A method and device for optimizing the resolution of a magnetic resonance fingerprinting dictionary

Abstract

The present invention discloses a method and device for optimizing the resolution of a magnetic resonance fingerprinting dictionary. The method includes: establishing the constraint conditions and objective function of a genetic algorithm, wherein the constraint conditions are established according to a preset magnetic resonance fingerprinting accuracy, and the objective function is established according to the step sizes of various tissue parameters of magnetic resonance fingerprinting; establishing the fitness function of the genetic algorithm according to the objective function of the genetic algorithm; based on the constraint conditions, using the genetic algorithm to perform an optimization calculation on the fitness function to obtain the step sizes of the various tissue parameters of the objective function corresponding to the optimal fitness function. The present invention solves the technical problem in the prior art that the resolution of the MRF dictionary is not high due to the manual setting of the intervals of the tissue parameters of the imaging object.

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and particularly to a method, device, electronic device and storage medium for optimizing the resolution of a magnetic resonance fingerprinting dictionary. Background Art

[0002] In magnetic resonance fingerprinting, it is necessary to design and construct the size of the MRF (magnetic resonance fingerprint) dictionary for the tissue parameters T1 (longitudinal relaxation time), T2 (transverse relaxation time), ρ (proton spin density) of the imaging object, all possible values, and the actual scanning sequence parameters TR (repetition time), FA (flip angle) and the system parameter ΔB 0 (magnetic field intensity difference caused by the inhomogeneity of the main magnetic field), and use the Bloch equation and computational simulation to calculate the signal evolution curve under the combination of tissue parameters and sequence parameters to form the MRF dictionary. Generally, the MRF dictionary determines the range of physiological parameters T1, T2, ρ according to the attributes of imaging physics, and then sets the interval of parameters (i.e., dictionary resolution) according to the requirements of imaging accuracy and imaging speed.

[0003] Currently, when generating the MRF dictionary, the parameter interval is generally set according to human experience. For example, the value range of T1 is [700 - 4200 ms], the step size is 5 ms, and there are 700 values in total; the value range of T2 is [50 - 2100 ms], the step size is 2 ms, and there are 1025 values in total; the value range of ρ is [0.6, 0.9], the step size is 0.05, and there are 7 values in total; so the combination of 3 parameters has a total of 700×1025×7 = 5022500 possibilities, and the specific combination forms are as Figure 1 shown.

[0004] The current parameter interval setting method (dictionary resolution setting method) results in a very large size of the MRF dictionary, a large number of combinations, and low resolution. When performing magnetic resonance fingerprinting, it will seriously affect the matching speed of the MRF dictionary, and thus the speed of magnetic resonance fingerprinting is not high. Summary of the Invention

[0005] The purpose of the present invention is to overcome the above technical deficiencies, and provide a method, device, electronic device and storage medium for optimizing the resolution of a magnetic resonance fingerprinting dictionary, so as to solve the technical problem of low resolution of the MRF dictionary caused by manually setting the interval of tissue parameters of the imaging object in the prior art.

[0006] To achieve the above technical purpose, the present invention takes the following technical solutions:

[0007] In the first aspect, the present invention provides a method for optimizing the resolution of a magnetic resonance fingerprinting dictionary, including the following steps:

[0008] Establish the constraint conditions and objective function of the genetic algorithm, wherein the constraint conditions are established according to the preset magnetic resonance fingerprint imaging accuracy, and the objective function is established according to the step sizes of the respective tissue parameters of the magnetic resonance fingerprint imaging;

[0009] Establish the fitness function of the genetic algorithm according to the objective function of the genetic algorithm;

[0010] Based on the constraint conditions, use the genetic algorithm to perform optimization calculation on the fitness function, and obtain the step sizes of the respective tissue parameters of the objective function corresponding to the optimal fitness function.

[0011] In some embodiments, the respective tissue parameters of the magnetic resonance fingerprint imaging at least include the longitudinal relaxation time, the transverse relaxation time, and the proton spin density.

[0012] In some embodiments, the method for establishing the objective function is as follows:

[0013] Establish an imaging speed function for each tissue parameter, wherein the imaging speed function is a function representing the mapping relationship between each tissue parameter and its step size;

[0014] Establish the objective function according to the imaging speed function of each tissue parameter.

[0015] In some embodiments, the establishing the objective function according to the imaging speed function of each tissue parameter includes:

[0016] Obtain the preset weights of each tissue parameter;

[0017] Establish the objective function according to the preset weights of each tissue parameter and the imaging speed function of each tissue parameter.

