Image Denoising Method, Electronic Device, and Storage Medium

Abstract

The present invention provides an image denoising method, an electronic device, and a storage medium. The method includes: obtaining a plurality of initial image sequences; obtaining the signal-to-noise ratios between all the initial image sequences; and performing denoising processing on the initial image sequences based on the signal-to-noise ratios to obtain a target denoised image sequence. The present invention can reduce the noise level of the initial image sequences while ensuring that the scanned contents of the finally obtained target denoised image sequences meet the consistency requirements.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and in particular to an image denoising method, an electronic device and a storage medium. Background Art

[0002] Image denoising has been widely applied to natural images and video processing. Traditional denoising methods include spatial domain filtering, transform domain filtering, partial differential equations, variational methods, morphological noise filtering, etc. In addition, methods for denoising images and videos using deep learning are emerging in an endless stream and show better effects than traditional methods. However, to obtain a better denoising effect using deep learning methods, a large number of noise-free images are usually required as supervised information to train the model, which is difficult for medical perfusion images. Most of the methods for denoising medical images based on traditional methods are applied to a single image sequence, such as CT plain scan, T1WI image, etc. For perfusion images, which contain several time series and are affected by contrast agents, the image quality is worse than that of single-sequence images, which also causes difficulties in denoising. Currently, the solutions for denoising CTP (CT Perfusion) images are mainly based on methods related to spatial domain filtering, and rarely combine the content of each time series for denoising. Therefore, the consistency of the scanned content information between each time series cannot be guaranteed, and at the same time, for data with different noise levels, the obtained image quality is uneven. Summary of the Invention

[0003] In order to solve the problem that the consistency of the scanned content information between each time series cannot be guaranteed when denoising perfusion images in the prior art, the present invention provides an image denoising method, an electronic device and a storage medium.

[0004] To achieve the above object, the present invention adopts the following technical solutions:

[0005] In a first aspect, the present invention provides an image denoising method, including:

[0006] Obtaining a plurality of initial image sequences;

[0007] Obtaining the signal-to-noise ratio between all the initial image sequences;

[0008] Based on the signal-to-noise ratio, performing denoising processing on the initial image sequences to obtain a target denoised image sequence.

[0009] Preferably, the performing denoising processing on the initial image sequences based on the signal-to-noise ratio to obtain a target denoised image sequence includes:

[0010] Performing denoising processing on the initial image sequences to obtain an initial denoised image sequence;

[0011] Use the initial denoised image sequence as the new initial image sequence, and return to perform the step of obtaining the signal-to-noise ratio between the initial image sequences until the signal-to-noise ratio obtained twice before and after satisfies a predetermined condition, and then obtain the target denoised image sequence.

[0012] Preferably, before obtaining the signal-to-noise ratio between all the initial image sequences, the method further includes:

[0013] Register each of the initial image sequences.

[0014] Preferably, before obtaining the signal-to-noise ratio between all the initial image sequences, the method further includes:

[0015] Perform normalization processing on each of the initial image sequences.

[0016] Preferably, the performing normalization processing on each of the initial image sequences includes:

[0017] Perform normalization processing on each of the initial image sequences according to the average value and standard deviation of the parameters of each initial image sequence.

[0018] Preferably, the obtaining the signal-to-noise ratio between all the initial image sequences includes:

[0019] Determine the first image region in each of the initial image sequences;

[0020] Determine the second image region in each of the initial image sequences according to the first image region;

[0021] Calculate the signal-to-noise ratio between all the initial image sequences according to the second image regions in each of the initial image sequences.

[0022] Preferably, the determining the first image region in each of the initial image sequences includes:

[0023] Obtain the maximum intensity projection data of each of the initial image sequences;

[0024] Determine the first image region in each of the initial image sequences according to the maximum intensity projection data of each of the initial image sequences.

[0025] Preferably, after obtaining the target denoised image sequence, the method further includes:

[0026] Perform inverse normalization processing on each of the target denoised image sequences.

[0027] In a second aspect, the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the above-described image denoising method is implemented.

[0028] In a third aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above-described image denoising method is implemented.

