Method, device and computer equipment for predicting fluctuations in pain rating

Abstract

The application relates to a pain level fluctuation prediction method and device and computer equipment, and relates to the technical field of machine learning. The prediction method comprises the following steps: obtaining a first pain data set of a target object; training a pre-obtained general pain model by using the first pain data set to obtain a personal pain model corresponding to the target object, wherein the general pain model is obtained by training a general pain data set; obtaining a second pain data set of the target object, and predicting the pain level fluctuation of the target object according to the personal pain model and the second pain data set to obtain personal pain level fluctuation data of the target object. The method can obtain personal pain level fluctuation data suitable for the target object, avoid the situation that the pain level fluctuation data is inaccurate due to the pain level selected by the target object or the pain level selected by medical staff, and improve the evaluation accuracy of the pain level fluctuation.

Description

Technical Field

[0001] This application relates to the field of physiological signal processing technology, and in particular to a method, apparatus, and computer device for predicting pain level fluctuations. Background Technology

[0002] Pain is an unpleasant subjective feeling and emotional experience associated with tissue damage or potential damage, and it is one of the most common symptoms in clinical practice. The most commonly used method for grading pain levels is the numerical rating scale, which uses numbers from 0 to 10 to represent the intensity of pain, with 0 representing no pain and 10 representing the most severe pain.

[0003] Pain level fluctuation assessment is a crucial aspect of pain management. For example, over a certain period after implementing pain interventions, it's necessary to assess pain level fluctuations to determine changes in pain intensity and the effectiveness of analgesics. Pain level fluctuations also need to be evaluated after any action or behavior anticipated to cause pain. However, due to the subjective nature of pain, current methods typically rely on patients selecting a number that best represents their pain level, or on healthcare professionals choosing a corresponding pain level based on the patient's description of the pain, thus determining pain level fluctuations. Therefore, the accuracy of current techniques for assessing pain level fluctuations is relatively poor. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, device, and computer equipment for predicting pain level fluctuations that can accurately predict the fluctuations in pain level, in order to address the aforementioned technical problems.

[0005] Firstly, this application provides a method for predicting fluctuations in pain levels. The prediction method includes:

[0006] Obtain the first pain dataset of the target object;

[0007] The first pain dataset is used to train a pre-obtained general pain model to obtain a personal pain model corresponding to the target object, wherein the general pain model is trained based on the general pain dataset.

[0008] A second pain dataset of the target object is obtained, and the pain level fluctuation of the target object is predicted based on the personal pain model and the second pain dataset to obtain the personal pain level fluctuation data of the target object.

[0009] In one embodiment, the first pain dataset includes vital sign parameters and pain levels of the target object at multiple sampling times. The step of training a pre-obtained general pain model using the first pain dataset to obtain a personal pain model corresponding to the target object includes:

[0010] The vital signs parameters and pain levels at the sampling time are input into the general pain model to obtain the test pain level fluctuation data output by the general pain model;

[0011] The first pain dataset contains multiple pain levels within a preset duration from the sampling time as the actual pain level fluctuation data.

[0012] The output value of the first loss function is obtained based on the test pain level fluctuation data and the actual pain level fluctuation data.

[0013] The weights of the general pain model are continuously adjusted based on the output value until the output value of the first loss function converges to a preset threshold.

[0014] Based on the weight values ​​of the general pain model corresponding to the convergence of the first loss function, the personal pain model corresponding to the target object is determined.

[0015] In one embodiment, obtaining the first pain dataset of the target object includes:

[0016] The device receives vital sign parameters and pain levels at multiple sampling times from a wearable device worn by the target object, wherein the pain level is input by the target object into the wearable device.

[0017] The first pain dataset is generated based on the vital signs parameters and pain levels received at each of the aforementioned sampling times.

[0018] In one embodiment, after obtaining the individual pain level fluctuation data of the target object, the prediction method further includes:

[0019] Determine whether the target subject is experiencing pain based on the individual pain level fluctuation data;

[0020] If the target object experiences pain, a prompt message will be output.

[0021] In one embodiment, after predicting the pain level fluctuation of the target object based on the personal pain model and the second pain dataset to obtain the personal pain level fluctuation data of the target object, the prediction method further includes:

[0022] The medical data of the target subject, which was obtained in advance, is adjusted using the individual pain level fluctuation data.