[0018] In some embodiments, the based on the constraint conditions, using the genetic algorithm to perform optimization calculation on the fitness function, and obtaining the step sizes of the respective tissue parameters of the objective function corresponding to the optimal fitness function includes:

[0019] Step 1: Generate an initial population, the initial population is composed of a number of individuals, and the individual is a random step size of one tissue parameter or a set of random step sizes of at least two tissue parameters;

[0020] Step 2: Calculate the individual fitness of the initial population according to the fitness function;

[0021] Step 3: Perform genetic selection on the individuals of the initial population according to the individual fitness;

[0022] Step 4: Perform a crossover operation on the selected individuals to obtain two new individuals;

[0023] Step 5: Perform mutation operation on the selected individuals to obtain new individuals;

[0024] Step 6: After repeating Steps 2 to 5 for a preset number of iterations, obtain the optimal individuals that meet the objective function and constraint conditions, and use the step sizes of the respective tissue parameters of the optimal individuals as the step sizes of the respective tissue parameters of the objective function corresponding to the optimal fitness function.

[0025] In some embodiments, the roulette wheel selection method is used to perform genetic selection on the individuals of the initial population.

[0026] In some embodiments, the preset number of iterations is 100 to 600.

[0027] In a second aspect, the present invention further provides a magnetic resonance fingerprint imaging dictionary resolution optimization device, including:

[0028] An objective function establishment module, configured to establish the constraint conditions and objective function of the genetic algorithm, wherein the constraint conditions are established according to a preset magnetic resonance fingerprint imaging accuracy, and the objective function is established according to the step sizes of the respective tissue parameters of the magnetic resonance fingerprint imaging;

[0029] A fitness function establishment module, configured to establish the fitness function of the genetic algorithm according to the objective function of the genetic algorithm;

[0030] An optimization module, configured to perform optimization calculation on the fitness function by using the genetic algorithm based on the constraint conditions, and obtain the step sizes of the respective tissue parameters of the objective function corresponding to the optimal fitness function.

[0031] In a third aspect, the present invention further provides an electronic device, including: a processor and a memory;

[0032] The memory stores a computer-readable program executable by the processor;

[0033] When the processor executes the computer-readable program, the steps in the magnetic resonance fingerprint imaging dictionary resolution optimization method as described above are implemented.

[0034] In a fourth aspect, the present invention further provides a computer-readable storage medium, where the computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps in the magnetic resonance fingerprint imaging dictionary resolution optimization method as described above.

[0035] Compared with the prior art, the method, device, electronic device and storage medium for optimizing the dictionary resolution of magnetic resonance fingerprint imaging provided by the present invention first establish the constraint conditions and objective function of the genetic algorithm. Among them, the constraint conditions are established according to the preset magnetic resonance fingerprint imaging accuracy, which limit the minimum accuracy of magnetic resonance fingerprint imaging. The objective function is established according to the step sizes of various tissue parameters of magnetic resonance fingerprint imaging, which reflects the speed of magnetic resonance fingerprint imaging. Then, according to the objective function of the genetic algorithm, the fitness function of the genetic algorithm is determined. Finally, according to the constraint conditions, the genetic algorithm is used to perform optimization calculation on the fitness function. The objective function corresponding to the finally obtained optimal fitness function is the optimal step size of each tissue parameter. In the embodiment of the present invention, the genetic algorithm is used to optimize the step sizes (dictionary resolutions) of various tissue parameters. With the imaging accuracy as the constraint condition and the imaging speed as the fitness function, the set of final solutions will minimize the combination number of tissue parameters (dictionary size) under the condition of meeting the imaging accuracy, thereby optimizing the imaging speed and avoiding the problem of low MRF dictionary resolution caused by manually setting the step sizes of tissue parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is a schematic diagram of MRF dictionary generation in the prior art;

[0037] Figure 2 is a flowchart of an embodiment of the method for optimizing the dictionary resolution of magnetic resonance fingerprint imaging provided by the present invention;

[0038] Figure 3 is a flowchart of an embodiment of step S300 in the method for optimizing the dictionary resolution of magnetic resonance fingerprint imaging provided by the present invention;

[0039] Figure 4 is a schematic diagram of an embodiment of the device for optimizing the dictionary resolution of magnetic resonance fingerprint imaging provided by the present invention;

[0040] Figure 5 is a schematic diagram of the operating environment of an embodiment of the program for optimizing the dictionary resolution of magnetic resonance fingerprint imaging of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0041] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0042] Magnetic resonance fingerprinting (MRF) technology is a new method of quantitative magnetic resonance imaging (MRI) that can simultaneously obtain multiple tissue parameters in a single acquisition. It has the ability to repeatedly and quantitatively measure tissue parameters, which can be used for more objective tissue diagnosis, comparison of scan data collected at different locations and different time points, longitudinal tracking of individual patients, and development of imaging biomarkers. Magnetic resonance fingerprinting technology can be used to evaluate the brain, prostate, liver, heart, musculoskeletal imaging and perfusion measurements, as well as to measure microvascular characteristics through magnetic resonance vessel fingerprinting.

