Method for assessing recovery state of nasopharyngeal carcinoma tumor patient and related device

Abstract

The embodiment of the application relates to the technical field of data processing, and provides a recovery state evaluation method for a nasopharyngeal carcinoma tumor patient and related devices, the method comprising: determining complication information of a user to be evaluated according to medical data of the user to be evaluated to obtain a complication information set; constructing a first knowledge graph according to elements in the complication information set and the medical data; optimizing the first knowledge graph according to an affected part image of the user to be evaluated to obtain a second knowledge graph; extracting historical sign detection data of the user to be evaluated to obtain a historical sign detection data set; and performing recovery state evaluation on the user to be evaluated according to elements in the historical sign detection data set, current sign detection data of the user to be evaluated and the second knowledge graph to obtain a target evaluation result; and the efficiency of evaluating the recovery state of the user to be evaluated is improved.

Description

Technical Field

[0001] This application relates to the field of data processing technology, specifically to a method and related apparatus for assessing the recovery status of nasopharyngeal carcinoma patients. Background Technology

[0002] The assessment of the recovery status of nasopharyngeal carcinoma patients after surgery, radiotherapy, chemotherapy, targeted therapy or immunotherapy is an important part of clinical follow-up, efficacy determination and prognosis.

[0003] Existing methods for assessing the recovery status of nasopharyngeal carcinoma patients typically involve collecting vital signs and medical images of the patient to be assessed, followed by a physician evaluating the patient's recovery status based on these data. This approach is inefficient in assessing the patient's recovery status. Summary of the Invention

[0004] This application provides a method and related apparatus for assessing the recovery status of nasopharyngeal carcinoma patients. The method assesses the recovery status of the patient by using elements from a historical vital sign detection data set, the current vital sign detection data of the patient, and images of the affected area of ​​the patient, thereby improving the efficiency of assessing the recovery status of the patient.

[0005] A first aspect of this application provides a method for assessing the recovery status of a nasopharyngeal carcinoma patient, the method comprising:

[0006] Based on the medical data of the user to be evaluated, determine the complication information of the user to be evaluated, and obtain a set of complication information;

[0007] Construct the first knowledge graph based on elements in the complication information set and medical data;

[0008] The first knowledge graph is optimized based on the images of the affected areas of the users to be evaluated to obtain the second knowledge graph;

[0009] Extract the historical vital sign detection data of the user to be evaluated to obtain the historical vital sign detection data set;

[0010] Based on the elements in the historical vital sign detection data set, the current vital sign detection data of the user to be evaluated, and the second knowledge graph, the recovery status of the user to be evaluated is assessed to obtain the target assessment result.

[0011] In this example, complication information of the user to be evaluated is determined based on the user's medical data, resulting in a complication information set. A first knowledge graph is constructed based on the elements in the complication information set and the medical data. The first knowledge graph is then optimized based on the image of the affected area of ​​the user to be evaluated, resulting in a second knowledge graph. Historical vital sign detection data of the user to be evaluated is extracted, resulting in a historical vital sign detection data set. The recovery status of the user to be evaluated is assessed based on the elements in the historical vital sign detection data set, the current vital sign detection data of the user to be evaluated, and the second knowledge graph, resulting in the target assessment result. Therefore, by assessing the recovery status of the user to be evaluated using the elements in the historical vital sign detection data set, the current vital sign detection data of the user to be evaluated, and the image of the affected area of ​​the user to be evaluated, the inefficiency of assessing the recovery status of the user to be evaluated, which requires collecting the vital sign data and medical images of the user to be evaluated and then having a doctor assess the recovery status based on the collected vital sign data and medical images, is avoided. This improves the efficiency of assessing the recovery status of the user to be evaluated.

[0012] In one possible implementation, a method for constructing a first knowledge graph based on elements in a complication information set and medical data includes:

[0013] Based on medical data, the probability of occurrence of each element in the complication information set is determined to obtain the first probability information set.

[0014] Determine the occurrence time information corresponding to each element in the complication information set to obtain the occurrence time information set;

[0015] Determine the severity of each element in the complication information set to obtain the severity information set.

[0016] The first knowledge graph is constructed based on elements from the complication information set, medical data, elements from the first probability of occurrence information set, elements from the occurrence time information set, and elements from the severity information set.

[0017] In one possible implementation, a method for optimizing a first knowledge graph based on images of the affected area of ​​a user to be evaluated to obtain a second knowledge graph includes:

[0018] The images corresponding to the preset recognition targets are extracted from the images of the affected areas of the users to be evaluated to obtain a first image set, wherein the preset recognition targets are the wound areas and lesion areas of the users to be evaluated.

[0019] Extract the center point corresponding to each element in the first image set to obtain the center point information set;

[0020] Based on the elements in the center point information set and the preset image size information, image segmentation is performed on the image of the affected area of ​​the user to be evaluated to obtain a second image set;

[0021] Extract the texture feature information corresponding to each element in the second image set to obtain the texture feature information set;

[0022] Extract the color feature information corresponding to each element in the second image set to obtain the color feature information set;

[0023] The optimization coefficients corresponding to the directed edges of the first knowledge graph are determined based on the elements in the texture feature information set and the elements in the color feature information set, thus obtaining the first optimization coefficient set.

[0024] The directed edges of the first knowledge graph are optimized based on the elements in the first set of optimization coefficients to obtain the first set of directed edges;

[0025] The second knowledge graph is obtained by replacing the corresponding directed edges in the first knowledge graph with the elements in the first set of directed edges.

[0026] In one possible implementation, a method for assessing the recovery status of the user to be assessed based on elements from a historical vital sign detection dataset, the current vital sign detection data of the user to be assessed, and a second knowledge graph, to obtain a target assessment result, includes:

[0027] Perform time-series feature analysis on the elements in the historical vital sign detection dataset to obtain the first time-series feature information set;

[0028] Based on the elements in the first time-series feature information set, determine the time-series correlation coefficient between the elements in the historical vital sign detection data set and the nodes in the second knowledge graph, and obtain the time-series correlation coefficient.

[0029] The second knowledge graph is optimized based on the temporal correlation coefficient to obtain the third knowledge graph;

[0030] Calculate the similarity between the temporal feature information of the current vital sign detection data of the user to be evaluated and the temporal feature information of the third knowledge graph to obtain a similarity information set;

[0031] Based on the elements in the similarity information set, the current vital sign detection data of the user to be evaluated, and the third knowledge graph, the recovery status of the user to be evaluated is assessed to obtain the target evaluation result.

[0032] In one possible implementation, a method for assessing the recovery state of the user to be assessed based on elements in a similarity information set, the current vital sign detection data of the user to be assessed, and a third knowledge graph, to obtain a target assessment result, includes:

[0033] The recovery status of the user to be evaluated is scored based on the elements in the similarity information set to obtain the first score information;

[0034] The target complication information set is obtained by extracting elements from the complication information set whose occurrence probability information is greater than a preset occurrence probability information threshold based on the third knowledge graph.

[0035] Based on the current vital sign detection data of the user to be evaluated, the occurrence time node corresponding to each element in the target complication information set is predicted to obtain the first prediction result set;

[0036] The target evaluation result is obtained by concatenating the corresponding elements in the first score information and the first prediction result set.

