A cardiac magnetic resonance imaging enhancement optimization method for patients with heart failure
By determining the end time and duration of the respiratory cycle, grouping and weight calculation, the motion artifacts of cardiac magnetic resonance images are eliminated, and the artifact problems of cardiac magnetic resonance images during breathing are solved, improving the clarity and interpretation accuracy of the image.
Patent Information
- Application Number
- CN202411156329.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-22
- Publication Date
- 2024-11-12
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Cardiac magnetic resonance images are prone to motion artifacts during the patient's breathing, resulting in poor image effects and difficult to completely eliminate, affecting the interpretation of heart state.
By acquiring heart images at continuous moments, determining the accuracy and duration of the end of the breathing cycle, grouping heart images, calculating differences and grayscale features between pixel blocks, image enhancement is performed based on the weight of each pixel point, and motion pseudo-extinguishing film.
Effectively eliminate the motion artifacts of cardiac magnetic resonance images, reduce the impact of artifacts on cardiac state discrimination, and make the image better present the patient's heart state.
Smart Images

Figure CN118674641B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of image data processing, and in particular to a cardiac magnetic resonance image enhancement and optimization method for patients with heart failure. Background Art
[0002] Heart failure is caused by cardiac systolic dysfunction or diastolic dysfunction, which results in the heart being unable to meet the basic metabolic needs of the whole body. Currently, the main method for obtaining cardiac images is to obtain cardiac images through magnetic resonance imaging.
[0003] During the process of collecting cardiac images through MRI, the patient's breathing may cause certain motion artifacts in the cardiac MRI images, resulting in poor presentation of the cardiac MRI images. In addition, during the MRI test, it is difficult for the patient to hold his breath all the time, making it difficult to eliminate motion artifacts in the cardiac MRI images. Therefore, the cardiac MRI images need to be enhanced and optimized so that the images can better present the condition of the patient's heart. Summary of the invention
[0004] The present invention provides a cardiac magnetic resonance image enhancement optimization method for patients with heart failure to solve the existing problems.
[0005] A cardiac magnetic resonance imaging enhancement optimization method for patients with heart failure of the present invention adopts the following technical solution:
[0006] An embodiment of the present invention provides a method for enhancing and optimizing cardiac magnetic resonance imaging of a patient with heart failure, the method comprising the following steps:
[0007] Collect several cardiac images of the patient at consecutive moments;
[0008] Determine the accuracy of the end time of each respiratory cycle according to the area of the heart image corresponding to the start and end time of each respiratory cycle, determine the actual end time of each respiratory cycle according to the accuracy of the end time of each respiratory cycle, determine the duration of each respiratory cycle according to the actual end time of each respiratory cycle, and thus determine the sample period according to the duration of each respiratory cycle;
[0009] The cardiac images are grouped according to the areas of the corresponding cardiac images at the same time in different sample periods to obtain a plurality of cardiac image groups;
[0010] Dividing the cardiac image into a number of pixel blocks, determining the difference between the pixel blocks in the cardiac image of the cardiac image group according to the distance relationship between the central pixel points of any two pixel blocks, the number of pixel blocks in the cardiac image and the grayscale value of the pixel points in each pixel block, and determining the weight of each pixel point in each pixel block of the cardiac image in the cardiac image group according to the difference between any pixel block in the cardiac image of the cardiac image group and other pixel blocks in the cardiac image and the grayscale characteristics of the pixel points in the pixel block;
[0011] The cardiac image is enhanced based on the weight of each pixel point in each pixel block of the cardiac image in each cardiac image group.
[0012] Furthermore, the accuracy of the end time of the respiratory cycle is determined according to the area of the heart image corresponding to the start and end time of each respiratory cycle, including the specific method of:
[0013]
[0014] In the formula, Indicates The accuracy of the end time of each breathing cycle; Indicates The heart image area at the end of a respiratory cycle, Indicates the heart image area at the end of the first respiratory cycle; Indicates the heart image area at the start of the second respiratory cycle; Indicates The heart image area at the start of each respiratory cycle; Represents the normalization function.