[0043] The MRF dictionary is a collection of signal evolutions that uses the Bloch equations to simulate all possible tissue properties that can be measured within the physiological range. The Bloch equations are as follows:

[0044]

[0045] Among them, the unknown is M. Given the tissue parameters T1 and T2 in advance, M, that is, the signal evolution curve, can be obtained by computer simulation.

[0046] The MRF dictionary is unique for each MRF sequence design and can be generated for multiple subjects at once, or unique for each scanned subject when subject-specific parameters (such as heart rate in cardiac imaging) are used to generate the dictionary. When the dictionary includes multiple parameters and has a wider measurement range and / or a smaller step size, the overall size of the dictionary is larger. In pattern matching, each voxel-based signal evolution generated during MRF acquisition is compared with the dictionary of known signal evolutions to find the closest match. The T1 and T2 values used to generate the dictionary entry are assigned to that voxel. This process is repeated for each voxel to generate a fully core-paired quantitative map for each slice. Techniques such as non-uniform fast Fourier transform are used to reconstruct the unacquired MRF raw data to generate an accurate tissue property map.

[0047] The method, device, equipment, or computer-readable storage medium for optimizing the resolution of the magnetic resonance fingerprint imaging dictionary according to the present invention can be used in a magnetic resonance imaging system (MRI system). The method, device, equipment, or computer-readable storage medium according to the present invention can either be integrated with the above system or be relatively independent.

[0048] In this embodiment, a method for optimizing the resolution of the magnetic resonance fingerprint imaging dictionary is provided, which can be executed by a magnetic resonance imaging system (MRI system), specifically by one or more processors of the system. Figure 2 is a flowchart of the method for optimizing the resolution of the magnetic resonance fingerprint imaging dictionary provided by the embodiment of the present invention. Please refer to Figure 2 , the method for optimizing the resolution of the magnetic resonance fingerprint imaging dictionary includes the following steps:

[0049] S100. Establish the constraint conditions and objective function of the genetic algorithm, where the constraint conditions are established according to the preset magnetic resonance fingerprint imaging accuracy, and the objective function is established according to the step sizes of the respective tissue parameters of the magnetic resonance fingerprint imaging;

[0050] S200. Establish the fitness function of the genetic algorithm according to the objective function of the genetic algorithm;

[0051] S300. Based on the constraint conditions, use the genetic algorithm to perform optimization calculation on the fitness function, and obtain the step sizes of the respective tissue parameters of the objective function corresponding to the optimal fitness function.

[0052] In this embodiment, first, the constraint conditions and objective function of the genetic algorithm are established. The constraint conditions are established according to the preset magnetic resonance fingerprint imaging accuracy, which limits the minimum accuracy of the magnetic resonance fingerprint imaging. The objective function is established according to the step sizes of the respective tissue parameters of the magnetic resonance fingerprint imaging, which reflects the imaging speed of the magnetic resonance fingerprint imaging. Then, according to the objective function of the genetic algorithm, the fitness function of the genetic algorithm is determined. Finally, based on the constraint conditions, the genetic algorithm is used to perform optimization calculation on the fitness function. The objective function corresponding to the finally obtained optimal fitness function is the optimal step size of each tissue parameter. Among them, the process of the optimization calculation is an iterative process of finding the optimal calculation result of the fitness function. In the embodiment of the present invention, the genetic algorithm is used to optimize the step sizes (dictionary resolutions) of the respective tissue parameters, with the imaging accuracy as the constraint condition and the imaging speed as the fitness function. The set of final solutions will minimize the number of combinations of tissue parameters (the size of the MRF dictionary) under the condition of meeting the imaging accuracy, thereby optimizing the imaging speed and avoiding the problem of low MRF dictionary resolution caused by manually setting the step sizes of the tissue parameters.

[0053] In some embodiments, in step S100, each tissue parameter of magnetic resonance fingerprinting imaging at least includes longitudinal relaxation time T1, transverse relaxation time T2, and proton spin density ρ. Since the generation of the MRF dictionary is related to the above three tissue parameters, any one of the above three parameters can be optimized to reduce the number of combinations of tissue parameters, or two or three of the above three parameters can be optimized simultaneously to minimize the influence of humans. Similarly, the objective function can set the step sizes of two of the tissue parameters as constants and set the other tissue parameter as an unknown. Of course, the step sizes of two or three of the tissue parameters can also be set as unknowns to solve for the step sizes of two or three tissue parameters. The specific selection method depends on the actual imaging requirements, and the embodiments of the present invention do not limit this. In some preferred embodiments, the step sizes of the three tissue parameters in the objective function are all unknowns. In other words, the embodiments of the present invention optimize the step sizes of the three tissue parameters simultaneously to minimize the influence of humans and maximize the resolution of the dictionary, thereby greatly improving the matching speed of the dictionary. Of course, it should be understood that in the embodiments of the present invention, each tissue parameter of magnetic resonance fingerprinting imaging is not limited to longitudinal relaxation time T1, transverse relaxation time T2, and proton spin density ρ. In other embodiments, the tissue parameters may also include other parameters, such as intensity difference, etc., and the embodiments of the present invention do not limit this.