[0023] In one embodiment, the process of acquiring medical data includes:

[0024] Obtain the third pain dataset of the target object;

[0025] The third pain dataset is input into the general pain model to obtain the general pain level fluctuation data output by the general pain model;

[0026] The medical data is determined based on the general pain level fluctuation data.

[0027] In one embodiment, before training a pre-obtained general pain model using a first pain dataset, the prediction method further includes:

[0028] The initial model is trained using a general pain dataset to obtain candidate pain models;

[0029] Candidate pain models were validated using a general pain validation set;

[0030] If the validation is successful, the candidate pain model will be adopted as the general pain model.

[0031] If the validation fails, continue training the candidate pain model until the validation succeeds.

[0032] In one embodiment, an initial model is trained using a general pain dataset to obtain a candidate pain model, including:

[0033] Assign the same initial values ​​to the weights in the initial model;

[0034] Obtain the pre-defined hyperparameters;

[0035] Based on hyperparameters and a general pain dataset, the initial model after assignment is trained to obtain candidate pain models.

[0036] In one embodiment, the vital signs parameters include the deviation of the target subject's vital signs when pain occurs compared to when pain does not occur.

[0037] Secondly, this application also provides a device for predicting pain level fluctuations. The prediction device includes:

[0038] The data acquisition module is used to acquire the first pain dataset of the target object;

[0039] The model training module is used to train a pre-obtained general pain model using the first pain dataset to obtain a personal pain model corresponding to the target object. The general pain model is trained based on the general pain dataset.

[0040] The data prediction module is used to acquire the second pain dataset of the target object, and predict the pain level fluctuation of the target object based on the personal pain model and the second pain dataset to obtain the personal pain level fluctuation data of the target object.

[0041] In one embodiment, the model training module is further configured to input the vital signs parameters and pain level at the sampling time into the general pain model to obtain the test pain level fluctuation data output by the general pain model;

[0042] The first pain dataset contains multiple pain levels within a preset duration from the sampling time as the actual pain level fluctuation data.

[0043] The output value of the first loss function is obtained based on the test pain level fluctuation data and the actual pain level fluctuation data.

[0044] The weight values ​​of the general pain model are adjusted based on the output values;

[0045] Based on the weight values ​​when the loss function converges, a personal pain model corresponding to the target object is obtained.

[0046] In one embodiment, the data acquisition module is further configured to receive vital sign parameters and pain levels at multiple sampling times sent by the wearable device worn by the target object, wherein the pain level is input by the target object into the wearable device;

[0047] The first pain dataset is generated based on the vital signs parameters and pain levels received at each of the aforementioned sampling times.

[0048] In one embodiment, the prediction device further includes:

[0049] The early warning module is used to determine whether the target object is experiencing pain based on the individual's pain level fluctuation data;

[0050] If the target object experiences pain, a prompt message will be output.

[0051] In one embodiment, the prediction device further includes:

[0052] The adjustment module is used to adjust the pre-acquired medical data of the target object using the individual pain level fluctuation data.

[0053] In one embodiment, the adjustment module is further configured to acquire a third pain dataset of the target object;

[0054] The third pain dataset is input into the general pain model to obtain the general pain level fluctuation data output by the general pain model;

[0055] The medical data is determined based on the general pain level fluctuation data.

[0056] In one embodiment, the model training module is further configured to train the initial model using a general pain dataset to obtain a candidate pain model;

[0057] Candidate pain models were validated using a general pain validation set;

[0058] If the validation is successful, the candidate pain model will be adopted as the general pain model.

[0059] If the validation fails, continue training the candidate pain model until the validation succeeds.

[0060] In one embodiment, the model training module is also used to assign the same initial values ​​to the weights in the initial model;

[0061] Obtain the pre-defined hyperparameters;

[0062] Based on hyperparameters and a general pain dataset, the initial model after assignment is trained to obtain candidate pain models.

[0063] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the above-described method for predicting pain level fluctuations.

