A method and device for predicting renal physiological age based on multimodal MRI images

By using multimodal MRI imaging and deep learning technology, a kidney physiological age prediction model was constructed, which solved the problems of insufficient early kidney injury identification and insufficient information integration in existing technologies. This enabled early diagnosis and personalized assessment, reduced radiation risk, and improved diagnostic accuracy and efficiency.

CN120013900BActive Publication Date: 2026-11-17PEKING UNION MEDICAL COLLEGE HOSPITAL
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202510092156.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-01-21
Publication Date
2026-11-17
Estimated Expiration
2045-01-21

AI Technical Summary

Technical Problem

Existing kidney health assessment technologies struggle to identify microscopic lesions in their early stages and lack integration of information on kidney structure and function, leading to inadequate diagnosis and limited personalized assessment. Furthermore, imaging methods carry radiation risks or rely on subjective evaluation.

Method used

By combining multimodal MRI images with deep learning technology, a kidney physiological age prediction model is constructed through semantic segmentation and radiomics feature extraction. This model automatically assesses kidney health status and uses a neural network model to integrate radiomics parameters for accurate prediction.

Benefits of technology

It enables early detection of kidney damage, provides personalized diagnosis, reduces radiation risk, improves diagnostic accuracy and efficiency, and supports individualized treatment decisions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120013900B_ABST
    Figure CN120013900B_ABST
Patent Text Reader

Abstract

This invention proposes a method and apparatus for predicting renal physiological age based on multimodal MRI images. The method includes: semantic segmentation of renal MRI images to obtain a cortical-medullary segmentation mask; high-throughput radiomics feature extraction of the cortical-medullary portion of the renal MRI images; combination and screening of radiomics parameters to obtain the radiomics parameter with the highest correlation to renal physiological age prediction as the final radiomics parameter; and inputting the final radiomics parameter as a feature into a neural network model to obtain the predicted age. This invention automatically segments different tissue regions of the kidney, extracts high-throughput radiomics features, and predicts "renal physiological age" based on these features.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the fields of image semantic segmentation, multimodal image analysis, and artificial intelligence, and particularly to a method and device for predicting the physiological age of the kidneys based on multimodal MRI images. Background Technology

[0002] Current technologies for assessing kidney health primarily rely on two main categories of methods: biomarkers and imaging examinations. Biomarkers provide information on kidney function status by measuring serum creatinine (sCr) and estimated glomerular filtration rate (eGFR). sCr is a metabolic product reflecting the kidney's filtration capacity, while eGFR, by incorporating factors such as age, sex, and body surface area, provides a more accurate estimate of kidney function. Imaging examinations include ultrasound, CT scans, and MRI. Ultrasound utilizes real-time dynamic imaging, is non-invasive and radiation-free, and is often used for the initial assessment of kidney morphology and abnormalities (such as stones, cysts, and masses). CT scans provide high-resolution cross-sectional images, suitable for analyzing complex anatomical structures, and are particularly advantageous in the qualitative and quantitative assessment of kidney tumors. MRI uses strong magnetic fields and radiofrequency waves to generate high soft tissue contrast images, enabling the detection of subtle differences between tissues and the assessment of renal blood flow and lesion characteristics. These methods, through their respective technical means, assess kidney health from multiple perspectives and are often used in combination to ensure diagnostic accuracy and comprehensiveness.

[0003] Insufficient timeliness of biomarkers: While sCr and eGFR are commonly used markers of kidney function, they typically only show significant changes when substantial kidney damage occurs. This lag means that in many cases, especially in the early stages of chronic kidney disease, patients may have unexplained kidney damage. Furthermore, because these markers are influenced by muscle mass, diet, and other individual differences, they can lead to false negative or false positive results in some populations, limiting their effectiveness in personalized assessment.

[0004] Limitations of Imaging Examinations: Ultrasound, a traditional imaging method, has limited visibility of fine structures due to its low resolution. While CT scans offer high anatomical detail, their radiation risks limit their application in young patients requiring long-term follow-up or pregnant women. Furthermore, CT is limited in assessing kidney function, being primarily used for detecting structural abnormalities. MRI, while radiation-free and providing excellent soft tissue contrast, often relies on subjective assessment in traditional MRI analysis methods, lacking quantitative functional evaluation. Current artificial intelligence technology focuses on simple segmentation of kidney images; analysis of kidney health status through kidney structure remains largely manual.

[0005] Lack of early lesion recognition capability: Current assessment methods often only provide reliable information in the middle and late stages of kidney disease, lacking the ability to actively monitor and identify subtle pathological changes in the kidneys. This leads to insufficient intervention in the early stages of the disease, missing the optimal treatment window.

[0006] Limited integrative analytical capabilities: Current biomarker and imaging techniques are often performed in isolation, failing to effectively integrate information from different data sources to provide a comprehensive view of kidney health. This lack of integrative approach leads to an insufficient overall understanding of the dynamic changes in kidney disease, limiting comprehensive management and decision-making in complex pathological states, especially in cases of multiple coexisting diseases (such as diabetes, hypertension, and chronic kidney disease).