[0054] In some embodiments, to ensure that the accuracy of magnetic resonance fingerprinting imaging is not affected after optimizing the resolution of the MRF dictionary, constraint conditions are set for the genetic algorithm. The constraint conditions are established based on the preset magnetic resonance fingerprinting imaging accuracy, where the preset magnetic resonance fingerprinting imaging accuracy depends on actual requirements and can be set by the user independently, and the embodiments of the present invention do not limit this. When setting the constraint conditions, in order to reflect the relationship between the step sizes of each tissue parameter and the imaging accuracy, an imaging accuracy function needs to be established first. The constraint condition is that the calculated value of the imaging accuracy function is greater than the preset magnetic resonance imaging accuracy. In a specific embodiment, the imaging accuracy equation is: P(x, y, z), which reflects the relationship between the step sizes of each tissue parameter and the imaging accuracy. Correspondingly, the constraint condition is: P(x, y, z) ≥ the preset magnetic resonance imaging accuracy. Only the step sizes of the tissue parameters that meet this constraint condition can be used as the final required step sizes of each tissue parameter.

[0055] In some embodiments, the objective function reflects the speed of magnetic resonance fingerprinting imaging. The objective of the embodiments of the present invention is to optimize the step sizes of various tissue parameters to maximize the speed of magnetic resonance fingerprinting imaging. Therefore, the objective function is a function related to the step sizes of various tissue parameters. In specific implementation, first, a function related to each tissue parameter and its step size is established. This function represents the speed of magnetic resonance fingerprinting imaging, specifically T1(x), T2(y), ρ(z), which is related to the combined magnitude of the tissue parameters T1, T2, ρ. Here, x, y, z are the step sizes of the tissue parameters T1, T2, ρ respectively. Then, an objective function is established based on these three functions. In one embodiment, the method for establishing the objective function is as follows:

[0056] Establish an imaging speed function for each tissue parameter, where the imaging speed function is a function representing the mapping relationship between each tissue parameter and its step size;

[0057] Establish an objective function based on the imaging speed functions of each tissue parameter.

[0058] In this embodiment, first, imaging speed functions T1(x), T2(y), ρ(z) of each tissue parameter are established. These three imaging speed functions all affect the final imaging speed, but the degrees of influence on the imaging speed function are different. Therefore, when establishing the objective function, the degrees of influence of these three imaging speed functions on the final imaging speed need to be considered. In one embodiment, establishing the objective function according to the imaging speed functions of each tissue parameter includes:

[0059] Obtain the preset weights of each tissue parameter;

[0060] Establish the objective function according to the preset weights of each tissue parameter and the imaging speed functions of each tissue parameter.

[0061] In this embodiment, the weights of each tissue parameter are assigned according to the degrees of influence of the three imaging speed functions, and a preset weight is given to the three tissue parameters. By combining the preset weights with each imaging speed function, it can be ensured that the final objective function has the highest correlation with the imaging speed, and can most reflect the speed of magnetic resonance fingerprinting imaging. Thus, after optimizing the step sizes of each tissue parameter, the obtained step sizes are values that can maximize the speed of magnetic resonance fingerprinting imaging.

[0062] In one embodiment, the objective function is:

[0063] argmax(f(x,y,z)),

[0064] where f(x,y,z) = T1(x,w 1 )*T2(y,w 2 )*ρ(z,w3 )。

[0065] In this embodiment, the objective of the objective function is to maximize the imaging speed, and w 1 , w 2 , w 3 respectively represent the preset weights of the tissue parameters longitudinal relaxation time T1, transverse relaxation time T2, and proton spin density ρ. Among them, w 1 +w 2 +w 3 = 100%. T1(x, w 1 ), T2(y, w 2 ), ρ(z, w 3 ) respectively represent the imaging speed functions of each tissue parameter after adding the preset weights. When the step sizes x, y, z of the tissue parameters T1, T2, ρ obtained after optimization are substituted into the f(x, y, z), and the value of f(x, y, z) is maximized, then the f(x, y, z) obtained at this time is the required objective function, which can maximize the imaging speed.

[0066] In some embodiments, in step S200, after obtaining the objective function of the genetic algorithm, the fitness function of the genetic algorithm can be established. Specifically, the fitness function is f(x, y, z) = T1(x, w 1 ) * T2(y, w 2 ) * ρ(z, w 3 ). By performing optimization calculations on the fitness function, the step sizes of each tissue parameter corresponding to the optimal fitness function can be obtained.