[0037] A second aspect of this application provides a device for assessing the recovery status of nasopharyngeal carcinoma patients, the device comprising:

[0038] The determination unit is used to determine the complication information of the user to be evaluated based on the user's medical data, and to obtain a set of complication information.

[0039] A building unit is used to construct the first knowledge graph based on elements in the complication information set and medical data.

[0040] An optimization unit is used to optimize the first knowledge graph based on the image of the affected area of ​​the user to be evaluated, so as to obtain a second knowledge graph;

[0041] The extraction unit is used to extract the historical vital sign detection data of the user to be evaluated, and obtain a set of historical vital sign detection data.

[0042] The evaluation unit is used to evaluate the recovery status of the user to be evaluated based on elements in the historical vital sign detection data set, the current vital sign detection data of the user to be evaluated, and the second knowledge graph, so as to obtain the target evaluation result.

[0043] In one possible implementation, the building unit is specifically used for:

[0044] Based on medical data, the probability of occurrence of each element in the complication information set is determined to obtain the first probability information set.

[0045] Determine the occurrence time information corresponding to each element in the complication information set to obtain the occurrence time information set;

[0046] Determine the severity of each element in the complication information set to obtain the severity information set.

[0047] The first knowledge graph is constructed based on elements from the complication information set, medical data, elements from the first probability of occurrence information set, elements from the occurrence time information set, and elements from the severity information set.

[0048] In one possible implementation, the optimization unit is specifically used for:

[0049] The images corresponding to the preset recognition targets are extracted from the images of the affected areas of the users to be evaluated to obtain a first image set, wherein the preset recognition targets are the wound areas and lesion areas of the users to be evaluated.

[0050] Extract the center point corresponding to each element in the first image set to obtain the center point information set;

[0051] Based on the elements in the center point information set and the preset image size information, image segmentation is performed on the image of the affected area of ​​the user to be evaluated to obtain a second image set;

[0052] Extract the texture feature information corresponding to each element in the second image set to obtain the texture feature information set;

[0053] Extract the color feature information corresponding to each element in the second image set to obtain the color feature information set;

[0054] The optimization coefficients corresponding to the directed edges of the first knowledge graph are determined based on the elements in the texture feature information set and the elements in the color feature information set, thus obtaining the first optimization coefficient set.

[0055] The directed edges of the first knowledge graph are optimized based on the elements in the first set of optimization coefficients to obtain the first set of directed edges;

[0056] The second knowledge graph is obtained by replacing the corresponding directed edges in the first knowledge graph with the elements in the first set of directed edges.

[0057] In one possible implementation, the evaluation unit is specifically used for:

[0058] Perform time-series feature analysis on the elements in the historical vital sign detection dataset to obtain the first time-series feature information set;

[0059] Based on the elements in the first time-series feature information set, determine the time-series correlation coefficient between the elements in the historical vital sign detection data set and the nodes in the second knowledge graph, and obtain the time-series correlation coefficient.

[0060] The second knowledge graph is optimized based on the temporal correlation coefficient to obtain the third knowledge graph;

[0061] Calculate the similarity between the temporal feature information of the current vital sign detection data of the user to be evaluated and the temporal feature information of the third knowledge graph to obtain a similarity information set;

[0062] Based on the elements in the similarity information set, the current vital sign detection data of the user to be evaluated, and the third knowledge graph, the recovery status of the user to be evaluated is assessed to obtain the target evaluation result.

[0063] In one possible implementation, the evaluation unit is specifically used to evaluate the recovery state of the user to be evaluated based on elements in the similarity information set, the current vital sign detection data of the user to be evaluated, and the third knowledge graph, to obtain the target evaluation result.

[0064] The recovery status of the user to be evaluated is scored based on the elements in the similarity information set to obtain the first score information;

[0065] The target complication information set is obtained by extracting elements from the complication information set whose occurrence probability information is greater than a preset occurrence probability information threshold based on the third knowledge graph.

[0066] Based on the current vital sign detection data of the user to be evaluated, the occurrence time node corresponding to each element in the target complication information set is predicted to obtain the first prediction result set;

[0067] The target evaluation result is obtained by concatenating the corresponding elements in the first score information and the first prediction result set.

[0068] A third aspect of this application provides a terminal including a processor, an input device, an output device, and a memory, wherein the processor, input device, output device, and memory are interconnected, wherein the memory is used to store a computer program, the computer program including program instructions, and the processor is configured to invoke the program instructions to execute the step instructions as described in the first aspect of this application.

[0069] A fourth aspect of this application provides a computer-readable storage medium storing a computer program for electronic data interchange, wherein the computer program causes a computer to perform some or all of the steps described in the first aspect of this application.

[0070] A fifth aspect of this application provides a computer program product, wherein the computer program product includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps described in the first aspect of this application. The computer program product may be a software installation package. Attached Figure Description

[0071] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0072] Figure 1 This application provides a flowchart illustrating a method for assessing the recovery status of nasopharyngeal carcinoma patients.

[0073] Figure 2 A schematic diagram of a knowledge graph is provided for an embodiment of this application;

[0074] Figure 3 This is a schematic diagram of a second image extraction method provided in an embodiment of this application;

[0075] Figure 4 This is a schematic diagram of a fourth image extraction method provided in an embodiment of this application;

[0076] Figure 5 This is a schematic diagram of the structure of a terminal provided in an embodiment of this application;

[0077] Figure 6 This application provides a schematic diagram of the structure of a device for assessing the recovery status of nasopharyngeal carcinoma patients. Reference numerals: 601-Determining unit, 602-Constructing unit, 603-Optimizing unit, 604-Extracting unit, 605-Evaluating unit. Detailed Implementation

[0078] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0079] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0080] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application can be combined with other embodiments.

[0081] To better understand the method for assessing the recovery status of nasopharyngeal carcinoma patients provided in this application, a brief introduction to existing methods for assessing the recovery status of nasopharyngeal carcinoma patients is first provided below. Existing methods for assessing the recovery status of nasopharyngeal carcinoma patients typically involve collecting vital sign data and medical images of the patient to be assessed, followed by a physician evaluating the patient's recovery status based on these data. This process is inefficient in assessing the patient's recovery status.

[0082] To address the aforementioned technical problems, this application provides a method for assessing the recovery status of nasopharyngeal carcinoma patients. The method involves determining the complication information of the user based on their medical data, obtaining a complication information set, constructing a first knowledge graph based on elements in the complication information set and the medical data, optimizing the first knowledge graph based on images of the affected area of ​​the user, obtaining a second knowledge graph, extracting historical vital sign detection data of the user, obtaining a historical vital sign detection data set, and assessing the user's recovery status based on elements in the historical vital sign detection data set, the user's current vital sign detection data, and the second knowledge graph, obtaining a target assessment result. This method avoids the inefficiency of assessing the user's recovery status by collecting vital sign data and medical images, and then having a doctor assess the user's recovery status based on the collected data. Therefore, it improves the efficiency of assessing the user's recovery status.

[0083] Please see Figure 1 , Figure 1 This application provides a flowchart illustrating a method for assessing the recovery status of nasopharyngeal carcinoma patients. Figure 1 As shown, the method includes:

[0084] Step 101: Determine the complication information of the user to be evaluated based on the user's medical data to obtain a set of complication information.