[0015] Furthermore, the cardiac images are grouped according to the areas of the cardiac images corresponding to the same time in different sample periods to obtain a plurality of cardiac image groups, including the specific method of:
[0016] According to the areas of the corresponding heart images at the same time in different sample periods, the possibility of the heart images at the same time in different sample periods being the same group is determined;
[0017] The possibility that the cardiac images at the same time in any two sample periods are in the same group is calculated, a second threshold is preset, and the cardiac images at the same time whose possibility of being in the same group is greater than the preset second threshold are recorded as the same group to obtain several cardiac image groups.
[0018] Furthermore, the specific calculation method of the possibility that the cardiac images at the same time in different sample periods are in the same group is:
[0019]
[0020] In the formula, Indicates The sample period and In the sample period The probability that the cardiac images at each moment are of the same group; Indicates In the sample period The cardiac image area at the moment; Indicates In the sample period The cardiac image area at the moment; Indicates In the sample period The cardiac image area at the moment; Indicates In the sample period The cardiac image area at the moment; Indicates taking the absolute value; Represents the normalization function.
[0021] Further, the determining of the difference between the pixel blocks in the cardiac images of the cardiac image group includes a specific method as follows:
[0022]
[0023] In the formula, Indicates In the cardiac image set Heart image a difference between a pixel block and other pixel blocks in the heart image; Indicates In the cardiac image set The number of pixel blocks in the heart image; Indicates In the cardiac image set Heart image The pixel block is related to the first pixel in the heart image. The Euclidean distance between the center pixels of a pixel block; Indicates In the cardiac image set Heart image The average gray value of the pixels in the pixel block; Indicates In the cardiac image set Heart image The average gray value of the pixels in the pixel block; Indicates taking the absolute value; Represents a natural constant.
[0024] Furthermore, the determining of the weight of each pixel point in each pixel block of the cardiac image in the cardiac image group includes the following specific methods:
[0025] determining a weight of each pixel block of the cardiac image in the cardiac image group according to a difference between any pixel block in the cardiac image of the cardiac image group and other pixel blocks in the cardiac image and the number of grayscale values of pixels in the pixel block in the cardiac image;
[0026] The obtained weight of each pixel block is used as the weight of each pixel point in the pixel block.
[0027] Furthermore, the specific calculation method of the weight of each pixel block of the cardiac image in the cardiac image group is:
[0028]
[0029] In the formula, Indicates In the cardiac image set Heart image The weight of the pixel block; Indicates In the cardiac image set a difference between a ℎth pixel block of a heart image and other pixel blocks in the heart image; Indicates In the cardiac image set The number of gray values of pixels in the ℎth pixel block of a heart image; Indicates Of all the cardiac images in the cardiac image group The average value of the number of gray values of pixels in a pixel block; Indicates Of all the cardiac images in the cardiac image group The average value of the gradient value of the pixel points in the pixel block; Represents the normalization function.
[0030] Furthermore, the cardiac image is enhanced based on the weight of each pixel point in each pixel block of the cardiac image in each cardiac image group, including the specific method of:
[0031]
[0032] In the formula, Indicates The first pixel block of all cardiac images in the cardiac image group The gray value of each pixel after enhancement; Indicates In the cardiac image set The first pixel in the ℎth pixel block of the heart image The weight of each pixel; Indicates The number of cardiac images in a cardiac image group; Indicates In the cardiac image set Heart image The pixel block The original gray value of each pixel;
[0033] An enhanced cardiac image is obtained according to the enhanced grayscale values of the pixels of all cardiac images in each cardiac image group.
[0034] Furthermore, the specific method of determining the actual end time of each respiratory cycle includes:
[0035] Preset the first threshold value, when the When the accuracy of the end time of the first respiratory cycle is greater than or equal to the preset first threshold, the The current end time of the breathing cycle is taken as the The actual end time of a respiratory cycle; When the accuracy of the end time of the first respiratory cycle is less than the preset first threshold and greater than 0, the The moment before the end of the first respiratory cycle is taken as the The actual end time of a respiratory cycle; When the accuracy of the end time of the first breathing cycle is less than or equal to 0, The next moment after the end of the first breathing cycle is taken as the The actual end time of a breathing cycle.
