Device, method and equipment for classifying liver perfusion states in on-orbit special-cause environment

Abstract

The embodiment of the invention provides a liver perfusion state classification device, method and equipment in an in-orbit special cause environment. The device comprises an acquisition module used for acquiring ultrasonic data of a user; the ultrasonic data comprises blood vessel data acquired by the first ultrasonic data acquisition equipment aiming at the abdominal trunk and portal vein of the user and ultrasonic elastography data acquired by the second ultrasonic data acquisition equipment aiming at the liver of the user; the processing module is used for extracting an index value of a target blood vessel index of the user from the blood vessel data; extracting an index value of a liver elasticity modulus index of the user from the ultrasonic elastography data; according to the index value of the target blood vessel index of the user and the index value of the liver elasticity modulus index of the user, the liver perfusion state of the user is classified through a liver perfusion state classification model. According to the technical scheme provided by the embodiment of the invention, the evaluation of the liver perfusion state can be realized in a special scene.

Description

[0001] Cross-references

[0002] This application claims priority to Chinese patent application number 2025101269121, the entire contents of which are incorporated by reference into this application. Technical Field

[0003] The present application relates to the field of computer technology, and in particular to an apparatus, method, and device for classifying liver perfusion status under an on-orbit specific environment. Background Art

[0004] Liver perfusion imaging (LPI) is primarily used to assess the blood flow status of the liver. It can accurately reproduce the actual state of the liver's vasculature and assist in the assessment of liver function. Currently, LPI is often performed using large-scale equipment such as computed tomography (CT) and magnetic resonance imaging (MRI). The results are then used to assess liver perfusion status.

[0005] However, this type of inspection equipment is often large and installed in fixed locations such as hospitals. Therefore, it cannot be used to complete liver perfusion status inspections in some outdoor emergency situations or on-orbit environments, resulting in the inability to assess liver blood flow and liver function. Therefore, a new solution is needed. Summary of the Invention

[0006] In view of the above problems, the present application proposes an on-orbit liver perfusion status classification device, method and equipment in a special environment to solve the above problems or at least partially solve the above problems.

[0007] In a first aspect of the present application, a device for classifying liver perfusion status for use in an on-orbit specific environment is provided, comprising:

[0008] an acquisition module, configured to acquire ultrasound data of a user, wherein the ultrasound data includes vascular data of the user's celiac trunk and portal vein acquired by a first ultrasound data acquisition device and ultrasound elastography data of the user's liver acquired by a second ultrasound data acquisition device;

[0009] a processing module, configured to: extract an index value of the user's target blood vessel index from the blood vessel data; extract an index value of the user's liver elastic modulus index from the ultrasound elastography data; determine input information of a pre-trained liver perfusion state classification model based on the index value of the user's target blood vessel index and the index value of the user's liver elastic modulus index; and input the input information into the liver perfusion state classification model to classify the user's liver perfusion state using the liver perfusion state classification model;

[0010] The liver perfusion status classification model is trained to classify the liver perfusion status.

[0011] A second aspect of the present application provides a method for classifying liver perfusion status, comprising:

[0012] Acquire ultrasound data of the user; the ultrasound data includes: vascular data of the user's celiac trunk and portal vein collected by a first ultrasound data acquisition device and ultrasound elastography data of the user's liver collected by a second ultrasound data acquisition device;

[0013] extracting an index value of a target blood vessel index of the user from the blood vessel data;

[0014] extracting an index value of the user's liver elastic modulus index from the ultrasonic elastography data;

[0015] determining input information of a pre-trained liver perfusion state classification model according to an index value of the user's target blood vessel index and an index value of the user's liver elastic modulus index;

[0016] inputting the input information into the liver perfusion state classification model to classify the liver perfusion state of the user using the liver perfusion state classification model;

[0017] The liver perfusion status classification model is trained to classify the liver perfusion status.

[0018] A third aspect of the present application provides a method for training a liver perfusion state classification model, comprising:

[0019] Acquire a training sample, wherein the training sample includes: an index value of a target blood vessel index of a sample user, an index value of a liver elastic modulus index of the sample user, and a real liver perfusion state category of the sample user;

[0020] determining input information of a liver perfusion state classification model to be trained according to the index value of the target blood vessel index of the sample user and the index value of the liver elastic modulus index of the sample user;

[0021] Inputting the input information into the liver perfusion state classification model to classify the liver perfusion state of the user using the liver perfusion state classification model to obtain a predicted liver perfusion state category of the sample user;

[0022] Optimizing parameters in the liver perfusion state classification model with the goal of minimizing the difference between the true liver perfusion state category and the predicted liver perfusion state category;

[0023] The liver perfusion status classification model is trained to classify the liver perfusion status.

[0024] The fourth aspect of the present application provides an electronic device. The electronic device includes: a memory and a processor, wherein:

[0025] The memory is used to store programs;

[0026] The processor is coupled to the memory and is used to execute the program stored in the memory to implement the methods provided in the second and third aspects above.

[0027] In a fifth aspect of the present application, a computer-readable storage medium storing a computer program is provided, and when the computer program is executed by a computer, the methods provided in the second and third aspects above can be implemented.

[0028] In a sixth aspect of the present application, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the methods provided in the second and third aspects above.

