System, method and computer-accessible medium for diffusion MRI without shells

Inventors

COELHO, Santiago • Fieremans, Els • Novikov, Dmitry S.

Assignees

New York University

Interested in licensing this patent?

MTEC can help explore whether this patent might be available for licensing for your application.

Publication Number

US-12638533-B2

Patent

Publication Date

2026-05-26

Expiration Date


Abstract

Exemplary system, method and computer arrangement for determining rotational invariants, fiber orientations, and scalar parameters of fiber tracts (e.g., compartment fractions, which can relate to intra/extra-cellular space volumes; compartment diffusivities; relaxation rates; exchange rates between compartments; characteristics of structural disorder such as axonal beading) from a general diffusion MRI acquisition is described. For example, gradient directions may not necessarily be arranged in so-called shells, and an acquisition may vary spatially. Furthermore, each acquisition can be undersampled in the k-space. A procedure can also be included for receiving information related to the at least one image. Another procedure can be provided for decoupling tissue and protocol parameters based on a singular value decomposition. A further procedure can be provided for grouping singular vectors into multiplets based on symmetries. Still further procedures can be provided for forming rotational invariants and/or for a parameter estimation.

Core Innovation

The invention determines a plurality of tissue parameters of at least one tissue using at least one diffusion magnetic resonance image based on a plurality of scan parameters. A subset of imaging voxels that contains at least one voxel in at least one image is selected, and protocol-specific and tissue-specific components of a signal are generated in the selected subset of imaging voxels.

The invention generates at least one rotational invariant of the signal based on the at least one tissue-specific component of the signal in the at least one voxel of the selected subset of imaging voxels. The rotational invariant is generated by estimating the L2 norm across the index m for given q and n, and assigning signs to such norms.

In related aspects, a fiber orientation distribution is generated based on at least one of the components and is grouped into multiplets according to symmetries. Protocol-specific and tissue-specific components are used in conjunction with the multiplet grouping to generate the rotational invariants, and the resulting rotational invariant generation is used to determine tissue parameters.

Claims Coverage

The independent claims cover selecting a subset of imaging voxels from diffusion magnetic resonance image information, generating protocol-specific and tissue-specific signal components, and generating rotational invariants from tissue-specific components. Additional independent coverage specifies rotational invariant construction using an L2 norm across index m with sign assignment and grouping a fiber orientation distribution into multiplets according to symmetries. In total, the consolidated claim set presents five inventive features.

Voxel subset selection and signal component generation from dMR image

Receiving information related to at least one diffusion magnetic resonance image of the at least one tissue, based on a plurality of scan parameters, and selecting a subset of imaging voxels that contains at least one voxel in at least one image to generate protocol-specific and tissue-specific components of a signal in the selected subset of imaging voxels.

Undersampled diffusion directions when scan parameters are modified

Generating protocol-specific and tissue-specific components of a signal in the selected subset of imaging voxels, wherein, when at least one scan parameter is modified, diffusion directions are undersampled.

Rotational invariant generation based on tissue-specific signal components

Generating the at least one rotational invariant of the signal based on the at least one tissue-specific component of the signal in the at least one voxel of the selected subset of imaging voxels.

L2-norm-based rotational invariant construction with sign assignment

Generating the at least one rotational invariant of the tissue based on the at least one component by estimating the L2 norm across the index m for given q and n, and assigning signs to such norms.

Fiber orientation distribution grouped into multiplets according to symmetries

Generating a fiber orientation distribution based on at least one of the components and grouping it into multiplets according to symmetries.

Overall, the claims cover creating protocol-specific and tissue-specific signal components from a selected voxel subset and generating rotational invariants from the tissue-specific components. The inventive features further include L2 norm-based invariant construction with sign assignment and symmetry-based multiplet grouping of a fiber orientation distribution.

Stated Advantages

Enables determining a plurality of tissue parameters of at least one tissue from diffusion magnetic resonance image information.

Generates rotational invariants of the signal based on tissue-specific components.

Generates rotational invariants using L2 norm across index m with sign assignment.

Groups a fiber orientation distribution into multiplets according to symmetries.

Rotational invariants are generated in the presence of undersampling of diffusion directions when at least one scan parameter is modified.

Documented Applications

Determining a plurality of tissue parameters of at least one tissue using at least one diffusion magnetic resonance image based on a plurality of scan parameters.

The method is applied to at least one tissue including at least one of white matter, gray matter, muscle tissue, prostate tissue, or kidney tissue.

JOIN OUR MAILING LIST

Stay Connected with MTEC

Keep up with active and upcoming solicitations, MTEC news and other valuable information.