[0036] Furthermore, the method of determining the sample period according to the duration of each respiratory period includes:
[0037] According to the actual end time of each respiratory cycle, determine the duration of each respiratory cycle and The duration of a respiratory cycle is recorded as , get several Value, for all The values are constructed from small to large to form a sequence of breathing cycle durations. , count each The number of times the value appears will be the one with the highest number of occurrences. The number of breathing cycles corresponding to the value is taken as the sample cycle.
[0038] The beneficial effects of the technical solution of the present invention are: according to the area of the heart image corresponding to the start and end time of each respiratory cycle, the accuracy of the end time of the respiratory cycle is determined, according to the accuracy of the end time of each respiratory cycle, the actual end time of each respiratory cycle is determined, according to the actual end time of each respiratory cycle, the duration of each respiratory cycle is determined, and thus the sample period is determined according to the duration of each respiratory cycle, which is used to group the heart images, and the heart images are grouped according to the area of the heart image corresponding to the same time in different sample periods to obtain a number of heart image groups, and the heart images are divided into a number of pixel blocks, and according to the distance relationship between the central pixel points of any two pixel blocks, The number of pixel blocks in the cardiac image and the grayscale value of the pixel points in each pixel block are used to determine the difference between the pixel blocks in the cardiac image of the cardiac image group. The weight of each pixel point in each pixel block of the cardiac image in the cardiac image group is determined according to the difference between any pixel block in the cardiac image of the cardiac image group and other pixel blocks in the cardiac image and the grayscale characteristics of the pixel points in the pixel block. The method is used to determine the amount of cardiac information contained in each pixel point. Based on the weight of each pixel point in each pixel block of the cardiac image in each cardiac image group, the cardiac image is enhanced to eliminate motion artifacts in the cardiac magnetic resonance image and reduce the influence of motion artifacts on the judgment of the cardiac state, so that the image better presents the patient's cardiac state. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0040] Figure 1 The present invention is a flowchart of the steps of a cardiac magnetic resonance image enhancement optimization method for patients with heart failure. DETAILED DESCRIPTION
[0041] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation method, structure, characteristics and effects of a cardiac magnetic resonance imaging enhancement optimization method for heart failure patients proposed by the present invention are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.
[0042] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
[0043] The specific scheme of the cardiac magnetic resonance imaging enhancement and optimization method for heart failure patients provided by the present invention is described in detail below with reference to the accompanying drawings.
[0044] See also Figure 1 , which shows a flowchart of a method for enhancing and optimizing cardiac magnetic resonance imaging of a patient with heart failure provided by an embodiment of the present invention, the method comprising the following steps:
[0045] Step S001: Acquire a number of cardiac images of the patient at consecutive moments.
[0046] It should be noted that in order to enhance and optimize the cardiac magnetic resonance image of the patient, it is first necessary to obtain the cardiac magnetic resonance image of the patient and obtain the heart area in the magnetic resonance image.
[0047] Specifically, a patient is scanned using a magnetic resonance scanner to obtain cardiac magnetic resonance images of the patient at each moment within a preset time period, to obtain a number of continuous time-series cardiac magnetic resonance images, the cardiac magnetic resonance images are denoised and semantically segmented, and the heart area in the cardiac magnetic resonance images is obtained and recorded as a cardiac image.
[0048] It should be noted that, based on experience, the preset time period is five minutes, which can be adjusted according to actual conditions and is not specifically limited in this embodiment.
[0049] It should be noted that denoising and semantic segmentation are existing technologies and will not be described in detail here.
[0050] So far, a heart image is obtained through the above method.
[0051] Step S002, determines the accuracy of the end time of each respiratory cycle based on the area of the heart image corresponding to the start and end time of each respiratory cycle, determines the actual end time of each respiratory cycle based on the accuracy of the end time of each respiratory cycle, determines the duration of each respiratory cycle based on the actual end time of each respiratory cycle, and thus determines the sample period based on the duration of each respiratory cycle.