[0029] By adopting the above technical solutions, the present invention has the following beneficial effects:

[0030] The present invention obtains the signal-to-noise ratio between initial image sequences, and based on the signal-to-noise ratio, performs denoising processing on the initial image sequences to obtain a target denoised image sequence. Since the signal-to-noise ratio between the initial image sequences can measure the consistency of the scanned content information between each time series, denoising processing based on this signal-to-noise ratio can reduce the noise level of the initial image sequences while ensuring that the scanned content of each final target denoised image sequence meets the consistency requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a schematic flowchart of the image denoising method provided in Embodiment 1 of the present invention;

[0032] Figure 2 It is a schematic diagram of the test result curve of the image denoising method provided in Embodiment 1;

[0033] Figure 3 It is a comparison diagram of the test results of the image denoising method provided in Embodiment 1 and the existing RAPID method;

[0034] Figure 4 It is a structural block diagram of the image denoising system provided in Embodiment 2 of the present invention;

[0035] Figure 5 It is a hardware architecture diagram of the electronic device provided in Embodiment 3 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0036] 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. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0037] The terms used in this invention are for the purpose of describing specific embodiments only and are not intended to limit the disclosure. The singular forms "a", "the", and "said" used in this disclosure and the appended claims are also intended to include the plural forms unless the context clearly dictates otherwise. It should also be understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0038] The perfusion process refers to the process in which blood flows from arteries to the capillary network and then drains into veins. To measure this process, a medium must be used to replace the blood so that people can track the flow process of the medium through external instruments. Iodine contrast agents are commonly used in CT perfusion imaging, and Gd-DTPA contrast agents are commonly used as the medium in MRI perfusion imaging. Through perfusion data, people can obtain the tissue blood flow conditions through some physical models such as the slope method, deconvolution method, and neural networks.

[0039] However, in the actual clinical scanning process, due to the influence of random factors during imaging, image noise and artifacts are inevitably generated. In addition, the acquisition of perfusion images usually lasts for a long time, and several sequences are acquired to obtain the dynamic changes of the entire perfusion process. Noise and artifacts will also be caused between multiple sequences due to the influence of motion and random noise, affecting further diagnosis.

[0040] In addition, due to factors such as differences in perfusion image acquisition equipment, contrast agent types, and different individual differences, there are relatively large differences in the noise levels of perfusion images obtained from different cases, causing certain difficulties in obtaining diagnostic results using traditional methods or deep learning methods subsequently, and also affecting the stability of the algorithm. Therefore, it is necessary to denoise perfusion images.

[0041] Embodiment 1

[0042] This embodiment provides an image denoising method for denoising the image sequence of perfusion images. As Figure 1 shown, the method specifically includes the following steps:

[0043] S1, obtaining a plurality of initial image sequences.

[0044] In this embodiment, the plurality of initial image sequences refer to the image sequence of perfusion images. The perfusion image can be, for example, the CTP image X obtained by scanning with a CT device, X ∈ (T × P × W × H), where T represents the number of CT images collected during the CT scan, and P, W, and H respectively represent the number of layers, width, and height of the CT image collected at each moment. Among them, according to the tissue organs of perfusion imaging, CTP images can include brain CTP images, liver CTP images, heart CTP images, etc.

[0045] S2. Obtain the signal-to-noise ratio between all the initial image sequences.

[0046] In this embodiment, the signal-to-noise ratio SNR between all the initial image sequences can be obtained by the following formula (1):

[0047] SNR = S / N (1)

[0048] Wherein, S represents the signal of all the initial image sequences (i.e., the image content, assuming that the image content collected at each time remains unchanged), and N represents the noise of all the initial image sequences (i.e., the random noise and artifacts generated in the image content during the scanning process over time).

[0049] S3. Based on the signal-to-noise ratio, perform denoising processing on the initial image sequences to obtain a target denoised image sequence.

[0050] In this embodiment, any suitable denoising algorithm (including but not limited to spatial domain filtering, transform domain filtering, partial differential equations, variational methods, morphological noise filtering, neural networks, etc.) can be used to perform denoising processing on the initial image sequences until the signal-to-noise ratio meets the predetermined requirements.

