Emotional health data processing system, method, and computer device

Abstract

This specification provides an emotional health data processing system, method, and computer equipment. Based on this method, before implementation, continuous sample detection data from a relatively long testing period of the target subjects can be collected and utilized. By jointly using logistic regression analysis, a generalized linear mixed-effects model, Poisson regression analysis, and a latent class mixed-effects model, a preset prediction model suitable for emotional health prediction can be trained. In specific implementation, the acquisition and processing modules first acquire and determine the key characteristic change trajectory of the target subject's blood cells during the first time period based on the blood cell detection data of the target subject during that first time period. Then, the determination module uses the preset prediction model to determine the target subject's emotional health prediction result based on the key characteristic change trajectory of the target subject's blood cells during the first time period. This allows for efficient and intelligent automatic determination of the target subject's emotional health status.

Description

Technical Field

[0001] This specification belongs to the field of artificial intelligence technology, and in particular relates to emotional health data processing systems, methods and computer devices. Background Technology

[0002] As living standards continue to improve, people are paying more and more attention to their emotional health. However, based on existing methods, it is often difficult to accurately predict a user's emotional health in advance.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This specification provides an emotional health data processing system, method, and computer device that can efficiently and intelligently predict the emotional health status of a target subject automatically.

[0005] This manual provides an emotional health data processing system, including:

[0006] The acquisition module is used to acquire blood cell test data of the target object in the first time period; wherein, the blood cell test data in the first time period includes blood cell test data at multiple time points;

[0007] The processing module is used to process the blood cell detection data of the target object in the first time period and determine the key feature change trajectory of the blood cells of the target object in the first time period.

[0008] The determination module is used to determine the emotional health prediction result of the target object based on the key feature change trajectory of blood cells of the target object in the first time period using a preset prediction model.

[0009] In one embodiment, the preset prediction model is trained in the following manner:

[0010] Obtain sample detection data of the sample test object based on the sample test time period;

[0011] The sample detection data is processed according to the preset processing rules to obtain sample training data that meets the requirements.

[0012] Using the sample training data, a preset prediction model that meets the requirements is trained by combining logistic regression analysis, generalized linear mixed effects model, Poisson regression analysis, and latent class mixed effects model.

[0013] In one embodiment, using the sample training data, a pre-defined prediction model that meets the requirements is trained by jointly using logistic regression analysis, a generalized linear mixed-effects model, Poisson regression analysis, and a latent class mixed-effects model, including:

[0014] The sample training data is split into a first dataset, a second dataset, and a third dataset;

[0015] Based on auxiliary reference data, several initial models based on the latent class mixed effects model were constructed;

[0016] Multiple initial models were trained using the first dataset, resulting in multiple intermediate models.

[0017] The second dataset was used to test multiple intermediate models; and based on the model test results, the target intermediate model that meets the requirements was selected from the multiple intermediate models.

[0018] Using a third dataset, logistic regression analysis, generalized linear mixed-effects model, and Poisson regression analysis are used in combination to adjust the target intermediate model in multiple rounds, resulting in a preset prediction model that meets the requirements.

[0019] In one embodiment, using a third dataset, logistic regression analysis, a generalized linear mixed-effects model, and Poisson regression analysis are used in combination to perform multiple rounds of adjustments on the target intermediate model to obtain a preset prediction model that meets the requirements, including:

[0020] The target intermediate model for the current round is adjusted by combining logistic regression analysis, generalized linear mixed effects model, and Poisson regression analysis in the following manner:

[0021] Obtain the target intermediate model from the previous round, and use the target intermediate model from the previous round to process the sample training data of the current round in the third dataset to obtain the prediction result of the current round.

[0022] Based on the prediction results of the current round and the primary and secondary labels of the sample training data of the current round, check whether the target intermediate model of the previous round meets the requirements;

[0023] If the target intermediate model of the previous round is determined to be unsuitable, the sample training data of the current round in the third dataset is processed by the first auxiliary model based on logistic regression analysis, the second auxiliary model based on generalized linear mixed effects model, and the third auxiliary model based on Poisson regression analysis, respectively, to obtain the corresponding first auxiliary result, second auxiliary result, and third auxiliary result of the current round.

