A method and apparatus for correction of super-overflow of a heart blood pool for quantification of myocardial blood flow

Abstract

The application discloses a method and device for correcting super-overflow of a heart blood pool in myocardial blood flow quantification. A myocardial time activity curve, a left ventricular average myocardial time activity curve and a heart blood pool time activity curve are generated according to a dynamic target heart map. A super-overflow area of the heart blood pool is divided based on the myocardial time activity curve, and the myocardial time activity curve in the super-overflow area of the heart blood pool is corrected. The application can effectively determine the myocardial part affected by the super-overflow effect of the heart blood pool through decomposition and analysis of the myocardial time curve characteristics. Meanwhile, the super-overflow effect of the heart blood pool of the part is corrected before dynamic modeling, so that the nonlinear fitting degree of the myocardial time activity curve and the chamber model is improved in the dynamic calculation process, the correctness of the dynamic parameters generated from the nonlinear fitting process and the rationality of the distribution are ensured, and the accuracy of the myocardial blood flow quantification is improved.

Description

Technical Field

[0001] This invention relates to the field of cardiac nuclear medicine, and more specifically, to a method and apparatus for correcting super-overflow of cardiac blood pools for quantitative myocardial blood flow. Background Technology

[0002] In the field of nuclear cardiology, positron emission tomography (PET) or single photon emission computed tomography (SPECT) for quantifying myocardial blood flow is the commonly used method. The accuracy of myocardial blood flow quantification can be affected by various interfering factors, including the spillover effect of the blood pool on the myocardium. However, previous correction methods cannot effectively correct for all cases of the blood pool spillover effect, especially in the early stages after PET or SPECT perfusion drug injection. Due to the limited spatial resolution of PET or SPECT images, when the perfusion drug undergoes rapid filling and clearing cycles in the blood pool area, radiation from the blood pool will spill over into the myocardial region. Furthermore, the degree of spillover into the myocardium can increase significantly due to reduced myocardial uptake capacity. For myocardial regions with low drug uptake capacity, the spillover from the blood pool will create a blood-pool ultra-spillover effect. Traditional methods can only estimate the degree of cardiac blood pool overflow during kinetic modeling by introducing cardiac blood pool overflow parameters based on the kinetic relationship between the myocardial time-activity curve and the cardiac blood pool, combined with the limited resolution features of PET and SPECT images. This is achieved through nonlinear curve fitting. However, when excessive cardiac blood pool overflow leads to a super-overflow effect, the kinetic correlation between the myocardial time-activity curve and the cardiac blood pool is disrupted. The characteristics of the myocardial time-activity curve no longer conform to the expectations of the kinetic model, and the significant difference in their correlation makes effective nonlinear fitting impossible, potentially even leading to divergence. Forcibly modifying the original kinetic model by introducing additional correction parameters or correcting general cardiac blood pool overflow terms increases the instability of nonlinear fitting. While this might mitigate the impact of super-overflow effect on myocardial blood flow quantification in the super-overflow effect region, it could also cause the kinetic model to collapse in myocardial regions without super-overflow effects due to limitations in the applicability of correction parameters, resulting in more complex and unexpected results, ultimately leading to nonlinear fitting failure. Summary of the Invention

[0003] In view of this, the present invention provides a method and device for correcting the super overflow of the cardiac blood pool in quantitative myocardial blood flow, which solves the problem of low accuracy of nonlinear fitting results in existing quantitative myocardial blood flow modeling methods.

[0004] To achieve the above objectives, the following solution is proposed: A method for correcting cardiac blood pool overflow in myocardial blood flow quantification includes: Dynamic images are normalized and superimposed to form static images; A dynamic short axis diagram is generated based on the translation vector and rotation angle values ​​generated from static images in the cardiac coordinate system. The target pixel value is obtained by sampling the dynamic short axis map based on the 3D myocardial sampling matrix generated from the static short axis map; A dynamic bullseye image is generated by arranging the target pixel values ​​in a circular pattern using a bullseye image matrix. The myocardial time-activity curve, the left ventricular mean myocardial time-activity curve, and the cardiac blood pool time-activity curve are generated based on the dynamic target map. The myocardial blood pool super-overflow region was delineated based on the myocardial time-activity curve, and the myocardial time-activity curve within the super-overflow region of the myocardial blood pool was corrected. The corrected myocardial time-activity curve was fitted to the myocardial hemodynamic model, and myocardial blood flow was quantified based on the kinetic parameters.

[0005] Preferably, the process of image normalization and overlay of dynamic images includes: By arranging dynamic images using time and spatial coordinates, a super-stack image arrangement of dynamic frames and image volumes is formed, thus revealing the distribution of radiation signals in the body's organs. Based on the distribution of radiation signals, dynamic frames in which myocardial uptake tends to stabilize are extracted, and the normalization coefficient of each dynamic frame is calculated based on the time length of the dynamic frame. According to the spatial coordinates corresponding to each pixel, the pixel value is multiplied by the normalization coefficient of the corresponding dynamic frame and then superimposed to form a set of static images.