[0007] Specifically, existing technologies have proposed a kidney age prediction method based on CT images, which predicts age by measuring the volume of the renal cortex and medulla from CT images. However, the radiation risks associated with acquiring CT images limit the application scenarios of this technology. Therefore, there is an urgent need for a kidney age prediction method based on non-radiation images, such as a kidney age prediction method based on MRI images. However, inputting MRI images into this kidney age prediction method will bring the following three technical problems: First, MRI images are not the first choice for measuring the volume of the renal cortex and medulla because CT images have high spatial resolution, which can clearly show the fine structure and morphology of the kidney, facilitating accurate volume measurement. CT images can be analyzed in detail through multiplanar reconstruction and three-dimensional imaging. Therefore, inputting MRI images into a kidney age prediction method based on CT images will lead to a decrease in prediction accuracy. Second, MRI images contain far more information than CT images. For example, MRI images have richer soft tissue contrast, multi-parameter imaging, and functional and metabolic information, while CT images can only provide limited soft tissue contrast, mainly focusing on tissue density. MRI's various imaging sequences and functional imaging techniques provide more comprehensive and diverse biological tissue information for radiomics analysis, while CT is relatively lacking in soft tissue detail identification and diversity information. Therefore, directly applying CT image-based methods for kidney age prediction would result in information waste. However, MRI images contain a wealth of information, not all of which is relevant to kidney age prediction. The current technical challenge is how to integrate and filter information from MRI images to improve the accuracy of AI models in predicting kidney age and make the kidney age output by AI models interpretable. Summary of the Invention

[0008] This application aims to address the shortcomings of existing kidney health assessment technologies in effectively identifying early kidney damage and integrating kidney structural and functional information for comprehensive evaluation. Traditional biomarkers such as sCr and eGFR often only show changes when kidney function declines significantly, lacking sensitivity for early, microscopic kidney lesions. While existing imaging methods can provide structural information, they fail to effectively quantify and evaluate the functional status of the kidneys. This invention proposes an automated model that completes the entire process of "automatic segmentation - omics extraction - kidney health assessment".

[0009] To address the shortcomings of existing technologies, such as Figure 5 As shown, this invention proposes a method for predicting renal physiological age based on multimodal MRI images, including:

[0010] The initial step involves acquiring multiple pairs of kidney MRI images from healthy individuals. Each pair of kidney MRI images includes a kidney T1-weighted image and a kidney T1-mapped image, and each pair of kidney MRI images is labeled with the corresponding actual age.

[0011] The semantic segmentation step involves performing semantic segmentation on each pair of kidney MRI images to obtain the first cortico-medullary segmentation mask of the kidney weighted image and the second cortico-medullary segmentation mask of the kidney mapping image.

[0012] The feature extraction step involves acquiring radiomics data from the kidney T1-weighted image and the kidney T1-mapped image based on the first corticome segmentation mask and the second corticome segmentation mask, thereby obtaining a first set of radiomics parameters for the kidney T1-weighted image and a second set of radiomics parameters for the kidney T1-mapped image.

[0013] The combined training step involves combining the first set of radiomics parameters and the second set of radiomics parameters, and filtering the combined results to obtain the radiomics parameter with the highest correlation to the prediction of renal physiological age as the final omics parameter. The final omics parameter is then used as the feature input of the renal MRI image pair into the neural network model to obtain the predicted age. A loss function is constructed using the predicted age and the actual age labeled in the renal MRI image pair to train the neural network model, thus obtaining the renal physiological age prediction model.

[0014] The age prediction step involves semantically segmenting the kidney MRI image pair for which the physiological age to be predicted is to extract the final omics parameters. The kidney physiological age prediction model then predicts the physiological age of the kidney based on these final omics parameters. If the difference between the predicted physiological age and the actual age in the kidney MRI image pair exceeds a threshold, kidney damage is considered to be present.

[0015] The method for predicting the physiological age of the kidney based on multimodal MRI images includes first-order radiomics parameters and higher-order radiomics parameters in both the first group of radiomics parameters and the second group of radiomics parameters. The first group of radiomics parameters also includes shapemics parameters.

[0016] The first-order radiomics parameters include the mean and median voxel gray values ​​of the renal cortex and medulla regions. The higher-order radiomics parameters include the gray-level co-occurrence matrix, gray-level dependence matrix, gray-level run matrix, gray-level size region matrix, and neighborhood gray-level difference matrix of the renal cortex and medulla regions. Shapemics parameters include the volume of the renal cortex and medulla regions, the long axis of the kidney, surface smoothness, and surface area-to-volume ratio.

[0017] The method for predicting renal physiological age based on multimodal MRI images, wherein the process of filtering the combined results in the combined training step includes:

[0018] The first set of radiomics parameters and the second set of radiomics parameters are classified to obtain a kidney shape parameter dataset, a T1-weighted image feature parameter dataset, and a T1-mapped image feature parameter dataset. The above three are then combined to obtain multiple datasets.

[0019] Using multiple datasets as input, the model predicts the physiological age of the kidneys as output. The neural network model is trained using an interpretable machine learning method, and the dataset with the best prediction performance is selected as the final omics parameter.

[0020] The age prediction step also includes outputting the final omics parameters of the renal MRI image to be predicted for physiological age, as interpretable features for predicting its renal physiological age.

