Cellular functional imaging method for detecting difference in glucose metabolism after pancreatic and renal combined transplantation

Abstract

The present application relates to the technical field of cell imaging detection, and particularly relates to a cell function imaging detection method for postoperative glucose metabolism difference of pancreas-kidney combined transplantation, comprising: acquiring weighted images of a patient under a plurality of b values, performing image segmentation and grid division on the weighted images to obtain a plurality of pancreas voxels and kidney reference voxels, determining perfusion fractions and diffusion coefficients of unit voxels according to variation degrees of corresponding signal intensities of the unit voxels under different degrees of b values, so as to determine matching relationships between the pancreas voxels and the kidney reference voxels; and using distribution conditions of the perfusion fractions and the diffusion coefficients of the kidney reference voxels and combining the diffusion coefficients of the kidney reference voxels to calculate effective cell retention rates and cell function efficiencies, so as to determine a prognosis grade of the patient and provide a risk prompt. The present application improves the cell function evaluation effect after the pancreas-kidney combined transplantation.

Description

Technical Field

[0001] This invention relates to the field of cell imaging detection technology, specifically to a cell function imaging detection method for differences in glucose metabolism after pancreas-kidney combined transplantation. Background Technology

[0002] Combined pancreas-kidney transplantation is a primary treatment for end-stage diabetic nephropathy. The postoperative glucose metabolism status of the transplanted pancreas directly determines whether the recipient can be weaned off exogenous insulin. In clinical practice, transplanted pancreas often exhibits early functional impairment due to ischemia-reperfusion injury or rejection. This impairment is typically caused by two pathological mechanisms: irreversible ischemic necrosis due to microcirculatory disruption, and reversible functional inhibition due to cellular edema. Clinically, there is an urgent need for non-invasive differentiation between these two states to guide decisions on anticoagulation or anti-inflammatory therapy.

[0003] In post-pancreas-kidney transplant functional monitoring, the core deficiency of existing imaging techniques lies in their inability to achieve non-invasive and precise classification of the graft's glucose metabolism status, particularly in the difficulty of specifically distinguishing between reversible cellular function suppression and irreversible tissue necrosis. This problem stems from the inherent limitations of conventional methods such as magnetic resonance diffusion-weighted imaging in clinical applications: their signals are easily affected by systemic hemodynamic fluctuations, resulting in a loss of local specificity; furthermore, the lack of endogenous reference points in pancreatic tissue leads to a lack of quantitative assessment of microstructural damage, thus failing to provide reliable decision-making basis for clinical medicine. Summary of the Invention

[0004] This invention provides a cellular function imaging method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation, in order to solve the existing problem that existing imaging techniques cannot non-invasively and accurately distinguish between reversible cellular function inhibition and irreversible tissue necrosis of the transplanted pancreas due to interference from systemic hemodynamics and lack of endogenous reference, resulting in poor quantitative assessment of cellular function.

[0005] The cellular functional imaging method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation of the present invention adopts the following technical solution: One embodiment of the present invention provides a cellular functional imaging method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation, the method comprising the following steps: Obtain patients in several types A weighted image with values, where each position in the weighted image corresponds to a value in each... Signal strength at the specified value; By segmenting the weighted image and dividing it into grids, several pancreas voxels and kidney reference voxels are obtained, collectively referred to as unit voxels; based on the different degrees of influence of any unit voxel... The perfusion fraction and diffusion coefficient of the unit voxel are determined by the degree of change of the corresponding signal intensity under the value. Combined with the discrete levels corresponding to the perfusion fraction and diffusion coefficient of the kidney reference voxel, the matching relationship between the pancreatic voxel and the kidney reference voxel is determined based on the distance between the pancreatic voxel and the kidney reference voxel with respect to the perfusion fraction and diffusion coefficient. The distribution of perfusion fraction and diffusion coefficient of renal reference voxels that are matched with all pancreatic voxels within a preset spatial range is used, and combined with the diffusion coefficient of renal reference voxels, to calculate the patient's effective cell retention rate and cell function efficiency. Several outcome levels were set, and the patient's outcome level was determined by the effective cell retention rate and cell function efficiency, and risk warnings were given.

[0006] Optionally, the specific method for obtaining the matching relationship between the pancreatic voxel and the kidney reference voxel is as follows: Based on the discrete values ​​of perfusion fraction and diffusion coefficient for all kidney reference voxels, and combined with the differences in perfusion fraction and diffusion coefficient between pancreatic voxels and kidney reference voxels, a state difference matrix is ​​obtained. Matching analysis was performed on pancreatic voxels and kidney reference voxels based on the state difference matrix to obtain the matching relationship between pancreatic voxels and kidney reference voxels.

[0007] Optionally, the specific method for obtaining the state difference matrix is ​​as follows: Calculate the variance of all kidney reference voxels in the perfusion fraction dimension, denoted as the reference perfusion dispersion; similarly, calculate the variance of all kidney reference voxels in the diffusion coefficient dimension, denoted as the reference diffusion dispersion. A combination of any pancreatic voxel and any kidney reference voxel is denoted as a voxel combination. By combining the reference system perfusion dispersion and the reference system diffusion dispersion, the weighted distance between the corresponding pancreatic voxel and kidney reference voxel in the voxel combination is calculated, thereby obtaining the state difference matrix formed by the weighted distances corresponding to all voxel combinations.

[0008] Optionally, the specific method for performing matching analysis between pancreatic voxels and kidney reference voxels based on the state difference matrix to obtain the matching relationship between pancreatic voxels and kidney reference voxels includes: Using the pancreas voxel set and the kidney reference voxel set as the two vertex sets of a weighted bipartite graph, and the state difference matrix as the edge weight matrix, the minimum weight matching problem of the weighted bipartite graph is solved by the maximum matching algorithm to obtain the kidney reference voxel that has a matching relationship with any pancreas voxel.

[0009] Optionally, the specific method for obtaining the effective cell retention rate is as follows: The perfusion fraction and diffusion coefficient of any unit voxel are combined to form a vector, which is denoted as the physiological state vector of the corresponding unit voxel; the feature neighborhood set of any pancreatic voxel is obtained, and the feature neighborhood set contains the physiological state vectors of several kidney reference voxels; the local homogeneity destruction index of the corresponding pancreatic voxel is calculated according to the distribution of the physiological state vectors in the feature neighborhood set. Based on the distribution of perfusion fraction and local homogeneity destruction index of all pancreatic protons, pancreatic protons were screened, and the effective cell retention rate of the patient was obtained based on the difference in quantity before and after screening.

