Generalized likelihood ratio tests for change detection in diffusion tensor images

The automatic analysis of subtle changes between MRI scans is an important tool for monitoring disease evolution. Several methods have already been proposed to detect changes in serial conventional MRI but few works tackle this issue in the context of diffusion tensor imaging. Existing methods are limited to the detection of changes between scalar images characterizing the diffusion properties, such as Fractional Anisotropy or Mean Diffusivity. In this paper we introduce a new statistical test for detecting changes between tensor images. The test is based on a Gaussian diffusion model. Results on synthetic and real images demonstrate the ability of the test to bring useful, complementary information, with respect to scalar only clues.