CT image reconstruction method and device based on quantum annealing and medium

Abstract

The invention relates to a CT image reconstruction method and device based on quantum annealing and a medium, and the method comprises the steps: collecting projection data y through a CT, and obtaining a projection matrix P of a CT system; using quantum bits to describe an image x to be reconstructed, and projecting the image x through the CT system matrix P to obtain projection data Px containing variables; obtaining a to-be-optimized expression based on the projection data Px containing the variables and the real projection data y; compiling the expression to be optimized to obtain a problem Hamiltonian used for calculation of a quantum computer; and substituting the Hamiltonian of the problem into a quantum computer to carry out quantum annealing calculation, and carrying out repeated iteration until the reconstructed image meets the precision requirement, thereby obtaining the reconstructed image. Compared with the prior art, efficient image reconstruction can be achieved, the image reconstruction process can be accelerated through high parallelism of quantum calculation, and the imaging efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image reconstruction, and in particular to a CT image reconstruction method based on quantum annealing. Background Art

[0002] With the continuous development of medical imaging technology, the quality of CT images has gradually improved, making the images clearer. This progress is due to the continuous innovation of hardware and software technology, such as the use of more advanced X-ray detectors and higher-resolution image reconstruction algorithms in photon counting CT. The application of these technologies enables doctors to observe and analyze the patient's anatomical structure and lesions more accurately during the diagnosis process.

[0003] However, as image quality improves, some problems have emerged. For example, current photon counting CT has the following problems: 1. CT scans generate a large amount of data, which requires powerful computing resources and storage space for processing and storage. 2. Although modern CT equipment has significantly improved the speed of image reconstruction, in some complex cases, the reconstruction time is still long. 3. The radiation dose problem. CT scans require the use of X-rays, and patients may be exposed to higher radiation doses, especially those who undergo frequent scans. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a CT image reconstruction method, device and medium based on quantum annealing.

[0005] The purpose of the present invention can be achieved by the following technical solutions:

[0006] As a first aspect of the present invention, a CT image reconstruction method based on quantum annealing is provided, the method steps comprising:

[0007] Collect projection data y through CT to obtain the projection matrix P of the CT system;

[0008] The image x to be reconstructed is described by quantum bits, and is projected through the CT system matrix P to obtain projection data Px containing variables;

[0009] An expression to be optimized is obtained based on the projection data Px containing variables and the real projection data y;

[0010] Compile the expression to be optimized to obtain the problem Hamiltonian for quantum computer calculation;

[0011] The problem Hamiltonian is brought into a quantum computer for quantum annealing calculations, and it is iterated repeatedly until the reconstructed image meets the accuracy requirements to obtain a reconstructed image.

[0012] As a preferred technical solution, the reconstructed image x is described as follows using a corresponding number of quantum bits according to the resolution:

[0013]

[0014] Where: qma x Indicates the number of quantum bits used for each image pixel; k is used to convert the reconstructed value from an integer to a real number; σ q，j represents quantum bit; d j It is a parameter related to the value of the quantum bit and changes with the iteration process.

[0015] As a preferred technical solution, the expression to be optimized is:

[0016] min|Px-y| 2

[0017] Where: P represents the CT system matrix; x represents the reconstructed image; y represents the projection data.

[0018] As a preferred technical solution, the Hamiltonian H for the problem calculated by the quantum computer p It is expressed as:

[0019] H p =|Px-y| 2 .

[0020] As a preferred technical solution, the method of bringing the problem Hamiltonian into a quantum computer for quantum annealing calculation is specifically as follows:

[0021] H(s)=(1-s)H0+H p

[0022] Here, s is a parameter that changes from 0 to 1 over time. During quantum annealing, the system evolves from a ground state with a known Hamiltonian H0, and eventually reaches the target Hamiltonian H when s = 1. p ground state.

[0023] As a preferred technical solution, in the quantum annealing process, update the parameters:

[0024]

[0025]

[0026] Among them, c j is a constant used to adjust the step size during the iteration; k j It is used to convert the reconstructed value from an integer to a real number, which will change continuously during the iteration process; q max Indicates the number of quantum bits used for each image pixel.

[0027] As a preferred technical solution, determining whether the reconstructed image meets the accuracy requirement is specifically as follows: reprojecting the iteratively obtained reconstructed image and checking whether |Px-y| meets the accuracy requirement.