[0010] Optionally, the specific method for obtaining the feature neighborhood set of any pancreatic voxel is as follows: Obtain the geometric center coordinates of each pancreatic voxel in the anatomical alignment reference frame, and denote them as the anatomical centroid coordinates; Any pancreatic voxel is designated as the target pancreatic voxel. The anatomical centroid coordinates of each pancreatic voxel are obtained. Based on the Euclidean distance between the target pancreatic voxel and the anatomical centroid coordinates of the pancreatic voxels in the pancreatic dataset, all pancreatic voxels in the pancreatic dataset whose Euclidean distance from the target pancreatic voxel does not exceed the preset neighborhood radius are obtained and designated as the neighboring pancreatic voxels of the target pancreatic voxel. The set of all neighboring pancreatic voxels of the target pancreatic voxel is denoted as the anatomical neighborhood set of the target pancreatic voxel. For each neighboring pancreatic voxel in the anatomical neighborhood set, the set of physiological state vectors of the kidney reference voxels that have matching relationships with all neighboring pancreatic voxels is obtained and denoted as the feature neighborhood set of the target pancreatic voxel.

[0011] Optionally, the specific method for calculating the local homogeneity disruption index of the corresponding pancreatic voxel based on the distribution of physiological state vectors in the feature neighborhood set includes: Using irrigation fraction as the horizontal axis and diffusion coefficient as the vertical axis, a two-dimensional rectangular coordinate system is established and denoted as the physiological state coordinate system. The elements in the physiological state vector are used as the coordinates of the corresponding coordinate axes in the physiological state coordinate system. This yields the data points of all kidney reference voxels in the feature neighborhood set of the target pancreatic voxel in the physiological state coordinate system, forming a corresponding scatter plot, denoted as the feature neighborhood scatter plot of the target pancreatic voxel. Obtain the geometric center of all data points in the feature neighborhood scatter plot, and denote it as the feature center point of the target pancreatic somatic cell. Based on the Euclidean distance between each data point in the feature neighborhood scatter plot and the feature center point, and using the perfusion dispersion and diffusion dispersion of the reference system as weights, calculate the local homogeneity destruction index of the target pancreatic somatic cell.

[0012] Optionally, the screening of pancreatic protons based on the distribution of perfusion fraction and local homogeneity disruption index of all pancreatic protons includes the following specific methods: The perfusion fraction and local homogeneity destruction index of all pancreatic protons were statistically analyzed. Based on the distribution of the perfusion fraction and local homogeneity destruction index of pancreatic protons, the boundary parameters were determined. The boundary parameters include the perfusion loss boundary parameter and the structural disorder boundary parameter. Iterate through all pancreatic somatic cells and denote the pancreatic somatic cells that satisfy the conditions of perfusion fraction being less than the perfusion loss threshold parameter and local homogeneity destruction index being greater than the structural disorder threshold parameter as ineffective pancreatic somatic cells.

[0013] Optionally, the specific method for obtaining the cell functional efficiency is as follows: The set of all pancreatic voxels except for the ineffective pancreatic voxels is denoted as the pancreatic survival set; the diffusion coefficient of the kidney reference voxel that has a matching relationship with any pancreatic voxel is denoted as the matching diffusion coefficient of the pancreatic voxel. The difference between the diffusion coefficient of any pancreatic voxel in the pancreatic survival set and the matching diffusion coefficient is denoted as the diffusion difference of the corresponding pancreatic voxel. The average diffusion difference of all pancreatic voxels in the pancreatic survival set is used as the edema offset. Based on the edema offset and the dispersion of the diffusion coefficient of the kidney reference voxel, the patient's cell function efficiency is calculated.

[0014] Optionally, the specific method for obtaining the pancreatic voxel and the kidney reference voxel is as follows: In the anatomical alignment reference frame, the region of interest (ROI) of the transplanted pancreas in an arbitrary weighted image is obtained using an image segmentation algorithm, denoted as the first ROI. The first ROI is then divided into regions with side lengths of... From several cubic meshes, extract all meshes that are completely located within the first region of interest and denote them as pancreatic voxels, thus obtaining... One pancreatic somatic cell; among which These are the preset mesh parameters; The region of interest (ROI) of the transplanted kidney cortex in an arbitrary weighted image is obtained using an image segmentation algorithm and denoted as the second ROI. The second ROI is then divided into regions with sides of length [missing information]. Several cubic grids were formed, and samples were extracted from them using a uniform sampling method. A cubic grid, denoted as the kidney reference voxel.

[0015] The beneficial effects of the technical solution of this invention are as follows: By constructing a dynamic physiological reference system based on homologous organs and globally optimized matching, specific restoration of the local microcirculatory state of the graft is achieved. Compared with conventional detection methods that rely on absolute signal values, this embodiment of the invention uses the transplanted kidney as a real-time physiological internal standard. Through voxel matching, it offsets systemic signal fluctuations in the recipient caused by postoperative stress, drug effects, or changes in blood volume. This allows the detection results to reflect the local pathological changes of the graft itself, avoiding false positives caused by systemic factors. Furthermore, this embodiment of the invention, by utilizing the distribution of perfusion fraction and diffusion coefficient of kidney reference voxels that are matched with all pancreatic voxels within a preset spatial range, and combining this with the diffusion coefficient of the kidney reference voxels for analysis, can clearly distinguish whether the signal abnormality is caused by physical disintegration of the microvascular bed or by cellular edema. By distinguishing between loss of reserves and decreased efficiency, it can accurately predict the insulin independence time of the recipient after edema subsides, improving the assessment effect of cellular function after pancreas-kidney combined transplantation. 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 some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the steps in the method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation according to the present invention. Figure 2 This is a structural block diagram of the cellular functional imaging detection system for differences in glucose metabolism after pancreas-kidney combined transplantation according to the present invention. Detailed Implementation

[0018] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the cellular functional imaging detection method for glucose metabolism differences after pancreas-kidney combined transplantation proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0019] Unless otherwise defined, 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 pertains.

[0020] The following describes in detail, with reference to the accompanying drawings, the specific scheme of the cellular functional imaging detection method for glucose metabolism differences after pancreas-kidney combined transplantation provided by the present invention.

[0021] Please see Figure 1 The diagram illustrates a flowchart of a cellular functional imaging method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation according to an embodiment of the present invention. The method includes the following steps: Step S001: Obtain patient information in several types Weighted image under the value.

[0022] It should be noted that, due to the dynamic changes in hemodynamics within the transplant recipient, both transplanted organs must be covered simultaneously within a single scan cycle to ensure signal comparability between the transplanted pancreas and kidney at the same physiological moment. Furthermore, respiratory motion can cause anatomical misalignment between sequences with different diffusion sensitivities, leading to inconsistencies in signal sources at the voxel level. Therefore, establishing a unified anatomical alignment reference frame and performing rigid transformation correction can eliminate the interference of motion artifacts on the calculation of microscopic parameters.

