Abstract:
The diffusion tensor imaging (DTI) assumption that water diffusion follows a Gaussian diffusion model is not always satisfied. Violations of this model can lead to tensor estimates that give misleading anisotropy indices. An alternative approach is presented to characterize diffusion anisotropy that assumes no a priori knowledge of the underlying diffusion process. Specifically, Shannon’s entropy, measuring the information content in a voxel’s diffusion attenuation values (measured across the applied diffusion-sensitizing gradient directions) is demonstrated. This measure provides additional contrast between different tissue types, especially between cerebrospinal fluid and gray matter, as compared with FA.
Reference:
abstract N. Metwalli, S. LaConte, X. Hu. Entropy-based characterization of diffusion anisotropy. In Proceedings 15th Scientific Meeting, International Society for Magnetic Resonance in Medicine, Berlin, page 1518, 2007. [bibtex]
Bibtex Entry:
@inproceedings{Berlin1518,
Author = {Metwalli, N. and LaConte, S. and Hu, X.},
Title ={Entropy-based characterization of diffusion anisotropy},
BookTitle = {Proceedings 15th Scientific Meeting, International Society for Magnetic Resonance in Medicine, Berlin},
Pages = {1518},
Abstract = {The diffusion tensor imaging (DTI) assumption that water diffusion follows a Gaussian diffusion model is not always satisfied. Violations of this model can lead to tensor estimates that give misleading anisotropy indices. An alternative approach is presented to characterize diffusion anisotropy that assumes no a priori knowledge of the underlying diffusion process. Specifically, Shannon’s entropy, measuring the information content in a voxel’s diffusion attenuation values (measured across the applied diffusion-sensitizing gradient directions) is demonstrated. This measure provides additional contrast between different tissue types, especially between cerebrospinal fluid and gray matter, as compared with FA.},
Keywords = {Berlin1518},
Year = {2007} }