[0085] Specifically, this can involve determining the type of treatment received by the user under evaluation through medical data, obtaining treatment type information; this information includes surgery, radiation therapy, and chemotherapy; and extracting all possible complications arising from the treatment methods corresponding to the treatment type information, resulting in a complication information set. Specifically, the medical data of the user under evaluation may include relevant medical data on nasopharyngeal carcinoma for patients with nasopharyngeal carcinoma.

[0086] Step 102: Construct the first knowledge graph based on the elements in the complication information set and medical data.

[0087] Specifically, a first knowledge graph can be constructed by extracting the probability of occurrence, time of occurrence, severity information, and medical data corresponding to the elements in the complication information set.

[0088] Step 103: Optimize the first knowledge graph based on the affected area images of the user to be evaluated to obtain the second knowledge graph.

[0089] Specifically, this can be achieved by extracting images corresponding to preset recognition targets from the affected area images of the user to be evaluated, obtaining a first image set; then, by segmenting the affected area images of the user to be evaluated based on the center points corresponding to the elements in the first image set, obtaining a second image set; finally, by determining the optimization coefficients corresponding to the directed edges of the first knowledge graph based on the texture and color feature information corresponding to the elements in the second image set, obtaining a first optimization coefficient set; and then, by optimizing the first knowledge graph based on the elements in the first optimization coefficient set, obtaining a second knowledge graph. Specifically, the affected area images can include T1 plain scan, T2 plain scan, and T1 enhanced images of the specific affected area of ​​a nasopharyngeal carcinoma patient, or images acquired by inserting a camera device into the nasal cavity.

[0090] Step 104: Extract the historical vital sign detection data of the user to be evaluated to obtain a set of historical vital sign detection data.

[0091] Specifically, historical vital sign data of the user to be evaluated can be extracted from a pre-set vital sign data database to obtain a historical vital sign data set. Each time feature detection data of the user to be evaluated is collected, it is stored in the pre-set vital sign data database.

[0092] Step 105: Based on the elements in the historical vital sign detection data set, the current vital sign detection data of the user to be evaluated, and the second knowledge graph, evaluate the recovery status of the user to be evaluated and obtain the target evaluation result.

[0093] Specifically, this can be achieved by performing time-series analysis on elements in the historical vital sign detection dataset and the second knowledge graph, and determining the time-series correlation coefficients between elements in the historical vital sign detection dataset and nodes in the second knowledge graph based on the time-series analysis results, thus obtaining a set of time-series correlation coefficients; using the time-series correlation coefficients to optimize the second knowledge graph, thus obtaining a third knowledge graph; and then evaluating the recovery status of the user to be evaluated based on the current vital sign detection data and the third knowledge graph, thus obtaining the target evaluation result.

[0094] By determining the complication information of the user to be evaluated based on their medical data, a complication information set is obtained. A first knowledge graph is constructed based on the elements in the complication information set and the medical data. The first knowledge graph is then optimized based on the image of the affected area of ​​the user to be evaluated, resulting in a second knowledge graph. Historical vital sign detection data of the user to be evaluated is extracted, resulting in a historical vital sign detection data set. The recovery status of the user to be evaluated is assessed based on the elements in the historical vital sign detection data set, the current vital sign detection data of the user to be evaluated, and the second knowledge graph, resulting in the target assessment result. Therefore, by assessing the recovery status of the user to be evaluated using the elements in the historical vital sign detection data set, the current vital sign detection data of the user to be evaluated, and the image of the affected area of ​​the user to be evaluated, the efficiency of assessing the recovery status of the user to be evaluated is improved, thus avoiding the low efficiency caused by collecting the vital sign data and medical images of the user to be evaluated and then having a doctor assess the recovery status based on the collected vital sign data and medical images.

[0095] In one possible implementation, a method for constructing a first knowledge graph based on elements in a complication information set and medical data includes:

[0096] Step A1: Determine the probability of occurrence for each element in the complication information set based on medical data to obtain the first probability information set;

[0097] Step A2: Determine the occurrence time information corresponding to each element in the complication information set to obtain the occurrence time information set;

[0098] Step A3: Determine the severity of each element in the complication information set to obtain the severity information set;

[0099] Step A4: Construct the first knowledge graph based on the elements in the complication information set, medical data, the elements in the first probability of occurrence information set, the elements in the occurrence time information set, and the elements in the severity information set.

[0100] Specifically, the process can involve determining the baseline probability of occurrence for each element in the complication information set based on medical data in a baseline probability database, thus obtaining a baseline probability information set; extracting user information corresponding to the user to be evaluated, thus obtaining target user information; where the target user information includes the age and underlying disease information of the user to be evaluated; determining the personalized correction coefficient corresponding to each element in the complication information set based on the target user information in a pre-defined personalized correction coefficient user information mapping table, thus obtaining a personalized correction coefficient set; performing personalized correction on the corresponding elements in the baseline probability information set based on the elements in the personalized correction coefficient set, thus obtaining a first sub-probability information set; using common causal association probability calculation methods (such as Bayesian network conditional causal probability calculation method, Markov random field calculation method, etc.) to calculate the corresponding association probability information between each element in the complication information set based on the elements in the first sub-probability information set, thus obtaining a second sub-probability information set; where the second sub-probability information can be understood as, for example, the probability of the second complication occurring after the first complication occurs; and determining the elements in the first and second sub-probability information sets as the first probability information, thus obtaining the first probability information set.

[0101] The baseline probability of occurrence in the baseline probability database can be determined by extracting the average probability of occurrence of each complication when different users receive the treatment corresponding to the medical data. The preset personalized correction coefficient user information mapping table records a general personalized correction coefficient, which corresponds to the user type; each user type has its own corresponding personalized correction coefficient. The user type can be determined through the user's medical information. For example, if the medical information is for nasal treatment, the user type is a nasal tumor user; similarly, if the medical information is for throat treatment, the user type is a throat tumor user. Because the wound of a nasal tumor user is only exposed to air after treatment, while the wound of a throat tumor user is exposed to saliva, phlegm, and refluxed stomach acid, the probability of complications is higher for throat tumor users than for nasal tumor users. Therefore, the personalized correction coefficient for nasal tumor users is lower than that for throat tumor users.

[0102] This can be achieved by extracting the occurrence time of each complication for different users when receiving treatment corresponding to the medical data, resulting in Q sets of reference time nodes; extracting the time nodes after different users receive the treatment corresponding to the medical data (i.e., the time nodes corresponding to the completion of the treatment), resulting in Q sets of initial time node information; calculating the duration between the elements in the Q sets of reference time node information and the corresponding elements in the Q sets of initial time node information, resulting in Q sets of first duration information; and calculating the average value of each of the Q sets of first duration information as the occurrence time information for each element in the complication information set, thus obtaining the occurrence time information set. Here, Q represents the number of complications.

[0103] A pre-defined expert evaluation model (such as a Bayesian network model or a random forest model) can be used to assess the severity of each element in the complication information set, thus obtaining a severity information set.

[0104] The pre-defined expert assessment model can be an assessment model constructed by combining relevant expert experience data, which can be obtained from the expert's assessment results on complications (e.g., severity assessment results). For example, the complication-severity assessment results are shown in Table 1 below.

[0105] Table 1. Results of Complication Severity Assessment

[0106] A 1 B 3 C 2 D 4

[0107] Specifically, A, B, C, and D represent complications, while 1, 2, 3, and 4 represent severity assessment values. The severity assessment value 2 is higher than the severity assessment value 1. The higher the severity assessment value, the more severe the corresponding complication.