[0067] Step S300 is to optimize the fitness function to obtain the optimal step sizes of each tissue parameter that satisfy the constraint conditions. In some embodiments, please refer to Figure 3 , step S300 specifically includes:

[0068] S310. Generate an initial population. The initial population is composed of several individuals, and an individual is a random step size of one tissue parameter or a set of random step sizes of at least two tissue parameters;

[0069] S320. Calculate the individual fitness of the initial population according to the fitness function;

[0070] S330. Perform genetic selection on the individuals of the initial population according to the individual fitness;

[0071] S340. Perform a crossover operation on the selected individuals to obtain two new individuals;

[0072] S350. Perform a mutation operation on the selected individuals to obtain new individuals;

[0073] After repeating S320 to S350 for a preset number of iterations in S360, an optimal individual that satisfies the objective function and constraint conditions is obtained, and the step sizes of the respective tissue parameters of the optimal individual are used as the step sizes of the respective tissue parameters of the objective function corresponding to the optimal fitness function.

[0074] In this embodiment, a genetic algorithm is used to optimize the fitness function. The genetic algorithm (GA) was first proposed by John Holland in the United States in the 1970s. This algorithm is designed based on the evolutionary laws of organisms in nature. It is a computational model that simulates the natural selection and genetic mechanism of Darwin's theory of biological evolution, and is a method for searching for the optimal solution by simulating the natural evolution process. This algorithm uses mathematical methods and computer simulation operations to convert the problem-solving process into processes similar to chromosome gene crossover and mutation in biological evolution. When solving relatively complex combinatorial optimization problems, compared with some conventional optimization algorithms, it can usually obtain better optimization results more quickly. The genetic algorithm has been widely applied in fields such as combinatorial optimization, machine learning, signal processing, adaptive control, and artificial life.

[0075] Among them, in step S310, the individual can be the random step size of one of the tissue parameters, such as the step size of the longitudinal relaxation time T1, or a set of random step sizes of two or more tissue parameters. For example, an individual includes the step size of the longitudinal relaxation time T1, the step size of the transverse relaxation time T2, and the step size of the proton spin density ρ. The embodiments of the present invention do not limit this. Preferably, in this embodiment, the individual is a set of the step sizes of the longitudinal relaxation time T1, the transverse relaxation time T2, and the proton spin density ρ.

[0076] Step S320 is to calculate the individual fitness to facilitate the screening of individuals. In specific implementation, the parameters in the individual can be directly substituted into the fitness function to obtain the individual fitness of each individual.

[0077] Step S330 is for genetic selection of individuals, weeding out the unfit and selecting the fit. Individuals with higher fitness have a greater probability of being passed on to the next generation, while those with lower fitness have a smaller probability. In some embodiments, in step S330, the roulette wheel selection method is used to perform genetic selection on the individuals in the initial population. The roulette wheel selection method is the simplest and most commonly used selection method. In this method, the selection probability of each individual is proportional to its fitness value. The greater the fitness, the greater the selection probability. Therefore, after obtaining the fitness of each individual through step S320, through the roulette wheel selection method, genetic selection of individuals can be performed based on individual fitness, and individuals with higher fitness have a higher probability of being selected. The purpose of selection is to directly pass on the optimized individuals to the next generation or generate new individuals through pairing and crossover and then pass them on to the next generation.

[0078] Steps S340 and S350 are for performing crossover and mutation operations on the selected individuals. The purpose of crossover is to generate new individuals by replacing and recombining parts of the structures of two parent individuals. Through crossover, the search ability of the genetic algorithm is improved by leaps and bounds. Mutation is to change the gene values at certain gene loci of the individual strings in the population. Its purpose is to endow the genetic algorithm with local random search ability. When the genetic algorithm is close to the neighborhood of the optimal solution through the crossover operator, the use of this local random search ability of the mutation operator can accelerate convergence to the optimal solution, and it can also keep the population diversity of the genetic algorithm to prevent premature convergence. At this time, the convergence probability should take a larger value.

[0079] In some embodiments, in step S360, the preset number of iterations is 100 - 600. In this embodiment, the preset number of iterations is 500. Under the condition of ensuring sufficient iterative calculation, it can also obtain better optimization results quickly and speed up the calculation process.

[0080] It should be understood that although Figures 2 to 3 the steps in the flowchart are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders.

[0081] Based on the above magnetic resonance fingerprint imaging dictionary resolution optimization method, the embodiments of the present invention also correspondingly provide a magnetic resonance fingerprint imaging dictionary resolution optimization device 400. Please refer to Figure 4 . The magnetic resonance fingerprint imaging dictionary resolution optimization device 400 includes an objective function establishment module 410, a fitness function establishment module 420, and an optimization module 430.

[0082] The objective function establishment module 410 is used to establish the constraint conditions and objective function of the genetic algorithm. Among them, the constraint conditions are established according to the preset magnetic resonance fingerprint imaging accuracy, and the objective function is established according to the step sizes of the respective tissue parameters of the magnetic resonance fingerprint imaging.