[0024] Based on the prediction results of the current round, the first auxiliary result of the current round, the second auxiliary result of the current round, and the third auxiliary result of the current round, the model adjustment rules for the current round for the target intermediate model of the previous round are determined;

[0025] Based on the model adjustment rules of the current round, adjust the target intermediate model of the previous round to obtain the target intermediate model of the current round.

[0026] In one embodiment, while processing the blood cell detection data of the target object for a first time period and determining the key characteristic change trajectory of the blood cells of the target object for the first time period, the acquisition module is also used to acquire auxiliary reference data of the target object; wherein, the auxiliary reference data includes at least one of the following: age, gender, lifestyle habits, and medical history;

[0027] The processing module is also used to construct a joint feature data set of the target object based on the key feature change trajectory of blood cells in the first time period of the target object and auxiliary reference data;

[0028] The determining module is also used to determine the emotional health prediction result of the target object by processing the joint feature data group of the target object using a preset prediction model.

[0029] In one embodiment, the key characteristics of the blood cells include: key characteristics of red blood cells and key characteristics of white blood cells;

[0030] The key characteristics of the red blood cells include at least one of the following: red blood cell count, red blood cell distribution width (SD), red blood cell distribution width (CV), hematocrit, mean corpuscular hemoglobin concentration, mean corpuscular hemoglobin content, mean corpuscular volume, and hemoglobin concentration.

[0031] The key characteristics of the white blood cells include at least one of the following: white blood cell count, neutrophil count, neutrophil percentage, lymphocyte count, lymphocyte percentage, eosinophil count, eosinophil percentage, basophil count, basophil percentage, monocyte count, monocyte percentage, T lymphocyte CD3 count, T lymphocyte CD3 percentage, T lymphocyte CD4 count, T lymphocyte CD4 percentage, T lymphocyte CD8 count, T lymphocyte CD8 percentage, and T lymphocyte CD4 / CD8 ratio.

[0032] In one embodiment, the key characteristics of the blood cells further include: key characteristics of platelets and / or composite characteristics among multiple types of blood cells;

[0033] The key characteristics of platelets include at least: platelet count;

[0034] The composite characteristics among the various blood cell types include at least one of the following: the ratio of monocytes to lymphocytes, the ratio of neutrophils to lymphocytes, and the ratio of platelets to lymphocytes.

[0035] In one embodiment, a prompting module is also included;

[0036] If, based on the emotional health prediction results of the target object, it is determined that the target object has an emotional health risk, the prompting module is used to generate emotional health risk prompt information; and based on the key feature change trajectory of blood cells of the target object in the first time period, a target emotional health adjustment plan that matches the current emotional health status of the target object is determined.

[0037] Furthermore, after determining a target emotional health adjustment plan that matches the current emotional health status of the target object, the acquisition module is also used to acquire blood cell test data of the target object for a second time period; wherein, the second time period is the period after the target emotional health adjustment plan is started.

[0038] The processing module is also used to determine the key feature change trajectory of blood cells in the target object during the second time period based on the blood cell detection data of the target object during the second time period.

[0039] The determining module is also used to determine risk change data about the emotional health of the target user in the second time period based on the key feature change trajectory of blood cells in the second time period of the target object and the key feature change trajectory of blood cells in the first time period of the target object;

[0040] The prompting module is also used to update the target emotional health adjustment plan based on the risk change data of the second time period, the key feature change trajectory of the target object's blood cells in the second time period, and the key feature change trajectory of the target object's blood cells in the first time period.

[0041] This manual also provides a method for processing emotional health data, including:

[0042] Obtain blood cell test data for the target object in the first time period; wherein, the blood cell test data in the first time period includes blood cell test data at multiple time points;

[0043] Process the blood cell detection data of the target object in the first time period to determine the key feature change trajectory of the blood cells of the target object in the first time period;

[0044] Using a pre-defined prediction model, the target object's emotional health prediction result is determined based on the key characteristic change trajectory of blood cells in the first time period.

[0045] This specification also provides a computer device including a processor and a memory for storing processor-executable instructions, wherein the processor executes the instructions to implement the relevant steps of the emotional health data processing method.