[0108] This can be achieved by using medical data as the master node of the first knowledge graph; using elements from the complication information set as slave nodes of the first knowledge graph to obtain a slave node set; determining elements from the occurrence time information set and the severity information set as node label information corresponding to the elements in the slave node set to obtain a first node label information set; determining the length information of the corresponding directed edges in the first knowledge graph based on the elements in the first occurrence probability information set in a preset directed edge length mapping table to obtain a first length information set; using elements from the first node label information set as labels for the corresponding slave nodes, and then setting the length between the master node and the corresponding slave node (a directed edge pointing from the master node to the slave node indicates that the slave node was induced by the master node) and the length of the directed edges between slave nodes (a directed edge pointing from one slave master node to another slave node indicates that the pointed slave node was induced by the master node from the directed edge) based on the elements in the first length information set, thereby obtaining the first knowledge graph.

[0109] The preset probability-directed edge length mapping table includes the mapping relationship between the probability of occurrence and the directed edge length information; the larger the first probability of occurrence, the smaller the value of its corresponding directed edge length information; the smaller the value of the first probability of occurrence, the larger the value of its corresponding directed edge length information.

[0110] Please see Figure 2 , Figure 2 This is a schematic diagram of a knowledge graph provided in an embodiment of this application. Directed edges are used to represent the probability that the occurrence of one node will lead to the occurrence of another node. The node pointed to by the arrow of the directed edge is the node pointed to by the directed edge. For example, if a directed edge points from node A to node B, it represents the probability that the occurrence of the complication corresponding to node A will lead to the occurrence of the complication corresponding to node B. This is a one-way edge.

[0111] In this example, the first knowledge graph is constructed by extracting the probability of occurrence, the time of occurrence, and the severity of occurrence of elements in the complication information set. This improves the accuracy of the first knowledge graph, thereby improving the accuracy of the recovery status assessment of the user to be assessed.

[0112] In one possible implementation, a method for optimizing a first knowledge graph based on images of the affected area of ​​a user to be evaluated to obtain a second knowledge graph includes:

[0113] Step B1: Extract the image corresponding to the preset recognition target from the image of the affected area of ​​the user to be evaluated to obtain the first image set. The preset recognition target is the wound area and lesion area of ​​the user to be evaluated.

[0114] Step B2: Extract the center point corresponding to each element in the first image set to obtain the center point information set;

[0115] Step B3: Perform image segmentation on the affected area image of the user to be evaluated based on the elements in the center point information set and the preset image size information to obtain a second image set;

[0116] Step B4: Extract the texture feature information corresponding to each element in the second image set to obtain the texture feature information set;

[0117] Step B5: Extract the color feature information corresponding to each element in the second image set to obtain a color feature information set;

[0118] Step B6: Determine the optimization coefficients corresponding to the directed edges of the first knowledge graph based on the elements in the texture feature information set and the elements in the color feature information set, and obtain the first optimization coefficient set;

[0119] Step B7: Optimize the directed edges of the first knowledge graph based on the elements in the first set of optimization coefficients to obtain the first set of directed edges;

[0120] Step B8: Replace the corresponding directed edges in the first knowledge graph with the elements in the first directed edge set to obtain the second knowledge graph.

[0121] Specifically, the nasopharyngeal carcinoma region delineation model can identify the corresponding wound and lesion areas of the user being evaluated from images of the affected area, thus obtaining a first image set. The nasopharyngeal carcinoma region delineation model is a pre-trained model specifically used for extracting wound and lesion areas. During the training phase, the nasopharyngeal carcinoma region delineation model can be based on sample images obtained through manual annotation of relevant affected area images from multiple users. This sample image is then combined with general model training methods to adjust the initial model, resulting in a model suitable for extracting wound and lesion areas. This model is trained using conventional techniques and will not be described in detail here.

[0122] This can be achieved by extracting the coordinate position information corresponding to each pixel in each element of the first image set to obtain k sets of first coordinate position information; where k represents the number of elements in the first image set; and by calculating the coordinate position information corresponding to the center point of each element in the first image set based on the elements in the k sets of first coordinate position information to obtain a set of center point information.

[0123] Specifically, the coordinate position information corresponding to the center point of each element in the first image set can be calculated based on the k elements in the first coordinate position information set, as shown in the following formula, to obtain the center point information set:

[0124]

[0125] In the formula This represents the information of the i-th center point in the center point information set; This represents the number of elements in the i-th set of first coordinate position information among k sets of first coordinate position information, which can be understood as the number of pixels in the i-th element of the first image set; This represents the x-coordinate of the m-th element in the i-th set of k first coordinate position information sets, which can be understood as the x-coordinate of the m-th pixel in the i-th element of the first image set. This represents the ordinate of the m-th element in the i-th set of k first coordinate position information sets, which can be understood as the ordinate of the m-th pixel in the i-th element of the first image set.

[0126] A second image set can be obtained by extracting images corresponding to n regions from the image of the affected area of ​​the user to be evaluated, according to a preset image size. The preset image size can be determined by user input or by system default. The image size corresponding to each element in the second image set is the preset image size, and the center point corresponding to each element in the second image set is the pixel corresponding to the element in the center point information set. The elements in the second image set include the corresponding elements in the first image set. Here, n is the number of center points. Please refer to [link to relevant documentation]. Figure 3 , Figure 3 This is a schematic diagram of a second image extraction method provided in an embodiment of this application.

[0127] Specifically, before performing the second image extraction, a first image size information set can be obtained by extracting the image size corresponding to each element in the first image set. If there is an image size information in the first image size information set that is greater than or equal to a preset image size information, the preset image size information is adjusted so that all elements in the first image size information set are smaller than the adjusted preset image size information. The adjusted preset image size and the elements in the center point information set are used to extract the image of the affected area of ​​the user to be evaluated, thereby ensuring that the adjusted preset image size is greater than the image size corresponding to any element in the first image set, thus obtaining the second image set.

[0128] Specifically, a common grayscale image conversion method is used to extract the grayscale image corresponding to the second image, resulting in the third image; then, by using a sliding window of a preset window size, image extraction is performed on the third image to obtain s fourth images; where the window size of the sliding window can be expressed as... C represents the side length of the sliding window. The preset window size can be determined by user input or by the system default; please refer to [link to relevant documentation]. Figure 4 , Figure 4 This application provides a schematic diagram of a fourth image extraction method; calculating the gray-level gradient information corresponding to each of the s fourth images to obtain s gray-level gradient information; extracting the gray-level value corresponding to the center pixel of each of the s fourth images to obtain s first gray-level value information; determining the other pixels in each of the s fourth images, excluding the center pixel, as neighborhood pixels to obtain s sets of neighborhood pixels; extracting the gray-level value corresponding to each element in the s sets of neighborhood pixels to obtain s sets of second gray-level value information; extracting the standard deviation of the gray-level value corresponding to each of the s fourth images to obtain s first gray-level standard deviations; and based on the s... The gray-level difference dispersion corresponding to each of the s fourth images is calculated from the first gray-level value information, resulting in s gray-level difference dispersions. The gray-level average values ​​corresponding to the s sets of second gray-level value information are calculated respectively, resulting in s gray-level average values. Based on the gray-level gradient information in the s gray-level gradient information, the first gray-level standard deviation in the s first gray-level standard deviation, the gray-level difference dispersion in the s gray-level difference dispersion, the first gray-level information in the s first gray-level value information, and the gray-level average value in the s gray-level average values, the texture feature information corresponding to the second image is calculated. Thus, texture features can be extracted from all the second images in the second image set to obtain a texture feature information set.