[0052] It should be noted that when using a magnetic resonance imaging device to collect a patient's cardiac image, the movement of the lungs and diaphragm caused by the patient's breathing causes the position of the heart to change continuously, resulting in motion artifacts in the magnetic resonance image and blurring it, affecting the doctor's interpretation of the image. In addition, the cardiac magnetic resonance images collected under different breathing states of the patient will show different characteristics. The patient's breathing state will directly affect the size of the heart in the collected magnetic resonance image. The patient's breathing usually has a certain regularity and shows periodic changes, and sometimes there will be a disorder in the respiratory frequency. Therefore, the preset time period can be divided into periods based on the area size of the heart image at different times and the patient's breathing cycle.
[0053] Specifically, first, one inhalation and exhalation of the patient is regarded as a respiratory cycle, and the accuracy of the end time of the respiratory cycle is determined according to the area of the heart image corresponding to the start and end time of each respiratory cycle.
[0054] As an embodiment, the specific calculation method of the accuracy of the end time of the breathing cycle is:
[0055]
[0056] In the formula, Indicates The accuracy of the end time of each breathing cycle; Indicates The heart image area at the end of a respiratory cycle, Indicates the heart image area at the end of the first respiratory cycle; Indicates the heart image area at the start of the second respiratory cycle; Indicates The heart image area at the start of each respiratory cycle; Represents the normalization function.
[0057] It should be noted that under normal circumstances, at the end of any cycle, the heart size will return to its minimum, and then continue to grow with the start of the next respiratory cycle. In two adjacent respiratory cycles, the heart size will be almost the same at the end of the previous respiratory cycle and the start of the next respiratory cycle. Indicates The reciprocal of the difference in the cardiac image area between the first respiratory cycle and the end of the first respiratory cycle. The smaller the difference between the heart image area at the end of the first respiratory cycle and the end of the first respiratory cycle, that is, The larger the value, the The greater the accuracy of the end time of a breathing cycle, It represents the difference in the cardiac image area between the end of the first respiratory cycle and the beginning of the second respiratory cycle. The smaller the value, the more similar the size of the heart image at the end of the first respiratory cycle and the start of the second respiratory cycle, which means the accuracy of the end of the first respiratory cycle is higher. Indicates The end of the breathing cycle and the The difference in heart area at the start of each respiratory cycle, The smaller the value, the The end of the breathing cycle and the The more similar the heart sizes are at the beginning of each respiratory cycle, the The higher the accuracy of the end time of a respiratory cycle, the The larger the value, The larger the value.
[0058] Then, the accuracy of the end time of all breathing cycles is calculated. The accuracy of the end time of each breathing cycle When The current end time of the breathing cycle is taken as the The actual end time of a respiratory cycle; The accuracy of the end time of each breathing cycle and When The moment before the end of the first respiratory cycle is taken as the The actual end time of a respiratory cycle; The accuracy of the end time of each breathing cycle When The next moment after the end of the first breathing cycle is taken as the The actual end time of a breathing cycle.
[0059] It should be noted that when The accuracy of the end time of each breathing cycle When The end time of each breathing cycle is accurate, so there is no need to adjust it; The accuracy of the end time of each breathing cycle and , explain After the end of the second respiratory cycle, the area of the heart image is still increasing, and the increase is greater than the increase in the first respiratory cycle, indicating that the current moment is already in a new respiratory cycle, so the first respiratory cycle is The moment before the end of the first respiratory cycle is taken as the The actual end time of a respiratory cycle; The accuracy of the end time of each breathing cycle , explain After the end of the first breathing cycle, the area of the heart image is still decreasing, indicating that the exhalation cycle has not yet ended. The next moment after the end of the first breathing cycle is taken as the The actual end time of a breathing cycle.
[0060] Finally, according to the actual end time of each respiratory cycle, the duration of each respiratory cycle is determined. The duration of a respiratory cycle is recorded as , get several Value, for all The values are constructed from small to large to form a sequence of breathing cycle durations. , count each The number of times the value appears will be the one with the highest number of occurrences. The number of breathing cycles corresponding to the value is taken as the sample cycle.