[0029] In the technical solution provided in the embodiment of the present application, the liver perfusion status classification model can classify the user's liver perfusion status by using the vascular data collected by the ultrasonic data acquisition equipment for the user's celiac trunk and portal vein and the ultrasonic elastic imaging data collected by the ultrasonic data acquisition equipment for the user's liver. In other words, when classifying the user's liver perfusion status, the liver perfusion status classification model does not rely on the magnetic resonance data of the liver, but relies on ultrasound data that is easily obtained in special scenarios. It can be seen that the technical solution provided in the embodiment of the present application can evaluate the user's liver perfusion status in an on-orbit special environment or an outdoor first aid scenario. Note: The on-orbit special environment refers to the special environment faced by astronauts during space flight, mainly including microgravity, radiation, circadian rhythm disorders, confined spaces, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0031] Figure 1a This is a structural block diagram of a liver perfusion status classification device provided in one embodiment of the present application;

[0032] Figure 1b This is a structural block diagram of a liver perfusion status classification method provided in one embodiment of the present application;

[0033] Figure 2 A flowchart of a model training method provided in one embodiment of the present application;

[0034] Figure 3 A schematic diagram of a liver perfusion status classification system provided in one embodiment of the present application;

[0035] Figure 4 A structural block diagram of an electronic device provided in one embodiment of the present application;

[0036] Figure 5 A flowchart of a model building method provided in one embodiment of the present application. DETAILED DESCRIPTION

[0037] In order to enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below based on the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0038] In addition, some of the processes described in the specification, claims and the above-mentioned figures of this application include multiple operations that appear in a specific order. These operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this document are used to distinguish different messages, devices, modules, etc., and do not represent the order of precedence, nor do they limit "first" and "second" to be different types.

[0039] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0040] In a normal liver, approximately three-quarters of its blood supply comes from the portal vein (specifically the hepatic portal vein), and approximately one-quarter comes from the hepatic artery. Under different physiological or pathological conditions, arterial and venous blood flow in the liver can change. For example, a relative increase in hepatic arterial blood flow may indicate liver malignancy or cirrhosis. Therefore, assessing liver perfusion status is important for improving the accuracy and specificity of liver disease diagnosis.

[0041] Currently, arterial spin labeling (ASL)-MRI has been applied to the detection of liver blood flow (LBF, i.e., the blood flow per 100g of liver tissue per unit time). Hepatic artery blood perfusion maps and portal vein blood perfusion maps can be obtained respectively through different labeling methods. Among them, the hepatic artery blood perfusion map is a liver blood perfusion map formed based on the blood supply of the hepatic artery, and the portal vein blood perfusion map is a liver blood perfusion map formed based on the blood supply of the portal vein. ASL-MRI technology uses labeled protons in the blood as endogenous contrast agents, which can realize non-invasive blood flow quantification.

[0042] However, due to the low popularity of magnetic resonance imaging equipment and poor operational convenience, spin artery labeling-MRI examinations are difficult to carry out in remote areas or special environments such as aerospace.

[0043] Considering that magnetic resonance data cannot be obtained in special scenarios, the embodiments of the present application provide a solution for predicting liver perfusion status based on ultrasound data to solve or partially solve the problem that liver perfusion status cannot be evaluated in special scenarios.

[0044] Figure 1a This is a structural block diagram of a liver perfusion status classification device provided in one embodiment of the present application. The liver perfusion status classification device may include but is not limited to: a device integrated in any terminal device such as a smart phone, tablet computer, PDA (Personal Digital Assistant), smart TV, laptop computer, desktop computer, smart wearable device, etc. Figure 1aAs shown, the apparatus may include: an acquisition module 101 and a processing module 102. Optionally, the apparatus further includes a display module for displaying the processing result (eg, classification result) of the apparatus, such as a screen in a terminal device.

[0045] The acquisition module 101 is used to acquire the ultrasound data of the user.

[0046] The ultrasound data includes: blood vessel data collected by a first ultrasound data acquisition device on the celiac trunk and portal vein of the user, and ultrasound elastic imaging data collected by a second ultrasound data acquisition device on the liver of the user.

[0047] It should be noted that, in practical applications, it is difficult to collect hepatic artery vascular data using ultrasound data acquisition equipment. Hepatic artery vascular data is related to celiac trunk vascular data, so celiac trunk vascular data can be collected to replace hepatic artery vascular data.

[0048] For example, the blood vessel data may include ultrasound spectrum data.

[0049] Exemplarily, the first ultrasound data acquisition device may be a pulsed Doppler ultrasound device.

[0050] Exemplarily, the second ultrasound data acquisition device may be an ultrasound device with an elastic imaging function.

[0051] Optionally, the first ultrasound data acquisition device and the second ultrasound data acquisition device may be two different ultrasound devices or the same ultrasound device. When the first ultrasound data acquisition device and the second ultrasound data acquisition device are the same ultrasound device, the same ultrasound device has both vascular data acquisition and ultrasound elastography data acquisition functions.

[0052] The first and second ultrasound data acquisition devices both acquire the above data in vitro and can send the acquired data to the acquisition module 101 .

[0053] The processing module 102 is configured to: extract an index value of the target blood vessel index of the user from the blood vessel data; and extract an index value of the liver elastic modulus index of the user from the ultrasound elastography data.

[0054] The index value of the user's target blood vessel index may be obtained by analyzing blood vessel data, such as ultrasonic spectrum data.

[0055] The index value of the liver elastic modulus index of the user may be obtained by analyzing the ultrasonic elastography data.

[0056] The target vascular indices include vascular indices related to the celiac trunk and portal vein.

[0057] The one or more vascular indicators related to the celiac trunk may include, but are not limited to, celiac trunk vascular morphology (e.g., vascular inner diameter) and celiac trunk hemodynamic parameters. The celiac trunk hemodynamic parameters may include, but are not limited to, peak systolic blood flow velocity, end-diastolic blood flow velocity, mean blood flow velocity, resistance index, pulsatility index, and systolic / diastolic ratio.

[0058] The one or more vascular indicators related to the portal vein may include, but are not limited to, portal vein morphology (e.g., vascular inner diameter) and portal vein hemodynamic parameters. Portal vein hemodynamic parameters may include, but are not limited to, maximum blood flow velocity and minimum blood flow velocity.

[0059] In addition, the blood flow of the celiac trunk and the blood flow of the portal vein can also be calculated. Furthermore, the hemodynamic parameters of the celiac trunk can also include the blood flow of the celiac trunk. Furthermore, the hemodynamic parameters of the portal vein can also include the blood flow of the portal vein.

[0060] Optionally, in order to improve the classification accuracy of the model, key arterial indices that affect the liver perfusion state can be screened from multiple vascular indices related to the celiac trunk and portal vein to serve as target arterial indices.