[0006] Preferably, the process of generating a dynamic minor axis diagram based on the translation vector and rotation angle values ​​generated from the static image in the cardiac coordinate system includes: The rotation axis in the static image was translated to a position close to the center of the myocardium, and the translation vector was recorded. The static image is rotated from the body coordinate system to the heart coordinate system by rotating the angle to obtain a static minor axis image, and the rotation angle value is recorded. The image volume of each dynamic frame of the dynamic image is translated according to the translation vector and rotated according to the rotation angle value to generate a dynamic short axis diagram.

[0007] Preferably, the process of sampling the dynamic short-axis diagram based on the 3D myocardial sampling matrix generated from the static short-axis diagram includes: Using the myocardial center of the static short axis diagram as the origin, the static short axis diagram is transformed from a rectangular coordinate system to a spherical coordinate system to generate a polar diagram of the static short axis. Find the maximum myocardial uptake value in the radial direction at each rotation angle in the polar diagram of the static short axis, and obtain the polar diagram of the maximum myocardial uptake value and the 3D myocardial sampling matrix. 3D myocardial sampling is performed on the dynamic short axis graph of each dynamic frame using a 3D myocardial sampling matrix to obtain the target pixel value.

[0008] Preferably, the process of arranging the target pixel values ​​in a circular pattern using the bullseye matrix includes: Based on the correspondence between the 3D sampling matrix and the 2D bullseye map, the rotation angles in the 3D sampling matrix are arranged in a ring on the 2D bullseye map and geometrically correlated to generate the bullseye map matrix. The target pixel values ​​sampled from each dynamic frame are arranged in a ring shape in the target image matrix to obtain the dynamic target image.

[0009] Preferably, the process of generating myocardial time-activity curves, left ventricular mean myocardial time-activity curves, and cardiac blood pool time-activity curves based on dynamic target maps includes: The dynamic target map is arranged by time and space coordinates, and a corresponding myocardial time-activity curve is generated for each set of spatial coordinates. The mean myocardial time-activity curve of the left ventricle is obtained by summing and averaging all the myocardial time-activity curves within the target area. Using the geometry of the myocardium as a reference, a 3D rectangular box was placed in the left atrium as the blood pool sampling box, and the corresponding blood pool time-activity curve was obtained by sampling. Preferably, the process of delineating the super-overflow region of the cardiac blood pool based on the myocardial time-activity curve includes: Each group of myocardial time-activity curves is decomposed into a combination of the left ventricular mean myocardial time-activity curve and the cardiac blood pool time-activity curve. The curves of each myocardial time-activity curve are combined and curve-fitted to obtain the combination coefficient; Determine whether the combination coefficient of the myocardial time-activity curve of each pixel is higher than the preset combination coefficient threshold. If the value is higher, then the target image position of that pixel is determined to have a super overflow effect of the blood pool.

[0010] Preferably, the process of correcting the myocardial time-activity curve within the super-overflow region of the cardiac blood pool includes: In the super-overflow region of the cardiac blood pool, reference values ​​and adjustment variables were determined based on the left ventricular mean myocardial time-activity curve and myocardial uptake of the myocardial time-activity curve, respectively. The adjustment coefficient for each pixel within the super overflow region of the cardiovascular pool is calculated based on the reference variable and the adjustment variable. The left ventricular mean myocardial time-activity curve is adjusted based on the adjustment coefficient of each pixel and then updated to the myocardial time-activity curve of that pixel.

[0011] Preferably, the method further includes: marking the blood flow status based on the quantitative results of myocardial blood flow through an established blood flow status system.

[0012] A cardiac blood pool super-overflow correction device for myocardial blood flow quantification includes: The image processing unit performs image normalization and overlay on dynamic images to form static images; The image conversion unit generates a dynamic short axis diagram based on the translation vector and rotation angle values ​​generated from the static image in the cardiac coordinate system. The image sampling unit samples the dynamic short axis image based on the 3D myocardial sampling matrix generated from the static short axis image to obtain the target pixel value; The image generation unit arranges the target pixel values ​​in a circular pattern using a bullseye matrix to generate a dynamic bullseye map. The curve generation unit generates myocardial time-activity curves, left ventricular mean myocardial time-activity curves, and cardiac blood pool time-activity curves based on the dynamic target map. The curve correction unit divides the super-overflow region of the cardiac blood pool based on the myocardial time-activity curve and corrects the myocardial time-activity curve within the super-overflow region of the cardiac blood pool. The data analysis unit fits the corrected myocardial time-activity curve with the myocardial hemodynamic model and quantifies myocardial blood flow based on the kinetic parameters.