[0083] The fitness function establishment module 420 is used to establish the fitness function of the genetic algorithm according to the objective function of the genetic algorithm.

[0084] The optimization module 430 is used to perform optimization calculations on the fitness function by using the genetic algorithm based on the constraint conditions, and obtain the step sizes of the respective tissue parameters of the objective function corresponding to the optimal fitness function.

[0085] In this embodiment, first, the constraint conditions and objective function of the genetic algorithm are established. Among them, the constraint conditions are established according to the preset magnetic resonance fingerprint imaging accuracy, which limits the minimum accuracy of the magnetic resonance fingerprint imaging. The objective function is established according to the step sizes of the respective tissue parameters of the magnetic resonance fingerprint imaging, which reflects the imaging speed of the magnetic resonance fingerprint imaging. Then, according to the objective function of the genetic algorithm, the fitness function of the genetic algorithm is determined. Finally, according to the constraint conditions, the genetic algorithm is used to perform optimization calculations on the fitness function. The objective function corresponding to the finally obtained optimal fitness function is the optimal step size of the respective tissue parameters. In the embodiment of the present invention, the genetic algorithm is used to optimize the step sizes (dictionary resolutions) of the respective tissue parameters, with the imaging accuracy as the constraint condition and the imaging speed as the fitness function. The set of final solutions will minimize the number of combinations of tissue parameters (the size of the MRF dictionary) under the condition of satisfying the imaging accuracy, thereby optimizing the imaging speed and avoiding the problem of low MRF dictionary resolution caused by manually setting the step sizes of the tissue parameters.

[0086] In some embodiments, the respective tissue parameters of the magnetic resonance fingerprint imaging at least include the longitudinal relaxation time, the transverse relaxation time, and the proton spin density.

[0087] In some embodiments, the objective function establishment module 410 is specifically used for:

[0088] Establish an imaging speed function for each tissue parameter, where the imaging speed function is a function representing the mapping relationship between each tissue parameter and its step size;

[0089] Establish an objective function according to the imaging speed function of each tissue parameter.

[0090] In some embodiments, the objective function establishment module 410 is further used for:

[0091] Obtain the preset weights of each tissue parameter;

[0092] Establish an objective function according to the preset weights of each tissue parameter and the imaging speed function of each tissue parameter.

[0093] In some embodiments, the optimization module 430 is specifically configured to implement the following steps:

[0094] Step 1: Generate an initial population. The initial population is composed of a number of individuals, and an individual is a random step size of an organizational parameter or a set of random step sizes of at least two organizational parameters;

[0095] Step 2: Calculate the individual fitness of the initial population according to the fitness function;

[0096] Step 3: Perform genetic selection on the individuals of the initial population according to the individual fitness;

[0097] Step 4: Perform a crossover operation on the selected individuals to obtain two new individuals;

[0098] Step 5: Perform a mutation operation on the selected individuals to obtain new individuals;

[0099] Step 6: After repeating steps 2 to 5 for a preset number of iterations, obtain the optimal individual that satisfies the objective function and the constraint conditions, and use the step sizes of the respective organizational parameters of the optimal individual as the step sizes of the respective organizational parameters of the objective function corresponding to the optimal fitness function.

[0100] In some embodiments, the roulette wheel selection method is used to perform genetic selection on the individuals of the initial population.

[0101] In some embodiments, the preset number of iterations is 100 - 600.

[0102] As Figure 5 shown, based on the above-mentioned magnetic resonance fingerprint imaging dictionary resolution optimization method, the present invention also correspondingly provides an electronic device, which may be a computing device such as a mobile terminal, a desktop computer, a notebook, a palm computer, and a server. The electronic device includes a processor 10, a memory 20, and a display 30. Figure 5 Only some components of the electronic device are shown, but it should be understood that it is not required to implement all the shown components, and more or fewer components may be alternatively implemented.

[0103] The memory 20 may be an internal storage unit of the electronic device in some embodiments, such as a hard disk or memory of the electronic device. The memory 20 may also be an external storage device of the electronic device in other embodiments, such as a plug-in hard disk equipped on the electronic device, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. Further, the memory 20 may also include both an internal storage unit and an external storage device of the electronic device. The memory 20 is used to store application software installed on the electronic device and various types of data, such as program codes installed on the electronic device. The memory 20 may also be used to temporarily store data that has been output or will be output. In one embodiment, a magnetic resonance fingerprint imaging dictionary resolution optimization program 40 is stored on the memory 20, and the magnetic resonance fingerprint imaging dictionary resolution optimization program 40 can be executed by the processor 10, so as to implement the magnetic resonance fingerprint imaging dictionary resolution optimization method of various embodiments of the present invention.

[0104] The processor 10 may be a central processing unit (CPU), a microprocessor or other data processing chips in some embodiments, and is used to run program codes stored in the memory 20 or process data, such as executing the magnetic resonance fingerprint imaging dictionary resolution optimization method, etc.