[0061] In an optional embodiment, the processing module 102 is further configured to: extract the index value of the target blood vessel index of the user from the blood vessel data according to the index screening result;

[0062] The index screening result includes the target blood vessel index obtained through the index screening;

[0063] The indicator screening result is based on sample data, and uses the Extreme Gradient Boosting (Xgboost) algorithm to screen out the target vascular indicator related to the liver perfusion indicator from multiple vascular indicators, and the multiple vascular indicators are related to the celiac trunk and the portal vein.

[0064] The liver perfusion index may include but is not limited to: a first liver blood perfusion volume formed by blood supply from the hepatic artery, a second liver blood perfusion volume formed by blood supply from the portal vein, and a ratio between the first liver blood perfusion volume and the second liver blood perfusion volume.

[0065] The XgBoost algorithm was used as the base classifier for feature selection / screening (i.e., the vascular indicators mentioned above). XgBoost is a decision tree-based gradient boosting algorithm that optimizes model performance by continuously adding new decision trees. It offers advantages such as high efficiency, flexibility, and ease of tuning. In XgBoost, each feature is assigned an importance score, which indicates its contribution to the model. This score is used to assess the importance of each feature and select the features that contribute most to the model. An XGBoost model was constructed using the first hepatic blood flow perfusion volume (LBF-A) supplied by the hepatic artery, the second hepatic blood flow perfusion volume (LBF-P) supplied by the portal vein, and the ratio between the first and second hepatic blood flow perfusion volumes (LBF-A / LBF-P) as model outputs. Based on the feature importance assessment results, key features that contribute most to the three output variables were retained for use in constructing the final hepatic perfusion status classification model.

[0066] In some embodiments, the target vessel index to be extracted may be stored in advance in a storage module (e.g., a memory) of the device. Thus, the subsequent processing module 101 may obtain the target vessel index to be extracted from the storage module of the device and complete data extraction based on the target vessel index.

[0067] The processing module 102 is further configured to determine input information of a pre-trained liver perfusion state classification model according to the index value of the user's target blood vessel index and the index value of the user's liver elastic modulus index.

[0068] The liver perfusion status classification model is trained to classify the liver perfusion status.

[0069] Among them, the liver perfusion classification model can be implemented based on machine learning models such as XGboost, support vector machine, naive Bayes, and deep learning models.

[0070] The liver perfusion state classification model may be trained based on training samples, including the target blood vessel index value of the sample user, the liver elastic modulus index value of the sample user, and the actual liver perfusion state category of the sample user.

[0071] In actual applications, multiple liver perfusion status categories can be divided according to the range of change of the liver perfusion index of the sample data, for example, it can include three categories: reduced, normal, and increased, or it can include: normal, slightly reduced, moderately reduced, severely reduced, and / or normal, slightly increased, moderately increased, and severely increased. The specific division will be set according to the actual data change distribution, and the embodiments of the present application do not make specific restrictions on this. For example, based on the range of change of the liver perfusion index of the sample data, the threshold values ​​corresponding to the multiple liver perfusion status categories (for example, the upper and lower limits of the liver perfusion index) can be determined, and the sample users can be classified based on the threshold value to determine the user's true liver perfusion status category. The liver perfusion status category is used to indicate the liver perfusion status or to indicate the level of the liver perfusion status.

[0072] In an optional embodiment, the parameters of the liver perfusion state classification model are optimized to minimize the difference between the predicted liver perfusion state category of the sample user and the actual liver perfusion state category; the predicted liver perfusion state category is predicted by the liver perfusion state classification model based on the index value of the target blood vessel index and the index value of the liver elastic modulus index of the sample user. The specific training method will be described in detail in the following embodiments.

[0073] In some embodiments, the input information may include: an index value of the user's target blood vessel index and an index value of the user's liver elastic modulus index.

[0074] The processing module 102 is further configured to input the input information into the liver perfusion state classification model, so as to classify the liver perfusion state of the user using the liver perfusion state classification model.

[0075] Among them, the liver perfusion status classification model can extract features from the input information and then classify the user's liver perfusion status based on the extracted features.

[0076] When the liver perfusion status classification model is a deep learning model architecture that combines a convolutional neural network and a fully connected layer, the convolutional neural network can be used to extract features from the input information to obtain features, and then the fully connected layer can be used to classify the user's liver perfusion status based on the features.

[0077] In the technical solution provided in the embodiment of the present application, the liver perfusion state classification model uses the vascular data collected by the ultrasonic data acquisition device for the user's celiac trunk and portal vein, as well as the ultrasonic elastic imaging data collected by the ultrasonic data acquisition device for the user's liver, to classify the user's liver perfusion state. In other words, when classifying the user's liver perfusion state, the liver perfusion state classification model does not rely on the magnetic resonance data of the liver, but relies on ultrasound data that is easily obtained in special scenarios. It can be seen that the technical solution provided in the embodiment of the present application can evaluate the user's liver perfusion state in some special scenarios, such as aerospace scenarios and outdoor first aid scenarios.

[0078] The following describes a method for determining the true liver perfusion status category of a sample user. The method includes the following steps 201 to 206:

[0079] 201. Obtain a first liver blood perfusion map, a second liver blood perfusion map, and a liver structure map of the sample user.

[0080] The first hepatic blood perfusion map is a hepatic blood perfusion map formed based on the blood supply from the hepatic artery, and the second hepatic blood perfusion map is a hepatic blood perfusion map formed based on the blood supply from the portal vein.

[0081] The first liver blood perfusion map, the second liver blood perfusion map, and the liver structure map of the sample user may be obtained based on data collected by a computer tomography device and / or a magnetic resonance imaging device.

[0082] Optionally, the liver structure image may be an image of the user's liver acquired by a computed tomography device or an image of the user's liver acquired by a magnetic resonance imaging device.

[0083] For example, the structural image is obtained by performing a magnetic resonance scan of the liver using T1-weighted imaging to obtain morphological and structural information of the liver. In other words, the structural image includes morphological and structural information of the liver.

[0084] Exemplarily, the first liver blood perfusion map and the second liver blood perfusion map are determined based on liver MRI data of the sample user, wherein the liver MRI data may include: ASL sequences acquired by a magnetic resonance imaging device.