[0021] The method for predicting renal physiological age based on multimodal MRI images, wherein the multiple datasets include:

[0022] The kidney shape parameter dataset, T1-weighted image feature parameter dataset, and T1-mapped image feature parameter dataset are combined in pairs and / or in triplicate to obtain the multiple datasets. The first set of radiomics parameters and the second set of radiomics parameters also include corticomedullary differentiation (CMD) features. The more prominent the CMD features, the more normal and healthy the kidneys are.

[0023] like Figure 6 As shown, this invention also proposes a device for predicting the physiological age of the kidneys based on multimodal MRI images, including:

[0024] The initial module acquires multiple pairs of kidney MRI images from healthy individuals. Each pair of kidney MRI images includes a kidney T1-weighted image and a kidney T1-mapped image, and each pair of kidney MRI images is labeled with the corresponding physiological age.

[0025] The semantic segmentation module performs semantic segmentation on each pair of kidney MRI images to obtain the first cortico-medullary segmentation mask of the kidney weighted image and the second cortico-medullary segmentation mask of the kidney mapping image.

[0026] The feature extraction module acquires radiomics data of the kidney T1-weighted image and the kidney T1-mapped image based on the first corticome segmentation mask and the second corticome segmentation mask, and obtains a first set of radiomics parameters of the kidney T1-weighted image and a second set of radiomics parameters of the kidney T1-mapped image.

[0027] The combined training module combines the first set of radiomics parameters and the second set of radiomics parameters, and filters the combined results to obtain the radiomics parameter with the highest correlation to the prediction of renal physiological age as the final omics parameter. The final omics parameter is used as the feature input of the renal MRI image pair into the neural network model to obtain the predicted age. The predicted age and the physiological age labeled by the renal MRI image pair are used to construct a loss function to train the neural network model and obtain the renal physiological age prediction model.

[0028] The age prediction module extracts the final omics parameters after semantic segmentation of the kidney MRI image pair to be predicted for physiological age. The kidney physiological age prediction model predicts the kidney physiological age based on the final omics parameters of the kidney MRI image pair to be predicted for physiological age.

[0029] The kidney physiological age prediction device based on multimodal MRI images, wherein the first set of radiomics parameters and the second set of radiomics parameters both include first-order radiomics parameters and higher-order radiomics parameters, and the first set of radiomics parameters also includes shapemics parameters;

[0030] The first-order radiomics parameters include the mean and median voxel gray values ​​of the renal cortex and medulla regions. The higher-order radiomics parameters include the gray-level co-occurrence matrix, gray-level dependence matrix, gray-level run matrix, gray-level size region matrix, and neighborhood gray-level difference matrix of the renal cortex and medulla regions. Shapemics parameters include the volume of the renal cortex and medulla regions, the long axis of the kidney, surface smoothness, and surface area-to-volume ratio.

[0031] The renal physiological age prediction device based on multimodal MRI images, wherein the process of filtering the combined results in the combined training module includes:

[0032] The first set of radiomics parameters and the second set of radiomics parameters are classified to obtain a kidney shape parameter dataset, a T1-weighted image feature parameter dataset, and a T1-mapped image feature parameter dataset. The above three are then combined to obtain multiple datasets.

[0033] Using multiple datasets as input, the model predicts the physiological age of the kidneys as output. The neural network model is trained using an interpretable machine learning method, and the dataset with the best prediction performance is selected as the final omics parameter.

[0034] The age prediction module also includes outputting the final omics parameters of the renal MRI image to be predicted for the physiological age of the kidney, as an interpretable feature for predicting its physiological age of the kidney.

[0035] The present invention also proposes an electronic device, including the aforementioned kidney physiological age prediction device, which may be connected to an information display device for displaying the kidney physiological age using user-set display parameters, attributes, or through an artificial intelligence model.

[0036] The present invention also proposes a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the described kidney physiological age prediction methods.

[0037] The present invention also proposes a computer program product comprising a computer program, wherein the computer program, when executed by a processor, implements the steps of any of the described kidney physiological age prediction methods.

[0038] As can be seen from the above solutions, the advantages of the present invention are:

[0039] Compared with existing technologies, this invention has significant beneficial effects in kidney health assessment:

[0040] 1. Enhanced early detection capabilities: Advanced radiomics and deep learning technologies can capture microscopic changes in early kidney injury, providing opportunities for intervention before clinical symptoms appear.

[0041] 2. Comprehensive assessment and personalized diagnosis: Combining structural and functional information, it provides a comprehensive assessment of kidney health, supporting more accurate diagnosis and individualized treatment decisions in clinical practice.

[0042] 3. No radiation risk and safety: MRI imaging data is obtained without radiation risk and can be repeatedly applied to users who need frequent monitoring, making it especially suitable for sensitive groups such as children and pregnant women.

[0043] 4. Improve diagnostic accuracy and efficiency: Automated image analysis methods reduce reliance on human experience, lower subjective errors, and improve diagnostic efficiency and consistency.

[0044] 5. Promoting clinical practice in prevention and early treatment: By providing a comparison of renal physiological age with the user's actual age (KAG), it helps identify high-risk groups and guides early treatment measures, thereby effectively delaying further deterioration of renal function. Attached Figure Description

[0045] Figure 1 This is a structural diagram of the model of the present invention;

[0046] Figure 2 Image of kidney semantic segmentation results;

[0047] Figure 3 A comparison chart of data prediction performance;

[0048] Figure 4 A diagram illustrating the risk of cardiovascular events resulting from changes in renal age differences over time;

[0049] Figure 5 This is a flowchart of the method of the present invention;

[0050] Figure 6 This is a block diagram of the device of the present invention;

[0051] Figure 7 This is a schematic diagram of the structure of the first electronic device of the present invention;

[0052] Figure 8 This is a schematic diagram of the application environment structure of the first electronic device of the present invention;

[0053] Figure 9 This is a schematic diagram of the structure of the second electronic device of the present invention.