[0061] It should be noted that based on experience The value is 0.9, which can be adjusted according to actual conditions and is not specifically limited in this embodiment.
[0062] So far, the sample period is obtained through the above method.
[0063] Step S003: grouping the cardiac images according to the areas of the cardiac images corresponding to the same time in different sample periods to obtain a plurality of cardiac image groups.
[0064] It should be noted that the motion artifacts caused by breathing may cause the details of the magnetic resonance image to be lost. It is difficult for patients to maintain consistent breathing in all cycles, and there may be different situations such as fast exhalation and slow exhalation. Therefore, it is necessary to analyze the cardiac image performance at each moment in multiple cycles and group the cardiac images at each moment in all sample cycles.
[0065] Specifically, first, according to the areas of the corresponding heart images at the same time in different sample periods, the possibility that the heart images at the same time in different sample periods belong to the same group is determined.
[0066] As an embodiment, a specific method for calculating the possibility that the corresponding cardiac images at the same time in different sample periods belong to the same group is as follows:
[0067]
[0068] In the formula, Indicates The sample period and In the sample period The probability that the cardiac images at each moment are of the same group; Indicates In the sample period The cardiac image area at the moment; Indicates In the sample period The cardiac image area at the moment; Indicates In the sample period The cardiac image area at the moment; Indicates In the sample period The cardiac image area at the moment; Indicates taking the absolute value; Represents the normalization function.
[0069] It should be noted that Indicates The sample period and In the sample period The difference in the cardiac image area at the moment The smaller the value of The sample period and In the sample period The smaller the difference in the areas of the heart images at the time instant is, the more likely it is that the heart images are of the same group. Indicates In the sample period Heart images of the moment, and In the sample period Heart image of the moment and the Changes in heart images at all times, The larger the value, the Samples and In the sample period, Moment and The smaller the difference in the heart image area at each moment, the The sample period and In the sample period The greater the possibility that the heart images at the same moment are of the same group, The larger the value.
[0070] Then, the probability that the heart images at the same time in any two sample periods belong to the same group is calculated, and several value, preset the second threshold ,when When the cardiac images are recorded as the same group, a plurality of cardiac image groups are obtained.
[0071] It should be noted that based on experience The value is 0.9, which can be adjusted according to actual conditions and is not specifically limited in this embodiment.
[0072] So far, several cardiac image groups are obtained through the above method.
[0073] Step S004, dividing the cardiac image into a number of pixel blocks, determining the difference between the pixel blocks in the cardiac image of the cardiac image group according to the distance relationship between the central pixel points of any two pixel blocks, the number of pixel blocks in the cardiac image and the grayscale value of the pixel points in each pixel block, and determining the weight of each pixel point in each pixel block of the cardiac image in the cardiac image group according to the difference between any pixel block in the cardiac image of the cardiac image group and other pixel blocks in the cardiac image and the grayscale characteristics of the pixel points in the pixel block.
[0074] Step (4.1), dividing the cardiac image into a number of pixel blocks, and determining the difference between the pixel blocks in the cardiac image of the cardiac image group according to the distance relationship between the central pixel points of any two pixel blocks, the number of pixel blocks in the cardiac image, and the grayscale value of the pixel points in each pixel block.
[0075] It should be noted that the cardiac images in the cardiac image group are cardiac images with similar areas at the same time in different sample periods. Since the heart beats continuously, motion artifacts may appear at every moment, so it is necessary to analyze the performance of the cardiac images in the same cardiac image group.
[0076] Specifically, the preset heart image The heart image is divided into several pixel blocks by taking the pixel points as a whole, obtaining the Euclidean distance between the central pixel points of any two pixel blocks, and determining the difference between any pixel block in the heart image of the heart image group and other pixel blocks in the heart image according to the Euclidean distance between the central pixel points of any two pixel blocks, the number of pixel blocks in the heart image, and the grayscale value of the pixel points in each pixel block.