[0051] The present invention obtains the signal-to-noise ratio between the initial image sequences, and based on the signal-to-noise ratio, performs denoising processing on the initial image sequences to obtain a target denoised image sequence. Since the signal-to-noise ratio between the initial image sequences can measure the consistency of the scanned content information between each time series, denoising processing based on this signal-to-noise ratio can, while reducing the noise level of the initial image sequences, make the scanned content of each final obtained target denoised image sequence meet the consistency requirements.

[0052] In an alternative embodiment, step S3 is specifically implemented by the following steps:

[0053] First, perform denoising processing on the initial image sequences to obtain an initial denoised image sequence;

[0054] Then, use the initial denoised image sequence as a new initial image sequence, and return to step S2 to obtain the signal-to-noise ratio between all the initial image sequences until the signal-to-noise ratio obtained twice before and after meets the predetermined conditions, and then obtain the target denoised image sequence.

[0055] Specifically, assume that the signal-to-noise ratio between all the initially acquired image sequences in the previous acquisition is SNR1, and the signal-to-noise ratio between all the initially acquired image sequences in the subsequent acquisition is SNR2. Then the predetermined condition can be set as SNR2 - SNR1 < δ, where δ is the threshold of the signal-to-noise ratio change of the perfusion image sequence and can be set according to specific circumstances. For example, δ is taken as 0.5. If the signal-to-noise ratio change after two denoising processes is less than this threshold, it is considered that the noise level of the perfusion image sequence meets the predetermined condition, and the image content of each sequence meets the consistency requirement, and the denoising process ends.

[0056] This embodiment uses the signal-to-noise ratio between different sequences of perfusion images as an evaluation index for the image quality of iterative denoising, which can effectively ensure the consistency of signals between sequences. At the same time, by adopting an iterative method for denoising, for data with different noise levels, results with similar image quality can be obtained, effectively ensuring that the perfusion images of different devices and different cases converge to the same signal-to-noise ratio level, and further ensuring the stability of the results of subsequent diagnostic algorithms.

[0057] In an alternative embodiment, before obtaining the signal-to-noise ratio between all the initially acquired image sequences in step S2, the method further includes: registering each of the initially acquired image sequences.

[0058] Specifically, taking one or more of the initially acquired image sequences as reference data, other initially acquired image sequences are registered to the reference data. In this embodiment, registration means seeking one or a series of spatial transformations for a medical image to make the corresponding points on it coincide spatially with those on another or multiple images (i.e., reference data).

[0059] In this embodiment, the specific manner of registration can be implemented with reference to any suitable existing registration method, and will not be described in detail. The parameters of each registered image sequence are denoted as X reg =[x1,x2,…,x T .

[0060] This embodiment can reduce the sequence differences caused by motion through registration and improve the accuracy of subsequent data processing.

[0061] In an alternative embodiment, before obtaining the signal-to-noise ratio between all the initially acquired image sequences in step S2, the method further includes: normalizing each of the initially acquired image sequences.

[0062] This embodiment can reduce the influence of individual differences and the brightness generated by the contrast agent by normalizing the initially acquired image sequences.

[0063] In an alternative embodiment, the normalization processing of each of the initial image sequences includes: normalizing each of the initial image sequences according to the mean and standard deviation of the parameters of each of the initial image sequences.

[0064] Specifically, first obtain the mean of the parameters of each of the initial image sequences, denoted as At the same time, obtain the standard deviation of the parameters of each of the initial image sequences, denoted as sd1, sd2, …, sd T , and then normalize each of the initial image sequences according to the following formula (2) to obtain the normalization processing results x′1, x′2, …, x′ of each of the initial image sequences T :

[0065]

[0066] where T represents the number of initial image sequences, and x i represents the parameter of the i-th image sequence after registration.

[0067] In this embodiment, by normalizing each of the initial image sequences according to the mean and standard deviation of the parameters of each of the initial image sequences, the processed results after normalization conform to the standard normal distribution.

[0068] In an alternative embodiment, step S2 obtains the signal-to-noise ratio between all the initial image sequences through the following steps:

[0069] First, determine the first image region in each of the initial image sequences.

[0070] In this embodiment, the first image region may, for example, be a blood vessel region and a region near the blood vessel that is greatly affected by the contrast agent. The content in these regions changes greatly (i.e., is significantly enhanced) during the perfusion process, which will affect the accuracy of the signal-to-noise ratio.