[0085] 202. Determine a segmentation mask for segmenting the liver according to the liver structure map.

[0086] In some embodiments, the segmentation mask may be manually drawn. For example, a liver structure diagram is displayed on an interface, and a segmentation mask for segmenting the liver is determined based on a drawing operation on the liver structure diagram on the interface.

[0087] In other embodiments, the target liver structure image is segmented using a trained liver segmentation model to obtain a segmentation mask for segmenting the liver.

[0088] The liver segmentation model may be a deep learning model. In an optional embodiment, the liver segmentation model may be a deep learning model with an encoding-decoding structure.

[0089] The training samples of the liver segmentation model may include: a sample liver structure image and an expected segmentation result.

[0090] The desired segmentation result can be obtained through manual annotation, which is not specifically limited in the embodiments of the present application.

[0091] In some embodiments, the sample liver structure map can be input into the liver segmentation model to obtain the actual segmentation result of the liver segmentation model, so as to minimize the difference between the expected segmentation result and the actual segmentation result and optimize the parameters of the liver segmentation model.

[0092] Optionally, the first liver blood perfusion map, the second liver blood perfusion map and the liver structure map of the sample user are obtained by image registration of the original first liver blood perfusion map, the original second liver blood perfusion map and the original liver structure map of the sample user.

[0093] Image registration refers to the process of spatially aligning two or more images so that they have consistent anatomical structures or features in the same spatial coordinate system.

[0094] Image registration is achieved by finding a spatial transformation relationship that aligns the reference image and the image to be registered, ensuring that the same anatomical point or feature point in the image has the same position in both images. This alignment is achieved by calculating the transformation parameters between the two images, which typically include operations such as translation, rotation, and scaling.

[0095] Using image registration technology, the original first liver blood perfusion map, the original second liver blood perfusion map, and the original liver structure map are spatially aligned to obtain the first liver blood perfusion map, the second liver blood perfusion map, and the liver structure map. The same anatomical feature point of the liver has the same position or coordinates in the first liver blood perfusion map, the second liver blood perfusion map, and the liver structure map. The liver can have multiple anatomical feature points, such as vascular branch points.

[0096] In an optional embodiment, the original first and second liver blood perfusion maps can be converted to the coordinate system of the original liver structure map to obtain the first and second liver blood perfusion maps. In this embodiment, the liver structure map is also the original liver structure map. In other words, image registration between the original first and second liver blood perfusion maps and the original liver structure map can be achieved by simply performing coordinate conversion on the original first and second liver blood perfusion maps.

[0097] Optionally, a non-rigid registration algorithm based on feature point matching can be used to transform the original first and second liver blood perfusion maps into the coordinate system of the original liver structure map, thereby obtaining the first and second liver blood perfusion maps. Feature points are also anatomical feature points. The transformation process involves a non-rigid transformation, meaning that the distance between two points in one image will change after being transformed into another image.

[0098] 203. Determine a liver region in the first liver blood perfusion map and the second liver blood perfusion map according to the segmentation mask.

[0099] 204. Eliminate abnormally distributed voxel values ​​in the liver region of the first liver blood perfusion map, and determine a first liver blood perfusion volume of the sample user formed by blood supply through the hepatic artery based on an average value of the remaining voxel values ​​in the liver region of the first liver blood perfusion map.

[0100] In an optional embodiment, a distribution of voxel values ​​within the liver region of the first hepatic blood perfusion map is determined, the distribution including a mean value μ and a standard deviation σ; and based on the distribution, voxel values ​​with abnormal distribution within the liver region of the first hepatic blood perfusion map are identified. Exemplarily, voxel values ​​outside the range [μ-3σ, μ+3σ] are identified as voxel values ​​with abnormal distribution.

[0101] The remaining voxel values ​​in the liver region of the first liver blood perfusion map refer to other voxel values ​​in the liver region of the first liver blood perfusion map except for the voxel values ​​with abnormal distribution.

[0102] 205. Eliminate abnormally distributed voxel values ​​in the liver region of the second liver blood perfusion map, and determine the second liver blood perfusion volume of the sample user formed by portal vein blood supply based on the average value of the remaining voxel values ​​in the liver region of the second liver blood perfusion map.

[0103] In an optional embodiment, a distribution of voxel values ​​within the liver region of the second hepatic blood perfusion map is determined, the distribution including a mean value μ and a standard deviation σ; and based on the distribution, voxel values ​​with abnormal distribution within the liver region of the second hepatic blood perfusion map are identified. Exemplarily, voxel values ​​outside the range [μ-3σ, μ+3σ] are identified as voxel values ​​with abnormal distribution.

[0104] The remaining voxel values ​​in the liver region of the second hepatic blood perfusion map refer to other voxel values ​​in the liver region of the second hepatic blood perfusion map except for the voxel values ​​with abnormal distribution.

[0105] 206. Determine a real liver perfusion status category of the sample user based on a first liver blood perfusion volume of the sample user formed by blood supply through the hepatic artery and a second liver blood perfusion volume of the sample user formed by blood supply through the portal vein.

[0106] Based on the first hepatic blood perfusion volume of the sample user formed by the hepatic artery blood supply and the second hepatic blood perfusion volume of the sample user formed by the portal vein blood supply, the ratio of the first hepatic blood perfusion volume of the sample user formed by the hepatic artery blood supply to the second hepatic blood perfusion volume of the sample user formed by the portal vein blood supply is determined; based on the first hepatic blood perfusion volume of the sample user formed by the hepatic artery blood supply, the second hepatic blood perfusion volume of the sample user formed by the portal vein blood supply and the ratio, the actual liver perfusion state category of the sample user is determined by comparing with the threshold value described above.

[0107] In actual applications, ASL images have low quality and are prone to abnormal interference. In addition, there are large blood vessels and bile ducts inside the liver, which means that there is interference from large blood vessels and bile ducts in the perfusion map. In order to eliminate these interferences, the embodiment of the present application eliminates the abnormal values ​​caused by these interferences and determines the final first liver blood flow perfusion volume and second liver blood flow perfusion volume based on the remaining values, which helps to improve the accuracy of the first liver blood flow perfusion volume and the second liver blood flow perfusion volume, thereby improving the accuracy of the training samples, and further improving the training effectiveness and classification accuracy of the liver perfusion state classification model.

