System and method for magnetic resonance fingerprinting in the presence of inhomogeneous magnetic fields

1. A method for using a magnetic resonance imaging (MRI) system to estimate quantitative parameters of a subject in a presence of inhomogeneous magnetic fields, the steps of the method comprising:

(a) estimating optimized acquisition parameters that are optimized to direct the MRI system to generate a plurality of different signal evolutions that maximize discrimination between magnetic resonance signal patterns in a minimized number of repetition time (TR) periods as a function of at least one relaxation time parameter and magnetic field homogeneity;

(b) acquiring data with the MRI system by directing the MRI system to perform a plurality of pulse sequences using the optimized acquisition parameters, the acquired data representing the plurality of different signal evolutions;

(c) computing quantitative parameters of the subject by comparing the acquired data with a dictionary database comprising a plurality of different signal templates, wherein each signal template includes values associated with flip angle homogeneity.

2. The method as recited in claim 1 , wherein step (a) includes estimating the acquisition parameters by minimizing an objective function that simulates the acquisition parameters and computes a matrix that is based on estimated values of the acquisition parameters and the quantitative parameters to be estimated.

3. The method as recited in claim 2 , wherein the objective function minimized in step (a) includes a penalty term that maintains a signal-to-noise ratio (SNR) of the plurality of different signal evolutions above a minimum threshold value.

4. The method as recited in claim 2 , wherein step (a) includes selecting initial estimates of the acquisition parameters and forming the matrix based on the initial estimates.

5. The method as recited in claim 4 , wherein the initial estimates of the acquisition parameters are selected by randomly generating values for the acquisition parameters.

6. The method as recited in claim 2 , wherein the computed matrix comprises a first matrix that defines a dot product between a second matrix and a transpose of the second matrix, wherein the second matrix includes estimates of the acquisition parameters, and simulated values for the quantitative parameters.

7. The method as recited in claim 1 , wherein at least some of the plurality of pulse sequences sample k-space along a Cartesian trajectory.

8. The method as recited in claim 7 , wherein at least some of the plurality of pulse sequences are echo planar imaging (EPI) pulse sequences that sample k-space along a Cartesian trajectory.

9. The method as recited in claim 8 , wherein the EPI pulse sequences are segmented EPI pulse sequences that sample less than a full extent of k-space.

10. The method as recited in claim 8 , wherein the EPI pulse sequences are spin-echo EPI pulse sequences.

11. The method as recited in claim 1 , wherein step (c) includes reconstructing images from the acquired data and comparing the reconstructed images to the dictionary database.

12. The method as recited in claim 1 , wherein step (c) includes computing simultaneously with the quantitative parameters, a radio frequency coil transmission homogeneity profile by comparing the acquired data with the values associated with flip angle homogeneity in the dictionary database.

13. The method as recited in claim 1 , wherein the optimized acquisition parameters estimated in step (a) are symmetric with respect to time, such that in each pulse sequence performed in step (b) nuclear spin magnetization is at least partially rewound in that pulse sequence.