[0054] Figure label:

[0055] A - First electronic device;

[0056] B- A device for predicting the physiological age of the kidneys based on multimodal MRI images;

[0057] C-Data acquisition equipment;

[0058] D-Information display device;

[0059] 1000 - Second electronic device;

[0060] Ⅰ-Computational Unit;

[0061] II-ROM;

[0062] III-RAM;

[0063] N-bus;

[0064] V-Interface;

[0065] VI - Input Unit;

[0066] VII - Output Unit;

[0067] VIII - Storage medium;

[0068] IX - Communication Unit. Detailed Implementation

[0069] It should be noted that, in this application, relational terms such as "first" and "second" are used merely 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.

[0070] In the absence of further restrictions, an element defined by the phrase "comprising a..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0071] The processor described in this invention is the control center of an electronic device. It can be a single processor or a collective term for multiple processing elements. For example, it can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement embodiments of this invention, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0072] Alternatively, the processor can perform various functions of the electronic device by running or executing software programs stored in memory and by calling data stored in memory.

[0073] In a specific implementation, as one example, the processor may include one or more CPUs. Each of these processors may be a single-core processor or a multi-core processor. Here, "processor" can refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions). Electronic devices may include servers, desktop computers, laptops, smartphones, tablets, embedded computers, etc., where the embedded computer includes vehicles and robots, etc.

[0074] The memory is used to store the software program that executes the solution of the present invention, and the execution is controlled by the processor. For specific implementation methods, please refer to the above method embodiments, which will not be repeated here.

[0075] It should be noted that the structure of the electronic device shown in the accompanying drawings of this invention does not constitute a limitation thereof. The actual knowledge structure recognition device may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0076] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0077] It should also be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0078] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0079] It should also be understood that, in various embodiments of the present invention, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0080] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0081] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0082] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0083] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0084] In conducting research on kidney health assessment, this invention revealed significant limitations in the accurate detection of early kidney injury using existing technologies. These limitations primarily stem from the fact that traditional methods rely solely on biomarkers such as serum creatinine (sCr) and estimated glomerular filtration rate (eGFR). These markers typically fail to effectively reflect early or subtle damage before patients exhibit significant renal function decline. Furthermore, while imaging techniques can provide some structural information, they lack integration with functional status (biomarkers). Moreover, these methods fail to adequately consider the microscopic changes in kidney structure, a factor particularly crucial for the early diagnosis and prevention of chronic kidney disease.

[0085] After in-depth research, the inventors realized that the key to overcoming this deficiency lies in combining structural and functional information of the kidneys, and obtaining a more comprehensive assessment through integrated analysis of multimodal imaging technology and machine learning algorithms. Specifically, the inventors discovered that multimodal MRI images can provide rich structural and tissue information, which can be transformed into quantitative data through radiomics feature extraction to reflect microscopic changes in kidney health.

[0086] Therefore, based on this, the present invention develops a novel solution using deep learning technology: This invention trains a complex neural network model, KAGE-NET, which automatically segments different tissue regions of the kidney, extracts detailed radiomics features, and predicts the "kidney physiological age" based on these features. These features include shape features and first-order statistical features (such as the mean and variance of volume, length, and brightness), as well as higher-order features (gray-level co-occurrence matrix, gray-level run matrix, etc.), thus more accurately reflecting the kidney's physiological age (K-AGE). By combining with traditional biomarkers, this model can transcend the limitations of current diagnostic methods, providing a perspective on kidney health. The concept of the kidney physiological age gap (KAG) is proposed, using the difference between the "kidney physiological age" and the actual age to reflect the degree of accelerated kidney aging: a positive value represents accelerated kidney aging, while a negative value indicates delayed aging. Ultimately, this results in a comprehensive diagnostic tool integrating structural and functional assessments, improving the accuracy of kidney health assessment and early detection capabilities.

[0087] To achieve the above-mentioned technical effects, the present invention proposes the following key technical points:

[0088] Key Point 1: Automated segmentation and radiomics feature extraction using multimodal MRI images combined with deep learning. By automatically segmenting various tissue regions of the kidney (such as the cortex and medulla) and extracting relevant radiomics features (such as volume, surface area, and pixel value statistics), the structural changes and functional status of the kidney can be quantitatively assessed, improving the ability to detect early kidney injury.

[0089] Key Point 2: Predictive Model for Kidney Physiological Age (K-AGE). A model for predicting kidney physiological age is constructed by integrating multiple radiomics features using machine learning algorithms. The predicted kidney physiological age can be compared with the user's biological age (actual age), and the kidney age difference (KAG) index is used to achieve a personalized assessment of the degree of kidney aging and health status. It is important to note that the actual kidney age of the user being tested in this invention is the user's biological age. For example, if the biological age of a user with normal, healthy kidneys is 20 years old, then their kidney biological age should also be 20 years old, and the KAG is 20 minus 20, which equals 0. If the age of a user with kidney disease is 20 years old, then their kidney biological age is 30 years old, and the KAG is 30 minus 20, which equals 10.