[0064] The aforementioned method, apparatus, and computer equipment for predicting pain level fluctuations train a pre-obtained general pain model using a first pain dataset of the target subject to obtain a personal pain model suitable for the target subject. Then, based on the personal pain model and a second pain dataset of the target subject, the pain level fluctuations of the target subject are predicted to obtain personal pain level fluctuation data suitable for the target subject. This avoids the situation where the pain level fluctuation data is determined based on the pain level selected by the target subject or the pain level selected by medical staff, which leads to inaccurate pain level fluctuation data, thus improving the accuracy of pain level fluctuation assessment. Attached Figure Description

[0065] Figure 1 This is an application environment diagram of a method for predicting pain level fluctuations in one embodiment;

[0066] Figure 2 This is a flowchart illustrating a method for predicting pain level fluctuations in one embodiment;

[0067] Figure 3 This is a flowchart illustrating a method for predicting pain level fluctuations in another embodiment;

[0068] Figure 4 This is a flowchart illustrating a method for predicting pain level fluctuations in another embodiment;

[0069] Figure 5 This is a structural block diagram of a pain level fluctuation prediction device in one embodiment;

[0070] Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0071] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0072] The method for predicting pain level fluctuations provided in this application can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or placed on the cloud or other network servers. In a specific implementation, terminal 102 can communicate with server 104 via Narrow Band Internet of Things (NB-IoT). Terminal 102 can periodically collect the target object's first pain dataset, such as vital sign parameters. The target object can also upload its current pain level by operating terminal 102. Simultaneously, terminal 102 will also collect and upload the target object's current vital signs. After obtaining the target object's first pain dataset, server 104 uses the first pain dataset to train a pre-obtained general pain model to obtain a personal pain model suitable for the target object. Then, based on the personal pain model, it predicts the target object's pain level fluctuations to obtain personal pain level fluctuation data suitable for the target object. Terminal 102 can be, but is not limited to, various portable wearable devices, such as smartwatches, smart bracelets, and head-mounted devices. Server 104 can be implemented using a standalone server or a server cluster composed of multiple servers.

[0073] In one embodiment, such as Figure 2 As shown, a method for predicting pain level fluctuations is provided, which can be applied to... Figure 1 Taking the server in the example, the following steps may be included:

[0074] Step 202: Obtain the first pain dataset of the target object.

[0075] The target object refers to the object for which pain level fluctuation prediction is needed, such as an individual or other test subjects. The first pain dataset refers to the set of pain-related data of the target object within a first time period, such as the target object's vital signs, gender, age, indoor and outdoor environment, clinical diagnosis, type of surgery, postoperative time, health status, and pain level when pain occurs.

[0076] The server can directly obtain the target object's first pain dataset from the terminal, or it can read the target object's first pain dataset from the data storage system. Taking a wearable device as the terminal and the first pain dataset including vital sign parameters and pain levels at multiple sampling times as an example, the server can receive vital sign parameters and pain levels at multiple sampling times sent by the wearable device worn by the target object. Among them, the pain level is input by the target object into the wearable device, and then the server generates the first pain dataset based on the received vital sign parameters and pain levels at each sampling time.

[0077] Step 204: Train the pre-obtained general pain model using the first pain dataset to obtain a personal pain model corresponding to the target object. The general pain model is trained based on the general pain dataset.

[0078] It should be understood that a general pain dataset refers to a collection of pain-related data for a general subject at different times, and a general pain model can predict fluctuations in the pain level of a general subject. Because of the universality of the general pain model, in this embodiment, the first pain dataset of the target subject is used to correct the general pain model through machine learning, generating a personalized pain model suitable for each target subject.

[0079] In a specific implementation, the first pain dataset may include the vital signs parameters and pain levels of the target object at multiple sampling times. In this embodiment, the vital signs parameters and pain levels at the sampling times can be input into a general pain model to obtain the test pain level fluctuation data output by the general pain model; multiple pain levels within a preset time period from the sampling time in the first pain dataset are taken as the real pain level fluctuation data; the output value of the first loss function is obtained based on the test pain level fluctuation data and the real pain level fluctuation data at all times; the weight values ​​of the general pain model are adjusted according to the output values; and a personal pain model corresponding to the target object is obtained based on the weight values ​​when the loss function converges.

[0080] Among them, vital sign parameters can include the deviation of vital signs when the target subject is in pain and when he / she is not in pain, and vital signs can include respiration, body temperature, pulse, blood pressure, etc.

[0081] Pain level fluctuation data is a dataset composed of multiple pain levels over a period of time. Taking a sampling period of time t, the pain level at the current sampling time T1 is 1, the pain level at sampling time T1+t is 2, and the pain level at sampling time T1+2t is 3 as an example, the pain level fluctuation data at the current sampling time T1 can be 1-2-3.

[0082] Real pain level fluctuation data refers to a dataset consisting of multiple pain levels within a preset time period from a certain sampling time in the first pain dataset, such as all pain levels within 3 days from a certain sampling time in the first pain dataset.