[0129] The texture feature information set can be obtained by extracting the texture feature information set for each element in the second image set using the method shown in the following formula:

[0130]

[0131] In the formula The i-th element in the texture feature information set can be understood as the texture feature information corresponding to the i-th element in the second image set. This represents the j-th element among the s standard deviations of the first grayscale values ​​corresponding to the i-th element in the second image set; This represents the j-th element among the s grayscale gradient information corresponding to the i-th element in the second image set; This represents the j-th element among the s first grayscale values ​​corresponding to the i-th element in the second image set; This represents the j-th element among the s average gray levels corresponding to the i-th element in the second image set; This represents the j-th gray-level difference dispersion among the s gray-level difference dispersions corresponding to the i-th element in the second image set; This represents the operation of exponential functions; This indicates the preset window size, which can be understood as the number of pixels that can be contained within a sliding window of the preset window size; This represents the coordinates of the pixel corresponding to the j-th element among the s first grayscale values ​​corresponding to the i-th element in the second image set. ; This represents the grayscale change rate in the x-direction of the j-th element in the s fourth images corresponding to the i-th element in the second image set. This represents the grayscale change rate in the y-direction of the j-th element in the s fourth images corresponding to the i-th element in the second image set.

[0132] One can use a general color feature extraction method to extract the color feature information corresponding to each element in the second image set, thus obtaining a color feature information set.

[0133] Because the color and texture of tissues in areas of infection and lesion change when the user being evaluated develops complications such as infection or lesions, the color and texture of tissues in normal areas differ from those in areas of infection or lesions. Furthermore, the vital signs of the user being evaluated do not show significant changes in the early stages of complications. Therefore, the current vital sign data of the user being evaluated cannot accurately predict the probability of occurrence of elements in the complication information set. The second image, however, carries a larger size than the first image and thus contains information about the surrounding tissues. Therefore, by combining the texture and color features of the elements in the second image set, the optimization coefficients corresponding to the predicted probability of occurrence of elements in the complication information set based on the current vital sign data of the user being evaluated (the length of the directed edge in the first knowledge graph) can be determined, resulting in the first set of optimization coefficients.

[0134] Specifically, the optimization coefficients corresponding to the directed edges of the first knowledge graph can be determined based on the elements in the texture feature information set and the color feature information set, as shown in the following formula, thus obtaining the first set of optimization coefficients:

[0135]

[0136] In the formula Represents the i-th first optimization coefficient; This represents the i-th first occurrence probability information in the first occurrence probability information set; The i-th element in the second probability information set can be understood as the probability information of occurrence corresponding to the elements in the complication information set after correction based on the elements in the texture feature information set and the elements in the color feature information set. It can be understood as the correlation coefficient mapped by the fused feature vector after fusing the elements in the texture feature information set and the elements in the color feature information set, representing the corresponding optimization information of color features and texture features; m represents the number of second images in the second image set; Represents the logarithmic function with base 10; The preset feature fusion weights can be determined using empirical or historical values. The j-th element in the color feature information set can be understood as the color feature information corresponding to the j-th second image in the second image set. The j-th element in the texture feature information set can be understood as the texture feature information corresponding to the j-th second image in the second image set. express and The mapping coefficient between them can be determined based on historical or empirical values. For example, its value can be 1. Specifically, it can be understood that the mapping relationship when mapping the probability of occurrence to the length of the directed edge is a mapping relationship with the same specifications. For example, the specifications of the directed edge length mapped by the first probability of occurrence are consistent with those of the directed edge length mapped by the second probability of occurrence.

[0137] Specifically, the length of the first directed edge can be obtained by calculating the product between the first length information of the directed edge in the first knowledge graph and the corresponding element in the optimization coefficient set, and then the first set of directed edges can be obtained.

[0138] Specifically, a second knowledge graph can be obtained by replacing the directed edges in the first knowledge graph with the corresponding elements in the first set of directed edges.

[0139] In this example, the first knowledge graph is optimized by extracting the texture and color feature information corresponding to each element in the second image set to obtain the second knowledge graph, thereby improving the accuracy of the second knowledge graph and thus improving the accuracy when evaluating the recovery status of the user to be evaluated.

[0140] In one possible implementation, a method for assessing the recovery status of the user to be assessed based on elements in a historical vital sign detection dataset, the current vital sign detection data of the user to be assessed, and a second knowledge graph, to obtain a target assessment result, includes:

[0141] Step C1: Perform time-series feature analysis on the elements in the historical vital sign detection data set to obtain the first time-series feature information set;

[0142] Step C2: Determine the temporal correlation coefficient between the elements in the historical vital sign detection data set and the nodes in the second knowledge graph based on the elements in the first temporal feature information set, and obtain the temporal correlation coefficient;

[0143] Step C3: Optimize the second knowledge graph based on the temporal correlation coefficient to obtain the third knowledge graph;

[0144] Step C4: Calculate the similarity between the temporal feature information of the current vital sign detection data of the user to be evaluated and the temporal feature information of the third knowledge graph to obtain a similarity information set;

[0145] Step C5: Based on the elements in the similarity information set, the current vital sign detection data of the user to be evaluated, and the third knowledge graph, evaluate the recovery status of the user to be evaluated to obtain the target evaluation result.

[0146] Since the probability of each complication in the second knowledge graph is determined by the baseline probability, which is the average probability of each complication occurring when different users receive the treatment corresponding to the medical data, and different users have different recovery abilities, the occurrence time information determined by averaging reflects the average recovery ability of different users. In practical application scenarios, users with strong recovery abilities have a lower probability of developing complications than users with weak recovery abilities. Since users' vital sign detection data can reflect the strength of their recovery abilities, the second knowledge graph can be corrected using elements from the historical vital sign detection data set to further improve the accuracy of the knowledge graph.

[0147] Specifically, this can be achieved by extracting the collection time information corresponding to the elements in the historical vital sign detection data set, thus obtaining a collection time information set; sorting the elements in the historical vital sign detection data set according to the chronological order of the elements in the collection time information set, thus obtaining a historical vital sign detection data sequence; and using common time series analysis methods (such as sliding window statistics, autocorrelation analysis, etc.) to extract the time series features corresponding to the historical vital sign detection data sequence, thus obtaining a first time series feature information set. The first time series feature information can be understood as the fluctuation information of the elements in the historical vital sign detection data set over time.

[0148] This can be achieved by extracting the node label information corresponding to each slave node in the second knowledge graph to obtain a set of second node label information; then, by using a general temporal analysis method to perform temporal analysis on the occurrence time information of each element in the set of second node label information, the temporal feature information corresponding to each slave node in the second knowledge graph can be extracted to obtain the temporal feature information of the knowledge graph; where the temporal feature information of the knowledge graph can be understood as the fluctuation information of the occurrence time information of each element in the set of second node label information over time.