[0023] Specifically, in order to implement the cellular functional imaging detection method for differences in glucose metabolism after pancreas-kidney combined transplantation proposed in this embodiment, it is first necessary to obtain data on the patient's glucose metabolism in several... The weighted image under the value is calculated as follows: First, obtain patient images from a medical image database in several categories. The weighted image is acquired using a 3.0T magnetic resonance scanner in conjunction with a phased array coil. During the acquisition process, the corresponding scanning range simultaneously covers the transplanted pancreas and transplanted kidney located in the abdominal cavity.

[0024] For example, in one specific embodiment, in order to separate the microcirculation perfusion and water molecule diffusion effects, a diffusion-sensitive gradient field is set. The value sequence is In other embodiments, adjustments may be made according to actual circumstances, and the embodiments of the present invention are not specifically limited.

[0025] Then, after the collection is complete, select Using the unweighted, diffusion-free image as the anatomical alignment reference frame, a rigid registration algorithm based on maximizing mutual information is employed, which integrates all other non-zero images. The weighted image of the values ​​is registered frame by frame to an anatomically aligned reference frame to generate a corrected four-dimensional signal matrix. This four-dimensional signal matrix contains several voxels, and each voxel has unique physical coordinates and a corresponding signal intensity sequence. For any voxel, the signal intensity sequence of the voxel contains each... The signal intensity of the corresponding voxel at the given value.

[0026] It should be noted that, regarding the aforementioned The value represents a parameter used in magnetic resonance imaging (MRI) to control the degree of diffusion weighting, and its unit is 1000 ppm. Its magnitude is determined by the gradient magnetic field strength, duration, and interval, directly affecting the sensitivity to the diffusion motion of water molecules. Since the value is a well-known concept, it will not be described further in this embodiment.

[0027] Thus, through the above methods, several types of... Weighted image under the value.

[0028] Step S002: By performing image segmentation and meshing on the weighted image, several pancreas voxels and kidney reference voxels are obtained, collectively referred to as unit voxels; based on the different degrees of any unit voxel... The perfusion fraction and diffusion coefficient of the unit voxel are determined by measuring the degree of change in the corresponding signal intensity. The discrete levels corresponding to the perfusion fraction and diffusion coefficient of the kidney reference voxel are combined with the distance between the pancreatic voxel and the kidney reference voxel regarding the perfusion fraction and diffusion coefficient. The matching relationship between the pancreatic voxel and the kidney reference voxel is determined based on the distance between them with respect to the perfusion fraction and diffusion coefficient.

[0029] It should be noted that the subsequent separation of pathological features relies on a one-to-one mapping relationship between the pancreas and the kidney, which requires that the number of elements in the two sets must be strictly equal. Given that the cortical volume of a transplanted kidney is usually significantly larger than the parenchymal volume of a transplanted pancreas, this embodiment of the invention uses the pancreas as the lesion observation object and the kidney as the physiological reference object. Therefore, a strategy of using the lesion observation object as the quantitative benchmark and performing restricted sampling on the physiological reference object is adopted to further establish a quantitatively aligned dataset while preserving all anatomical information of the pancreas.

[0030] As a preferred embodiment, the method for obtaining the matching relationship between the pancreatic voxel and the kidney reference voxel is as follows: Step S201: By performing image segmentation and grid division on the weighted image, several pancreas voxels and kidney reference voxels are obtained respectively.

[0031] As an optional embodiment, the method of obtaining several pancreas voxels and kidney reference voxels by performing image segmentation and mesh division on the weighted image includes the following specific methods: First, in the anatomical alignment reference frame, the region of interest (ROI) of the transplanted pancreas in an arbitrary weighted image is obtained using an image segmentation algorithm, denoted as the first ROI. The first ROI is then divided into regions with side lengths of... From several cubic meshes, extract all meshes that are completely located within the first region of interest and denote them as pancreatic voxels, thus obtaining... Each pancreatic voxel is identified; the geometric center coordinates of each pancreatic voxel in the anatomical alignment reference frame are obtained, denoted as the anatomical centroid coordinates, where... These are the preset mesh parameters.

[0032] Then, the region of interest (ROI) of the transplanted kidney cortex in an arbitrary weighted image is obtained using an image segmentation algorithm, denoted as the second ROI; the second ROI is then divided into regions with side lengths of... Several cubic grids were formed, and samples were extracted from them using a uniform sampling method. A cubic grid, denoted as the kidney reference voxel.

[0033] Additionally, specifically, when the number of available grids in the second region of interest is less than... In such cases, resampling (repeated sampling) is used to supplement the sample size. Individual elements; when the number is greater than At that time, uniform downsampling was used. Individual chromatids are used to obtain the kidney reference chromatid.

[0034] In other embodiments, the difference can also be supplemented by interpolation generation. Individual voxels, so that the number of kidney reference voxels is indivual.

[0035] In one specific embodiment, threshold segmentation is performed in the weighted image using the OTSU algorithm to obtain a first interest threshold and a second interest region.

[0036] It should be noted that this embodiment uses mesh parameters. In other embodiments, adjustments may be made as appropriate, but this embodiment does not impose specific limitations. In addition, during the process of obtaining kidney reference voxels, the cortical volume of the transplanted kidney is usually significantly larger than the parenchymal volume of the transplanted pancreas. In this embodiment of the invention, the pancreas is used as the lesion observation object and the kidney is used as the physiological reference object. Therefore, when obtaining kidney reference voxels, the cube network in the second region of interest is sampled so that the number of kidney reference voxels obtained is consistent with the number of pancreatic voxels, that is, the number of lesion observation objects (pancreas) is used as the quantitative benchmark.

[0037] Step S202, based on the different degrees of any unit voxel The degree of change in the corresponding signal intensity is used to determine the perfusion fraction and diffusion coefficient of the unit voxel.

[0038] It should be noted that the microscopic pathological state of transplanted organs is mainly determined by two orthogonal physiological dimensions: the blood supply capacity of the microvascular bed and the density of the extracellular space. The raw magnetic resonance signal is a mixture and superposition of these two dimensions, and it is difficult to distinguish between microcirculatory disturbances and tissue edema directly using the raw signal. By utilizing the sensitivity of the low b-value range to pseudo-diffusion motion of blood in capillaries and the sensitivity of the high b-value range to Brownian motion of water molecules, the perfusion fraction and diffusion coefficient can be separated, thereby constructing a physiological state vector that accurately describes the microscopic function of the transplanted organ.