[0013] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects: This invention effectively identifies myocardial regions affected by the super-overflow effect of the cardiac blood pool by decomposing and analyzing the characteristics of myocardial time-activity curves. Simultaneously, before kinetic modeling, the super-overflow effect of the cardiac blood pool in these regions is corrected, thereby improving the nonlinear fitting degree between the myocardial time-activity curve and the chamber model during kinetic calculations. This ensures the correctness and rationality of the kinetic parameters generated by the nonlinear fitting process, thus improving the accuracy of myocardial blood flow quantification.

[0014] This invention addresses the issue of cardiac blood pool super-overflow effect by decomposing, analyzing, and correcting the myocardial time-activity curve before kinetic modeling. This effectively avoids the uncertainty in the cardiac blood pool super-overflow effect region caused by introducing additional correction parameter sets to correct the cardiac blood pool super-overflow effect during kinetic modeling.

[0015] The method of this invention not only solves the problem of the super-overflow effect of the cardiac blood pool, but also retains the original characteristics of the kinetic model. It can maximize the original application scope of the kinetic model without increasing its complexity, and has high applicability. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0017] Figure 1 A flowchart of a method for correcting cardiac blood pool super-overflow for quantifying myocardial blood flow, provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the static image generation process provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the dynamic image conversion process provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the 3D sampling matrix generation process provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of dynamic target image generation provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the time-activity curve extraction process provided in an embodiment of the present invention; Figure 7 This is a schematic diagram of the extraction process of the super overflow effect region of the cardiac blood pool provided in an embodiment of the present invention; Figure 8 This is a schematic diagram of myocardial time-activity curve correction provided in an embodiment of the present invention; Figure 9 This is a comparison chart of the results before and after correction for super overflow of the cardiac blood pool under resting state, provided in an embodiment of the present invention. Figure 10 This is a comparison chart of the results before and after correction for super overflow of the cardiac blood pool under load, provided in an embodiment of the present invention. Figure 11 This is a schematic diagram of the myocardial blood flow state system provided in an embodiment of the present invention; Figure 12 A schematic diagram of the cardiac blood pool super overflow correction device for myocardial blood flow quantification provided in an embodiment of the present invention; Figure 13 The hardware structure block diagram of the cardiac blood pool super overflow correction device for myocardial blood flow quantification provided in the embodiments of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] First, combined Figure 1 The method for correcting cardiac blood pool super-overflow for myocardial blood flow quantification provided in the embodiments of the present invention is described below, such as... Figure 1 As shown, the process is as follows: Step S01: Normalize and overlay the dynamic images to form a static image.

[0020] Specifically, such as Figure 2 As shown, dynamic images are processed by normalization and overlay to produce a set of static images.

[0021] First, dynamic images (dynamic PET or SPECT images) are arranged in four dimensions, including time coordinates (t) and spatial coordinates (x, y, z), to form a hyperstack image arrangement of dynamic frames and image volumes, in order to show the distribution of radiation signals in the body organs after myocardial blood flow drug injection.

[0022] Then, based on the distribution of radiation signals, dynamic frames where myocardial uptake tends to stabilize are extracted to define the image volume stacking range of the dynamic frames. Furthermore, the normalization coefficient of each dynamic frame within the image stacking range is calculated based on the time length of the dynamic frames.

[0023] Finally, within the image overlay range, the image volume of each dynamic frame is multiplied by the pixel value and the normalization coefficient of the corresponding dynamic frame according to the spatial coordinates (x, y, z) of each pixel, and then overlaid to form a set of static images.

[0024] Step S02: Generate a dynamic short axis diagram based on the translation vector and rotation angle values ​​generated from the static image in the cardiac coordinate system.

[0025] Specifically, such as Figure 3As shown, the dynamic image is processed to generate a dynamic minor axis diagram. In the static image, the rotation axis is translated to the center of the myocardium. Then, the static image is rotated from the body coordinate system to the heart coordinate system by two rotation angles (θ, φ) to obtain a static minor axis image, and the translation vector and rotation angle values ​​are recorded. Finally, the translation vector and rotation angle values ​​are applied to the image volume of each dynamic frame of the dynamic image, translating the image volume of each dynamic frame according to the translation vector and rotating it according to the rotation angle values ​​to generate the dynamic minor axis diagram.

[0026] Step S03: Based on the 3D myocardial sampling matrix generated from the static short axis image, the dynamic short axis image is sampled to obtain the target pixel value.

[0027] Specifically, such as Figure 4 As shown, with the myocardial center of the static short axis plot as the origin, the static short axis plot is transformed from the rectangular coordinate system (x,y,z) to the spherical coordinate system (r,θ,φ) to generate a polar gram of the static short axis.

[0028] In the polar diagram of the static minor axis, for each rotation angle (θ, φ), the maximum myocardial uptake value is found in the radial direction, resulting in the polar diagram of the maximum myocardial uptake value and the 3D myocardial sampling matrix.