[0149] Since sorting the occurrence time information of each element in the second node label information set in chronological order can simulate the continuous deterioration of a user's condition with average recovery ability, and its changes over time until all complications occur, a positive correlation between the first temporal feature information and the knowledge graph temporal feature information indicates that the recovery ability of the user to be evaluated is weaker than that of the simulated user in the second knowledge graph. Conversely, a negative correlation indicates that the recovery ability of the user to be evaluated is stronger than that of the simulated user in the second knowledge graph. Therefore, the correlation coefficient between the first temporal feature information and the knowledge graph temporal feature information can be used to characterize the strength of the user's recovery ability relative to that of the simulated user in the second knowledge graph. This allows for the correction of the second knowledge graph, incorporating relevant features of the user to be evaluated, thereby improving the accuracy of the knowledge graph.

[0150] The correlation coefficient between the first temporal feature information and the knowledge graph temporal feature information is calculated using common correlation coefficient methods (such as Pearson correlation coefficient method, DTW (Dynamic Time Warping) correlation coefficient method, etc.). The second node label information is consistent with the first node label information. Specifically, when the correlation coefficient is greater than or equal to 0, it indicates a positive correlation between the first temporal feature information and the knowledge graph temporal feature information, suggesting that the user being evaluated has relatively weak recovery ability; conversely, when the correlation coefficient is less than 0, it indicates a negative correlation between the first temporal feature information and the knowledge graph temporal feature information, suggesting that the user being evaluated has relatively strong recovery ability.

[0151] The optimization weights corresponding to the elements in the second occurrence probability information set can be calculated using the temporal correlation coefficient to obtain the optimization weight information set. The elements in the optimization weight information set are then used to optimize the corresponding elements in the second occurrence probability information set to obtain the third occurrence probability information set. The optimization coefficients corresponding to the pairs in the first directed edge set are determined based on the elements in the third occurrence probability information set to obtain the second optimization coefficient set. The product between the elements in the first directed edge set and the corresponding elements in the second optimization coefficient set is calculated to obtain the second directed edge set. The elements in the second directed edge set are then used to replace the directed edges in the second knowledge graph to obtain the third knowledge graph.

[0152] Specifically, the third knowledge graph can be obtained by optimizing the second knowledge graph based on the temporal correlation coefficient using the method shown in the following formula:

[0153]

[0154] In the formula This represents the i-th element in the set of second optimization coefficients; This represents the i-th element in the third probability information set. The i-th element in the second probability information set can be understood as the probability information of occurrence corresponding to the elements in the complication information set after correction based on the elements in the texture feature information set and the elements in the color feature information set. This represents the i-th element in the set of optimization weight information; Represents the time-series correlation coefficient; express and The mapping coefficient between them can be determined through historical and empirical values; for example, it could be 1. and There is a mapping relationship between them. and The mapping relationships between them include linear mapping relationships and nonlinear mapping relationships.

[0155] This can be achieved by extracting the collection time corresponding to the current vital sign detection data of the user to be evaluated, thus obtaining a first collection time information set; using a general time series analysis method, extracting the time series feature information corresponding to the current vital sign detection data of the user to be evaluated from the elements in the first collection time information set, thus obtaining a second time series feature information set; and using a general similarity calculation method (such as Euclidean distance similarity calculation method) to calculate the similarity between the elements in the second time series feature information set and the corresponding elements in the knowledge graph time series feature information set, thus obtaining a similarity information set. The current vital sign detection data of the user to be evaluated consists of the user's body temperature, blood pressure, heart rate, and medical images collected within a preset collection duration and according to a preset collection cycle. The preset collection duration and preset collection cycle are set based on empirical values ​​or historical data.

[0156] The first score information can be obtained by scoring the recovery status of the user to be evaluated based on the elements in the similarity information set; the occurrence time node corresponding to each element in the complication information set can be predicted based on the current vital sign detection data of the user to be evaluated and the third knowledge graph; and the recovery status evaluation information corresponding to the user to be evaluated can be determined based on the prediction results and the first score information to obtain the target evaluation result.

[0157] In this example, by performing time-series analysis on elements in the historical vital sign detection dataset and the second knowledge graph, the temporal correlation coefficients between the elements in the historical vital sign detection dataset and the nodes in the second knowledge graph are determined, resulting in a set of temporal correlation coefficients. The second knowledge graph is then optimized based on these correlation coefficients to obtain a third knowledge graph. Furthermore, a similarity information set is obtained based on the similarity between the current vital sign detection data of the user to be evaluated and each node in the third knowledge graph. The recovery status of the user to be evaluated is then assessed based on the elements in the similarity information set to obtain the target assessment result, thereby improving the accuracy of assessing the recovery status of the user to be evaluated.

[0158] In one possible implementation, a method for assessing the recovery state of the user to be evaluated based on elements in a similarity information set, the current vital sign detection data of the user to be evaluated, and a third knowledge graph, to obtain a target evaluation result, includes:

[0159] Step D1: Score the recovery status of the user to be evaluated based on the elements in the similarity information set to obtain the first score information;

[0160] Step D2: Extract the elements in the complication information set whose occurrence probability information is greater than the preset occurrence probability information threshold based on the third knowledge graph to obtain the target complication information set;

[0161] Step D3: Based on the current vital sign detection data of the user to be evaluated, predict the occurrence time node corresponding to each element in the target complication information set to obtain the first prediction result set;

[0162] Step D4: Concatenate the corresponding elements in the first scoring information and the first prediction result set to obtain the target evaluation result.

[0163] Specifically, the first score can be obtained by scoring the recovery status of the user to be evaluated based on the elements in the similarity information set, as shown in the following formula:

[0164]

[0165] In the formula This indicates the first rating information; This indicates the number of elements in the similarity information set; This represents the score weight corresponding to the i-th element in the similarity information set; This represents the i-th element in the similarity information set; This indicates the first preset scoring weight, which can be determined through historical or empirical values. This indicates the second preset scoring weight, which can be determined through historical or empirical values, and satisfies... ; This represents the preset similarity information threshold, which can be determined through historical or empirical values.

[0166] This can be achieved by extracting the occurrence probability information corresponding to each node in the third knowledge graph to obtain a third occurrence probability information set; extracting elements in the third occurrence probability information set that are greater than a preset occurrence probability information threshold to obtain a fourth occurrence probability information set; and extracting the corresponding element in the complication information set for each element in the fourth occurrence probability information set to obtain the target complication information set. The preset occurrence probability information threshold is set based on empirical values ​​or historical data.

[0167] A general time-point prediction method can be used to predict the occurrence time of each element in the target complication information set based on the current vital sign data of the user to be evaluated, resulting in a first prediction result set. For example, the first prediction result could be: infection occurs, with an occurrence time of 1-3 days; sepsis occurs, with an occurrence time of 4-7 days.

[0168] The recovery status level information can be obtained by extracting the recovery status level corresponding to the first evaluation information from a preset rating information recovery status level mapping table; wherein the recovery status level information can be "good recovery status" or "deteriorating recovery status"; this is only an example of recovery status information and does not limit the specific content of the recovery status information; the recovery status level information is vectorized to obtain a first vector; the elements in the first prediction result set are vectorized to obtain a second vector set; the elements in the first vector and the second vector set are concatenated to obtain the target evaluation vector; and the semantic information corresponding to the target evaluation vector is extracted using a general semantic extraction method to obtain the target evaluation result.