[0039] As an optional embodiment, the method based on arbitrary unit voxels at different degrees... The method for determining the perfusion fraction and diffusion coefficient of a unit voxel by analyzing the degree of change in signal intensity under different values ​​includes: collectively referring to pancreatic voxels and renal reference voxels as unit voxels; pre-setting a first intensity range and a second intensity range; and determining the perfusion fraction and diffusion coefficient of any unit voxel under the first intensity range and the second intensity range. The rate of change of the signal intensity corresponding to the value is used to calculate the perfusion fraction and diffusion coefficient of the unit voxel.

[0040] It should be noted that, in a specific embodiment of the present invention, the first intensity range can be set as follows: The second intensity range can be In other embodiments, adjustments may be made as appropriate, and the embodiments of the present invention are not specifically limited.

[0041] As an optional embodiment, the perfusion fraction and diffusion coefficient of any unit voxel can be obtained, and the specific calculation method can be as follows:

[0042] in, Indicates the perfusion fraction of a unit voxel; Represents the diffusion coefficient of a unit voxel; Indicates the unit voxel in Signal strength when the value is 0; Indicates the unit voxel in Signal strength when the value is 50; Indicates the unit voxel in Signal strength when the value is 200; Indicates the unit voxel in Signal strength when the value is 800; This represents the logarithmic function with the natural constant as the base.

[0043] It should be noted that the perfusion fraction represents the proportion of microcirculatory blood volume within a tissue unit; this is achieved by analyzing the natural logarithm of the signal intensity. The slope of the value is obtained by linear fitting and used as a diffusion coefficient to characterize the apparent diffusion rate of water molecules. A decrease in the value indicates cell swelling or dense arrangement.

[0044] In addition, in other embodiments, those skilled in the art can also use a more accurate IVIM double-exponential model for multiple... Nonlinear fitting is performed on the value data to obtain the perfusion score. and diffusion coefficient However, in real-time intraoperative assessment or resource-constrained scenarios, the simplified formula mentioned above still has significant application value due to its computational efficiency and low hardware requirements.

[0045] Step S203: Combining the discrete levels corresponding to the perfusion fraction and diffusion coefficient of the kidney reference voxel, and based on the distance between the pancreatic voxel and the kidney reference voxel regarding the perfusion fraction and diffusion coefficient, determine the matching relationship between the pancreatic voxel and the kidney reference voxel.

[0046] It should be noted that systemic hemodynamic fluctuations in the transplant recipient often mask local pathological changes in the graft, and a single signal intensity cannot distinguish between the physical disintegration (necrosis) of the microcirculatory structure and the functional swelling (edema) of cell morphology. This invention constructs an individualized physiological reference system based on the transplanted kidney, utilizes bipartite graph global matching to eliminate systemic background interference, and further leverages the topological continuity of the lesion to specifically separate irreversible damage from reversible functional inhibition from the dimension of microstructural integrity by verifying the degree of aggregation of physical neighborhoods after mapping in the feature space. Furthermore, to eliminate the inherent differences in physical dimensions and numerical magnitudes between perfusion fraction and diffusion coefficient during feature matching, a normalization standard based on the recipient's own physiological state must first be established. This facilitates the subsequent use of the statistical distribution characteristics of the transplanted kidney data to provide adaptive weighting factors for subsequent difference measurements.

[0047] Furthermore, the specific method for determining the matching relationship between pancreatic voxels and kidney reference voxels includes: obtaining a state difference matrix based on the discreteness of perfusion fraction and diffusion coefficient corresponding to all kidney reference voxels, and combining the differences between pancreatic voxels and kidney reference voxels regarding perfusion fraction and diffusion coefficient; performing a matching analysis between pancreatic voxels and kidney reference voxels based on the state difference matrix to obtain the matching relationship between pancreatic voxels and kidney reference voxels.

[0048] As an optional embodiment, the specific method for obtaining the state difference matrix is ​​as follows: First, calculate the variance of all kidney reference voxels in the perfusion fraction dimension, denoted as the reference perfusion dispersion; similarly, calculate the variance of all kidney reference voxels in the diffusion coefficient dimension, denoted as the reference diffusion dispersion.

[0049] Then, the combination of any pancreatic voxel and any kidney reference voxel is denoted as a voxel combination. Combining the reference system perfusion dispersion and the reference system diffusion dispersion, the weighted distance between the corresponding pancreatic voxel and kidney reference voxel in the voxel combination is calculated, thereby obtaining the state difference matrix formed by the weighted distances corresponding to all voxel combinations.

[0050] Specifically, the size of the state difference matrix is ,in This represents the number of pancreatic reference voxels; additionally, since the number of kidney reference voxels is uniformly sampled to match the number of pancreatic reference voxels, the number of kidney reference voxels is also [missing information]. .

[0051] As an optional embodiment, the specific method for calculating the weighted distance can be:

[0052] in, Indicates the first The first pancreatic hormone and the first Weighted distance of each kidney reference voxel; Indicates the first The perfusion fraction of pancreatic somatostatin; Indicates the first The diffusion coefficient of pancreatic somatostatin; Indicates the first The perfusion fraction of a kidney reference voxel; Indicates the first The diffusion coefficient of a kidney reference voxel; Indicates the perfusion dispersion of the reference frame; Indicates the diffusion dispersion of the reference frame; This represents the preset hyperparameters.

[0053] It should be noted that, to prevent division by zero errors in subsequent calculations, this embodiment of the invention introduces a minimum machine precision value as a hyperparameter. In this embodiment, take The calculations can be adjusted according to the actual situation, and the embodiments of the present invention are not specifically limited. The weighted distance calculated above fully records all possible feature differences between the pancreas set to be detected and the physiological reference kidney set, and the differences have been adaptively normalized for the current physiological fluctuation range of the receptor, quantifying the standardized physiological distances corresponding to various combinations formed by all pancreas voxels and all kidney reference voxels.

[0054] It should be further explained that the lesions of transplanted pancreas usually exhibit a focal spatial distribution, while the recipient's systemic blood flow status has a global impact. In order to isolate common factors and lock in local heterogeneity, it is necessary to find an optimal mapping scheme in the global scope to minimize the overall characteristic differences. This forces each pancreatic voxel to determine the kidney reference object that is closest to its physiological state in the current systemic environment. This process establishes an individualized physiological twin relationship and eliminates non-pathological background differences introduced by different anatomical locations.