[0108] The total blood flow into the liver per unit time can be derived based on the ultrasonic data collected by the first ultrasonic data acquisition device, and the ASL liver blood perfusion value represents the blood flow into every 100g of liver tissue per unit time. Since the liver volume of different individuals is different, the correlation between the two is not exactly the same, which may affect the performance of the classification model. However, in extreme scenarios, it is impossible to obtain the size of the user's liver. Further research found that the liver volume has an individual correlation with the user's body surface area (BSA), and the body surface area can be calculated based on height, gender and weight. Inputting the body surface area into the classification model can eliminate differences in user physical signs to improve the robustness of the classification model.

[0109] Therefore, in some embodiments, the acquisition module 101 is further configured to: acquire the height, gender, and weight of the user. The processing module 102 is further configured to: determine the body surface area of ​​the user based on the height, gender, and weight of the user.

[0110] For example, for men, the following formula can be used for calculation:

[0111] BSA=0.0057×H+0.0121×W+0.0882 (1)

[0112] Among them, H is height and W is weight.

[0113] For women, the following formula can be used for calculation:

[0114] BSAf(H,W)=0.0073×H+0.0127×W+0.2106 (2)

[0115] Among them, H is height and W is weight.

[0116] In some embodiments, when the processing module 102 determines the input information of the pre-trained liver perfusion state classification model based on the index value of the user's target blood vessel index and the index value of the user's liver elastic modulus index, it is specifically configured to:

[0117] Input information of a pre-trained liver perfusion state classification model is determined according to the index value of the user's target blood vessel index, the index value of the user's liver elastic modulus index, and the user's body surface area.

[0118] For example, the input information may include an index value of the target blood vessel index of the user, an index value of the liver elastic modulus index of the user, and a body surface area of ​​the user.

[0119] Furthermore, it should be noted that when the input of the liver perfusion state classification model involves body surface area, the training samples for training the liver perfusion state classification model must also include the body surface area of ​​the sample user. This helps the liver perfusion state classification model learn the correlation between body surface area and liver perfusion state during the training phase.

[0120] Figure 1b This is a flow chart of a method for classifying liver perfusion status provided in one embodiment of the present application. The execution subject of the method may be a terminal device or a server device, and this embodiment of the present application does not specifically limit this. In the above-mentioned on-orbit special environment, the above-mentioned execution subject may be a terminal device. Figure 1b As shown, the method may include the following steps:

[0121] 301. Obtain the user's ultrasound data.

[0122] The ultrasound data includes: blood vessel data collected by a first ultrasound data acquisition device on the celiac trunk and portal vein of the user, and ultrasound elastic imaging data collected by a second ultrasound data acquisition device on the liver of the user.

[0123] 302. Extract an index value of the user's target blood vessel index from the blood vessel data.

[0124] 303. Extract an index value of the user's liver elastic modulus index from the ultrasonic elastography data.

[0125] 304. Determine input information of a pre-trained liver perfusion state classification model according to the index value of the user's target blood vessel index and the index value of the user's liver elastic modulus index.

[0126] 305. Input the input information into the liver perfusion state classification model to classify the liver perfusion state of the user using the liver perfusion state classification model.

[0127] The liver perfusion status classification model is trained to classify the liver perfusion status.

[0128] In the technical solution provided in the embodiment of the present application, the liver perfusion state classification model uses the vascular data collected by the ultrasonic data acquisition device for the user's celiac trunk and portal vein, as well as the ultrasonic elastic imaging data collected by the ultrasonic data acquisition device for the user's liver, to classify the user's liver perfusion state. In other words, when classifying the user's liver perfusion state, the liver perfusion state classification model does not rely on the magnetic resonance data of the liver, but relies on ultrasound data that is easily obtained in special scenarios. It can be seen that the technical solution provided in the embodiment of the present application can evaluate the user's liver perfusion state in some special scenarios, such as aerospace scenarios and outdoor first aid scenarios.

[0129] Optionally, the above method further includes:

[0130] Get the user's height, gender, and weight;

[0131] Determining the user's body surface area based on the user's height, gender, and weight;

[0132] Determining input information of a pre-trained liver perfusion state classification model according to the index value of the user's target blood vessel index and the index value of the user's liver elastic modulus index includes:

[0133] Input information of a pre-trained liver perfusion state classification model is determined according to the index value of the user's target blood vessel index, the index value of the user's liver elastic modulus index, and the user's body surface area.

[0134] Optionally, the above method further includes:

[0135] extracting the index value of the user's target blood vessel index from the blood vessel data according to the index screening result;

[0136] The index screening result includes the target blood vessel index obtained through the index screening;

[0137] The indicator screening result is based on sample data, and uses the extreme gradient boosting Xgboost algorithm to screen out the target vascular indicator related to the liver perfusion indicator from multiple vascular indicators, and the multiple vascular indicators are related to the celiac trunk and the portal vein.

[0138] Optionally, in the above method, the liver perfusion index includes: a first liver blood perfusion volume formed by blood supply from the hepatic artery, a second liver blood perfusion volume formed by blood supply from the portal vein, and a ratio between the first liver blood perfusion volume and the second liver blood perfusion volume.

[0139] Optionally, in the above method, the training samples of the liver perfusion state classification model include: the index value of the target blood vessel index of the sample user, the index value of the liver elastic modulus index of the sample user, and the real liver perfusion state category of the sample user;

[0140] The parameters of the liver perfusion state classification model are optimized with the goal of minimizing the difference between the predicted liver perfusion state category of the sample user and the true liver perfusion state category; the predicted liver perfusion state category is predicted by the liver perfusion state classification model based on the index value of the target blood vessel index of the sample user and the index value of the liver elastic modulus index of the sample user.