[0029] For each dynamic frame, the 3D myocardial sampling matrix is ​​applied to the dynamic short axis map to perform 3D myocardial sampling and obtain the corresponding target pixel value.

[0030] Step S04: Arrange the target pixel values ​​in a circular pattern using the target center map matrix to generate a dynamic target center map.

[0031] Specifically, such as Figure 5 As shown, based on the correspondence between the rotation angle (θ,φ) in the 3D sampling matrix and the 2D bullseye (r,θ), the rotation angle (θ,φ) in the 3D sampling matrix is ​​arranged in a ring in the 2D bullseye and geometrically correlated to generate the bullseye matrix.

[0032] For each dynamic frame, the target pixel values ​​sampled from each dynamic frame are arranged in a ring shape in the bullseye map using the bullseye map matrix to obtain the dynamic bullseye map.

[0033] Step S05: Generate myocardial time-activity curves, left ventricular mean myocardial time-activity curves, and cardiac blood pool time-activity curves based on the dynamic target map.

[0034] Specifically, the dynamic target image is arranged along three dimensions: time (t), spatial coordinates (x, y), and a corresponding myocardial time-activity curve is generated for each set of spatial coordinates (x, y). The acquisition time point corresponding to the PET or SPECT imaging after the injection of myocardial blood perfusion drugs represents the true time point of each dynamic frame.

[0035] The mean myocardial time-activity curve of the left ventricle is obtained by summing and averaging all the myocardial time-activity curves within the target area. Using the geometry of the myocardium as a reference, a 3D rectangular box was placed in the left atrium as the blood pool sampling box, and the corresponding blood pool time-activity curve was obtained. The acquisition time point corresponding to PET or SPECT imaging after myocardial blood flow perfusion drug injection represents the true time point of each dynamic frame. Figure 6 As shown, the dynamic target map reveals that in the transmural myocardial infarction region, myocardial uptake is higher than in the normal region in the early stage, and significantly lower than in the normal region in the later stage. Based on the time-activity curves, in the early stage, the peak myocardial uptake is significantly higher than the myocardial time-activity curve and the left ventricular mean time-activity curve in the normal region, exceeding 50% of the peak value of the blood pool time-activity curve. In the later stage, myocardial uptake is significantly lower than the myocardial time-activity curve and the left ventricular mean time-activity curve in the normal region, and is consistent with the blood pool time-activity curve, thus forming a blood pool super-overflow effect in the transmural myocardial infarction region.

[0036] Step S06: Divide the super-overflow region of the cardiac blood pool based on the myocardial time-activity curve, and correct the myocardial time-activity curve within the super-overflow region of the cardiac blood pool.

[0037] Specifically, each group of myocardial time-activity curves is decomposed into a combination of the left ventricular mean myocardial time-activity curve and the cardiac blood pool time-activity curve. Curve fitting is then performed on the combination of these myocardial time-activity curves to obtain the combination coefficient, TAC. 心肌 = a×TAC 心血池 +(1-a)×TAC 左室平均 , where 'a' is the combination coefficient.

[0038] In the target image, based on the target image position of the pixel corresponding to the myocardial time-activity curve of each group of (x, y) pixels, it is determined whether the combination coefficient of the myocardial time-activity curve of each pixel is higher than a preset combination coefficient threshold. If it is higher, then the target image position of that pixel has a cardiac blood pool super-overflow effect; if it is lower, then the target image position of that pixel does not have a cardiac blood pool super-overflow effect. Finally, the cardiac blood pool super-overflow region is delineated. Figure 7As shown, the myocardial time-activity curve is decomposed into a combination of the left ventricular mean myocardial time-activity curve and the cardiac blood pool time-activity curve. By applying a threshold of a combination coefficient, the cardiac blood pool super-overflow region is delineated. This delineated region is generally located in transmural myocardial infarction and marginal areas.

[0039] In the super-overflow region of the cardiac blood pool, reference values ​​and adjustment variables were determined based on the left ventricular mean myocardial time-activity curve and the myocardial uptake of the myocardial time-activity curve, respectively. Based on the left ventricular mean myocardial time-activity curve, myocardial uptake with a stable dynamic frame was used as the reference value, and myocardial uptake with a stable myocardial time-activity curve was used as the adjustment variable.

[0040] The adjustment coefficient for each pixel within the cardiac blood pool super-overflow region is calculated based on a reference variable and an adjustment variable (reference value / adjustment variable). This adjustment coefficient is then applied to adjust the left ventricular mean myocardial time-activity curve to generate a myocardial time-activity curve corrected for cardiac blood pool super-overflow. This corrected curve replaces the original myocardial time-activity curve for that pixel, thus restoring the correct myocardial time-activity curve. For example... Figure 8 As shown, in transmural myocardial infarction areas where there is a super-overflow effect of the cardiac pool, the myocardial time-activity curve of the transmural myocardial infarction is restored to its proper value after processing by the super-overflow correction step of the cardiac pool, thus obtaining the correct myocardial time-activity curve for this area. Step S07: Fit the corrected myocardial time-activity curve to the myocardial hemodynamic model, and quantify myocardial blood flow based on the kinetic parameters.