[0055] As an optional embodiment, the method of performing matching analysis between pancreatic voxels and kidney reference voxels based on the state difference matrix to obtain the matching relationship between pancreatic voxels and kidney reference voxels includes the following specific method: taking the pancreatic voxel set and the kidney reference voxel set as two vertex sets of a weighted bipartite graph, using the state difference matrix as the edge weight matrix, and using the maximum matching algorithm to solve the minimum weight matching problem of the weighted bipartite graph to obtain the kidney reference voxels that have a matching relationship with any pancreatic voxel.

[0056] It should be noted that, in a specific embodiment, the maximum matching algorithm can be the Hungarian Algorithm. Solving the weighted bipartite graph using the Hungarian Algorithm is essentially finding a one-to-one mapping function that minimizes the total state difference under that mapping function. Since the Hungarian Algorithm is an existing maximum matching algorithm, it is not within the scope of protection of this invention, and therefore will not be described in detail in this embodiment.

[0057] Thus, several pancreatic voxels and kidney reference voxels, as well as the matching relationships between pancreatic voxels and kidney reference voxels, were obtained through the above methods.

[0058] Step S003: Using the distribution of perfusion fraction and diffusion coefficient of kidney reference voxels that are matched with all pancreatic voxels within the preset spatial range, and in combination with the diffusion coefficient of the kidney reference voxels, calculate the patient's effective cell retention rate and cell function efficiency.

[0059] It should be noted that systemic hemodynamic fluctuations in the transplant recipient often mask local pathological changes in the graft, and a single signal intensity cannot distinguish between the physical disintegration (necrosis) of the microcirculatory structure and the functional swelling (edema) of cell morphology. This invention constructs an individualized physiological reference system based on the transplanted kidney, utilizes bipartite graph global matching to eliminate systemic background interference, and further leverages the topological continuity of the lesion. By verifying the degree of aggregation of physical neighborhoods after mapping in the feature space, it specifically separates irreversible damage from reversible functional inhibition from the dimension of microstructural integrity. Furthermore, the physical disintegration of the microvascular bed and cellular structure causes the magnetic resonance signal in the local area to lose spatial correlation and exhibit random noise characteristics, while cellular edema only causes an overall shift in signal values ​​but still maintains the topological continuity of the tissue structure. Therefore, by verifying whether physically adjacent pancreatic voxels remain aggregated after being mapped to the feature space, the integrity of the microstructure can be specifically identified using topological dispersion.

[0060] As an optional embodiment, the specific method for obtaining the patient's effective cell retention rate and cell functional efficiency is as follows: Step S301: Combine the perfusion fraction and diffusion coefficient of any unit voxel to form a vector, denoted as the physiological state vector of the corresponding unit voxel; obtain the feature neighborhood set of any pancreatic voxel, wherein the feature neighborhood set contains the physiological state vectors of several kidney reference voxels; calculate the local homogeneity destruction index of the pancreatic voxel based on the distribution of the physiological state vectors in the feature neighborhood set.

[0061] First, any pancreatic voxel is designated as the target pancreatic voxel. The anatomical centroid coordinates of each pancreatic voxel are obtained. Based on the Euclidean distance between the target pancreatic voxel and the anatomical centroid coordinates of pancreatic voxels in the pancreatic dataset, all pancreatic voxels in the pancreatic dataset whose Euclidean distance from the target pancreatic voxel does not exceed a preset neighborhood radius are obtained and designated as the neighboring pancreatic voxels of the target pancreatic voxel. The set of all neighboring pancreatic voxels of the target pancreatic voxel is designated as the anatomical neighborhood set of the target pancreatic voxel. For each neighboring pancreatic voxel in the anatomical neighborhood set, the set of physiological state vectors of the kidney reference voxels that have matching relationships with all neighboring pancreatic voxels is obtained and designated as the feature neighborhood set of the target pancreatic voxel.

[0062] It should be noted that in the process of obtaining the neighborhood pancreatic somatic cells of any given pancreatic somatic cell, the specific preset neighborhood range can be 26 neighborhoods. The 26-neighborhood used is a common representation of neighborhood ranges in three-dimensional matrices. Specifically, it represents the 26 adjacent pancreatic somatic cells surrounding the central pancreatic somatic cell in three-dimensional space, that is, the 8-neighborhood on the horizontal plane where the central pancreatic somatic cell is located, plus the 9 pancreatic somatic cells corresponding to the upper and lower adjacent planes of this horizontal plane, and thus further... The pancreatic somatic cells of 26 neighborhoods are obtained. In addition, when the pancreatic somatic cell at the center point is located at the boundary, there may be cases where it is impossible to obtain pancreatic somatic cells within the effective range of 26 neighborhoods. In this case, the number of neighboring pancreatic somatic cells actually obtained by that pancreatic somatic cell shall be used as the standard.

[0063] Then, based on the distribution of physiological state vectors in the feature neighborhood set, the local homogeneity disruption index of the corresponding pancreatic somatic cells is calculated.

[0064] As an optional embodiment, the method for calculating the local homogeneity destruction index of the corresponding pancreatic voxel based on the distribution of physiological state vectors in the feature neighborhood set includes: establishing a two-dimensional rectangular coordinate system, denoted as the physiological state coordinate system, by using irrigation fraction as the horizontal axis and diffusion coefficient as the vertical axis; using the elements in the physiological state vector as the coordinates of the corresponding axes in the physiological state coordinate system; thereby obtaining the data points corresponding to all kidney reference voxels in the feature neighborhood set of the target pancreatic voxel in the physiological state coordinate system, forming a corresponding scatter plot, denoted as the feature neighborhood scatter plot of the target pancreatic voxel; obtaining the geometric center of all data points in the feature neighborhood scatter plot, denoted as the feature center point of the target pancreatic voxel; and calculating the local homogeneity destruction index of the target pancreatic voxel based on the Euclidean distance between each data point in the feature neighborhood scatter plot and the feature center point, and using the perfusion dispersion and diffusion dispersion of the reference system as weights.

[0065] As an optional embodiment, the specific method for calculating the local homogeneity disruption index of the target pancreatic proton can be as follows:

[0066] in, The index representing the local homogeneity disruption of the target pancreatic somatic hormone; The set of data points formed in the feature neighborhood scatter plot representing the target pancreatic protamine; The number of data points in the feature neighborhood scatter plot representing the target pancreatic protamine; The scatter plot of the characteristic neighborhood of the target pancreatic somatic cell represents the first... Each data point corresponds to an irrigation score for a kidney reference voxel; The scatter plot of the characteristic neighborhood of the target pancreatic somatic cell represents the first... Each data point corresponds to the diffusion coefficient of a kidney reference voxel; This represents the irrigation score corresponding to the feature center point of the target pancreatic somatic cell. This represents the diffusion coefficient corresponding to the characteristic center point of the target pancreatic protease. Indicates the perfusion dispersion of the reference frame; Indicates the diffusion dispersion of the reference frame; This represents the preset hyperparameters.