[0105] The display 30 may be an LED display, a liquid crystal display, a touch liquid crystal display, and an OLED (Organic Light-Emitting Diode) toucher, etc. in some embodiments. The display 30 is used to display information in the magnetic resonance fingerprint imaging dictionary resolution optimization device and is used to display a visual user interface. Components 10-30 of the electronic device communicate with each other through a system bus.

[0106] In some embodiments, when the processor 10 executes the magnetic resonance fingerprint imaging dictionary resolution optimization program 40 in the memory 20, the following steps are implemented:

[0107] Establish the constraint conditions and objective function of the genetic algorithm, wherein the constraint conditions are established according to a preset magnetic resonance fingerprint imaging accuracy, and the objective function is established according to the step sizes of various tissue parameters of the magnetic resonance fingerprint imaging;

[0108] Establish a fitness function of the genetic algorithm according to the objective function of the genetic algorithm;

[0109] Based on the constraint conditions, use the genetic algorithm to perform an optimization calculation on the fitness function, and obtain the step sizes of various tissue parameters of the objective function corresponding to the optimal fitness function.

[0110] In some embodiments, each tissue parameter of magnetic resonance fingerprinting imaging at least includes longitudinal relaxation time, transverse relaxation time, and proton spin density.

[0111] In some embodiments, when the processor 10 executes the magnetic resonance fingerprinting imaging dictionary resolution optimization program 40 in the memory 20, the following steps are further implemented:

[0112] Establish an imaging speed function for each tissue parameter, where the imaging speed function is a function representing the mapping relationship between each tissue parameter and its step size;

[0113] Establish an objective function according to the imaging speed function of each tissue parameter.

[0114] In some embodiments, when the processor 10 executes the magnetic resonance fingerprinting imaging dictionary resolution optimization program 40 in the memory 20, the following steps are further implemented:

[0115] Obtain the preset weight of each tissue parameter;

[0116] Establish an objective function according to the preset weight of each tissue parameter and the imaging speed function of each tissue parameter.

[0117] In some embodiments, when the processor 10 executes the magnetic resonance fingerprinting imaging dictionary resolution optimization program 40 in the memory 20, the following steps are further implemented:

[0118] Step 1: Generate an initial population. The initial population is composed of several individuals, and an individual is a random step size of one tissue parameter or a set of random step sizes of at least two tissue parameters;

[0119] Step 2: Calculate the individual fitness of the initial population according to the fitness function;

[0120] Step 3: Perform genetic selection on the individuals of the initial population according to the individual fitness;

[0121] Step 4: Perform a crossover operation on the selected individuals to obtain two new individuals;

[0122] Step 5: Perform a mutation operation on the selected individuals to obtain new individuals;

[0123] Step 6: After repeating steps 2 to 5 for a preset number of iterations, obtain the optimal individual that satisfies the objective function and the constraint conditions, and use the step sizes of each tissue parameter of the optimal individual as the step sizes of each tissue parameter of the objective function corresponding to the optimal fitness function.

[0124] In some embodiments, the roulette wheel selection method is used to perform genetic selection on the individuals of the initial population.

[0125] In some embodiments, the preset number of iterations is 100 to 600.

[0126] In summary, for the method, device, electronic device, and storage medium for optimizing the dictionary resolution of magnetic resonance fingerprint imaging provided by the present invention, constraint conditions and an objective function of a genetic algorithm are first established. The constraint conditions are established according to a preset magnetic resonance fingerprint imaging accuracy, which limits the minimum accuracy of magnetic resonance fingerprint imaging. The objective function is established according to the step sizes of various tissue parameters of magnetic resonance fingerprint imaging, which reflects the speed of magnetic resonance fingerprint imaging. Then, according to the objective function of the genetic algorithm, the fitness function of the genetic algorithm is determined. Finally, according to the constraint conditions, the genetic algorithm is used to perform an optimization calculation on the fitness function, and the objective function corresponding to the optimal fitness function finally obtained is the optimal step size of each tissue parameter. In the embodiments of the present invention, the genetic algorithm is used to optimize the step sizes (dictionary resolution) of various tissue parameters, with the imaging accuracy as the constraint condition and the imaging speed as the fitness function. The set of final solutions will minimize the number of combinations of tissue parameters (dictionary size) under the condition of meeting the imaging accuracy, thereby optimizing the imaging speed and avoiding the problem of low MRF dictionary resolution caused by manually setting the step sizes of tissue parameters.

[0127] Certainly, those of ordinary skill in the art can understand that all or part of the processes of implementing the above method embodiments can be completed by instructing relevant hardware (such as a processor, a controller, etc.) through a computer program. The program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the above method embodiments. The storage medium can be a memory, a magnetic disk, an optical disc, etc.