[0083] Step 206: Obtain the second pain dataset of the target object, and predict the pain level fluctuation of the target object based on the personal pain model and the second pain dataset to obtain the personal pain level fluctuation data of the target object.

[0084] The second pain dataset refers to the set of pain-related data of the target object in a second time period, which is different from the first time period and is later than the first time period.

[0085] It should be understood that personal pain level fluctuation data refers to pain level fluctuation data applicable to the target individual. After determining the target individual's personal pain model, when the second pain dataset of the target individual is received from the wearable device, the second pain dataset is input into the target individual's personal pain model to obtain the target individual's personal pain level fluctuation data.

[0086] After obtaining the individual's pain level fluctuation data, this embodiment can also determine whether the target object is experiencing pain based on the individual's pain level fluctuation data; if the target object is experiencing pain, a prompt message is output.

[0087] Specifically, the first pain level in the individual pain level fluctuation data can be compared with the preset pain level. When the first pain level is greater than or equal to the preset pain level, it is determined that the target subject is experiencing pain. When the first pain level is less than the preset pain level, it is determined that the target subject is not experiencing pain.

[0088] Taking a sample time when the individual's pain level fluctuation data is 7-6-8, and the preset pain level is 5 as an example, since 7 is greater than 5, it can be determined that the target object is in pain.

[0089] It should be understood that individual pain level fluctuation data can accurately reflect the pain level of the target subject at a certain sampling time. When the target subject experiences pain, outputting prompt information can help doctors treat the target subject in a timely manner.

[0090] In the above-mentioned method for predicting pain level fluctuations, a pre-obtained general pain model is trained using the first pain dataset of the target object to obtain a personal pain model suitable for the target object. Then, the pain level fluctuations of the target object are predicted based on the personal pain model and the second pain dataset of the target object to obtain personal pain level fluctuation data suitable for the target object. This avoids the situation where the pain level fluctuation data is determined based on the pain level selected by the target object or the pain level selected by the medical staff, which leads to inaccurate pain level fluctuation data, thus improving the accuracy of pain level fluctuation assessment.

[0091] Figure 3 This is a flowchart illustrating a method for predicting pain level fluctuations in another embodiment, as shown below. Figure 3 As shown, after step 206, the prediction method may further include the following steps:

[0092] Step 208: Adjust the pre-acquired medical data of the target subject using individual pain level fluctuation data.

[0093] It should be understood that medical data refers to data related to the diagnosis and treatment of a target individual, such as treatment plans. In specific implementation, a third pain dataset of the target individual can be obtained; this third pain dataset can be input into a general pain model to obtain general pain level fluctuation data output by the general pain model; and medical data can be determined based on the general pain level fluctuation data.

[0094] The third pain dataset refers to the set of pain-related data of the target object in the third time period. The third time period is different from the first and second time periods, and the third time period is earlier than the first time period.

[0095] Taking hospitalized patients as the target group and treatment plans as the medical data as an example, when a patient is admitted, the third pain dataset is acquired and input into a general pain model to obtain general pain level fluctuation data output by the general pain model. Based on this general pain level fluctuation data, a preliminary treatment plan is determined for the patient. After admission, the first pain dataset sent by the patient's wearable device is received in real time. This first pain dataset is used to train the general model to obtain a personal pain model. Then, the second pain dataset sent by the patient's wearable device is received in real time and input into the patient's personal pain model for prediction, yielding the patient's personal pain level fluctuation data. Since the pain level fluctuation data at admission and after admission may differ, the corresponding treatment plans will also differ. Therefore, the initial treatment plan can be adjusted based on the predicted personal pain level fluctuation data obtained after admission to obtain a more suitable treatment plan for the patient.

[0096] This embodiment adjusts the pre-acquired medical data of the target subject by utilizing individual pain level fluctuation data, thereby obtaining more suitable medical data for the target subject and improving the practicality of the solution.

[0097] Figure 4 This is a flowchart illustrating a method for predicting pain level fluctuations in another embodiment, as shown below. Figure 4 As shown, the prediction method may also include the following steps:

[0098] Step 2011: Train the initial model using a general pain dataset to obtain a candidate pain model.

[0099] In a practical implementation, general pain data for general objects can be obtained. After labeling and cleaning the general pain data, it can be divided into a general pain dataset and a general pain validation set. The initial model can be trained using the general pain dataset to obtain candidate pain models.