[0077] As an embodiment, a specific method for calculating the difference between any pixel block in a cardiac image of a cardiac image group and other pixel blocks in the cardiac image is as follows:
[0078]
[0079] In the formula, Indicates In the cardiac image set Heart image a difference between a pixel block and other pixel blocks in the heart image; Indicates In the cardiac image set The number of pixel blocks in the heart image; Indicates In the cardiac image set Heart image The pixel block is related to the first pixel in the heart image. The Euclidean distance between the center pixels of a pixel block; Indicates In the cardiac image set Heart image The average gray value of the pixels in the pixel block; Indicates In the cardiac image set Heart image The average gray value of the pixels in the pixel block; Indicates taking the absolute value; Represents a natural constant.
[0080] It should be noted that Indicates In the cardiac image set Heart image pixel blocks and the The average gray value difference of pixels in a pixel block is The larger the value, the greater the grayscale difference between the two pixel blocks, corresponding to The larger the value, Indicates In the cardiac image set The Euclidean distance between the center points of the pixel blocks of the heart image is The larger the value, the larger the Euclidean distance, corresponding to The larger the value, The larger the value, the In the cardiac image set Heart image The greater the difference between the pixel block and other pixel blocks in the heart image.
[0081] It should be noted that according to experience The value of is 9, which can be adjusted according to actual conditions and is not specifically limited in this embodiment.
[0082] Step (4.2), determining the weight of each pixel point in each pixel block of the cardiac image in the cardiac image group according to the difference between any pixel block in the cardiac image of the cardiac image group and other pixel blocks in the cardiac image and the grayscale characteristics of the pixel points in the pixel block.
[0083] It should be noted that the greater the difference between any pixel block in the cardiac image and other pixel blocks, the more likely the pixel block is a normal area or an abnormal area. Therefore, the weight of any pixel block needs to be determined based on the difference between pixel blocks at the same position in different cardiac images.
[0084] Specifically, first, the weight of each pixel block of the cardiac image in the cardiac image group is determined according to the difference between any pixel block in the cardiac image of the cardiac image group and other pixel blocks in the cardiac image and the number of grayscale values of pixels in the pixel block in the cardiac image.
[0085] As an embodiment, the specific calculation method of the weight of each pixel block of the cardiac image in the cardiac image group is:
[0086]
[0087] In the formula, Indicates In the cardiac image set Heart image The weight of the pixel block; Indicates In the cardiac image set a difference between a ℎth pixel block of a heart image and other pixel blocks in the heart image; Indicates In the cardiac image set The number of gray values of pixels in the ℎth pixel block of a heart image; Indicates Of all the cardiac images in the cardiac image group The average value of the number of gray values of pixels in a pixel block; Indicates Of all the cardiac images in the cardiac image group The average value of the gradient value of the pixel points in the pixel block; Represents the normalization function.
[0088] It should be noted that Indicates In the cardiac image set The number of grayscale values of pixels in the ℎth pixel block of the heart image is equal to the number of grayscale values of the The difference between the average values of the number of grayscale values of pixels in the ℎth pixel block in all cardiac images in the cardiac image group, The smaller the value, the In the cardiac image set Heart image The number of grayscale values of pixels in the pixel block is The smaller the difference between the average values of the number of grayscale values of pixels in the ℎth pixel block in all cardiac images in the same cardiac image group, the smaller the difference between the average values of the number of grayscale values of pixels in the ℎth pixel block in all cardiac images in the same cardiac image group. The smaller the difference in grayscale values of the pixel blocks, the The smaller, The smaller the value.
[0089] Then, the weight of each pixel block is taken as the weight of each pixel point in the pixel block, which is recorded as .
[0090] So far, the weight of each pixel point in each pixel block of the cardiac image in the cardiac image group is obtained through the above method.
[0091] Step S005 : enhancing the cardiac image based on the weight of each pixel point in each pixel block of the cardiac image in each cardiac image group.
[0092] It should be noted that after obtaining the weight of each pixel block of each cardiac image in each cardiac image group, the cardiac image can be enhanced according to the weight, and the size of the weight can represent the importance of the information in the pixel block.
[0093] Specifically, first, the grayscale value of each pixel before the cardiac image is enhanced is recorded as the original grayscale value, and the grayscale value of each pixel after enhancement is determined according to the weight of each pixel block of each cardiac image in each cardiac image group and the original grayscale value of each pixel in the pixel block.
