Systems and methods for determining a resolution of a dictionary for a magnetic resonance fingerprinting reconstruction
The present disclosure is related to systems and methods for determining a resolution of a dictionary for a magnetic resonance fingerprinting (MRF) reconstruction. The method includes obtaining one or more tissue properties; determining an image resolution condition based on an image resolution of the MRF reconstruction; and determining, based on a fitness function and the image resolution condition, a target step size for each of the one or more tissue properties. The target step size for the each of the one more tissue properties is an optimal solution of the fitness function, which represents an imaging velocity of the MRF reconstruction. The target step size for the each of the one more tissue properties satisfies the image resolution condition.
1. A system for determining a resolution of a dictionary for a magnetic resonance fingerprinting (MRF) reconstruction, comprising:
at least one storage device including a set of instructions; and
at least one processor in communication with the at least one storage device, wherein when executing the set of instructions, the at least one processor is directed to cause the system to perform operations including:
obtaining one or more tissue properties;
determining an image resolution condition based on an image resolution of the MRF reconstruction; and
determining, based on a fitness function and the image resolution condition, a target step size for each of the one or more tissue properties, wherein the target step size for the each of the one more tissue properties is an optimal solution of the fitness function, which represents an imaging velocity of the MRF reconstruction, and the target step size for the each of the one more tissue properties satisfies the image resolution condition.
2. The system of claim 1 , wherein the determining, based on the fitness function and the image resolution condition, the target step size for the each of the one or more tissue properties includes:
generating a first data set including one or more candidate step sizes each of which corresponds to a tissue property of the one or more tissue properties;
determining a value of the fitness function based on the one or more candidate step sizes; and
determining the target step size for the each of the one or more tissue properties by performing a plurality of iterations on the first data set until the image resolution condition is satisfied.
3. The system of claim 2 , wherein the performing the plurality of iterations on the first data set includes:
selecting, from the first data set, a candidate step size combination;
transforming the candidate step size combination to obtain a second data set; and
updating the first data set based on the second data set.
4. The system of claim 3 , wherein the selecting, from the first data set, the candidate step size combination includes selecting the candidate step size combination based on an evaluation value of the candidate step size combination, wherein the evaluation value is associated with a candidate imaging velocity or a candidate image resolution contributed by the candidate step size combination.
5. The system of claim 3 , wherein the selecting, from the first data set, the candidate step size combination includes selecting the candidate step size combination based on an iteration count of iterations during which the candidate step size combination is within the first data set.
6. The system of claim 3 , wherein the transforming the candidate step size combination includes:
determining a target transformation process to transform the candidate step size combination based on a transformation probability; and
obtaining the second data set by performing the target transformation process on the candidate step size combination.
7. The system of claim 6 , wherein the transformation probability is determined based on an evaluation value of the candidate step size combination, wherein the evaluation value is associated with a candidate imaging velocity or a candidate image resolution contributed by the candidate step size combination.
8. The system of claim 6 , wherein the target transformation process is a crossover process, and the obtaining the second data set by performing the target transformation process on the candidate step size combination includes:
selecting, from the first data set, a second candidate step size combination based on a weighted evaluation value of the second candidate step size combination; and
obtaining the second data set by hybridizing the candidate step size combination and the second candidate step size combination.
9. The system of claim 3 , wherein the updating the first data set based on the second data set includes:
determining an evaluation value of the first data set;
determining a transformed evaluation value of the second data set; and
determining an updated first data set based on the evaluation value of the first data set and the transformed evaluation value of the second data set.
10. The system of claim 9 , wherein an evaluation value of the first data set or a transformed evaluation value of the second data set is determined based on the fitness function.
11. The system of claim 1 , wherein the fitness function includes a plurality of weights respectively corresponding to the one or more tissue properties.
12. The system of claim 1 , wherein the operations further includes determining a dictionary based on the target step size for the each of the one or more tissue properties.
13. A method for determining a resolution of a dictionary for a magnetic resonance fingerprinting (MRF) reconstruction, implemented on a computing device, the computing device including at least one processor, the method comprising:
obtaining one or more tissue properties;
determining an image resolution condition based on an image resolution of the MRF reconstruction; and
determining, based on a fitness function and the image resolution condition, a target step size for each of the one or more tissue properties, wherein the target step size for the each of the one more tissue properties is an optimal solution of the fitness function, which represents an imaging velocity of the MRF reconstruction, and the target step size for the each of the one more tissue properties satisfies the image resolution condition.
14. The method of claim 13 , wherein the determining, based on the fitness function and the image resolution condition, the target step size for the each of the one or more tissue properties includes:
generating a first data set including one or more candidate step sizes each of which corresponds to a tissue property of the one or more tissue properties;
determining a value of the fitness function based on the one or more candidate step sizes; and
determining the target step size for the each of the one or more tissue properties by performing a plurality of iterations on the first data set until the image resolution condition is satisfied.
15. The method of claim 14 , wherein the performing the plurality of iterations on the first data set includes:
selecting, from the first data set, a candidate step size combination;
transforming the candidate step size combination to obtain a second data set; and
updating the first data set based on the second data set.
16. The method of claim 15 , wherein the selecting, from the first data set, the candidate step size combination includes selecting the candidate step size combination based on an evaluation value of the candidate step size combination, wherein the evaluation value is associated with a candidate imaging velocity or a candidate image resolution contributed by the candidate step size combination.
17. The method of claim 15 , wherein the selecting, from the first data set, the candidate step size combination includes selecting the candidate step size combination based on an iteration count of iterations during which the candidate step size combination is within the first data set.
18. The method of claim 15 , wherein the transforming the candidate step size combination includes:
determining a target transformation process to transform the candidate step size combination based on a transformation probability; and
obtaining the second data set by performing the target transformation process on the candidate step size combination.
19. The method of claim 13 , wherein the fitness function includes a plurality of weights respectively corresponding to the one or more tissue properties.
20. A non-transitory computer readable medium, comprising executable instructions that, when executed by at least one processor, direct the at least one processor to perform a method for determining a resolution of a dictionary for a magnetic resonance fingerprinting (MRF) reconstruction, the method comprising:
obtaining one or more tissue properties;
determining an image resolution condition based on an image resolution of the MRF reconstruction; and
determining, based on a fitness function and the image resolution condition, a target step size for each of the one or more tissue properties, wherein the target step size for the each of the one more tissue properties is an optimal solution of the fitness function, which represents an imaging velocity of the MRF reconstruction, and the target step size for the each of the one more tissue properties satisfies the image resolution condition.