[0041] Specifically, based on the characteristics of drugs perfused into myocardial blood flow, a suitable kinetic model is selected, and the time-activity curve corrected for superoverflow in the myocardial blood pool is nonlinearly fitted to the kinetic model to obtain the nonlinear fit degree of superoverflow correction and the kinetic parameter K1 corrected for superoverflow in the myocardial blood pool along with other parameter groups (k2, k3, FBV). Based on the characteristics of the drugs perfused into myocardial blood flow, it is determined whether to further correct K1 using a myocardial extraction fraction correction curve associated with myocardial blood flow to obtain the myocardial blood flow value corrected for superoverflow in the myocardial blood pool. Drugs perfused into myocardial blood flow that do not require further myocardial extraction fraction correction include, but are not limited to: O15-water, NH3, and F18 labeled drugs. Drugs perfused into myocardial blood flow that require further myocardial extraction fraction correction include, but are not limited to: Rb-82, Tc99m-Sestamibi, and Tc99m-Tetrofosmin.

[0042] A myocardial hemodynamic model was constructed before fitting, and the model construction process is as follows: 1) Single-tissue double-chamber model: Based on the characteristics of drugs perfused in myocardial blood flow, the kinetic model considers the rate at which the drug enters cardiomyocytes from microvessels (K1) and the rate at which it returns to the microvessels after entering the cardiomyocytes (k2). The model formula is as follows:

[0043] For myocardial time-activity curve, The function is the cardiac blood pool time-activity curve, FBV is the cardiac blood pool overflow term, K1 is the first kinetic parameter, and k2 is the second kinetic parameter.

[0044] 2) Two-tissue three-chamber model: Based on the characteristics of drugs perfused in myocardial blood flow, the kinetic model considers the rate at which the drug enters cardiomyocytes from microvessels (K1), the rate at which it flows back to microvessels after entering cardiomyocytes (k2), and the rate at which the drug interacts with intracellular organs or functions after entering cardiomyocytes (k3). The model formula is as follows:

[0045] For myocardial time-activity curve, K1 represents the time-activity curve of the cardiac blood pool, FBV represents the overflow term of the cardiac blood pool, K1 represents the first kinetic parameter, k2 represents the second kinetic parameter, and k3 represents the third kinetic parameter.

[0046] Furthermore, in the cardiac blood pool super-overflow correction method for myocardial blood flow quantification according to embodiments of the present invention, after fitting the corrected myocardial time-activity curve with the myocardial hemodynamic model in step S07 and quantifying myocardial blood flow based on the kinetic parameters, the following steps may also be performed: Based on the quantitative results of myocardial blood flow, the blood flow status is indicated by the established blood flow status system.

[0047] Specifically, while ensuring the accuracy of myocardial blood flow quantification, a resting blood flow state system and a stress blood flow state system were established using the boundary values ​​of the myocardial blood flow range. Through multiple boundary values ​​of the myocardial blood flow range, resting myocardial blood flow and stress blood flow were each classified into excessively high, normal, slightly reduced, moderately reduced, and severely reduced states. Based on the myocardial blood flow quantification results, the current state of resting and stress myocardial blood flow for the test subject was determined and labeled.

[0048] Next, the feasibility of the cardiac blood pool super-overflow correction method for myocardial blood flow quantification is verified through comparative experiments in this embodiment of the invention. The experiments are as follows: Taking the infusion of NH3 into myocardial blood flow under resting conditions as an example, a suitable two-tissue three-compartment model for NH3 was used for curve fitting. The time-activity curve corrected for super-overflow in the cardiac blood pool was nonlinearly fitted to the kinetic model to obtain the nonlinearity of fit after super-overflow correction and the kinetic parameter K1 corrected for super-overflow in the cardiac blood pool along with other parameter groups (k2, k3, FBV). Simultaneously, the uncorrected time-activity curve was nonlinearly fitted to the kinetic model to obtain the uncorrected nonlinearity of fit and the uncorrected kinetic parameter K1 along with other parameter groups (k2, k3, FBV).