[0100] During training, the weights in the initial model can be assigned the same initial values; pre-set hyperparameters can be obtained; and the initial model with the assigned values ​​can be trained based on the hyperparameters and a general pain dataset to obtain a candidate pain model.

[0101] The hyperparameters can include learning rate, batch size, optimizer, number of training rounds, etc. A general pain dataset is distributed and input into the algorithm. Machine learning is performed in conjunction with the hyperparameters to start the iterative algorithm. In each iteration, the weight values ​​are automatically adjusted through backpropagation. The second loss function is recorded in each round of calculation. After multiple rounds of training, the second loss function converges continuously. The candidate pain model is obtained based on the weight values ​​when the second loss function converges.

[0102] Step 2012: Validate the candidate pain model using a general pain validation set.

[0103] In a practical implementation, a general pain validation set can be input into a candidate pain model to obtain the prediction results output by the candidate pain model. The prediction results can then be compared with the actual results in the general pain validation set to validate the candidate pain model.

[0104] Step 2013: If the verification is successful, the candidate pain model is adopted as the general pain model.

[0105] In one example, when the error between the predicted result and the true result in the general pain validation set is within a preset range (e.g., ±5%), the validation can be determined to be successful, and the candidate pain model at this time is the general pain model.

[0106] Step 2014: If the verification fails, continue training the candidate pain model until the verification passes.

[0107] In another example, if the error between the predicted result and the true result in the general pain validation set is not within a preset range (±5%), the validation is determined to fail, and the candidate pain model needs to be trained again.

[0108] In this embodiment, a candidate pain model is obtained by training the model using a general pain dataset. The candidate pain model is then validated using a general pain validation set to generate a general pain model applicable to a general population, which further improves the prediction accuracy of the individual pain level fluctuation data of the target population.

[0109] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0110] Based on the same inventive concept, this application also provides a pain level fluctuation prediction device for implementing the pain level fluctuation prediction method described above. The solution provided by this device is similar to the implementation described in the above method; therefore, the specific limitations of one or more pain level fluctuation prediction device embodiments provided below can be found in the limitations of the pain level fluctuation prediction method described above, and will not be repeated here.

[0111] In one embodiment, such as Figure 5 As shown, a device for predicting pain level fluctuations is provided, comprising: a data acquisition module 10, a model training module 20, and a data prediction module 30, wherein:

[0112] Data acquisition module 10 is used to acquire the first pain dataset of the target object;

[0113] The model training module 20 is used to train a pre-obtained general pain model using the first pain dataset to obtain a personal pain model corresponding to the target object. The general pain model is trained based on the general pain dataset.

[0114] The data prediction module 30 is used to acquire the second pain dataset of the target object, and predict the pain level fluctuation of the target object based on the personal pain model and the second pain dataset to obtain the personal pain level fluctuation data of the target object.

[0115] In one embodiment, the first human pain dataset includes vital sign parameters and pain levels of the target object at multiple sampling times. The model training module 20 is further used to input the vital sign parameters and pain levels at the sampling times into a general pain model to obtain test pain level fluctuation data output by the general pain model; to use multiple pain levels within a preset time period from the sampling time in the first pain dataset as real pain level fluctuation data; to obtain the output value of the first loss function based on the test pain level fluctuation data and the real pain level fluctuation data; to adjust the weight values ​​of the general pain model based on the output values; and to obtain a personal pain model corresponding to the target object based on the weight values ​​when the loss function converges.

[0116] In one embodiment, the data acquisition module 10 is further configured to receive vital sign parameters and pain levels at multiple sampling times sent by the wearable device worn by the target object, wherein the pain level is input by the target object into the wearable device; and generate a first pain dataset based on the received vital sign parameters and pain levels at each sampling time.

[0117] In one embodiment, the prediction device further includes an early warning module (not shown) for determining whether the target object is experiencing pain based on individual pain level fluctuation data; and outputting a prompt message if the target object is experiencing pain.

[0118] In one embodiment, the prediction device further includes an adjustment module (not shown) for adjusting the pre-acquired medical data of the target subject using individual pain level fluctuation data.

[0119] In one embodiment, the adjustment module is further configured to acquire a third pain dataset of the target object; input the third pain dataset into a general pain model to obtain general pain level fluctuation data output by the general pain model; and determine medical data based on the general pain level fluctuation data.

[0120] In one embodiment, the model training module 20 is further configured to train the initial model using a general pain dataset to obtain a candidate pain model; to validate the candidate pain model using a general pain validation set; if the validation passes, the candidate pain model is adopted as the general pain model; if the validation fails, the candidate pain model is trained until the validation passes.