[0067] It should be noted that the local homogeneity disruption index reflects the weighted dispersion of data points in the characteristic neighborhood scatter plot of the target pancreatic protamine relative to its geometric center (i.e., the characteristic center point of the target pancreatic protamine). The higher the value of the local homogeneity disruption index, the more dispersed the local region where the target pancreatic protamine is located is in the physiological characteristic space, suggesting that the microstructure has undergone physical disintegration or severe signal disturbance.

[0068] Step S302: Based on the distribution of perfusion fraction and local homogeneity destruction index of all pancreatic voxels, pancreatic voxels are screened. The effective cell retention rate of the patient is obtained based on the difference in quantity before and after screening. The cell function efficiency of the patient is calculated based on the difference between the diffusion coefficient of the screened pancreatic voxels and the diffusion coefficient of the renal reference voxels that have a matching relationship with the pancreatic voxels, and in combination with the dispersion of the diffusion coefficient of the renal reference voxels.

[0069] It should be noted that clinical treatment decisions rely on accurately distinguishing between irreversible necrotic tissue and reversible edematous tissue. By combining the numerical residuals of perfusion fraction with the topological characteristics of the local homogeneity destruction index, a dual-verification logic can be constructed: first, clearly necrotic areas are eliminated, and then the degree of functional inhibition in the remaining viable areas is quantified.

[0070] First, pancreatic protons were screened based on the distribution of perfusion fraction and local homogeneity destruction index of all pancreatic protons. The effective cell retention rate of the patient was obtained based on the difference in quantity before and after screening.

[0071] As an optional embodiment, the step of screening pancreatic somatic cells based on the distribution of perfusion fraction and local homogeneity destruction index of all pancreatic somatic cells, and obtaining the patient's effective cell retention rate based on the difference in quantity before and after screening, includes the following specific methods: 1) Statistically calculate the perfusion fraction and local homogeneity destruction index of all pancreatic somatic cells, and determine boundary parameters based on the distribution of the perfusion fraction and local homogeneity destruction index of pancreatic somatic cells respectively. These boundary parameters include a perfusion loss boundary parameter and a structural disorder boundary parameter. 2) Traverse all pancreatic somatic cells and record those that meet the conditions of a perfusion fraction less than the perfusion loss boundary parameter and a local homogeneity destruction index greater than the structural disorder boundary parameter as failed pancreatic somatic cells. 3) Count the number of failed pancreatic somatic cells among all pancreatic somatic cells and record it as the pancreatic failure amount. 4) Obtain the patient's effective cell retention rate based on the difference between the pancreatic failure amount and the number of all pancreatic somatic cells.

[0072] As an optional embodiment, the perfusion loss limit parameter and the structural disorder limit parameter are obtained by: calculating the mean and standard deviation of the perfusion fraction of all pancreatic protons, and using the difference between the mean and standard deviation as the perfusion loss limit; and combining the median and absolute median difference of the local homogeneity disruption index of all pancreatic protons to obtain the structural disorder limit parameter.

[0073] For example, the specific calculation method for the perfusion loss threshold parameter can be: ,in Indicates the perfusion loss limit parameter; This represents the mean perfusion fraction of all pancreatic protons; This represents the preset first limit coefficient; This represents the standard deviation of the perfusion fraction of all pancreatic protons.

[0074] For example, the specific calculation method for the structural disorder limit parameter can be as follows: ,in, Indicates the structural disorder limit parameter; This represents the median of the local homogeneity disruption index for all pancreatic somatosensory cells; This represents the preset second limit coefficient; This represents the absolute median difference in the local homogeneity disruption index of all pancreatic somatic cells.

[0075] It should be noted that the perfusion loss threshold parameter is used to determine significant ischemia, while the structural disorder threshold parameter is used to determine structural disintegration. Furthermore, in one specific embodiment, the first threshold coefficient can be set to 2, and the second threshold coefficient can be set to 1.5. In other embodiments, these can be adjusted according to actual conditions, and this embodiment of the invention does not impose specific limitations. Specifically, when the calculated result of the perfusion loss threshold parameter is negative, in other embodiments, the calculation method for the perfusion loss threshold parameter can also be: ,in Indicates the perfusion loss limit parameter; This represents the mean perfusion fraction of all pancreatic protons. This indicates a preset infusion adjustment parameter; in one specific embodiment, the infusion adjustment parameter can be preset to 3, and in other embodiments it can be adjusted according to the actual situation. This embodiment of the invention does not impose specific limitations.

[0076] As an optional embodiment, the specific method for calculating the effective cell retention rate can be: ,in This indicates the patient's effective cell retention rate; Indicates the amount of pancreatic failure; This indicates the amount of pancreatic protons.

[0077] The lower the effective cell retention rate, the more likely the graft is to lose its functional basis. This represents the permanent loss due to damage to the microcirculation. Cell stock.

[0078] Then, based on the difference between the diffusion coefficient of the screened pancreatic voxel and the diffusion coefficient of the kidney reference voxel that has a matching relationship with the pancreatic voxel, and combined with the dispersion of the diffusion coefficient corresponding to the kidney reference voxel, the patient's cell function efficiency is calculated.

[0079] As an optional embodiment, the method for calculating the patient's cell function efficiency based on the difference between the diffusion coefficient of the screened pancreatic voxels and the diffusion coefficient of the renal reference voxels that have a matching relationship with the pancreatic voxels includes the following specific method: The set consisting of all pancreatic voxels except for ineffective pancreatic voxels is denoted as the pancreatic survival set; the diffusion coefficient of the renal reference voxel that has a matching relationship with any pancreatic voxel is denoted as the matching diffusion coefficient of the pancreatic voxel; the difference between the diffusion coefficient and the matching diffusion coefficient of any pancreatic voxel in the pancreatic survival set is denoted as the diffusion difference of the corresponding pancreatic voxel; the average diffusion difference of all pancreatic voxels in the pancreatic survival set is used as the edema offset; and the patient's cell function efficiency is calculated based on the edema offset and the dispersion of the corresponding diffusion coefficient of the renal reference voxel.

[0080] As an optional embodiment, the specific method for calculating the cell functional efficiency can be:

[0081] in, Indicates cell function efficiency; Indicates the edema offset; This represents the standard deviation of the diffusion coefficients for all kidney reference voxels; This represents the absolute value function.