[0028] As a preferred technical solution, when the annealing process is completed, the state of the quantum bit corresponds to the problem Hamiltonian H p The approximate ground state of the quantum bit is obtained by reading the state of the quantum bit, and the solution x of the optimization problem is obtained, that is, the reconstructed image.

[0029] As a second aspect of the present invention, a CT image reconstruction device based on quantum annealing is provided, comprising a memory, a processor, and a program stored in the memory, wherein the processor implements any of the above-described quantum annealing-based CT image reconstruction methods when executing the program.

[0030] As a third aspect of the present invention, a storage medium is provided, on which a program is stored, characterized in that when the program is executed, the CT image reconstruction method based on quantum annealing as described above is implemented.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1) The present invention converts the CT reconstruction problem into an optimization problem and adopts a quantum annealing optimization algorithm based on quantum mechanics. It uses the properties of quantum bit superposition and entanglement to more effectively jump out of the local minimum and find the global optimal solution, thereby improving the quality of the reconstructed image. At the same time, after reconstructing the projection data containing noise, it has better image quality than the traditional reconstruction method.

[0033] 2) The present invention can achieve efficient image reconstruction, and the high parallelism of quantum computing can accelerate the image reconstruction process and improve imaging efficiency.

[0034] 3) Compared with the classical iterative reconstruction algorithm, the present invention only requires fewer projection angles to complete accurate reconstruction, reducing the frequency of collecting CT projection data, thereby reducing the radiation dose received by the patient. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a flow chart of the implementation of the quantum annealing-based CT image reconstruction method of the present invention. DETAILED DESCRIPTION

[0036] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0037] Example 1

[0038] The invention realizes the requirements of combining the iterative reconstruction algorithm of classical CT with the quantum algorithm, converting the loss function in the iterative reconstruction into a Hamiltonian that can be used for quantum computing, evolving the Hamiltonian through a quantum annealing algorithm or a quantum approximate optimization algorithm, obtaining the final result through measurement, and finally converting the optimized result into a reconstructed image. In the process, the loss function needs to be encoded first, a model that can be used for quantum computer calculation is generated, and a quantum computer needs to be calculated. Figure 1 As shown, the specific steps include:

[0039] The first step is to collect projection data y through CT for subsequent reconstruction.

[0040] The second step is to obtain the projection matrix P of the CT system and the initial values ​​of the given parameters.

[0041] Among them, c j k is a constant, which is mainly used to adjust the step size in the iteration process. If it is too large, the optimal solution will be missed, and if it is too small, the iteration time will be longer. j It is to transform the reconstructed value from an integer to a real number, which will change continuously during the iteration process. max Indicates the number of quantum bits used for each image pixel;

[0042] The third step is to describe the image x to be reconstructed using the corresponding number of quantum bits according to the resolution:

[0043]

[0044] Where: q，j represents a quantum bit, which has a value of 0 or 1. As the quantum annealing process continues, it will get a certain value; d j It is a number related to the value of the quantum bit, and it will also change continuously with the iteration process. When the reconstruction result becomes more and more accurate, the d value tends to be determined.

[0045] The reconstructed image x described by the quantum bit is projected through the CT system matrix P to obtain projection data Px containing variables;

[0046] Subtract the projection data Px containing variables from the real data y to get the expression to be optimized min|Px-y| 2 .

[0047] The fourth step is to compile the expression to be optimized to obtain the problem Hamiltonian H that can be used for quantum computer calculation. p :

[0048] Hp =|Px-y| 2

[0049] Step 5: Bring the problem Hamiltonian into the quantum computer for calculation:

[0050] H(s)=(1-s)H0+H p

[0051] Here, s is a parameter that changes from 0 to 1 over time; during quantum annealing, the system evolves from a known ground state (corresponding to the ground state of H0). Finally, when s = 1, the state of the system H(s) is the target Hamiltonian H p ground state.

[0052] During quantum annealing, update the parameters:

[0053]

[0054] Step 6. Project the reconstructed image again to check whether |Px-y| meets the accuracy requirements. If so, the reconstructed image is obtained; if not, the parameters updated in step 5 are brought into step 3 for recalculation.

[0055] Repeat steps 3 to 6 until the accuracy requirement is met. When the annealing process is completed, the state of the quantum bit corresponds to the problem Hamiltonian H. p By reading the states of these quantum bits, we can get the solution x to the optimization problem, that is, the reconstructed image x.