[0094] As an embodiment, the specific method for calculating the enhanced grayscale value of each pixel in each pixel block of all cardiac images of each cardiac image group is as follows:
[0095]
[0096] In the formula, Indicates The first pixel block of all cardiac images in the cardiac image group The gray value of each pixel after enhancement; Indicates In the cardiac image set The first pixel in the ℎth pixel block of the heart image The weight of each pixel; Indicates The number of cardiac images in a cardiac image group; Indicates In the cardiac image set Heart image The pixel block The original gray value of each pixel.
[0097] It should be noted that Indicates The sum of the weights of the i-th pixel in the ℎ-th pixel block of all cardiac images in the cardiac image group, Indicates In the cardiac image set The first pixel in the ℎth pixel block of the heart image The gray value of the pixel after enhancement, The larger the value, the The larger the value of .
[0098] Then, an enhanced cardiac image is obtained according to the enhanced grayscale values of the pixels of all cardiac images in each cardiac image group.
[0099] At this point, this embodiment is completed.
[0100] It should be noted that the The model is only used to represent negative correlation and constrain the output of the model to be in In the specific implementation, it can be replaced by other models with the same purpose. This embodiment is only based on The model is described as an example without any specific limitation. is the input to the model.
[0101] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A cardiac magnetic resonance imaging enhancement optimization method for patients with heart failure, characterized in that: The method comprises the following steps: Collect several cardiac images of the patient at consecutive moments; Determine the accuracy of the end time of each respiratory cycle according to the area of the heart image corresponding to the start and end time of each respiratory cycle, determine the actual end time of each respiratory cycle according to the accuracy of the end time of each respiratory cycle, determine the duration of each respiratory cycle according to the actual end time of each respiratory cycle, and thus determine the sample period according to the duration of each respiratory cycle; The cardiac images are grouped according to the areas of the corresponding cardiac images at the same time in different sample periods to obtain a plurality of cardiac image groups; Dividing the cardiac image into a number of pixel blocks, determining the difference between the pixel blocks in the cardiac image of the cardiac image group according to the distance relationship between the central pixel points of any two pixel blocks, the number of pixel blocks in the cardiac image and the grayscale value of the pixel points in each pixel block, and determining the weight of each pixel point in each pixel block of the cardiac image in the cardiac image group according to the difference between any pixel block in the cardiac image of the cardiac image group and other pixel blocks in the cardiac image and the grayscale characteristics of the pixel points in the pixel block; enhancing the cardiac image based on the weight of each pixel point in each pixel block of the cardiac image in each cardiac image group; The method of determining the accuracy of the end time of the respiratory cycle according to the area of the heart image corresponding to the start and end time of each respiratory cycle includes the following specific methods: In the formula, A v Indicates the accuracy of the end time of the vth respiratory cycle; Sj v represents the heart image area at the end of the vth respiratory cycle, Sj1 represents the heart image area at the end of the first respiratory cycle; Sq2 represents the heart image area at the start of the second respiratory cycle; Sq v+1 represents the heart image area at the start of the v+1th respiratory cycle; norm() represents the normalization function; The specific method of determining the actual end time of each respiratory cycle includes: A first threshold is preset, and when the accuracy of the end time of the vth respiratory cycle is greater than or equal to the preset first threshold, the current end time of the vth respiratory cycle is taken as the actual end time of the vth respiratory cycle; when the accuracy of the end time of the vth respiratory cycle is less than the preset first threshold and greater than 0, the moment before the end time of the vth respiratory cycle is taken as the actual end time of the vth respiratory cycle; when the accuracy of the end time of the vth respiratory cycle is less than or equal to 0, the next moment after the end time of the vth respiratory cycle is taken as the actual end time of the vth respiratory cycle; The cardiac images are grouped according to the areas of the cardiac images corresponding to the same time in different sample periods to obtain a plurality of cardiac image groups, including the specific method of: According to the areas of the corresponding heart images at the same time in different sample periods, the possibility of the heart images at the same time