[0046] Based on the emotional health data processing system, method, and computer equipment provided in this specification, before implementation, continuous sample detection data from a relatively long sample testing period of the test subjects can be collected and utilized. By combining multiple algorithmic models such as logistic regression analysis, generalized linear mixed-effects model, Poisson regression analysis, and latent class mixed-effects model, a preset prediction model suitable for emotional health prediction can be trained. In specific implementation, the acquisition and processing modules first acquire and determine the key characteristic change trajectory of the target subject's blood cells during the first time period based on the blood cell detection data of the target subject. Then, the determination module uses the aforementioned preset prediction model to predict the target subject's emotional health prediction result based on the key characteristic change trajectory of the target subject's blood cells during the first time period. This effectively mines and fully utilizes the data information carried by the target subject's blood cell detection data, efficiently and intelligently predicting the target subject's emotional health status automatically. Attached Figure Description

[0047] To more clearly illustrate the embodiments of this specification, the accompanying drawings used in the embodiments will be briefly introduced below. The drawings described below are only some embodiments recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a schematic diagram of the structural composition of an emotional health data processing system provided in one embodiment of this specification;

[0049] Figure 2 This is a schematic diagram of an embodiment of training a preset prediction model using the emotional health data processing system provided in the embodiments of this specification in a scenario example;

[0050] Figure 3 This is a schematic diagram of an embodiment of the emotional health data processing system provided in this specification for processing sample test data of a sample test subject in a scenario example;

[0051] Figure 4 This is a schematic diagram of an embodiment in which the emotional health data processing system provided in the embodiments of this specification is used to perform a round of model training and adjustment on a preset prediction model in a scenario example;

[0052] Figure 5 This is a schematic diagram of an embodiment in which the emotional health data processing system provided in the embodiments of this specification is used in conjunction with multiple algorithm models to train a preset prediction model in a scenario example;

[0053] Figure 6 This is a flowchart illustrating an embodiment of the emotional health data processing method provided in this specification.

[0054] Figure 7 This is a schematic diagram of the structural composition of a computer device provided in one embodiment of this specification. Detailed Implementation

[0055] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0056] It should be noted that the information and data related to users involved in the embodiments of this specification are all information and data authorized by the user or fully authorized by the relevant parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with relevant laws, regulations, and standards, and necessary confidentiality measures have been taken. They do not violate public order and good morals, and corresponding operation entry points are provided for users or relevant parties to choose to authorize or refuse.

[0057] It should also be noted that in the embodiments of this specification, certain software, components, models and other existing solutions in the industry may be mentioned. These should be regarded as exemplary and are only intended to illustrate the feasibility of implementing the technical solution of this application. However, it does not mean that the applicant has used or necessarily used the solution.

[0058] See Figure 1 As shown in the embodiments of this specification, an emotional health data processing system is provided, which may specifically include the following structural modules:

[0059] The acquisition module 101 can be used to acquire blood cell test data of the target object in the first time period; wherein, the blood cell test data in the first time period includes blood cell test data at multiple time points;

[0060] The processing module 102 can be used to process the blood cell detection data of the target object in the first time period and determine the key feature change trajectory of the blood cells of the target object in the first time period.

[0061] The determination module 103 can be used to determine the emotional health prediction result of the target object based on the key feature change trajectory of blood cells in the first time period of the target object using a preset prediction model.

[0062] The target object mentioned above can be understood as a biological object whose emotional health status is to be predicted. Specifically, the target object can be a human user or an animal, such as a monkey or a chimpanzee.

[0063] The aforementioned first time period can be understood as a continuous period of time (e.g., the most recent two months). Blood cell counts of the target subject are performed at at least two different time points within this first time period to obtain the target subject's blood cell count data for that first time period. Accordingly, the blood cell count data for the target subject's first time period can include blood cell count data from multiple time points, specifically at least two different time points.

[0064] Specifically, for example, the aforementioned acquisition module can be connected to the health center's physical examination system's physical examination database.

[0065] Accordingly, in practice, the target individuals can regularly visit the health center for a physical examination every week. During the examination, blood cell test data can be collected from the target individuals and stored in the physical examination system's database.

[0066] When it is necessary to predict the future (or current) emotional health status of a target, the emotional health data processing system can use the acquisition module to query the physical examination database of the physical examination system based on the target's object identifier, and find the blood cell test data of the target at multiple time points within the first time period (for example, blood cell test data collected weekly within the most recent 2 months), as the blood cell test data of the target for the first time period.