[0169] The mathematical expression for the target evaluation vector can be: In the formula, Q represents the target evaluation vector; Represents the first vector, which recovers the state level information; This represents the first element in the second vector set; This represents the second element in the second vector set; This represents the i-th element in the second vector set. The target assessment result could be, for example: a deterioration in recovery status, increased susceptibility to infection within 1-3 days, and increased susceptibility to sepsis within 4-7 days; the above target assessment results are for illustrative purposes only and do not limit the specific content of the target assessment results.

[0170] In this example, the recovery status of the user to be evaluated is assessed by using elements in the similarity information set and the current vital sign detection data of the user to be evaluated, thereby obtaining the target assessment result and improving the efficiency of assessing the recovery status of the user to be evaluated.

[0171] For examples consistent with the above embodiments, please refer to... Figure 5 , Figure 5 This is a schematic diagram of the structure of a terminal provided in an embodiment of this application, such as... Figure 5 As shown, it includes a processor, an input device, an output device, and a memory, which are interconnected. The memory is used to store a computer program, which includes program instructions. The processor is configured to call the program instructions. The program includes instructions for performing the following steps.

[0172] Based on the medical data of the user to be evaluated, determine the complication information of the user to be evaluated, and obtain a set of complication information;

[0173] Construct the first knowledge graph based on elements in the complication information set and medical data;

[0174] The first knowledge graph is optimized based on the images of the affected areas of the users to be evaluated to obtain the second knowledge graph;

[0175] Extract the historical vital sign detection data of the user to be evaluated to obtain the historical vital sign detection data set;

[0176] Based on the elements in the historical vital sign detection data set, the current vital sign detection data of the user to be evaluated, and the second knowledge graph, the recovery status of the user to be evaluated is assessed to obtain the target assessment result.

[0177] The above mainly describes the solutions of the embodiments of this application from the perspective of the method execution process. It is understood that, in order to achieve the above functions, the terminal includes the corresponding hardware structure and / or software modules for executing each function. Those skilled in the art should readily recognize that, in conjunction with the units and algorithm steps of the various examples described in the embodiments provided herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0178] This application embodiment can divide the terminal into functional units according to the above method example. For example, each function can be divided into a separate functional unit, or two or more functions can be integrated into one processing unit. The integrated unit can be implemented in hardware or as a software functional unit. It should be noted that the unit division in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0179] For those consistent with the above, please refer to Figure 6 , Figure 6 This application provides a schematic diagram of a device for assessing the recovery status of nasopharyngeal carcinoma patients. (See attached diagram.) Figure 6 As shown, the device includes:

[0180] The determining unit 601 is used to determine the complication information of the user to be evaluated based on the user's medical data, and obtain a set of complication information.

[0181] Construction unit 602 is used to construct the first knowledge graph based on elements in the complication information set and medical data;

[0182] The optimization unit 603 is used to optimize the first knowledge graph based on the image of the affected area of ​​the user to be evaluated, so as to obtain the second knowledge graph;

[0183] Extraction unit 604 is used to extract the historical vital sign detection data of the user to be evaluated, and obtain a set of historical vital sign detection data;

[0184] The evaluation unit 605 is used to evaluate the recovery status of the user to be evaluated based on elements in the historical vital sign detection data set, the current vital sign detection data of the user to be evaluated, and the second knowledge graph, so as to obtain the target evaluation result.

[0185] In one possible implementation, the building unit 602 is specifically used for:

[0186] Based on medical data, the probability of occurrence of each element in the complication information set is determined to obtain the first probability information set.

[0187] Determine the occurrence time information corresponding to each element in the complication information set to obtain the occurrence time information set;

[0188] Determine the severity of each element in the complication information set to obtain the severity information set.

[0189] The first knowledge graph is constructed based on elements from the complication information set, medical data, elements from the first probability of occurrence information set, elements from the occurrence time information set, and elements from the severity information set.

[0190] In one possible implementation, the optimization unit 603 is specifically used for:

[0191] The images corresponding to the preset recognition targets are extracted from the images of the affected areas of the users to be evaluated to obtain a first image set, wherein the preset recognition targets are the wound areas and lesion areas of the users to be evaluated.

[0192] Extract the center point corresponding to each element in the first image set to obtain the center point information set;

[0193] Based on the elements in the center point information set and the preset image size information, image segmentation is performed on the image of the affected area of ​​the user to be evaluated to obtain a second image set;

[0194] Extract the texture feature information corresponding to each element in the second image set to obtain the texture feature information set;

[0195] Extract the color feature information corresponding to each element in the second image set to obtain the color feature information set;

[0196] The optimization coefficients corresponding to the directed edges of the first knowledge graph are determined based on the elements in the texture feature information set and the elements in the color feature information set, thus obtaining the first optimization coefficient set.

[0197] The directed edges of the first knowledge graph are optimized based on the elements in the first set of optimization coefficients to obtain the first set of directed edges;

[0198] The second knowledge graph is obtained by replacing the corresponding directed edges in the first knowledge graph with the elements in the first set of directed edges.

[0199] In one possible implementation, the evaluation unit 605 is specifically used for:

[0200] Perform time-series feature analysis on the elements in the historical vital sign detection dataset to obtain the first time-series feature information set;

[0201] Based on the elements in the first time-series feature information set, determine the time-series correlation coefficient between the elements in the historical vital sign detection data set and the nodes in the second knowledge graph, and obtain the time-series correlation coefficient.

[0202] The second knowledge graph is optimized based on the temporal correlation coefficient to obtain the third knowledge graph;

[0203] Calculate the similarity between the temporal feature information of the current vital sign detection data of the user to be evaluated and the temporal feature information of the third knowledge graph to obtain a similarity information set;

[0204] Based on the elements in the similarity information set, the current vital sign detection data of the user to be evaluated, and the third knowledge graph, the recovery status of the user to be evaluated is assessed to obtain the target evaluation result.

[0205] In one possible implementation, the evaluation unit 605 is specifically used to: evaluate the recovery state of the user to be evaluated based on elements in the similarity information set, the current vital sign detection data of the user to be evaluated, and the third knowledge graph to obtain the target evaluation result;

[0206] The recovery status of the user to be evaluated is scored based on the elements in the similarity information set to obtain the first score information;

[0207] The target complication information set is obtained by extracting elements from the complication information set whose occurrence probability information is greater than a preset occurrence probability information threshold based on the third knowledge graph.

[0208] Based on the current vital sign detection data of the user to be evaluated, the occurrence time node corresponding to each element in the target complication information set is predicted to obtain the first prediction result set;

[0209] The target evaluation result is obtained by concatenating the corresponding elements in the first score information and the first prediction result set.

[0210] This application also provides a computer storage medium storing a computer program for electronic data interchange, which causes a computer to perform some or all of the steps of any of the methods for assessing the recovery status of nasopharyngeal carcinoma patients as described in the above method embodiments.

[0211] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program that causes a computer to perform some or all of the steps of any of the recovery status assessment methods for nasopharyngeal carcinoma patients described in the above method embodiments.

[0212] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0213] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0214] In the several embodiments provided in this application, it should be understood that the disclosed apparatus 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 system, 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 or other forms.

[0215] 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.

[0216] Furthermore, the functional units in the various embodiments of the application 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. The integrated unit can be implemented in hardware or as a software program module.