[0049] The K1 value corrected for superoverflow from the cardiac blood pool is adjusted to obtain the myocardial blood flow value corrected for superoverflow from the cardiac blood pool. The uncorrected K1 value is then adjusted for myocardial uptake fraction to obtain the uncorrected myocardial blood flow value. For example... Figure 9 As shown, in the region affected by the cardiac blood pool superoverflow without correction, the nonlinear fit is low. K1 is significantly overestimated, and its distribution is disrupted compared to myocardial perfusion. K2 exhibits both severe overestimation and a disrupted distribution, as does K3, which also shows significant overestimation and a disrupted distribution. FBV is also significantly overestimated. For NH3, K1 represents myocardial blood flow, exhibiting the same degree of overestimation and disrupted distribution as K1. After correction for cardiac blood pool superoverflow, the nonlinear fit is further improved, effectively resolving the K1 bias and repairing the disrupted distribution. This brings the K1 distribution into consistency with the myocardial perfusion distribution, thereby improving the accuracy and rationality of K1's distribution. Simultaneously, it effectively reduces the overestimation of K2, K3, and FBV, improving the overall distribution rationality.

[0050] Taking Tc99m-Sestamibi myocardial perfusion under load as an example, a single-tissue dual-chamber model of suitable Tc99m-Sestamibi was used for curve fitting. The above experimental fitting process was repeated, and the experimental results are as follows: Figure 10As shown, in the region affected by cardiac blood pool superoverflow without correction, although the nonlinear fit is acceptable, K1 is significantly overestimated, and the rationality of the K1 distribution is compromised compared to myocardial perfusion. k2 also shows severe overestimation and a disrupted distribution, and FBV also shows significant overestimation and an unreasonable distribution. For Tc99m-Sestamibi, K1 needs to be corrected using the uptake fraction to convert it into myocardial blood flow, and the converted myocardial blood flow shows overestimation and a disrupted distribution. After correction for cardiac blood pool superoverflow, the nonlinear fit is further improved, effectively resolving the overestimation of K1 and repairing the disrupted K1 distribution, making the K1 distribution consistent with the myocardial perfusion distribution. This improves the accuracy and rationality of K1 distribution, while effectively reducing the overestimation of k2 and FBV and improving the distribution rationality. After conversion into myocardial blood flow through correction using K1 and the uptake fraction, compared to the uncorrected version, the myocardial blood flow shows an effective reduction in overestimation and improved distribution rationality.

[0051] In summary, the cardiac blood pool super-overflow correction method for myocardial blood flow quantification of the present invention improves the accuracy of myocardial blood flow quantification.

[0052] Next, the application process of the cardiac blood pool super-overflow correction method for myocardial blood flow quantification in this embodiment of the invention is described as follows: 1) Under resting physiological conditions, dynamic PET or SPECT images undergo the following steps: dynamic image normalization and overlay, dynamic short-axis plot generation, 3D myocardial sampling, dynamic target plot generation, time-activity curve generation, time-activity curve analysis, cardiac pool super-overflow correction, kinetic modeling, myocardial blood flow quantification, and pre- and post-correction comparison. This allows for resting blood flow measurement under resting conditions and assessment of the reduction in myocardial blood flow deviation. The specific implementation process of each step is described in the aforementioned embodiment and will not be repeated here.

[0053] 2) Under physiological stress, the following steps are performed on dynamic PET or SPECT images: dynamic image normalization and overlay, dynamic short-axis plot generation, 3D myocardial sampling, dynamic target plot generation, time-activity curve generation, time-activity curve analysis, cardiac blood pool super-overflow correction, kinetic modeling, myocardial blood flow quantification, and pre- and post-correction comparison. These steps are used to measure blood flow under stress and assess the degree of reduction in myocardial blood flow deviation. The specific implementation process of each step is described in the aforementioned embodiment and will not be repeated here.

[0054] 3) While ensuring the accuracy of myocardial blood flow quantification, a resting blood flow state system and a stress blood flow state system were established by defining the boundaries of the myocardial blood flow range. For example... Figure 11As shown, by using the threshold values ​​of multiple myocardial blood flow ranges, resting myocardial blood flow and overloaded blood flow are respectively divided into excessively high state, normal state, slightly reduced state, moderately reduced state and severely reduced state.

[0055] The following describes the cardiac blood pool super-overflow correction device for myocardial blood flow quantification provided in the embodiments of the present invention. The cardiac blood pool super-overflow correction device for myocardial blood flow quantification described below can be referred to in correspondence with the cardiac blood pool super-overflow correction method for myocardial blood flow quantification described above.

[0056] Combination Figure 12 This paper introduces a cardiac blood pool super-overflow correction device for quantifying myocardial blood flow, such as... Figure 12 As shown, the cardiac blood pool super-overflow correction device for myocardial blood flow quantification may include: Image processing unit 100 performs image normalization and superposition on dynamic images to form static images; Image conversion unit 200 generates dynamic short axis diagrams based on translation vectors and rotation angles generated from static images in cardiac coordinate system; Image sampling unit 300 samples the dynamic short axis image based on the 3D myocardial sampling matrix generated from the static short axis image to obtain the target pixel value; The image generation unit 400 arranges the target pixel values ​​in a circular pattern using a bullseye matrix to generate a dynamic bullseye map. Curve generation unit 500 generates myocardial time-activity curves, left ventricular mean myocardial time-activity curves, and cardiac blood pool time-activity curves based on dynamic target maps. The curve correction unit 600 divides the super-overflow region of the cardiac blood pool based on the myocardial time-activity curve and corrects the myocardial time-activity curve within the super-overflow region of the cardiac blood pool. The data analysis unit 700 fits the corrected myocardial time-activity curve with the myocardial hemodynamic model and performs myocardial blood flow quantification based on the kinetic parameters.

