Voxelwise spectral profile modeling for use in multispectral magnetic resonance imaging
AssigneeThe Medical College of Wisconsin, Inc.
InventorsKevin M. Koch, Suryanarayanan Kaushik
Life of this patent
Estimating the term…
Abstract
Described here are systems and methods for using a magnetic resonance imaging (“MRI”) system to estimate parameters of spectral profiles contained in multispectral data acquired using multispectral imaging (“MSI”) techniques, such as MAVRIC. These spectral profile parameters are reliably extracted using an iterative perturbation theory technique and utilized in a number of different applications, including fat suppression, artifact correction, and providing accelerated data acquisitions.
Claims19
A method for multispectral magnetic resonance imaging (MRI), the steps of the method comprising:
- (a) providing to a computer system, multispectral data acquired from a subject with an MRI system, wherein the multispectral data contain at least some multispectral data that are undersampled in a spectral domain by acquiring multispectral data from fewer spectral bins relative to fully spectrally sampled multispectral data;
- (b) providing to the computer system, multispectral calibration data;
- (c) generating a spectral model based at least in part on the multispectral calibration data;
- (d) estimating missing multispectral data for spectral bins not sampled in the provided multispectral data using the spectral model;
- and
- (e) reconstructing an image of the subject from the provided multispectral data and the estimated missing multispectral data.
A method for accelerating a multispectral magnetic resonance imaging (MRI) data acquisition, the steps of the method comprising:
- (a) providing to a computer system, multispectral calibration data acquired from a subject with an MRI system;
- (b) estimating spectral profile parameters at each voxel in the multispectral calibration data, wherein the spectral profile parameters are based on modeling signals at each voxel using a spectral profile model;
- (c) acquiring undersampled multispectral data from the subject with the MRI system using a multispectral imaging (MSI) acquisition by acquiring multispectral data from fewer spectral bins relative to fully spectrally sampled multispectral data;
- (d) estimating missing multispectral data for spectral bins not sampled in the acquired multispectral data using the estimated spectral profile parameters;
- and
- (e) reconstructing an image of the subject from the undersampled multispectral data and the estimated missing multispectral data.
A method for suppressing fat signals in multispectral magnetic resonance images, the steps of the method comprising:
- (a) providing to a computer system, multispectral data acquired from a subject with an MRI system using a multispectral imaging (MSI) acquisition;
- (b) estimating spectral profile parameters at each voxel in the multispectral data, wherein the spectral profile parameters are based on modeling signals at each voxel using a spectral profile model and include at least a width of the spectral profile;
- (c) estimating a chemical shift fraction associated with signal contributions from fat spins using the estimated width of the spectral profile;
- (d) reconstructing an image of the subject from the multispectral data;
- and
- (e) suppressing signal contributions from fat spins in the reconstructed image using the estimated chemical shift fraction.
A method for identifying regions of signal pile-up artifacts from multispectral magnetic resonance images, the steps of the method comprising:
- (a) providing to a computer system, multispectral data acquired from a subject with an MRI system using a multispectral imaging (MSI) acquisition;
- (b) estimating spectral profile parameters at each voxel in the multispectral data, wherein the spectral profile parameters are based on modeling signals at each voxel using a spectral profile model and include at least a width of the spectral profile;
- and
- (c) identifying voxels associated with pile-up intensity artifacts based on the width of the spectral profile estimated for each voxel.
Dependencies are inferred from the claim text.