance of hemorrhage, which depends on multiple factors. The aim of this study was to establish MR appearance of cerebral hemorrhages in dogs using low magnetic field and the efficacy of T2*-Gradient echo sequence for hemorrhage detection. Eight clinically normal beagle dogs, weighing approximately 9kg each were used. After a baseline MR examination, an intracranial hematoma was produced. MR examination was performed just after development of hemorrhage model and then at 1 to 2 day intervals for 30 days using low field MR (0.25 T). Sagittal images were acquired to select reproducible slice positions for transverse images. Sequences include spin echo (SE) T2, fluid attenuated inversion recovery (FLAIR), short tau inversion recovery, SE T1, and T2*-Gradient echo (GRE) images. The acquired MR images were compared subjectively to evaluate the signal changes. Signal-to-noise ratio was also measured and compared. Signal of the lesion was significantly hypo-intense in STIR and hyper-intense in T1W at day 3 after hemo-rrhage creation. The signal intensity of the hemorrhage gradually decreased in T1W images from day 3 to day 20. On T2W and FLAIR images, signal intensity was hyper-intense compared to normal and decreased gradually. No significant hypo-intense signal was seen on T2*-GRE image during examination of hemorrhage. This study shows that signal changes in intracranial hemorrhage do not follow the guidelines for hemorrhage interpretation in T1W and T2W images using low field magnets, except acute stage. T2*-GRE imaging maybe less useful in hemorrhage detection.

Key words: Canine cerebral hemorrhages, Dog, Gradient echo sequences, Low field, Magnetic resonance, Signal