in different sample periods being the same group is determined; Calculate the possibility that the cardiac images at the same time in any two sample periods are in the same group, preset a second threshold, and record the cardiac images at the same time whose possibility of being in the same group is greater than the preset second threshold as the same group, to obtain a plurality of cardiac image groups; The specific calculation method of the possibility that the cardiac images at the same time in different sample periods are in the same group is: In the formula, represents the possibility that the cardiac images at the t-th moment in the u-th sample period and the s-th sample period are the same group; S u,t represents the heart image area at the tth moment in the uth sample cycle; S s,t represents the cardiac image area at the tth moment in the sth sample cycle; S u,t+1 represents the heart image area at the t+1th moment in the uth sample period; S s,t+1 represents the heart image area at the t+1th moment in the sth sample period; | | represents the absolute value; norm() represents the normalization function; The method of determining the difference between the pixel blocks in the cardiac images of the cardiac image group includes: Where I(k) n,h represents the difference between the hth pixel block of the nth cardiac image in the kth cardiac image group and other pixel blocks in the cardiac image; H(k) n represents the number of pixel blocks in the nth cardiac image in the kth cardiac image group; O(k) n,h,x represents the Euclidean distance between the h-th pixel block of the n-th cardiac image in the k-th cardiac image group and the central pixel point of the x-th pixel block in the cardiac image; G(k) n,h represents the average gray value of the pixel points in the hth pixel block of the nth cardiac image in the kth cardiac image group; G(k) n,x represents the average gray value of the pixel in the xth pixel block of the nth cardiac image in the kth cardiac image group; || represents the absolute value; e represents a natural constant; The method of determining the weight of each pixel point in each pixel block of the cardiac image in the cardiac image group includes: for: determining a weight of each pixel block of the cardiac image in the cardiac image group according to a difference between any pixel block in the cardiac image of the cardiac image group and other pixel blocks in the cardiac image and the number of grayscale values of pixels in the pixel block in the cardiac image; The obtained weight of each pixel block is used as the weight of each pixel point in the pixel block; The specific calculation method of the weight of each pixel block of the cardiac image in the cardiac image group is: Where σ(k) n,h I(k) represents the weight of the hth pixel block of the nth cardiac image in the kth cardiac image group; n,h represents the difference between the hth pixel block of the nth cardiac image in the kth cardiac image group and other pixel blocks in the cardiac image; M(k) n,h represents the number of gray values of pixels in the h-th pixel block of the n-th cardiac image in the k-th cardiac image group; represents the average value of the number of gray values of the pixels in the h-th pixel block in all cardiac images in the k-th cardiac image group; represents the average value of the gradient values of the pixels in the hth pixel block in all cardiac images in the kth cardiac image group; norm[ ] represents the normalization function; The weight of each pixel point in each pixel block of the cardiac image in each cardiac image group is used to calculate the weight of each pixel point in each pixel block of the cardiac image. The specific methods for enhancement are as follows: In the formula, G(k) h,i represents the enhanced gray value of the i-th pixel in the h-th pixel block of all cardiac images in the k-th cardiac image group; σ(k) n,h,i represents the weight of the i-th pixel in the h-th pixel block of the n-th cardiac image in the k-th cardiac image group; N(k) represents the number of cardiac images in the k-th cardiac image group; G(k) n,h,i represents the original gray value of the i-th pixel in the h-th pixel block of the n-th cardiac image in the k-th cardiac image group; An enhanced cardiac image is obtained according to the enhanced grayscale values of the pixels of all cardiac images in each cardiac image group.
2. A cardiac magnetic resonance imaging enhancement and optimization method for patients with heart failure according to claim 1, characterized in that: The specific method of determining the sample period according to the duration of each respiratory period is as follows: According to the actual end time of each respiratory cycle, the duration of each respiratory cycle is determined, and the duration of the vth respiratory cycle is recorded as T v , get several T v value, for all T v The values are constructed from small to large to construct a breathing cycle duration sequence {T1, T2...T v }, count each T in the sequence v The number of times the value appears, the T with the largest number of occurrences will be v The number of breathing cycles corresponding to the value is taken as the sample cycle.