[0057] The cardiac blood pool super overflow correction device for myocardial blood flow quantification provided in this embodiment of the invention can be applied to cardiac blood pool super overflow correction equipment for myocardial blood flow quantification. Figure 13 The hardware block diagram of the cardiac blood pool super overflow correction device for myocardial blood flow quantification is shown. (Refer to...) Figure 13 The hardware structure of the device may include: at least one processor 1, at least one communication interface 2, at least one memory 3, and at least one communication bus 4; In this embodiment of the invention, the number of processor 1, communication interface 2, memory 3, and communication bus 4 is at least one, and processor 1, communication interface 2, and memory 3 communicate with each other through communication bus 4. Processor 1 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. Memory 3 may include high-speed RAM, and may also include non-volatile memory, such as at least one disk storage device; The memory stores a program, which the processor can call. The program is used to implement the various processing steps in the aforementioned cardiac blood pool super overflow correction scheme for myocardial blood flow quantification.

[0058] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0059] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0060] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for correcting cardiac blood pool super-overflow in quantitative myocardial blood flow, characterized in that, include: Dynamic images are normalized and superimposed to form static images; A dynamic short axis diagram is generated based on the translation vector and rotation angle values ​​generated from static images in the cardiac coordinate system. The target pixel value is obtained by sampling the dynamic short axis map based on the 3D myocardial sampling matrix generated from the static short axis map; A dynamic bullseye image is generated by arranging the target pixel values ​​in a circular pattern using a bullseye image matrix. The myocardial time-activity curve, the left ventricular mean myocardial time-activity curve, and the cardiac blood pool time-activity curve are generated based on the dynamic target map. The myocardial blood pool super-overflow region was delineated based on the myocardial time-activity curve, and the myocardial time-activity curve within the super-overflow region of the myocardial blood pool was corrected. The corrected myocardial time-activity curve was fitted to the myocardial hemodynamic model, and myocardial blood flow was quantified based on the kinetic parameters. The process of delineating the super-overflow region of the cardiac blood pool based on the myocardial time-activity curve includes: Each group of myocardial time-activity curves is decomposed into a combination of the left ventricular mean myocardial time-activity curve and the cardiac blood pool time-activity curve. The curves of each myocardial time-activity curve are combined and curve-fitted to obtain the combination coefficient; Determine whether the combination coefficient of the myocardial time-activity curve of each pixel is higher than the preset combination coefficient threshold. If it is higher, the target image position of that pixel is determined to have a super overflow effect of the blood pool; The process of correcting the myocardial time-activity curve within the super-overflow region of the cardiac blood pool includes: In the super-overflow region of the cardiac blood pool, reference values ​​and adjustment variables were determined based on the left ventricular mean myocardial time-activity curve and myocardial uptake of the myocardial time-activity curve, respectively. Based on the left ventricular mean myocardial time-activity curve, myocardial uptake with a stable dynamic frame was used as the reference value, and myocardial uptake with a stable myocardial time-activity curve was used as the adjustment variable. The adjustment coefficient for each pixel within the super-overflow region of the cardiovascular pool is calculated based on the reference variable and the adjustment variable; the adjustment coefficient is obtained by comparing the reference variable and the adjustment variable. The left ventricular mean myocardial time-activity curve is adjusted based on the adjustment coefficient of each pixel and then updated to the myocardial time-activity curve of that pixel.

2. The method for correcting cardiac blood pool super-overflow for myocardial blood flow quantification according to claim 1, characterized in that, The process of normalizing and overlaying dynamic images includes: By arranging dynamic images using time and spatial coordinates, a super-stack image arrangement of dynamic frames and image volumes is formed, thus revealing the distribution of radiation signals in the body's organs. Based on the distribution of radiation signals, dynamic frames in which myocardial uptake tends to stabilize are extracted, and the normalization coefficient of each dynamic frame is calculated based on the time length of the dynamic frame. According to the spatial coordinates corresponding to each pixel, the pixel value is multiplied by the normalization coefficient of the corresponding dynamic frame and then superimposed to form a set of static images.