[0082] It should be noted that, if This indicates that no edema has occurred. ;like The cell function efficiency of the patient is determined by normalization calculation using a ratio function. This cell function efficiency quantifies the expected metabolic efficiency of surviving tissue after the elimination of edema effects. Specifically, to assess the significance of edema offset, this embodiment of the invention uses the standard deviation of the diffusion coefficient of the renal reference voxel as the physiological background noise benchmark of the receptor itself; and due to the edema offset... This reflects the degree of cellular edema in surviving tissue relative to a physiological baseline, i.e., the temporary functional inhibition of surviving cells due to edema compression or inflammatory response. A larger negative value indicates more severe edema. Therefore, by combining the standard deviation of the diffusion coefficients corresponding to all renal reference voxels and the edema offset, the degree of edema is determined. At that time, the patient's cellular function efficiency; the specific calculation method for cellular function efficiency indicates that when Much larger than the patient's physiological fluctuation range At that time, cell efficiency decreased significantly, but as edema subsided, the edema offset approached zero, and the corresponding cell functional efficiency approached [the expected value]. .

[0083] Thus, the effective cell retention rate and cell function efficiency were obtained through the above methods.

[0084] Step S004: Set several outcome levels, and determine the patient's outcome level by effective cell retention rate and cell function efficiency, and provide risk warnings.

[0085] It should be noted that, based on the biological principle of graft functional quality, namely that total metabolic capacity equals the product of the number of effective cells and the working efficiency of a single cell, this step constructs a predictive model that includes survival factors (i.e., effective cell retention rate) and efficiency factors (i.e., cell functional efficiency) to quantify the likelihood of patients being able to wean off exogenous insulin after edema subsides, thereby comprehensively assessing the true potential of the graft.

[0086] Specifically, firstly, by combining the effective cell retention rate and cell function efficiency, a comprehensive score for the patient is obtained. The comprehensive score is positively correlated with both the effective cell retention rate and cell function efficiency.

[0087] As an optional embodiment, the specific calculation method for the comprehensive score can be: ,in This indicates the overall score; Indicates the effective cell retention rate; Indicates cell function efficiency; This represents the sigmoid normalization function.

[0088] It should be noted that the range of values ​​for the comprehensive score output by the normalization function is as follows: to The higher the overall score, the greater the potential for the graft to recover its glucose metabolism function after the acute phase.

[0089] Then, several outcome levels are set, and the patient's outcome level is determined based on the comprehensive score.

[0090] As an optional embodiment, the specific method for dividing the prognostic levels according to the magnitude of the comprehensive score is as follows: preset a scoring threshold. and ,in At the same time, a preset retention threshold is set. and ,in ;when At that time, the outcome grade was Level 1, indicating that the patient was in the physiological compensation period, which showed that the graft microcirculation was intact and there was no significant edema, predicting that the recipient would be able to wean off exogenous insulin immediately after surgery without special intervention; when and At that time, the outcome grade was grade II, indicating that the patient was in a period of functional suppression. This indicated that although the graft retained a sufficient number of effective cells, it was currently suppressed by severe edema, suggesting that clinical treatment should be given for inflammation or dehydration. It was predicted that as the edema subsided, the patient would gradually regain insulin independence within 2-4 weeks. or At that time, the outcome was classified as Grade III, indicating functional failure, which suggests extensive irreversible microcirculatory damage in the graft and insufficient effective pancreatic function. The cell mass has fallen below the threshold for maintaining normal metabolism. This suggests that clinical consideration should include anticoagulation resuscitation or re-transplantation, and the recipient is predicted to be dependent on exogenous insulin for a long period.

[0091] For the preset scoring threshold and and retention threshold and In a specific embodiment, it can be set In other embodiments, the scoring threshold can be adjusted according to the actual situation. and and retention threshold and The specific value of is not specifically limited in the embodiments of the present invention.

[0092] This concludes the embodiment.

[0093] Please see Figure 2 The diagram illustrates a structural block diagram of a cell function imaging detection system for glucose metabolism differences after pancreas-kidney transplantation according to an embodiment of the present invention. The system includes a memory 202, a processor 201, and a computer program 2021 stored in the memory 202 and executable on the processor. When the processor 201 executes the computer program 2021, it implements steps S001 to S005 of the cell function imaging detection method for glucose metabolism differences after pancreas-kidney transplantation.

[0094] Furthermore, in an optional embodiment, the memory 202 described above may include read-only memory and random access memory, and provide instructions and data to the processor. The memory 202 may also include non-volatile random access memory. For example, the memory may also store device type information.

[0095] In one optional embodiment, the processor 201 described above can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. It is worth noting that the processor can be a processor supporting Advanced Reduced Instruction Set Machines (ARM) architecture.

[0096] 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 within the protection scope of the present invention.

Claims

1. A cellular functional imaging method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation, characterized in that, The method includes the following steps: Obtain patients in several types A weighted image with values, where each position in the weighted image corresponds to a value in each... Signal strength at the specified value; By segmenting the weighted image and dividing it into grids, several pancreas voxels and kidney reference voxels are obtained, collectively referred to as unit voxels; based on the different degrees of influence of any unit voxel... The perfusion fraction and diffusion coefficient of the unit voxel are determined by the degree of change of the corresponding signal intensity under the value. Combined with the discrete levels corresponding to the perfusion fraction and diffusion coefficient of the kidney reference voxel, the matching relationship between the pancreatic voxel and the kidney reference voxel is determined based on the distance between the pancreatic voxel and the kidney reference voxel with respect to the perfusion fraction and diffusion coefficient. The distribution of perfusion fraction and diffusion coefficient of renal reference voxels that are matched with all pancreatic voxels within a preset spatial range is used, and combined with the diffusion coefficient of renal reference voxels, to calculate the patient's effective cell retention rate and cell function efficiency. Several outcome levels were set, and the patient's outcome level was determined by the effective cell retention rate and cell function efficiency, and risk warnings were given.

2. The cellular functional imaging method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation according to claim 1, characterized in that, The specific method for obtaining the matching relationship between the pancreatic voxel and the kidney reference voxel is as follows: Based on the discrete values ​​of perfusion fraction and diffusion coefficient for all kidney reference voxels, and combined with the differences in perfusion fraction and diffusion coefficient between pancreatic voxels and kidney reference voxels, a state difference matrix is ​​obtained. Matching analysis was performed on pancreatic voxels and kidney reference voxels based on the state difference matrix to obtain the matching relationship between pancreatic voxels and kidney reference voxels.