[0071] Then, determine the region outside the first image region in each of the initial image sequences as the second image region.

[0072] Finally, calculate the signal-to-noise ratio between all the initial image sequences according to the second image regions in each of the initial image sequences.

[0073] Specifically, calculate the signal S and noise N of all the initial image sequences respectively through the following formulas (3) and (4):

[0074]

[0075] where T represents the number of initial image sequences, and x′ tRepresents the normalization result of each of the initial image sequences, V represents the first image region in each of the initial image sequences, 1 - V represents the second image region other than the first image region participating in the calculation, x′ t ×(1 - V) represents the normalization result of the second image region in the t-th image sequence, (x′ t - S) 2 ×(1 - V) represents the square of the difference between the normalization result of the second image region in the t-th image sequence and the signal S.

[0076] In this embodiment, by regarding each initial image sequence as a large number of time signals, except for the enhanced first image region such as blood vessels, the signal intensities of other tissues are considered to remain constant. Excluding the first image region when calculating the signal-to-noise ratio can avoid the interference of the enhanced part and make the obtained signal-to-noise ratio more accurate.

[0077] Random noise will make the image signals at each time point of the perfusion image inconsistent, thereby reducing the above signal-to-noise ratio index. In formula (4) above, S is the average value of all image sequences (excluding the first image region therein), representing the effective signal; N is the difference between each image sequence and the average value, representing the noise. The larger the value of S / N calculated by the signal-to-noise ratio calculation method described by the above formula, the smaller the noise, that is, the smaller the difference, which means the higher the consistency of the image content among the various sequences of the perfusion image.

[0078] In an alternative embodiment, the first image region in each of the initial image sequences is determined by the following steps:

[0079] First, obtain the maximum intensity projection data MIP of each of the initial image sequences.

[0080] Specifically, calculate the maximum intensity projection data MIP of each of the initial image sequences according to the following formula: MIP = max(X′, axis = 0), where X′ = x′1, x′2, …, x′ T ,x′ i represents the normalization result of the i-th image sequence after registration.

[0081] Then, determine the first image region in each of the initial image sequences according to the maximum intensity projection data of each of the initial image sequences.

[0082] Specifically, when the maximum intensity projection data MIP of a certain initial image sequence is greater than the preset density threshold θ, it is considered that there is a first image region in this sequence, and then the region where the density projection data exceeds θ in this sequence is determined as the first image region.

[0083] Since the first image region such as blood vessels will be significantly enhanced during perfusion, the density projection data corresponding to this region will be greater than the density projection data corresponding to other regions in the initial image sequence. Therefore, the first image region can be accurately determined through the above steps.

[0084] In an alternative embodiment, if each of the initial image sequences is normalized before step S2, then after obtaining the target denoised image sequences, the method further includes: performing denormalization processing on each of the target denoised image sequences.

[0085] Specifically, the denormalization processing of each target denoised image sequence can be performed through the following formula (5):

[0086]

[0087] where, o i represents the denormalization processing result of each of the target denoised image sequences, that is, the perfusion image data of each of the target denoised image sequences before normalization, represents the average value of the parameters of each of the target denoised image sequences, y' i represents the parameters of each of the target denoised image sequences, sd i represents the standard deviation of the parameters of each of the initial image sequences.

[0088] In this embodiment, by performing denormalization processing on each of the target denoised image sequences, the perfusion level before normalization can be restored for the target denoised image sequences.

[0089] Figure 2 shows the iterative process of the CTP image sequences with different noise levels processed by the image denoising method of this embodiment. Among them, the curves corresponding to "GOU" and "HAO" represent the denoising processes of the CTP image sequences with relatively serious noise and artifacts, and the remaining curves are the denoising processes of the CTP image sequences with general noise. From Figure 2 it can be seen that the CTP image sequences corresponding to "GOU" and "HAO" need to go through 7 and 6 iterations respectively to complete denoising, and the remaining CTP image sequences can complete denoising after 2 - 3 iterations. It can be seen that the CTP image sequences with serious noise require more iterations than the CTP image sequences with general noise, reflecting that for image sequences with different noise levels, different denoising intensities need to be adopted to obtain similar effects.