[0067] The aforementioned key characteristic changes in blood cells may specifically include changes in designated biomarkers of blood cells.

[0068] The aforementioned pre-set prediction model can be understood as an algorithm model that is pre-trained by combining multiple analytical algorithms such as logistic regression analysis, generalized linear mixed effects model, Poisson regression analysis, and latent class mixed effects model. It is capable of predicting the current or future emotional health status of an input object based on the change trajectory of key features of blood cells over a period of time and the change trend over a period of time, and outputting the corresponding emotional health prediction result.

[0069] The above-mentioned emotional health prediction results can be used to indicate the future (or current) emotional health status of a target individual. For example, these results may include: a probability value of happiness and a level of happiness. Alternatively, they may include: a probability value of depression risk and a level of depression. Or, they may include: a probability value of anxiety risk and a level of anxiety, etc. It should be noted that the emotional health prediction results listed above are merely illustrative. In practice, depending on the specific application scenario and processing needs, the emotional health prediction results may include other content. This specification does not limit this.

[0070] The aforementioned emotional health prediction results can specifically include: primary labels and secondary labels. The primary labels are associated with the probability value of the target emotion (e.g., depression or happiness); the secondary labels are associated with the severity level of the target emotion. Accordingly, based on the emotional health prediction results, the primary labels can be used to determine whether the target individual possesses the target emotion; and, if the target individual is confirmed to possess the target emotion, the secondary labels can be used to further determine the specific severity level of the target emotion. This allows for an accurate and comprehensive determination of the target individual's current or future emotional health status. The target emotion can specifically include negative emotions, such as depression and anxiety, as well as positive emotions, such as happiness and satisfaction. When a negative emotion is predicted for the target individual in the future or currently, it can be determined that the target individual faces an emotional health risk.

[0071] Specifically, the aforementioned processing and determining modules can be integrated and deployed on the server side of the health center. The server can specifically include a backend server responsible for data processing, capable of data transmission and processing. Specifically, the server can be, for example, an electronic device with data computing, storage, and network interaction capabilities. Alternatively, the server can be a software program running on the electronic device, providing support for data processing, storage, and network interaction. In this embodiment, the number of servers is not specifically limited. The server can be a single server, several servers, or a server cluster formed by several servers.

[0072] Accordingly, in practical implementation, the emotional health data processing system can first use the blood cell detection data of the target object in the first time period through the processing module to obtain the key feature change trajectory of the target object's blood cells in the first time period through data fitting processing; then, the determination module uses a preset prediction model to process the key feature change trajectory of the target object's blood cells in the first time period to obtain the emotional health prediction result of the target object. Furthermore, the emotional health data processing system can also determine the current or future emotional health status of the target object based on the above emotional health prediction result.

[0073] Furthermore, the aforementioned emotional health data processing system may also include an interactive module, such as a display screen and / or a voice player. Accordingly, in specific implementations, the emotional health data processing system can also utilize the interactive module to communicate information about the target object's current or future emotional health status to the target object or other relevant users.

[0074] Based on the above embodiments, by introducing and utilizing a preset prediction model, the relevant information carried by the blood cell test data of the target object can be effectively mined and fully utilized, thereby enabling the target object's emotional health status to be determined efficiently and intelligently without increasing the testing cost.

[0075] In some embodiments, when the above-mentioned processing module specifically processes the blood cell detection data of the target object in the first time period and determines the key feature change trajectory of the blood cells in the first time period of the target object, it may include: determining the discrete data and continuous data in the blood cell detection data of the target object in the first time period; and obtaining the processed data by performing corresponding encoding processing on the discrete data and vector mapping on the continuous data; then calculating the key features of blood cells at multiple time points within the first time period based on the processed data; and constructing a curve that reflects the continuous dynamic changes of the key features of blood cells within a time period through data fitting based on the key features of blood cells at multiple time points, as the key feature change trajectory of the blood cells in the first time period of the target object.

[0076] In this way, the continuous change trajectory of key features of blood cells can be used to replace the original relatively discrete key features of blood cells at multiple time points. By mining and utilizing the time-series relationship of key features of blood cells, more accurate predictions can be made.

[0077] In some embodiments, in addition to acquiring the key feature change trajectory of blood cells of the target object in the first time period, auxiliary reference data of the target object can also be acquired simultaneously.