[0090] Key Point 3: Potential Tools for Personalized Medicine. A comprehensive suite of tools has been developed for personalized care in clinical settings to quantify the complex impacts of various risk factors on kidney health. This has facilitated personalized treatment plans and efficacy monitoring for patients with chronic kidney disease (CKD) and metabolic syndrome.

[0091] To make the above-mentioned features and effects of the present invention clearer and easier to understand, specific embodiments are described below in conjunction with the accompanying drawings. This specification discloses one or more embodiments incorporating the features of the present invention. The disclosed embodiments are merely illustrative. The scope of protection of the present invention is not limited to the disclosed embodiments, but is defined by the appended claims.

[0092] This invention introduces an automated radiomics analysis method based on multimodal MRI imaging and artificial intelligence technology. By creating a model of the kidney's "physiological age (K-AGE)," it enables a comprehensive assessment of kidney health. This innovative technology can reflect subtle changes in kidney structure and function early and accurately, surpassing the limitations of traditional methods and providing information on the rate of kidney aging and potential lesions. Using this method, medical professionals can more accurately monitor and manage kidney disease, develop more personalized treatment strategies for users, and significantly improve the effectiveness and timeliness of clinical interventions. Figure 1 As shown, specifically, the method of the present invention includes:

[0093] Step 1:

[0094] Obtain the kidney T1-weighted and T1-mapped dataset (DICOM format) of healthy kidney users, along with the users' actual ages. Convert the kidney T1-weighted and T1-mapped dataset (DICOM format) to a machine-readable and processable NIFTI format. For example... Figure 2 As shown, kidney T1-weighted and kidney T1-mapped data are two different imaging modalities in MRI imaging.

[0095] Step 2:

[0096] The T1-weighted data was divided into training and testing sets. The corticomena of the left and right kidneys in the training set were manually labeled to establish an automatic segmentation model of the renal corticomena from T1-weighted images.

[0097] Step 3:

[0098] The images in the test set are automatically segmented to obtain automatic segmentation masks for the renal cortex and medulla.

[0099] Step 4:

[0100] The T1 weighted image mask is mapped one-to-one with the T1 mapping dataset. The two are then manually reviewed by a radiology expert to check if they overlap. If they do, proceed to step 6; otherwise, proceed to step 5.

[0101] Step 5:

[0102] Images that do not meet the requirements are manually labeled and corrected to establish a correspondence between the T1 weighted image mask and the T1 mapping dataset weighted image mask. This allows the T1 weighted image mask to correctly describe the renal cortex and medulla regions of the T1 mapping dataset, thereby improving the accuracy of subsequent image feature extraction and the accuracy of renal physiological age prediction.

[0103] Step 6:

[0104] The range of T1-weighted image values ​​will fluctuate with changes in the instrument, so it is necessary to standardize the T1-weighted data to obtain the T1-weighted standardized dataset and the T1-mapped dataset.

[0105] Step 7:

[0106] Radiomics data were collected for the T1 mapping dataset, including first-order radiomics parameters (mean, median, etc. of voxel gray values ​​in the renal cortex and medulla regions) and higher-order radiomics parameters (including gray co-occurrence matrix, gray dependence matrix, gray run matrix, gray size region matrix, neighborhood gray difference matrix, etc. in the renal cortex and medulla regions).

[0107] Step 8:

[0108] Radiomics data were collected for the T1-weighted normalized dataset, including shape features (kidney volume, long diameter, surface area-volume ratio, etc.), first-order radiomics parameters (mean, median, etc. of voxel gray values ​​in the renal cortex and medulla regions) and higher-order radiomics parameters (including gray-level co-occurrence matrix, gray-level dependence matrix, gray-level run matrix, gray-level size region matrix, neighborhood gray-level difference matrix, etc. in the renal cortex and medulla regions).

[0109] Step 9:

[0110] After screening, combining, cleaning, dimensionality reduction and standardization of two sets of T1 radiomics parameters, a variety of standard datasets of radiomics parameters were established.

[0111] Combination of kidney datasets: Based on their specific physical meaning, the features of kidney images are divided into three categories of radiomics datasets: shape parameter radiomics dataset (kidney volume, thickness, long diameter, etc.) (Shape), T1-weighted image feature parameter radiomics dataset (T1W), and T1 mapping influence feature parameter radiomics dataset (T1M); and four other radiomics datasets obtained by free combination: WShape (T1W+Shape), MSipe (T1M+Shape), T1 (T1W+T1M), and ALL (T1W+T1M+Shape).

[0112] Step 10:

[0113] A model set containing various neural network models such as random forest, support vector machine, and XGboost is obtained. The aforementioned radiomics dataset and the actual age are obtained. The radiomics dataset is input into the neural network in the model set to predict the physiological age of the kidneys. It is important to note that the actual age is not input into the neural network, but rather serves a supervisory role. Various different machine learning methods are used to iteratively integrate and learn the neural network models. By comparing the functional indicators, the neural network model with the best predictive performance (best functional indicators) in the model set is found, which is the final "physiological age of the kidneys" prediction model.