[0141] Optionally, the above method further includes:

[0142] Obtaining a first liver blood perfusion map, a second liver blood perfusion map, and a liver structure map of the sample user, wherein the first liver blood perfusion map is a liver blood perfusion map formed based on hepatic artery blood supply, and the second liver blood perfusion map is a liver blood perfusion map formed based on portal vein blood supply;

[0143] determining a segmentation mask for segmenting the liver according to the liver structure map;

[0144] determining a liver region in the first liver blood perfusion map and the second liver blood perfusion map according to the segmentation mask;

[0145] Eliminating abnormally distributed voxel values ​​in the liver region of the first hepatic blood perfusion map, and determining a first hepatic blood perfusion volume of the sample user formed by blood supply through the hepatic artery based on an average value of the remaining voxel values ​​in the liver region of the first hepatic blood perfusion map;

[0146] Eliminating abnormally distributed voxel values ​​in the liver region of the second hepatic blood perfusion map, and determining a second hepatic blood perfusion volume of the sample user formed by portal vein blood supply based on an average value of remaining voxel values ​​in the liver region of the second hepatic blood perfusion map;

[0147] The real liver perfusion state category of the sample user is determined according to the first liver blood perfusion volume formed by the hepatic artery blood supply of the sample user and the second liver blood perfusion volume formed by the portal vein blood supply of the sample user.

[0148] Optionally, the above method further includes:

[0149] Determining a distribution of voxel values ​​in the liver region of the first liver blood perfusion map, the distribution including: a mean value and a standard deviation;

[0150] According to the distribution, abnormal voxel values ​​are distributed within the liver region of the first liver blood perfusion map.

[0151] It should be noted that for any steps not fully described in detail in the methods provided in the embodiments of the present application, reference may be made to the corresponding contents in the aforementioned embodiments, and no further elaboration is required here. Furthermore, in addition to the aforementioned steps, the methods provided in the embodiments of the present application may also include some or all of the other steps in the aforementioned embodiments, for which reference may be made to the corresponding contents in the aforementioned embodiments, and no further elaboration is required here.

[0152] Figure 2 This is a flow chart of a method for training a liver perfusion state classification model provided in one embodiment of the present application. The execution subject of the method can be a terminal device or a server device, and this embodiment of the present application does not specifically limit this. Figure 2 As shown, the method may include the following steps:

[0153] 401. Obtain training samples.

[0154] The training sample includes: the index value of the target blood vessel index of the sample user, the index value of the liver elastic modulus index of the sample user, and the real liver perfusion state category of the sample user.

[0155] 402. Determine input information of a liver perfusion state classification model to be trained based on the index value of the target blood vessel index of the sample user and the index value of the liver elastic modulus index of the sample user.

[0156] 403. Input the input information into the liver perfusion state classification model, so as to classify the liver perfusion state of the user using the liver perfusion state classification model to obtain a predicted liver perfusion state category of the sample user.

[0157] 404. Optimize parameters in the liver perfusion state classification model with the goal of minimizing the difference between the true liver perfusion state category and the predicted liver perfusion state category.

[0158] The liver perfusion status classification model is trained to classify the liver perfusion status.

[0159] It should be noted that for any steps not fully described in detail in the methods provided in the embodiments of the present application, reference may be made to the corresponding contents in the aforementioned embodiments, and no further elaboration is required here. Furthermore, in addition to the aforementioned steps, the methods provided in the embodiments of the present application may also include some or all of the other steps in the aforementioned embodiments, for which reference may be made to the corresponding contents in the aforementioned embodiments, and no further elaboration is required here.

[0160] Figure 5 Schematic diagram of the method for constructing liver perfusion state classification provided in the embodiment of this application. Figure 5 As shown, the construction method includes:

[0161] S61 . Collect the sample user's hepatic artery ASL sequence, portal vein ASL sequence, and structural MRI data.

[0162] S62. Data Preprocessing

[0163] The sample user's hepatic artery ASL sequence, portal vein ASL sequence, and structural MRI data are preprocessed to obtain hepatic artery perfusion distribution features and portal vein perfusion distribution features. The hepatic artery perfusion distribution features and portal vein perfusion distribution features include: the sample user's first hepatic blood flow perfusion volume formed by the hepatic artery blood supply and the sample user's second hepatic blood flow perfusion volume formed by the portal vein blood supply.

[0164] Data preprocessing includes:

[0165] A first hepatic blood perfusion map of the sample user is obtained based on the hepatic artery ASL sequence; a second hepatic blood perfusion map of the sample user is obtained based on the portal vein ASL sequence; a liver structural map of the sample user is obtained based on the structural MRI data; and based on the first hepatic blood perfusion map, the second hepatic blood perfusion map, and the liver structural map, a first hepatic blood perfusion volume of the sample user formed by blood supply through the hepatic artery and a second hepatic blood perfusion volume of the sample user formed by blood supply through the portal vein are determined. The specific implementation of each step can be found in the corresponding content of the above embodiments and will not be repeated here.

[0166] S63. Analysis of liver filling status.

[0167] Specifically, the actual liver perfusion status category of the sample user is determined based on the first hepatic blood perfusion volume of the sample user, which is provided by the hepatic artery, and the second hepatic blood perfusion volume of the sample user, which is provided by the portal vein. The specific implementation of the determination step can be found in the corresponding content of the above embodiments and will not be repeated here.

[0168] S64. Collect ultrasound data of the celiac trunk and portal vein.

[0169] S65. Data preprocessing

[0170] The ultrasound data of the celiac trunk and portal vein are preprocessed to obtain the vascular morphology (such as the inner diameter of the vessel) and hemodynamic parameters.

[0171] S66. Key feature extraction.

[0172] The target vascular index related to the liver perfusion index is screened out from the vascular morphology and hemodynamic parameters. The specific screening process is referred to the corresponding content in the above embodiments and will not be repeated here.

[0173] S67. Perform liver ultrasound elastography and collect height and weight data.

[0174] S68. Data preprocessing

[0175] The liver ultrasound elastography data, height, and weight were processed to obtain the liver elastic modulus and body surface area.