[0217] If the integrated unit is implemented as a software program module and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory 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 this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0218] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage device, which may include: a flash drive, a read-only memory, a random access memory, a magnetic disk, or an optical disk, etc.

[0219] The embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for assessing the recovery status of nasopharyngeal carcinoma patients, characterized in that, The method includes: Based on the medical data of the user to be evaluated, determine the complication information of the user to be evaluated, and obtain a set of complication information; Construct the first knowledge graph based on elements in the complication information set and medical data; The first knowledge graph is optimized based on the images of the affected areas of the users to be evaluated to obtain the second knowledge graph; Extract the historical vital sign detection data of the user to be evaluated to obtain the historical vital sign detection data set; Based on the elements in the historical vital sign detection data set, the current vital sign detection data of the user to be evaluated, and the second knowledge graph, the recovery status of the user to be evaluated is assessed to obtain the target assessment result.

2. The method for assessing the recovery status of nasopharyngeal carcinoma patients according to claim 1, characterized in that, The construction of the first knowledge graph based on elements in the complication information set and medical data includes: Based on medical data, the probability of occurrence of each element in the complication information set is determined to obtain the first probability information set. Determine the occurrence time information corresponding to each element in the complication information set to obtain the occurrence time information set; Determine the severity of each element in the complication information set to obtain the severity information set. The first knowledge graph is constructed based on elements from the complication information set, medical data, elements from the first probability of occurrence information set, elements from the occurrence time information set, and elements from the severity information set.

3. The method for assessing the recovery status of nasopharyngeal carcinoma patients according to claim 2, characterized in that, The step of optimizing the first knowledge graph based on the affected area image of the user to be evaluated to obtain the second knowledge graph includes: The images corresponding to the preset recognition targets are extracted from the images of the affected areas of the users to be evaluated to obtain a first image set, wherein the preset recognition targets are the wound areas and lesion areas of the users to be evaluated. Extract the center point corresponding to each element in the first image set to obtain the center point information set; Based on the elements in the center point information set and the preset image size information, image segmentation is performed on the image of the affected area of ​​the user to be evaluated to obtain a second image set; Extract the texture feature information corresponding to each element in the second image set to obtain the texture feature information set; Extract the color feature information corresponding to each element in the second image set to obtain the color feature information set; The optimization coefficients corresponding to the directed edges of the first knowledge graph are determined based on the elements in the texture feature information set and the elements in the color feature information set, thus obtaining the first optimization coefficient set. The directed edges of the first knowledge graph are optimized based on the elements in the first set of optimization coefficients to obtain the first set of directed edges; The second knowledge graph is obtained by replacing the corresponding directed edges in the first knowledge graph with the elements in the first set of directed edges.

4. The method for assessing the recovery status of nasopharyngeal carcinoma patients according to claim 3, characterized in that, The process involves assessing the recovery status of the user based on elements from the historical vital sign detection data set, the current vital sign detection data of the user to be assessed, and the second knowledge graph, to obtain the target assessment result, including: Perform time-series feature analysis on the elements in the historical vital sign detection dataset to obtain the first time-series feature information set; Based on the elements in the first time-series feature information set, determine the time-series correlation coefficient between the elements in the historical vital sign detection data set and the nodes in the second knowledge graph, and obtain the time-series correlation coefficient. The second knowledge graph is optimized based on the temporal correlation coefficient to obtain the third knowledge graph; Calculate the similarity between the temporal feature information of the current vital sign detection data of the user to be evaluated and the temporal feature information of the third knowledge graph to obtain a similarity information set; Based on the elements in the similarity information set, the current vital sign detection data of the user to be evaluated, and the third knowledge graph, the recovery status of the user to be evaluated is assessed to obtain the target evaluation result.

5. The method for assessing the recovery status of nasopharyngeal carcinoma patients according to claim 4, characterized in that, The recovery status assessment of the user to be assessed is performed based on elements in the similarity information set, the current vital sign detection data of the user to be assessed, and the third knowledge graph to obtain the target assessment result, including: The recovery status of the user to be evaluated is scored based on the elements in the similarity information set to obtain the first score information; The target complication information set is obtained by extracting elements from the complication information set whose occurrence probability information is greater than a preset occurrence probability information threshold based on the third knowledge graph. Based on the current vital sign detection data of the user to be evaluated, the occurrence time node corresponding to each element in the target complication information set is predicted to obtain the first prediction result set; The target evaluation result is obtained by concatenating the corresponding elements in the first score information and the first prediction result set.

6. A device for assessing the recovery status of nasopharyngeal carcinoma patients, characterized in that, The device includes: The determination unit is used to determine the complication information of the user to be evaluated based on the user's medical data, and to obtain a set of complication information. A building unit is used to construct the first knowledge graph based on elements in the complication information set and medical data. An optimization unit is used to optimize the first knowledge graph based on the image of the affected area of ​​the user to be evaluated, so as to obtain a second knowledge graph; The extraction unit is used to extract the historical vital sign detection data of the user to be evaluated, and obtain a set of historical vital sign detection data. The evaluation unit is used to evaluate the recovery status of the user to be evaluated based on elements in the historical vital sign detection data set, the current vital sign detection data of the user to be evaluated, and the second knowledge graph, so as to obtain the target evaluation result.

7. The device for assessing the recovery status of nasopharyngeal carcinoma patients according to claim 6, characterized in that, The building unit is specifically used for: Based on medical data, the probability of occurrence of each element in the complication information set is determined to obtain the first probability information set. Determine the occurrence time information corresponding to each element in the complication information set to obtain the occurrence time information set; Determine the severity of each element in the complication information set to obtain the severity information set. The first knowledge graph is constructed based on elements from the complication information set, medical data, elements from the first probability of occurrence information set, elements from the occurrence time information set, and elements from the severity information set.

8. The device for assessing the recovery status of nasopharyngeal carcinoma patients according to claim 7, characterized in that, The optimization unit is specifically used for: The images corresponding to the preset recognition targets are extracted from the images of the affected areas of the users to be evaluated to obtain a first image set, wherein the preset recognition targets are the wound areas and lesion areas of the users to be evaluated. Extract the center point corresponding to each element in the first image set to obtain the center point information set; Based on the elements in the center point information set and the preset image size information, image segmentation is performed on the image of the affected area of ​​the user to be evaluated to obtain a second image set; Extract the texture feature information corresponding to each element in the second image set to obtain the texture feature information set; Extract the color feature information corresponding to each element in the second image set to obtain the color feature information set; The optimization coefficients corresponding to the directed edges of the first knowledge graph are determined based on the elements in the texture feature information set and the elements in the color feature information set, thus obtaining the first optimization coefficient set. The directed edges of the first knowledge graph are optimized based on the elements in the first set of optimization coefficients to obtain the first set of directed edges; The second knowledge graph is obtained by replacing the corresponding directed edges in the first knowledge graph with the elements in the first set of directed edges.

9. A terminal, characterized in that, The device includes a processor, an input device, an output device, and a memory, which are interconnected. The memory stores a computer program, which includes program instructions. The processor is configured to invoke the program instructions to perform the steps of the method for assessing the recovery status of nasopharyngeal carcinoma patients as described in any one of claims 1 to 5.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the method for assessing the recovery status of a nasopharyngeal carcinoma patient as described in any one of claims 1 to 5.

Images