[0114] Using the Shapely-based SHAP interpretable machine learning method, we found that the MShape and WShape datasets had the best prediction performance. Specifically, certain selected features showed significant predictive value for "renal physiological age." Figure 3 The median T1 intensity value of the cortex is one of the most significant features, exhibiting a wide range of SHAP values. Higher eigenvalues ​​(red) increase model predictions, while lower eigenvalues ​​(blue) decrease predictions. The difference in median T1 intensity values ​​between the medulla and cortex also shows a significant impact, with a wide range of SHAP values. Medullary volume and renal surface-to-volume ratio exhibit different patterns of influence.

[0115] Clinical studies have shown that an elevated median T1 value in the renal cortex typically indicates a higher degree of renal fibrosis. This fibrotic process not only increases the rigidity and density of renal tissue but also impairs glomerular filtration function, thereby accelerating renal aging. Therefore, an elevated median T1 value is logically correlated with an increased model-predicted value. Furthermore, in T1 images of normal healthy kidneys, the cortex and medulla are clearly distinguishable; this feature is called corticomedullary differentiation (CMD). A decrease in the difference in median T1 intensity between the medulla and cortex indicates indistinct CMD, another key signal of declining renal function. In cases of renal injury, the structural boundary between the cortex and medulla is no longer clear, reflecting pathological changes in the internal renal structure. Regarding renal shape parameters, a decrease in medullary volume and an increase in the renal surface area-to-volume ratio indicate renal atrophy and increased surface irregularity, both of which are biological markers of renal aging and disease progression. Decreased renal volume is usually accompanied by functional decline, while an increased surface area-to-volume ratio suggests renal surface irregularities that may be caused by fibrosis or other pathological changes.

[0116] After excluding participants from the training set, we recorded 254 cardiovascular events (IQR 12.9–14.4 years) during a median follow-up of 13.7 years. Figure 4As shown, participants in the top 10% or bottom 10% of the renal age difference (KAG) were defined as “extreme KAG”, and these individuals exhibited a significantly higher cumulative risk of cardiovascular events. Compared with individuals with normal KAG, the hazard ratio for cardiovascular events in the extreme KAG group was 1.45 (95% CI 1.10–1.90), after adjusting for age and sex, and further adjusting for traditional cardiovascular risk factors such as hypertension, diabetes, and overweight (defined as BMI greater than 25 kg / m²). 2 After adjusting for baseline eGFR, the hazard ratio was 1.37 (95% CI 1.04–1.80). This effect remained statistically significant even after adjusting for baseline eGFR.

[0117] The following are system embodiments corresponding to the above method embodiments. This embodiment can be implemented in conjunction with the above embodiments. The relevant technical details mentioned in the above embodiments are still valid in this embodiment, and will not be repeated here to reduce repetition. Accordingly, the relevant technical details mentioned in this embodiment can also be applied to the above embodiments.

[0118] This invention also proposes a device for predicting the physiological age of the kidneys based on multimodal MRI images, comprising:

[0119] The initial module acquires multiple pairs of kidney MRI images from healthy individuals. Each pair of kidney MRI images includes a kidney T1-weighted image and a kidney T1-mapped image, and each pair of kidney MRI images is labeled with the corresponding physiological age.

[0120] The semantic segmentation module performs semantic segmentation on each pair of kidney MRI images to obtain the first cortico-medullary segmentation mask of the kidney weighted image and the second cortico-medullary segmentation mask of the kidney mapping image.

[0121] The feature extraction module acquires radiomics data of the kidney T1-weighted image and the kidney T1-mapped image based on the first corticome segmentation mask and the second corticome segmentation mask, and obtains a first set of radiomics parameters of the kidney T1-weighted image and a second set of radiomics parameters of the kidney T1-mapped image.

[0122] The combined training module combines the first set of radiomics parameters and the second set of radiomics parameters, and filters the combined results to obtain the radiomics parameter with the highest correlation to the prediction of renal physiological age as the final omics parameter. The final omics parameter is used as the feature input of the renal MRI image pair into the neural network model to obtain the predicted age. The predicted age and the physiological age labeled by the renal MRI image pair are used to construct a loss function to train the neural network model and obtain the renal physiological age prediction model.

[0123] The age prediction module extracts the final omics parameters after semantic segmentation of the kidney MRI image pair to be predicted for physiological age. The kidney physiological age prediction model predicts the kidney physiological age based on the final omics parameters of the kidney MRI image pair to be predicted for physiological age.

[0124] The kidney physiological age prediction device based on multimodal MRI images, wherein the first set of radiomics parameters and the second set of radiomics parameters both include first-order radiomics parameters and higher-order radiomics parameters, and the first set of radiomics parameters also includes shapemics parameters;

[0125] The first-order radiomics parameters include the mean and median voxel gray values ​​of the renal cortex and medulla regions. The higher-order radiomics parameters include the gray-level co-occurrence matrix, gray-level dependence matrix, gray-level run matrix, gray-level size region matrix, and neighborhood gray-level difference matrix of the renal cortex and medulla regions. Shapemics parameters include the volume of the renal cortex and medulla regions, the long axis of the kidney, surface smoothness, and surface area-to-volume ratio.