[0090] Figure 3 shows the comparison chart of the denoising effects of the image denoising method of this embodiment and the RAPID method. The upper and lower rows are images with different parameters, and the noise levels in the left two columns and the right two columns are different. The left two columns have general noise, and the right two columns have serious noise. From Figure 3It can be seen that for image sequences with different parameters and different noise levels, the denoising effect achieved by using the method of this embodiment is close to, or even better than, that of the RAPID method.

[0091] Embodiment 2

[0092] As Figure 4 shown, the present invention provides an image denoising system, including: a sequence acquisition module 11, a signal-to-noise ratio acquisition module 12, and a denoising module 13. Among them, the sequence acquisition module 11 is used to acquire a plurality of initial image sequences; the signal-to-noise ratio acquisition module 12 is used to acquire the signal-to-noise ratio between all the initial image sequences; the denoising module 13 is used to perform denoising processing on the initial image sequences based on the signal-to-noise ratio to obtain a target denoised image sequence.

[0093] Preferably, the denoising module 13 is specifically configured to:

[0094] Perform denoising processing on the initial image sequences to obtain an initial denoised image sequence;

[0095] Use the initial denoised image sequence as a new initial image sequence, and return to execute the step of obtaining the signal-to-noise ratio between the initial image sequences until the signal-to-noise ratio obtained twice before and after meets a predetermined condition, and then obtain the target denoised image sequence.

[0096] Preferably, the system further includes: a registration module, configured to perform registration on each of the initial image sequences before obtaining the signal-to-noise ratio between all the initial image sequences.

[0097] Preferably, the system further includes: a normalization processing module, configured to perform normalization processing on each of the initial image sequences before obtaining the signal-to-noise ratio between all the initial image sequences.

[0098] Preferably, the normalization processing module is specifically configured to:

[0099] Perform normalization processing on each of the initial image sequences according to the average value and standard deviation of the parameters of each initial image sequence.

[0100] Preferably, the signal-to-noise ratio acquisition module 12 is specifically configured to:

[0101] Determine a first image region in each of the initial image sequences;

[0102] Determine the region outside the first image region in each of the initial image sequences as a second image region;

[0103] Calculate the signal-to-noise ratio between all the initial image sequences according to the second image regions in each of the initial image sequences.

[0104] Preferably, the signal-to-noise ratio acquisition module 12 determines the first image regions in each of the initial image sequences as follows:

[0105] Obtain the maximum intensity projection data of each of the initial image sequences;

[0106] Based on the maximum intensity projection data of each of the initial image sequences, determine the first image regions in each of the initial image sequences.

[0107] Preferably, the system further includes: a normalization processing module, configured to perform denormalization processing on each of the target denoised image sequences after obtaining the target denoised image sequences.

[0108] In the present invention, by obtaining the signal-to-noise ratio between the initial image sequences and performing denoising processing on the initial image sequences based on the signal-to-noise ratio to obtain the target denoised image sequences, since the signal-to-noise ratio between the initial image sequences can measure the consistency of the scanned content information between each time series, denoising processing based on this signal-to-noise ratio can, while reducing the noise level of the initial image sequences, ensure that the scanned content of each finally obtained target denoised image sequence meets the consistency requirements.

[0109] For the system embodiment of the present invention, since it basically corresponds to the method embodiment, reference can be made to the partial description of the method embodiment for relevant parts. The system embodiment described above is merely illustrative, where the units described as separate components may or may not be physically separated, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of the present invention. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0110] Embodiment 3

[0111] This embodiment provides an electronic device, which can be presented in the form of a computing device (for example, it can be a server device), including a memory, a processor, and a computer program stored on the memory and executable on the processor, where when the processor executes the computer program, the image denoising method provided in Embodiment 1 can be implemented.

[0112] Figure 5 Shows the schematic hardware structure of this embodiment, as Figure 5 shown, the electronic device 9 specifically includes:

[0113] At least one processor 91, at least one memory 92, and a bus 93 for connecting different system components (including the processor 91 and the memory 92), where:

[0114] The bus 93 includes a data bus, an address bus, and a control bus.

[0115] The memory 92 includes volatile memory, such as random access memory (RAM) 921 and / or cache memory 922, and may further include read-only memory (ROM) 923.