[0128] The specific embodiments of the present invention described above do not constitute a limitation on the protection scope of the present invention. Any other corresponding changes and deformations made according to the technical concept of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A method for optimizing the dictionary resolution of magnetic resonance fingerprint imaging, characterized in that, it includes the following steps: Establish the constraint conditions and objective function of the genetic algorithm, wherein the constraint conditions are established according to the preset magnetic resonance fingerprint imaging accuracy, and the objective function is established according to the step sizes of the various tissue parameters of magnetic resonance fingerprint imaging; According to the objective function of the genetic algorithm, establish the fitness function of the genetic algorithm; Based on the constraint conditions, use the genetic algorithm to perform optimization calculation on the fitness function, and obtain the step sizes of the various tissue parameters of the objective function corresponding to the optimal fitness function; The various tissue parameters of the magnetic resonance fingerprint imaging at least include longitudinal relaxation time, transverse relaxation time, and proton spin density; The step of using the genetic algorithm to perform optimization calculation on the fitness function based on the constraint conditions to obtain the step sizes of the various tissue parameters of the objective function corresponding to the optimal fitness function includes: Step 1: Generate an initial population, which is composed of several individuals, and the individual is a random step size of a tissue parameter or a set of random step sizes of at least two tissue parameters; Step 2: Calculate the individual fitness of the initial population according to the fitness function; Step 3: Perform genetic selection on the individuals of the initial population according to the individual fitness; Step 4: Perform crossover operation on the selected individuals to obtain two new individuals; Step 5: Perform mutation operation on the selected individuals to obtain new individuals; Step 6: After repeating steps 2 to 5 for a preset number of iterations, obtain the optimal individual that satisfies the objective function and constraint conditions, and use the step sizes of the various tissue parameters of the optimal individual as the step sizes of the various tissue parameters of the objective function corresponding to the optimal fitness function.

2. The method for optimizing the dictionary resolution of magnetic resonance fingerprint imaging according to claim 1, characterized in that, the method for establishing the objective function is: Establish the imaging speed function of each tissue parameter, wherein the imaging speed function is a function representing the mapping relationship between each tissue parameter and its step size; According to the imaging speed function of each tissue parameter, establish the objective function.

3. The method for optimizing the dictionary resolution of magnetic resonance fingerprint imaging according to claim 2, characterized in that, the step of establishing the objective function according to the imaging speed function of each tissue parameter includes: Obtain the preset weights of each tissue parameter; According to the preset weights of each tissue parameter and the imaging speed function of each tissue parameter, establish the objective function.

4. The method for optimizing the dictionary resolution of magnetic resonance fingerprint imaging according to claim 1, characterized in that, the roulette wheel selection method is used to perform genetic selection on the individuals of the initial population.

5. The method for optimizing the dictionary resolution of magnetic resonance fingerprint imaging according to claim 1, characterized in that, the preset number of iterations is 100 - 600.

6. A device for optimizing the dictionary resolution of magnetic resonance fingerprint imaging, characterized in that, it includes: An objective function establishment module, configured to establish the constraint conditions and the objective function of the genetic algorithm, wherein the constraint conditions are established according to a preset magnetic resonance fingerprint imaging accuracy, and the objective function is established according to the step sizes of the respective tissue parameters of the magnetic resonance fingerprint imaging; A fitness function establishment module, configured to establish the fitness function of the genetic algorithm according to the objective function of the genetic algorithm; An optimization module, configured to perform optimization calculation on the fitness function by using the genetic algorithm based on the constraint conditions, and obtain the step sizes of the respective tissue parameters of the objective function corresponding to the optimal fitness function; The respective tissue parameters of the magnetic resonance fingerprint imaging at least include the longitudinal relaxation time, the transverse relaxation time, and the proton spin density; The optimization module is specifically configured to implement the following steps: Step 1, generate an initial population, which is composed of a plurality of individuals, and an individual is a random step size of one tissue parameter or a set of random step sizes of at least two tissue parameters; Step 2, calculate the individual fitness of the initial population according to the fitness function; Step 3, perform genetic selection on the individuals of the initial population according to the individual fitness; Step 4, perform a crossover operation on the selected individuals to obtain two new individuals; Step 5, perform a mutation operation on the selected individuals to obtain new individuals; Step 6, after repeating Steps 2 to 5 for a preset number of iterations, obtain an optimal individual that satisfies the objective function and the constraint conditions, and use the step sizes of the respective tissue parameters of the optimal individual as the step sizes of the respective tissue parameters of the objective function corresponding to the optimal fitness function.

7. An electronic device, characterized in that it includes: a processor and a memory; a computer-readable program executable by the processor is stored on the memory; when the processor executes the computer-readable program, the steps in the magnetic resonance fingerprint imaging dictionary resolution optimization method according to any one of claims 1-5 are implemented.

8. A computer-readable storage medium, characterized in that the computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps in the magnetic resonance fingerprint imaging dictionary resolution optimization method according to any one of claims 1-5.

Images