[0078] The aforementioned supplementary reference data shall include at least one of the following: age, gender, lifestyle habits, medical history, etc.

[0079] Specifically, the aforementioned lifestyle habits may include one or more of the following: smoking, drinking, staying up late, etc. The aforementioned diseases may include one or more of the following: a history of diabetes, a history of hypertension, etc.

[0080] Furthermore, the aforementioned supplementary reference data may also include: basic body attribute parameters, such as body fat percentage (BMI).

[0081] In practice, the above processing module can construct a joint feature data set of the target object based on the key feature change trajectory of blood cells in the first time period of the target object and auxiliary reference data.

[0082] Furthermore, the aforementioned determining module can use a preset prediction model to process the joint feature data set of the target object and determine the emotional health prediction result of the target object.

[0083] The aforementioned determining module utilizes a preset prediction model to process the joint feature data set of the target object and determine the emotional health prediction result of the target object. Specifically, this may include: extracting corresponding trajectory features from the key feature change trajectory of the blood cells using the preset prediction model; extracting corresponding auxiliary features from auxiliary reference data in the joint feature data set of the target object; using a first classifier to output a first prediction result corresponding to the primary label based on the curve features and the auxiliary features; using a second classifier to output a second prediction result corresponding to the secondary label based on the first prediction result, the curve features, and the auxiliary features; and using an output layer to finally generate and output the emotional health prediction result of the target object based on the first and second prediction results.

[0084] Based on the above embodiments, by further introducing and utilizing auxiliary reference data of the target object as a covariate for prediction guidance and supplementation, the emotional health prediction results of the target object can be predicted more accurately.

[0085] In some embodiments, the key characteristics of the aforementioned blood cells may specifically include biomarkers of blood cells that can be directly or indirectly related to changes in the emotional health of the subject. The aforementioned blood cells may include red blood cells, white blood cells, platelets, etc.

[0086] Specifically, the key characteristics of blood cells may include: key characteristics of red blood cells, key characteristics of white blood cells, etc.

[0087] Specifically, the key characteristics of red blood cells mentioned above may include at least one of the following: red blood cell count (RBC), red blood cell distribution width SD (RDW-SD), red blood cell distribution width CV (RDW-CV), hematocrit (HCT), mean corpuscular hemoglobin concentration (MCHC), mean corpuscular hemoglobin content (MCH), mean corpuscular volume (MCV), hemoglobin concentration (Hb), etc.

[0088] The aforementioned red blood cell distribution width (SD) specifically refers to the standard deviation of red blood cell distribution width, a parameter reflecting the heterogeneity of red blood cell volume and used to reflect the uniformity of red blood cell size and shape. The aforementioned red blood cell distribution width (CV) specifically refers to the coefficient of variation of red blood cell distribution width, used to reflect the dispersion of red blood cell size.

[0089] Furthermore, the key characteristics of red blood cells mentioned above may also include: erythrocyte sedimentation rate (ESR), erythrocyte sedimentation rate equation K value (ESR-K), etc.

[0090] The key characteristics of the aforementioned white blood cells may include at least one of the following: key characteristics of granulocytes (e.g., neutrophils, eosinophils, basophils, etc.), key characteristics of lymphocytes (e.g., B lymphocytes, T lymphocytes, natural killer cells, etc.), key characteristics of monocytes (e.g., macrophages, dendritic cells, etc.), etc.

[0091] The key characteristics of the aforementioned white blood cells may specifically include at least one of the following: white blood cell count (WBC), neutrophil count (NeuC), neutrophil percentage (Neu%), lymphocyte count (LymC), lymphocyte percentage (Lym%), eosinophil count (EosC), eosinophil percentage (Eos%), basophil count (BasC), basophil percentage (Bas%), monocyte count (MonC), monocyte percentage (Mon%), and T lymphocyte CD3 count (CD3 count), T lymphocyte CD3 percentage (CD3%), T lymphocyte CD4 count (CD4 count), T lymphocyte CD4 percentage (CD4%), T lymphocyte CD8 count (CD8 count), T lymphocyte CD8 percentage (CD8%), and T lymphocyte CD4 / CD8 ratio.