[0116] The memory 92 further includes a program / utilities 925 having a set (at least one) of program modules 924. Such program modules 924 include, but are not limited to: an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include the implementation of a network environment.

[0117] The processor 91 executes various functional applications and data processing by running computer programs stored in the memory 92, such as the image denoising method provided in Embodiment 1 of the present invention.

[0118] The electronic device 9 can further communicate with one or more external devices 94 (such as a keyboard, a pointing device, etc.). Such communication can be carried out through an input / output (I / O) interface 95. Also, the electronic device 9 can communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 96. The network adapter 96 communicates with other modules of the electronic device 9 through the bus 93. It should be understood that although not shown in the figure, other hardware and / or software modules can be used in conjunction with the electronic device 9, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (redundant array of independent disks) systems, tape drives, and data backup storage systems, etc.

[0119] It should be noted that although several units / modules or sub-units / modules of the electronic device are mentioned in the above detailed description, this division is merely exemplary and not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, the features and functions of one unit / module described above can be further divided and embodied by multiple units / modules.

[0120] Embodiment 4

[0121] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the image denoising method provided in Embodiment 1.

[0122] Among them, the readable storage medium can more specifically include, but is not limited to: portable disks, hard disks, random access memories, read-only memories, erasable programmable read-only memories, optical storage devices, magnetic storage devices, or any suitable combination of the above.

[0123] In a possible implementation manner, the present invention can also be implemented in the form of a program product, which includes program code. When the program product runs on a terminal device, the program code is used to cause the terminal device to execute the image denoising method described in Embodiment 1.

[0124] Among them, the program code for executing the present invention can be written in any combination of one or more programming languages. The program code can be executed entirely on the user device, partially on the user device, executed as an independent software package, partially on the user device and partially on a remote device, or entirely on a remote device.

[0125] Although the specific implementation manners of the present invention have been described above, those skilled in the art should understand that this is only an example. The protection scope of the present invention is defined by the appended claims. Without departing from the principles and essence of the present invention, those skilled in the art can make various changes or modifications to these implementation manners, but these changes and modifications all fall within the protection scope of the present invention.

Claims

1. An image denoising method, characterized in that, Including: Obtain a plurality of initial image sequences; Obtain the signal-to-noise ratio between all the initial image sequences; Based on the signal-to-noise ratio, perform denoising processing on the initial image sequences to obtain a target denoised image sequence; The obtaining the signal-to-noise ratio between all the initial image sequences includes: Determine a first image region in each of the initial image sequences; Determine a second image region in each of the initial image sequences according to the first image region; Calculate the signal-to-noise ratio between all the initial image sequences according to the second image region in each of the initial image sequences.

2. The image denoising method according to claim 1, characterized in that, The performing denoising processing on the initial image sequences based on the signal-to-noise ratio to obtain a target denoised image sequence includes: Perform denoising processing on the initial image sequences to obtain an initial denoised image sequence; Use the initial denoised image sequence as a new initial image sequence, and return to execute the step of obtaining the signal-to-noise ratio between the initial image sequences until the signal-to-noise ratio obtained twice before and after meets a predetermined condition, and then obtain the target denoised image sequence.

3. The image denoising method according to claim 1, wherein Before obtaining the signal-to-noise ratio between all the initial image sequences, the method further includes: Perform registration on each of the initial image sequences.

4. The image denoising method according to claim 1, characterized in that, Before obtaining the signal-to-noise ratio between all the initial image sequences, the method further includes: Perform normalization processing on each of the initial image sequences.

5. The image denoising method according to claim 4, characterized in that, The performing normalization processing on each of the initial image sequences includes: Perform normalization processing on each of the initial image sequences according to the average value and standard deviation of the parameters of each initial image sequence.

6. The image denoising method according to claim 1, wherein The determining the first image region in each of the initial image sequences includes: Obtain the maximum density projection data of each of the initial image sequences; Determine the first image region in each of the initial image sequences according to the maximum density projection data of each of the initial image sequences.

7. The image denoising method according to claim 3, wherein After obtaining the target denoised image sequence, the method further includes: Perform inverse normalization processing on each of the target denoised image sequences.

8. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method according to any one of claims 1 to 7.

Images