3. The cellular functional imaging method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation according to claim 2, characterized in that, The specific method for obtaining the state difference matrix is ​​as follows: Calculate the variance of all kidney reference voxels in the perfusion fraction dimension, denoted as the reference perfusion dispersion; similarly, calculate the variance of all kidney reference voxels in the diffusion coefficient dimension, denoted as the reference diffusion dispersion. A combination of any pancreatic voxel and any kidney reference voxel is denoted as a voxel combination. By combining the reference system perfusion dispersion and the reference system diffusion dispersion, the weighted distance between the corresponding pancreatic voxel and kidney reference voxel in the voxel combination is calculated, thereby obtaining the state difference matrix formed by the weighted distances corresponding to all voxel combinations.

4. The cellular functional imaging method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation according to claim 2, characterized in that, The matching analysis based on the state difference matrix between pancreatic voxels and kidney reference voxels to obtain the matching relationship between pancreatic voxels and kidney reference voxels includes the following specific methods: Using the pancreas voxel set and the kidney reference voxel set as the two vertex sets of a weighted bipartite graph, and the state difference matrix as the edge weight matrix, the minimum weight matching problem of the weighted bipartite graph is solved by the maximum matching algorithm to obtain the kidney reference voxel that has a matching relationship with any pancreas voxel.

5. The cellular functional imaging method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation according to claim 1, characterized in that, The specific method for obtaining the effective cell retention rate is as follows: The perfusion fraction and diffusion coefficient of any unit voxel are combined to form a vector, which is denoted as the physiological state vector of the corresponding unit voxel; the feature neighborhood set of any pancreatic voxel is obtained, and the feature neighborhood set contains the physiological state vectors of several kidney reference voxels; the local homogeneity destruction index of the corresponding pancreatic voxel is calculated according to the distribution of the physiological state vectors in the feature neighborhood set. Based on the distribution of perfusion fraction and local homogeneity destruction index of all pancreatic protons, pancreatic protons were screened, and the effective cell retention rate of the patient was obtained based on the difference in quantity before and after screening.

6. The cellular functional imaging method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation according to claim 5, characterized in that, The specific method for obtaining the feature neighborhood set of any pancreatic voxel is as follows: Obtain the geometric center coordinates of each pancreatic voxel in the anatomical alignment reference frame, and denote them as the anatomical centroid coordinates; Any pancreatic voxel is designated as the target pancreatic voxel. The anatomical centroid coordinates of each pancreatic voxel are obtained. Based on the Euclidean distance between the target pancreatic voxel and the anatomical centroid coordinates of the pancreatic voxels in the pancreatic dataset, all pancreatic voxels in the pancreatic dataset whose Euclidean distance from the target pancreatic voxel does not exceed the preset neighborhood radius are obtained and designated as the neighboring pancreatic voxels of the target pancreatic voxel. The set of all neighboring pancreatic voxels of the target pancreatic voxel is denoted as the anatomical neighborhood set of the target pancreatic voxel. For each neighboring pancreatic voxel in the anatomical neighborhood set, the set of physiological state vectors of the kidney reference voxels that have matching relationships with all neighboring pancreatic voxels is obtained and denoted as the feature neighborhood set of the target pancreatic voxel.

7. The cellular functional imaging method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation according to claim 6, characterized in that, The specific method for calculating the local homogeneity disruption index of the corresponding pancreatic voxel based on the distribution of physiological state vectors in the feature neighborhood set includes: Using irrigation fraction as the horizontal axis and diffusion coefficient as the vertical axis, a two-dimensional rectangular coordinate system is established and denoted as the physiological state coordinate system. The elements in the physiological state vector are used as the coordinates of the corresponding coordinate axes in the physiological state coordinate system. This yields the data points of all kidney reference voxels in the feature neighborhood set of the target pancreatic voxel in the physiological state coordinate system, forming a corresponding scatter plot, denoted as the feature neighborhood scatter plot of the target pancreatic voxel. Obtain the geometric center of all data points in the feature neighborhood scatter plot, and denote it as the feature center point of the target pancreatic somatic cell. Based on the Euclidean distance between each data point in the feature neighborhood scatter plot and the feature center point, and using the perfusion dispersion and diffusion dispersion of the reference system as weights, calculate the local homogeneity destruction index of the target pancreatic somatic cell.

8. The cellular functional imaging method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation according to claim 5, characterized in that, The method for screening pancreatic protons based on the distribution of perfusion fraction and local homogeneity disruption index of all pancreatic protons includes the following specific methods: The perfusion fraction and local homogeneity destruction index of all pancreatic protons were statistically analyzed. Based on the distribution of the perfusion fraction and local homogeneity destruction index of pancreatic protons, the boundary parameters were determined. The boundary parameters include the perfusion loss boundary parameter and the structural disorder boundary parameter. Iterate through all pancreatic somatic cells and denote the pancreatic somatic cells that satisfy the conditions of perfusion fraction being less than the perfusion loss threshold parameter and local homogeneity destruction index being greater than the structural disorder threshold parameter as ineffective pancreatic somatic cells.

9. The cellular functional imaging method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation according to claim 8, characterized in that, The specific method for obtaining the cell functional efficiency is as follows: The set of all pancreatic voxels except for the ineffective pancreatic voxels is denoted as the pancreatic survival set; the diffusion coefficient of the kidney reference voxel that has a matching relationship with any pancreatic voxel is denoted as the matching diffusion coefficient of the pancreatic voxel. The difference between the diffusion coefficient of any pancreatic voxel in the pancreatic survival set and the matching diffusion coefficient is denoted as the diffusion difference of the corresponding pancreatic voxel. The average diffusion difference of all pancreatic voxels in the pancreatic survival set is used as the edema offset. Based on the edema offset and the dispersion of the diffusion coefficient of the kidney reference voxel, the patient's cell function efficiency is calculated.

10. The method for detecting differences in glucose metabolism after pancreas-kidney combined transplantation according to claim 1, characterized in that, The specific methods for obtaining the pancreatic voxel and the kidney reference voxel are as follows: In the anatomical alignment reference frame, the region of interest (ROI) of the transplanted pancreas in an arbitrary weighted image is obtained using an image segmentation algorithm, denoted as the first ROI. The first ROI is then divided into regions with side lengths of... From several cubic meshes, extract all meshes that are completely located within the first region of interest and denote them as pancreatic voxels, thus obtaining... One pancreatic somatic cell; among which These are the preset mesh parameters; The region of interest (ROI) of the transplanted kidney cortex in an arbitrary weighted image is obtained using an image segmentation algorithm and denoted as the second ROI. The second ROI is then divided into regions with sides of length [missing information]. Several cubic grids were formed, and samples were extracted from them using a uniform sampling method. A cubic grid, denoted as the kidney reference voxel.

Images