[0121] In one embodiment, the model training module 20 is further configured to assign the same initial values ​​to the weights in the initial model; obtain pre-set hyperparameters; and train the initial model after assignment based on the hyperparameters and a general pain dataset to obtain a candidate pain model.

[0122] The modules in the aforementioned pain level fluctuation prediction device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0123] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores data such as pain datasets for the first person, pain datasets for the second person, and historical vital sign parameters. The network interface is used for communication with external terminals via a network connection. When executed by the processor, the computer program implements a method for predicting pain level fluctuations.

[0124] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0125] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0126] 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, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0127] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0128] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0129] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for predicting fluctuations in pain levels, characterized in that, The prediction method includes: Obtain the first pain dataset of the target object; the first pain dataset refers to the set of pain-related data of the target object within a first time period; The first pain dataset is used to train a pre-obtained general pain model to obtain a personal pain model corresponding to the target object, wherein the general pain model is trained based on the general pain dataset; the first pain dataset is used to correct the general pain model. A second pain dataset of the target object is obtained, and the fluctuation of the target object's pain level is predicted based on the personal pain model and the second pain dataset to obtain the target object's personal pain level fluctuation data; the second pain dataset refers to the set of pain-related data of the target object in a second time period; the second time period is different from the first time period, and the second time period is later than the first time period; The first pain dataset includes the vital signs parameters and pain levels of the target object at multiple sampling times. The step of training a pre-obtained general pain model using the first pain dataset to obtain a personal pain model corresponding to the target object includes: The vital signs parameters and pain levels at the sampling time are input into the general pain model to obtain the test pain level fluctuation data output by the general pain model; The first pain dataset contains multiple pain levels within a preset duration from the sampling time as the actual pain level fluctuation data. The output value of the first loss function is obtained based on the test pain level fluctuation data and the actual pain level fluctuation data. The weight values ​​of the general pain model are continuously adjusted based on the output value until the output value of the first loss function converges to a preset threshold. Based on the weight values ​​of the general pain model corresponding to the convergence of the first loss function, a personal pain model corresponding to the target object is determined.

2. The prediction method according to claim 1, characterized in that, The acquisition of the first pain dataset of the target object includes: The device receives vital sign parameters and pain levels at multiple sampling times from a wearable device worn by the target object, wherein the pain level is input by the target object into the wearable device. The first pain dataset is generated based on the vital signs parameters and pain levels received at each of the aforementioned sampling times.

3. The prediction method according to claim 1, characterized in that, After obtaining the individual pain level fluctuation data of the target object, the prediction method further includes: Determine whether the target subject is experiencing pain based on the individual pain level fluctuation data; If the target object experiences pain, a prompt message will be output.

4. The prediction method according to any one of claims 1 to 3, characterized in that, After predicting the pain level fluctuation of the target object based on the personal pain model and the second pain dataset to obtain the personal pain level fluctuation data of the target object, the prediction method further includes: The medical data of the target subject, which was obtained in advance, is adjusted using the individual pain level fluctuation data.

5. The prediction method according to claim 4, characterized in that, The process of acquiring the medical data includes: Obtain the third pain dataset of the target object; The third pain dataset is input into the general pain model to obtain the general pain level fluctuation data output by the general pain model; The medical data is determined based on the general pain level fluctuation data.

6. The prediction method according to any one of claims 1 to 3, characterized in that, Before training the pre-obtained general pain model using the first pain dataset, the prediction method further includes: The initial model is trained using the general pain dataset to obtain candidate pain models; The candidate pain model was validated using a general pain validation set; If the verification is successful, the candidate pain model will be adopted as the general pain model. If the verification fails, the candidate pain model will continue to be trained until the verification passes.

7. The prediction method according to claim 6, characterized in that, The process of training the initial model using the general pain dataset to obtain candidate pain models includes: Assign the same initial values ​​to the weights in the initial model; Obtain the pre-defined hyperparameters; Based on the hyperparameters and the general pain dataset, the initial model after assignment is trained to obtain the candidate pain model.

8. The prediction method according to any one of claims 1 to 3, characterized in that, The vital signs parameters include the deviations of the target object's vital signs when the object is experiencing pain compared to when it is not experiencing pain.

9. A device for predicting pain level fluctuations, characterized in that, The prediction device is used to perform the steps of the method according to any one of claims 1 to 8.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 8.

Images