[0176] S69. Constructing a liver perfusion status classification model

[0177] Based on the analysis results of step S63, the key features extracted in step S66, the liver elastic modulus and the human body surface area, a liver perfusion state classification model is constructed.

[0178] The input parameters of the liver perfusion status classification model are the key features extracted from S66, the liver elastic modulus, and the human body surface area.

[0179] Based on the analysis results of step S63, the key features extracted in step S66, the liver elastic modulus, and the body surface area, training samples are prepared and used to train the liver perfusion state classification model. The specific training process can be found in the corresponding content of the above embodiments and will not be repeated here.

[0180] Figure 3 Schematic diagram of the liver perfusion status classification system provided in the embodiment of the present application. Figure 3 As shown, the system includes: a first ultrasonic data acquisition device 50, a second ultrasonic data acquisition device 51 and a liver perfusion status classification device 52. The first ultrasonic data acquisition device 50 and the second ultrasonic data acquisition device 51 are respectively connected to the liver perfusion status classification device 52 for communication.

[0181] The liver perfusion status classification device 52 may include an acquisition module 521 and a processing module 522. Optionally, the device 52 may further include a display module 523.

[0182] It should be noted here that the specific implementation and interaction of each device in the system provided in the embodiments of the present application can be found in the corresponding contents of the above embodiments and will not be repeated here.

[0183] Figure 4 FIG. 1 shows a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. Figure 4 As shown, the electronic device includes a memory 1101 and a processor 1102. The memory 1101 can be configured to store various other data to support operations on the electronic device. Examples of such data include instructions for any application or method operating on the electronic device. The memory 1101 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read only memory (EEPROM), erasable programmable read only memory (EPROM), programmable read only memory (PROM), read only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0184] The memory 1101 is used to store programs;

[0185] The processor 1102 is coupled to the memory 1101 and is configured to execute the program stored in the memory 1101 to implement the method provided by any of the above method embodiments.

[0186] Further, if Figure 4 As shown, the electronic device also includes: a communication component 1103, a display 1104, a power component 1105, an audio component 1106 and other components. Figure 4 Only some components are shown schematically, which does not mean that the electronic device only includes Figure 4 Components shown.

[0187] Accordingly, an embodiment of the present application further provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a computer, the method provided by any of the above method embodiments can be implemented.

[0188] An embodiment of the present application also provides a computer program product, including a computer program, which, when executed by a processor, can implement the method provided by any of the above method embodiments.

[0189] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0190] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, or of course, by hardware. Based on this understanding, the above technical solution, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM (Read Only Memory) / RAM (Random Access Memory), a disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or certain parts of the embodiments.

[0191] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A liver perfusion status classification device for use in an on-orbit special environment, characterized in that: include: An acquisition module is used to: acquire ultrasound data of a user; The ultrasound data includes: blood vessel data collected by a first ultrasound data acquisition device on the celiac trunk and portal vein of the user and ultrasound elastic imaging data collected by a second ultrasound data acquisition device on the liver of the user; A processing module, used to: extract an index value of the target blood vessel index of the user from the blood vessel data; extract an index value of the liver elastic modulus index of the user from the ultrasound elastic imaging data; determine input information of a pre-trained liver perfusion state classification model according to the index value of the target blood vessel index of the user and the index value of the liver elastic modulus index of the user; input the input information into the liver perfusion state classification model to classify the liver perfusion state of the user using the liver perfusion state classification model; The liver perfusion status classification model is trained to classify the liver perfusion status.

2. The device according to claim 1, characterized in that The acquisition module is further used to: acquire the height, weight and gender of the user; The processing module is further used to: determine the body surface area of ​​the user according to the height, weight and gender of the user; When the processing module determines the input information of the pre-trained liver perfusion state classification model according to the index value of the target blood vessel index of the user and the index value of the liver elastic modulus index of the user, it is specifically used to: Input information of a pre-trained liver perfusion state classification model is determined according to the index value of the target blood vessel index of the user, the index value of the liver elastic modulus index of the user, and the body surface area of ​​the user.

3. The device according to claim 1, characterized in that The processing module is used to extract the index value of the target blood vessel index of the user from the blood vessel data according to the index screening result; Wherein, the index screening result includes the target blood vessel index obtained through the index screening; The index screening result is based on sample data, and uses the extreme gradient boosting Xgboost algorithm to screen out the target vascular index related to the liver perfusion index from multiple vascular indexes, and the multiple vascular indexes are related to the celiac trunk and the portal vein.

4. The device according to claim 1, characterized in that The liver perfusion index includes: a first liver blood perfusion volume formed by blood supply from the hepatic artery, a second liver blood perfusion volume formed by blood supply from the portal vein, and a ratio between the first liver blood perfusion volume and the second liver blood perfusion volume.

5. The device according to any one of claims 1 to 4, characterized in that The training samples of the liver perfusion state classification model include: the index value of the target blood vessel index of the sample user, the index value of the liver elastic modulus index of the sample user, and the real liver perfusion state category of the sample user; The parameters of the liver perfusion state classification model are optimized with the goal of minimizing the difference between the predicted liver perfusion state category of the sample user and the true liver perfusion state category; the predicted liver perfusion state category is predicted by the liver perfusion state classification model based on the index value of the target blood vessel index of the sample user and the index value of the liver elastic modulus index of the sample user.

6. The device according to claim 5, characterized in that The method for determining the real liver perfusion status category of the sample user includes: Acquire a first liver blood perfusion map, a second liver blood perfusion map, and a liver structure map of the sample user, wherein the first liver blood perfusion map is a liver blood perfusion map formed based on hepatic artery blood supply, and the second liver blood perfusion map is a liver blood perfusion map formed based on portal vein blood supply; Determining a segmentation mask for segmenting the liver according to the liver structure map; determining a liver region in the first liver blood perfusion map and the second liver blood perfusion map according to the segmentation mask; Eliminating voxel values ​​with abnormal distribution in the liver region of the first liver blood perfusion map, and determining a first liver blood perfusion volume formed by blood supply through the hepatic artery of the sample user based on an average value of remaining voxel values ​​in the liver region of the first liver blood perfusion map; Eliminating abnormally distributed voxel values ​​in the liver region of the second liver blood perfusion map, and determining a second liver blood perfusion volume of the sample user formed by portal vein blood supply based on an average value of remaining voxel values ​​in the liver region of the second liver blood perfusion map; The real liver perfusion state category of the sample user is determined according to the first liver blood perfusion volume formed by the hepatic artery blood supply of the sample user and the second liver blood perfusion volume formed by the portal vein blood supply of the sample user.