[0056] The pseudo code of the CT reconstruction algorithm based on quantum annealing is as follows:

[0057]

[0058]

[0059] Example 2

[0060] As a second aspect of the present invention, the present application also provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned quantum annealing-based CT image reconstruction method. In addition to the above-mentioned processor, memory and interface, any device with data processing capability in which the device in the embodiment is located may also include other hardware according to the actual function of the device with data processing capability, which will not be described in detail.

[0061] Example 3

[0062] As a third aspect of the present invention, the present application also provides a computer-readable storage medium on which computer instructions are stored, and when the instructions are executed by a processor, the CT image reconstruction method based on quantum annealing as described above is implemented. The computer-readable storage medium can be an internal storage unit of any device with data processing capabilities as described in any of the aforementioned embodiments, such as a hard disk or a memory. The computer-readable storage medium can also be an external storage device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), an SD card, a flash card (Flash Card), etc. equipped on the device. Further, the computer-readable storage medium can also include both an internal storage unit of any device with data processing capabilities and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capabilities, and can also be used to temporarily store data that has been output or is to be output.

[0063] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A CT image reconstruction method based on quantum annealing, characterized in that: The method steps include: Collect projection data y through CT to obtain the projection matrix P of the CT system; The image x to be reconstructed is described by quantum bits, and is projected through the CT system matrix P to obtain projection data Px containing variables; An expression to be optimized is obtained based on the projection data Px containing variables and the real projection data y; Compile the expression to be optimized to obtain the problem Hamiltonian for quantum computer calculation; The problem Hamiltonian is brought into a quantum computer for quantum annealing calculations, and it is iterated repeatedly until the reconstructed image meets the accuracy requirements to obtain a reconstructed image.

2. The CT image reconstruction method based on quantum annealing according to claim 1, characterized in that: The reconstructed image x is described as follows using a corresponding number of quantum bits according to the resolution: Where: q max Indicates the number of quantum bits used for each image pixel; k j Used to convert the reconstructed value from an integer to a real number; σ q,j represents quantum bit; d j It is a parameter related to the value of the quantum bit and changes with the iteration process.

3. The CT image reconstruction method based on quantum annealing according to claim 1, characterized in that: The expression to be optimized is: min|Px-y| 2 Where: P represents the CT system matrix; x represents the reconstructed image; y represents the projection data.

4. The CT image reconstruction method based on quantum annealing according to claim 3, characterized in that: The Hamiltonian H for the problem calculated by the quantum computer p It is expressed as: H p =|Px-y| 2 。 5. The CT image reconstruction method based on quantum annealing according to claim 1, characterized in that: The specific method of bringing the problem Hamiltonian into a quantum computer for quantum annealing calculation is as follows: H(s)=(1-s)H0+H p Here, s is a parameter that changes from 0 to 1 over time. During quantum annealing, the system evolves from a ground state with a known Hamiltonian H0, and eventually reaches the target Hamiltonian H when s = 1. p ground state.

6. The method for CT image reconstruction based on quantum annealing according to claim 5, characterized in that: During the quantum annealing process, the parameters are updated: Among them, c j is a constant used to adjust the step size during the iteration; k j It is used to convert the reconstructed value from an integer to a real number, which will change continuously during the iteration process; q max Indicates the number of quantum bits used for each image pixel.

7. The CT image reconstruction method based on quantum annealing according to claim 1, characterized in that: Determining whether the reconstructed image meets the accuracy requirement is specifically as follows: reprojecting the iteratively obtained reconstructed image to check whether |Px-y| meets the accuracy requirement.

8. The CT image reconstruction method based on quantum annealing according to claim 1, characterized in that: When the annealing process is completed, the state of the quantum bit corresponds to the problem Hamiltonian H p The approximate ground state of the quantum bit is obtained by reading the state of the quantum bit, and the solution x of the optimization problem is obtained, that is, the reconstructed image.

9. A CT image reconstruction device based on quantum annealing, comprising a memory, a processor, and a program stored in the memory, characterized in that: When the processor executes the program, the CT image reconstruction method based on quantum annealing as described in any one of claims 1-8 is implemented.

10. A storage medium having a program stored thereon, characterized in that: When the program is executed, the quantum annealing-based CT image reconstruction method as described in any one of claims 1 to 8 is implemented.

Images