[0092] The T lymphocytes mentioned above specifically refer to mature lymphocytes in the thymus, which are divided into helper T cells (CD4+) and cytotoxic T cells (CD8+). CD3 specifically refers to the CD3 protein marker commonly carried on the surface of mature T lymphocytes; its value reflects the body's cellular immune function. CD4 specifically refers to the CD4 protein marker carried on the surface of helper T cells. CD8 specifically refers to the CD8 protein marker carried on the surface of cytotoxic T cells.

[0093] In some embodiments, the key characteristics of the blood cells may also include: key characteristics of platelets and / or composite characteristics among multiple types of blood cells, etc.

[0094] Among these, the key characteristics of platelets can include at least: platelet count (PLT), etc.

[0095] The aforementioned composite characteristics among multiple types of blood cells specifically refer to new indicator characteristics formed by the combination of key characteristics of different types of blood cells.

[0096] Specifically, the composite characteristics among the aforementioned multiple blood cell types include at least one of the following: monocyte-to-lymphocyte ratio (MLR), neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), etc.

[0097] It should be noted that the key characteristics of red blood cells mentioned above were chosen as the key characteristics of blood cells to be obtained and used because of the correlation between oxidative stress and changes in the emotional health of the subjects; furthermore, according to relevant studies, changes in red blood cell indicators (e.g., elevated RDW) may be related to hematopoietic dysfunction caused by oxidative damage. Therefore, by introducing and using key characteristics of red blood cells, it is possible to analyze and predict changes in the subjects' emotional health from the perspective of oxidative stress correlation.

[0098] The key characteristics of leukocytes mentioned above were chosen as the key blood cell characteristics to be obtained and used because of the correlation between inflammation and changes in the subject's emotional health, as well as the correlation between neuroendocrine interactions and changes in the subject's emotional health. For example, relevant studies have found that emotional health problems such as depression are associated with chronic low-grade inflammation, and using leukocyte subtypes (e.g., elevated CD4+ T cells, decreased CD8+ T cells) can reflect immune dysregulation. As another example, relevant studies have found that fluctuations in the monocyte / lymphocyte ratio (MLR) may affect blood-brain barrier permeability, indirectly leading to neuroinflammation. Therefore, by introducing and using key characteristics of leukocytes, changes in the subject's emotional health can be analyzed and predicted from the perspectives of inflammation correlation and neuroendocrine interaction correlation.

[0099] Furthermore, the selection of composite characteristics among multiple blood cell types as the key features for obtaining and using blood cells is based on the correlation between chronic inflammatory pathways and changes in the emotional health of the subjects. For example, relevant studies have found that emotional health problems such as depression are also associated with chronic inflammatory pathways. Therefore, by introducing and using composite characteristics among multiple blood cell types, it is possible to analyze and predict changes in the subjects' emotional health from the perspective of the correlation with chronic inflammatory pathways.

[0100] Of course, it should be noted that the key characteristics of blood cells listed above are only illustrative. In practice, other appropriate characteristics of blood cells may be introduced and used depending on the specific circumstances and processing requirements. This instruction manual does not limit this.

[0101] Based on the above embodiments, the key features of blood cells can be selectively acquired and used according to the blood cell detection data of the target object, thereby enabling a more accurate and comprehensive prediction of emotional health status by integrating multiple relevant perspectives.

[0102] In some embodiments, see Figure 2 As shown, the preset prediction model can be trained in the following manner:

[0103] S2-1: Obtain sample detection data of the sample test object based on the sample test time period;

[0104] S2-2: Process the sample detection data according to the preset processing rules to obtain sample training data that meets the requirements;

[0105] S2-3: Using the sample training data, a preset prediction model that meets the requirements is trained by combining logistic regression analysis, generalized linear mixed effects model, Poisson regression analysis and latent class mixed effects model.

[0106] Specifically, the test subjects mentioned above can be individuals who have prior knowledge and consent to participate in the test, and who have authorized the use of the relevant test data. The test data can include, at a minimum, the blood cell test data of the test subjects.

[0107] For different sample test objects, the sample test time period can be different; however, within the sample test time period for each sample test object, sample detection data must be collected at least twice at different times. The collection time of each sample detection data point within the sample test time period can be recorded as a test time point. Accordingly, the sample detection data for each sample test object based on the sample test time period can include sample detection data from multiple test time points.