7. The device according to claim 6, characterized in that The method for determining the real liver perfusion status category of the sample user also includes: Determine the distribution of voxel values ​​in the liver region of the first liver blood perfusion map, wherein the distribution includes: a mean value and a standard deviation; According to the distribution condition, abnormal voxel values ​​are distributed in the liver region of the first liver blood perfusion map.

8. A method for classifying liver perfusion status, characterized in that: include: Obtaining the user's ultrasound data; The ultrasound data includes: blood vessel data collected by a first ultrasound data acquisition device on the celiac trunk and portal vein of the user and ultrasound elastic imaging data collected by a second ultrasound data acquisition device on the liver of the user; extracting an index value of a target blood vessel index of the user from the blood vessel data; Extracting an index value of the liver elastic modulus index of the user from the ultrasonic elastic imaging data; Determining input information of a pre-trained liver perfusion state classification model according to an index value of the target blood vessel index of the user and an index value of the liver elastic modulus index of the user; Inputting the input information into the liver perfusion state classification model to classify the liver perfusion state of the user using the liver perfusion state classification model; The liver perfusion status classification model is trained to classify the liver perfusion status.

9. The method according to claim 8, characterized in that Also includes: Obtain the height, gender and weight of the user; Determining a body surface area of ​​the user according to the height, gender and weight of the user; Determining input information of a pre-trained liver perfusion state classification model according to the index value of the target blood vessel index of the user and the index value of the liver elastic modulus index of the user, including: Input information of a pre-trained liver perfusion state classification model is determined according to the index value of the target blood vessel index of the user, the index value of the liver elastic modulus index of the user, and the body surface area of ​​the user.

10. The method according to claim 8, characterized in that Also includes: extracting the index value of the target blood vessel index of the user from the blood vessel data according to the index screening result; Wherein, the index screening result includes the target blood vessel index obtained through the index screening; The index screening result is based on sample data, and uses the extreme gradient boosting Xgboost algorithm to screen out the target vascular index related to the liver perfusion index from multiple vascular indexes, and the multiple vascular indexes are related to the celiac trunk and the portal vein.

11. The method according to claim 8, characterized in that The liver perfusion index includes: a first liver blood perfusion volume formed by blood supply from the hepatic artery, a second liver blood perfusion volume formed by blood supply from the portal vein, and a ratio between the first liver blood perfusion volume and the second liver blood perfusion volume.

12. The method according to any one of claims 8 to 11, characterized in that The training samples of the liver perfusion state classification model include: the index value of the target blood vessel index of the sample user, the index value of the liver elastic modulus index of the sample user, and the real liver perfusion state category of the sample user; The parameters of the liver perfusion state classification model are optimized with the goal of minimizing the difference between the predicted liver perfusion state category of the sample user and the true liver perfusion state category; the predicted liver perfusion state category is predicted by the liver perfusion state classification model based on the index value of the target blood vessel index of the sample user and the index value of the liver elastic modulus index of the sample user.

13. The method according to claim 12, characterized in that Also includes: Acquire a first liver blood perfusion map, a second liver blood perfusion map, and a liver structure map of the sample user, wherein the first liver blood perfusion map is a liver blood perfusion map formed based on hepatic artery blood supply, and the second liver blood perfusion map is a liver blood perfusion map formed based on portal vein blood supply; Determining a segmentation mask for segmenting the liver according to the liver structure map; determining a liver region in the first liver blood perfusion map and the second liver blood perfusion map according to the segmentation mask; Eliminating voxel values ​​with abnormal distribution in the liver region of the first liver blood perfusion map, and determining a first liver blood perfusion volume formed by blood supply through the hepatic artery of the sample user based on an average value of remaining voxel values ​​in the liver region of the first liver blood perfusion map; Eliminating abnormally distributed voxel values ​​in the liver region of the second liver blood perfusion map, and determining a second liver blood perfusion volume of the sample user formed by portal vein blood supply based on an average value of remaining voxel values ​​in the liver region of the second liver blood perfusion map; The real liver perfusion state category of the sample user is determined according to the first liver blood perfusion volume formed by the hepatic artery blood supply of the sample user and the second liver blood perfusion volume formed by the portal vein blood supply of the sample user.

14. The method according to claim 13, characterized in that Also includes: Determine the distribution of voxel values ​​in the liver region of the first liver blood perfusion map, wherein the distribution includes: a mean value and a standard deviation; According to the distribution condition, abnormal voxel values ​​are distributed in the liver region of the first liver blood perfusion map.

15. A method for training a liver perfusion state classification model, characterized in that: include: Acquire a training sample, wherein the training sample includes: an index value of a target blood vessel index of a sample user, an index value of a liver elastic modulus index of the sample user, and a real liver perfusion state category of the sample user; Determining input information of a liver perfusion state classification model to be trained according to the index value of the target blood vessel index of the sample user and the index value of the liver elastic modulus index of the sample user; Inputting the input information into the liver perfusion state classification model, so as to classify the liver perfusion state of the user by using the liver perfusion state classification model, and obtain the predicted liver perfusion state category of the sample user; Optimizing the parameters in the liver perfusion state classification model with the goal of minimizing the difference between the true liver perfusion state category and the predicted liver perfusion state category; The liver perfusion status classification model is trained to classify the liver perfusion status.

16. An electronic device, characterized in that: include: A memory and a processor, wherein The memory is used to store programs; The processor is coupled to the memory, and is used to execute the program stored in the memory to implement the method described in any one of claims 8 to 15.

17. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a computer, the method according to any one of claims 8 to 15 can be implemented.

18. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 8 to 15 is implemented.

Images