Angular resolution enhancement technique for diffusion-weighted imaging (DWI) using predicted diffusion gradient directions

• Published in: NeuroImage (Orlando, Fla.) Vol. 183; pp. 836 - 846
• Main Authors: Lee, Mun Bae, Kim, Yeon Hyang, Jahng, Geon-Ho, Kwon, Oh-In
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.12.2018; Elsevier Limited
• Subjects: Algorithms; Angular resolution enhanced diffusion imaging; Brain - physiology; Brain Mapping - methods; Brain research; Dietary fiber; Diffusion; Diffusion Magnetic Resonance Imaging - methods; Diffusion MRI; Fourier transforms; Humans; Image Processing, Computer-Assisted - methods; Interpolation; Local diffusion pattern map; Magnetic resonance imaging; Methods; Multi-shell acquisition; Neuroimaging; NMR; Nuclear magnetic resonance; Spin orientation distribution function; Multi-shell acquisition; Local diffusion pattern map; Diffusion MRI; Spin orientation distribution function; Angular resolution enhanced diffusion imaging
• ISSN: 1053-8119; 1095-9572
• DOI: 10.1016/j.neuroimage.2018.08.072

Anisotropic diffusion MRI techniques using single-shell or multi-shell acquisitions have been proposed as a means to overcome some limitations imposed by diffusion tensor imaging (DTI), especially in complex models of fibre orientation distribution in voxels. A long acquisition time for the angular resolution of diffusion MRI is a major obstacle to practical clinical implementations. In this paper, we propose a novel method to improve angular resolution of diffusion MRI acquisition using given diffusion gradient (DG) directions. First, we define a local diffusion pattern map of diffusion MR signals on a single shell in given DG directions. Using the local diffusion pattern map, we design a prediction scheme to determine the best DG direction to be synthesized within a nearest neighborhood DG directions group. Second, the local diffusion pattern map and the spherical distance on the shell are combined to determine a synthesized diffusion signal in the new DG direction. Using the synthesized and measured diffusion signals on a single sphere, we estimate a spin orientation distribution function (SDF) with human brain data. Although the proposed method is applied to SDF, a basic idea is to increase the angular resolution using the measured diffusion signals in various DG directions. The method can be applicable to different acquired multi-shell data or diffusion spectroscopic imaging (DSI) data. We validate the proposed method by comparing the recovered SDFs using the angular resolution enhanced diffusion signals with the recovered SDF using the measured diffusion data. The developed method provides an enhanced SDF resolution and improved multiple fiber structure by incorporating synthesized signals. The proposed method was also applied neurite orientation dispersion and density imaging (NODDI) using multi-shell acquisitions. •Novel method to improve angular resolution of DWI signals is proposed.•The method is based on local diffusion pattern maps of diffusion MR signals on a single and/or multi shells.•The method provides an enhanced SDF resolution and improved multiple fiber structures.•The proposed angular resolution improvement is achieved without increasing MR scan time.