[0126] The renal physiological age prediction device based on multimodal MRI images, wherein the process of filtering the combined results in the combined training module includes:

[0127] The first set of radiomics parameters and the second set of radiomics parameters are classified to obtain a kidney shape parameter dataset, a T1-weighted image feature parameter dataset, and a T1-mapped image feature parameter dataset. The above three are then combined to obtain multiple datasets.

[0128] Using multiple datasets as input, the model predicts the physiological age of the kidneys as output. The neural network model is trained using an interpretable machine learning method, and the dataset with the best prediction performance is selected as the final omics parameter.

[0129] The age prediction module also includes outputting the final omics parameters of the renal MRI image to be predicted for the physiological age of the kidney, as an interpretable feature for predicting its physiological age of the kidney.

[0130] like Figure 7 As shown, in another embodiment of the present invention, a first electronic device A is also provided, including the aforementioned kidney physiological age prediction device.

[0131] like Figure 8 As shown, the first electronic device A can also be connected to the data acquisition device C and the information display device D through a wired or wireless information transmission scheme. The data acquisition device C is used to acquire and obtain the kidney MRI images to be predicted for physiological age, and the information display device D is used to display the kidney physiological age prediction results, interpretable features, semantic segmentation results, etc. obtained by the present invention.

[0132] The information display device D can process and organize the data output by the first electronic device A based on an information display mechanism to improve the readability of the data. This information display mechanism can be manually preset, for example, visualizing the data output by the first electronic device A. It can be based on user-defined display parameters and / or attributes, such as the data range to be displayed and font, color, and scrolling options. It can present the user with their specified key information, such as kidney physiological age prediction results, interpretable features, and semantic segmentation results, allowing users to access this information more quickly without needing to access secondary pages or scroll through pages, thus saving user effort. Alternatively, this information display mechanism can be an artificial intelligence (AI) display model, which can learn the user's key information interests based on previous usage habits, such as viewing time, number of clicks, and number of edits, and automatically present the user with rich and necessary key information.

[0133] The present invention also provides a computer program product, which includes a computer program that can be stored on a readable storage medium. When the computer program is executed by a processor, the computer is able to execute the kidney physiological age prediction method provided by the above methods.

[0134] In another embodiment of the present invention, a storage medium VIII is provided for storing a computer program that performs the renal physiological age prediction method. It should be understood that the storage medium in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which serves as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DRRAM).

[0135] Figure 9 A schematic block diagram of a second electronic device 1000 that can be used to implement embodiments of the present invention is shown. The second electronic device 1000 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The second electronic device 1000 can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein. The second electronic device 1000 may be the same as or different from the first electronic device A.

[0136] The second electronic device 1000 includes a computing unit I, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory II (ROM) or a computer program loaded from storage medium VIII into random access memory (RAM) III. The RAM III may also store various programs and data required for the operation of the device 1000. The computing unit I, ROM II, and RAM III are interconnected via bus IV. An input / output (I / O) interface V is also connected to bus IV.

[0137] Multiple components in the second electronic device 1000 are connected to I / O interface V, including: input unit VI, such as a keyboard, mouse, etc.; output unit VII, such as various types of displays, speakers, etc.; storage medium VIII, such as a disk, optical disk, etc.; and communication unit IX, such as a network card, modem, wireless transceiver, etc. Communication unit IX allows the second electronic device 1000 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0138] The computing unit I can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of computing unit I include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit I performs the various methods and processes described above, such as method steps S1-S5. For example, in some embodiments, the methods can be implemented as computer software programs tangibly contained in a machine-readable medium, such as storage medium VIII. In some embodiments, part or all of the computer program can be loaded and / or installed on device 1000 via ROM II and / or communication unit IX. When the computer program is loaded into RAM III and executed by computing unit I, one or more steps of the methods described above can be performed. Alternatively, in other embodiments, computing unit I can be configured to perform methods by any other suitable means (e.g., by means of firmware).

[0139] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be applied to various fields suitable for the present invention. For those skilled in the art, other modifications can be easily made. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and illustrations shown and described herein.

Claims

1. A method for predicting renal physiological age based on multimodal MRI images, characterized in that, include: The initial step involves acquiring multiple modalities of kidney MRI images from healthy individuals. Each kidney MRI image pair includes a kidney T1-weighted image and a kidney T1-mapped image, and each kidney MRI image pair is labeled with the corresponding actual age. The semantic segmentation step involves performing semantic segmentation on each pair of kidney MRI images to obtain the first cortico-medullary segmentation mask of the kidney weighted image and the second cortico-medullary segmentation mask of the kidney mapping image. The feature extraction step involves performing multimodal feature extraction on the kidney T1-weighted image and the kidney T1-mapped image based on the first corticome segmentation mask and the second corticome segmentation mask, to obtain the first set of radiomics parameters of the kidney T1-weighted image and the second set of radiomics parameters of the kidney T1-mapped image. The combined training step involves combining the first set of radiomics parameters and the second set of radiomics parameters, and filtering the combined results to obtain the radiomics parameter with the highest correlation to the prediction of renal physiological age as the final omics parameter. The final omics parameter is then used as the feature input of the renal MRI image pair into the neural network model to obtain the predicted age. A loss function is constructed using the predicted age and the actual age labeled in the renal MRI image pair to train the neural network model, resulting in the renal physiological age prediction model. The age prediction step involves semantically segmenting the kidney MRI image pair to be predicted for physiological age and extracting the final omics parameters. The kidney physiological age prediction model then predicts the kidney physiological age based on the final omics parameters of the kidney MRI image pair to be predicted for physiological age. The process of filtering the combined results in this combined training step includes: The first set of radiomics parameters and the second set of radiomics parameters are classified to obtain a kidney shape parameter dataset, a T1-weighted image feature parameter dataset, and a T1-mapped image feature parameter dataset. The above three are then combined to obtain multiple datasets. Using multiple datasets and the actual age as input, the model predicts the physiological age of the kidneys as output. The neural network model is trained using an interpretable machine learning method, and the dataset with the best prediction performance is selected as the final omics parameter. The age prediction step also includes outputting the final omics parameters of the renal MRI image to be predicted for physiological age, as interpretable features for predicting its renal physiological age.