3. The method for correcting cardiac blood pool super-overflow for myocardial blood flow quantification according to claim 1, characterized in that, The process of generating a dynamic minor axis diagram based on translation vectors and rotation angles generated from static images in the cardiac coordinate system includes: The rotation axis in the static image was translated to a position close to the center of the myocardium, and the translation vector was recorded. The static image is rotated from the body coordinate system to the heart coordinate system by rotating the angle to obtain a static minor axis image, and the rotation angle value is recorded. The image volume of each dynamic frame of the dynamic image is translated according to the translation vector and rotated according to the rotation angle value to generate a dynamic short axis diagram.

4. The method for correcting cardiac blood pool super-overflow for myocardial blood flow quantification according to claim 3, characterized in that, The process of sampling dynamic short-axis plots based on a 3D myocardial sampling matrix generated from static short-axis plots includes: Using the myocardial center of the static short axis diagram as the origin, the static short axis diagram is transformed from a rectangular coordinate system to a spherical coordinate system to generate a polar diagram of the static short axis. Find the maximum myocardial uptake value in the radial direction at each rotation angle in the polar diagram of the static short axis, and obtain the polar diagram of the maximum myocardial uptake value and the 3D myocardial sampling matrix. 3D myocardial sampling is performed on the dynamic short axis graph of each dynamic frame using a 3D myocardial sampling matrix to obtain the target pixel value.

5. The method for correcting cardiac blood pool super-overflow for myocardial blood flow quantification according to claim 1, characterized in that, The process of arranging target pixel values ​​in a circular pattern using a bullseye matrix includes: Based on the correspondence between the 3D sampling matrix and the 2D bullseye map, the rotation angles in the 3D sampling matrix are arranged in a ring on the 2D bullseye map and geometrically correlated to generate the bullseye map matrix. The target pixel values ​​sampled from each dynamic frame are arranged in a ring shape in the target image matrix to obtain the dynamic target image.

6. The method for correcting cardiac blood pool super-overflow for myocardial blood flow quantification according to claim 1, characterized in that, The process of generating myocardial time-activity curves, left ventricular mean myocardial time-activity curves, and cardiac blood pool time-activity curves based on dynamic target maps includes: The dynamic target map is arranged by time and space coordinates, and a corresponding myocardial time-activity curve is generated for each set of spatial coordinates. The mean myocardial time-activity curve of the left ventricle is obtained by summing and averaging all the myocardial time-activity curves within the target area. Using the geometry of the myocardium as a reference, a 3D rectangular box was placed in the left atrium as the blood pool sampling box, and the corresponding blood pool time-activity curve was obtained by sampling.

7. The method for correcting cardiac blood pool super-overflow for myocardial blood flow quantification according to claim 1, characterized in that, Also includes: Based on the quantitative results of myocardial blood flow, the blood flow status is indicated by the established blood flow status system.

8. A cardiac blood pool super-overflow correction device for quantifying myocardial blood flow, characterized in that, include: The image processing unit performs image normalization and overlay on dynamic images to form static images; The image conversion unit generates a dynamic short axis diagram based on the translation vector and rotation angle values ​​generated from the static image in the cardiac coordinate system. The image sampling unit samples the dynamic short axis image based on the 3D myocardial sampling matrix generated from the static short axis image to obtain the target pixel value; The image generation unit arranges the target pixel values ​​in a circular pattern using a bullseye matrix to generate a dynamic bullseye map. The curve generation unit generates myocardial time-activity curves, left ventricular mean myocardial time-activity curves, and cardiac blood pool time-activity curves based on the dynamic target map. The curve correction unit divides the super-overflow region of the cardiac blood pool based on the myocardial time-activity curve and corrects the myocardial time-activity curve within the super-overflow region of the cardiac blood pool. The data analysis unit fits the corrected myocardial time-activity curve with the myocardial hemodynamic model and performs myocardial blood flow quantification based on the kinetic parameters. The process of delineating the super-overflow region of the cardiac blood pool based on the myocardial time-activity curve includes: Each group of myocardial time-activity curves is decomposed into a combination of the left ventricular mean myocardial time-activity curve and the cardiac blood pool time-activity curve. The curves of each myocardial time-activity curve are combined and curve-fitted to obtain the combination coefficient; Determine whether the combination coefficient of the myocardial time-activity curve of each pixel is higher than the preset combination coefficient threshold. If it is higher, the target image position of that pixel is determined to have a super overflow effect of the blood pool; The process of correcting the myocardial time-activity curve within the super-overflow region of the cardiac blood pool includes: In the super-overflow region of the cardiac blood pool, reference values ​​and adjustment variables were determined based on the left ventricular mean myocardial time-activity curve and myocardial uptake of the myocardial time-activity curve, respectively. The adjustment coefficient for each pixel within the super overflow region of the cardiovascular pool is calculated based on the reference variable and the adjustment variable. The left ventricular mean myocardial time-activity curve is adjusted based on the adjustment coefficient of each pixel and then updated to the myocardial time-activity curve of that pixel.

Images