2. The method for predicting renal physiological age based on multimodal MRI images as described in claim 1, characterized in that, Both the first set of radiomics parameters and the second set of radiomics parameters include first-order radiomics parameters and higher-order radiomics parameters. The first set of radiomics parameters also includes shape omics parameters. The first-order radiomics parameters include the mean and median voxel gray values ​​of the renal cortex and medulla regions; the higher-order radiomics parameters include the gray co-occurrence matrix, gray dependence matrix, gray run matrix, gray size region matrix, and neighborhood gray difference matrix of the renal cortex and medulla regions; and the shapemics parameters include the volume of the renal cortex and medulla regions, the long diameter of the kidney, surface smoothness, and surface area-volume ratio.

3. The method for predicting renal physiological age based on multimodal MRI images as described in claim 1, characterized in that, These datasets include: The kidney shape parameter dataset, T1-weighted image feature parameter dataset, and T1-mapped image feature parameter dataset are combined in pairs and / or in triplicate to obtain the multiple datasets.

4. A device for predicting renal physiological age based on multimodal MRI images, characterized in that, include: The initial module acquires multiple modal pairs of kidney MRI images from healthy individuals. Each pair of kidney MRI images includes a kidney T1-weighted image and a kidney T1-mapped image, and each pair of kidney MRI images is labeled with the corresponding actual age. The semantic segmentation module performs semantic segmentation on each pair of kidney MRI images to obtain the first cortico-medullary segmentation mask of the kidney weighted image and the second cortico-medullary segmentation mask of the kidney mapping image. The feature extraction module performs multimodal feature extraction on the kidney T1-weighted image and the kidney T1-mapped image based on the first corticome segmentation mask and the second corticome segmentation mask, to obtain the first set of radiomics parameters of the kidney T1-weighted image and the second set of radiomics parameters of the kidney T1-mapped image. The combined training module combines the first set of radiomics parameters and the second set of radiomics parameters, and filters the combined results to obtain the radiomics parameter with the highest correlation to the prediction of kidney physiological age as the final omics parameter. The final omics parameters are used as features of the kidney MRI image pair and input into the neural network model to obtain the predicted age. The predicted age and the actual age labeled in the kidney MRI image pair are used to construct a loss function to train the neural network model and obtain the kidney physiological age prediction model. The age prediction module extracts the final omics parameters after semantic segmentation of the kidney MRI image pair to be predicted for physiological age. The kidney physiological age prediction model predicts the kidney physiological age based on the final omics parameters of the kidney MRI image pair to be predicted for physiological age. The process of filtering the combined results in this combined training module includes: The first set of radiomics parameters and the second set of radiomics parameters are classified to obtain a kidney shape parameter dataset, a T1-weighted image feature parameter dataset, and a T1-mapped image feature parameter dataset. The above three are then combined to obtain multiple datasets. Using multiple datasets as input, the model predicts the physiological age of the kidneys as output. The neural network model is trained using an interpretable machine learning method, and the dataset with the best prediction performance is selected as the final omics parameter. The age prediction module also includes outputting the final omics parameters of the renal MRI image to be predicted for the physiological age of the kidney, as an interpretable feature for predicting its physiological age of the kidney.

5. The renal physiological age prediction device based on multimodal MRI images as described in claim 4, characterized in that, Both the first set of radiomics parameters and the second set of radiomics parameters include first-order radiomics parameters and higher-order radiomics parameters. The first set of radiomics parameters also includes shape omics parameters. The first-order radiomics parameters include the mean and median voxel gray values ​​of the renal cortex and medulla regions; the higher-order radiomics parameters include the gray co-occurrence matrix, gray dependence matrix, gray run matrix, gray size region matrix, and neighborhood gray difference matrix of the renal cortex and medulla regions; the shape-mics parameters include the volume of the renal cortex and medulla regions, the long diameter of the kidney, surface smoothness, and surface area-volume ratio.

6. An electronic device, characterized in that, The device includes the renal physiological age prediction device as described in claim 4 or 5, wherein the electronic device is connected to an information display device, which is used to display the renal physiological age using user-set display parameters, attributes, or through an artificial intelligence model.

7. A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the kidney physiological age prediction method according to any one of claims 1-3.

8. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the kidney physiological age prediction method according to any one of claims 1-3.

Citation Information

Patent Citations

  • One-stop multi-delay artery spin marking kidney perfusion measurement method and system

    CN116309317A

  • KR20240176643A