[0108] The aforementioned logistic regression analysis can specifically refer to a generalized linear regression analysis model, often used in data mining. By introducing and using logistic regression analysis, greater emphasis can be placed on multivariate classification involving complex multiple variables (classifications with more than two categories), enabling more targeted discovery and utilization of data patterns for multivariate classification within complex data. Specifically, the aforementioned logistic regression analysis can be ordinal multivariate logistic regression analysis.

[0109] The Generalized Linear Mixed Model (GLMM) mentioned above is a type of hierarchical generalized linear model, and it is an algorithmic model based on statistical distribution. By introducing and using the GLMM, we can pay more attention to binary classification (classification of two categories) involving complex multivariates, and more effectively mine and utilize data patterns for binary classification in complex data.

[0110] The Poisson regression analysis mentioned above specifically refers to a generalized linear model based on the Poisson distribution, often used for analyzing the relationship between count-type dependent and independent variables. By introducing and using Poisson regression analysis, the time-dependent relationships between data can be more effectively uncovered, and data patterns based on time dependence can be identified in a targeted manner.

[0111] The aforementioned Latent Class Mixed Model (LCMM) specifically refers to an algorithmic model that simultaneously considers fixed effects (systematic influencing factors) and random effects (individual random variations) to classify potential classes. By introducing and using the Latent Class Mixed Model, it is more suitable for processing trajectory-type data and can also effectively take into account the data patterns of both binary and multi-class classification.

[0112] Based on the above embodiments, a relatively good preset prediction model can be obtained by combining various different model algorithms such as logistic regression analysis, generalized linear mixed effects model, Poisson regression analysis, and latent class mixed effects model.

[0113] Specifically, the above-mentioned acquisition of sample detection data of sample test objects based on the sample test time period can be implemented by obtaining the sample detection data of the current sample test object in the sample test objects in the following way:

[0114] S1: Collect auxiliary reference data for the current sample test object;

[0115] S2: According to the preset tracking and collection rules, collect blood cell detection data of the current sample test subject at multiple test time points during the sample test period; and conduct emotion tests related to the target emotion of interest on the current sample test subject to obtain blood cell detection data and emotion test results of the sample test subject at multiple test time points;

[0116] S3: Based on the emotion test results, determine the primary and secondary labels for multiple test time points of the current sample test subject; and concatenate the primary and secondary labels for multiple test time points with the blood cell detection data corresponding to the same test time point to obtain the detection data for multiple test time points; wherein, the primary label includes positive and negative labels, and the secondary label includes an empty label and multiple degree labels;

[0117] S4: Arrange the detection data of the multiple test time points in sequence according to the test time points, and combine them with the auxiliary reference data of the current sample test object to obtain the sample detection data of the current sample test object.

[0118] Specifically, taking depression as the target emotion, the above-mentioned emotion test can be an SDS-based emotion test.

[0119] Specifically, the SDS (Self-rating depression scale) mentioned above can refer to the self-rating depression scale.

[0120] Accordingly, in practice, a physical examination can be conducted on the current sample test subjects at the test time to obtain relevant blood cell test data. Simultaneously, after the physical examination, an SDS-based questionnaire can be sent to the current sample test subjects, and their feedback can be collected. Based on the feedback, the presence and degree of the target emotion in the current sample test subjects can be assessed to determine their emotional test results.

[0121] The primary labels mentioned above can be binary labels, including positive and negative labels. Positive labels can be used to indicate the presence of a target emotion, while negative labels can be used to indicate the absence of a target emotion.

[0122] The aforementioned secondary label can be a multi-dimensional label, including an empty label and multiple degree labels. The empty label can be used to indicate that the degree level of the target emotion is 0, that is, there is no target emotion label; the multiple degree labels can be used to indicate different degree levels of the target emotion.

[0123] The sample detection data of the current sample test object can also carry the time information of the test time point of each detection data.

[0124] After obtaining the sample detection data of the sample test objects based on the sample test time period, the discrete and continuous data in the sample detection data can be identified; and by performing corresponding encoding processing on the discrete data and vector mapping on the continuous data, the processed data can be obtained to facilitate subsequent data processing.

[0125] Following the above method, sample detection data based on the sample testing time period of multiple sample test objects can be accurately obtained.

[0126] In some embodiments, see Figure 3 As shown, the sample detection data is processed according to the preset processing rules to obtain sample training data that meets the requirements. In specific implementation, the sample detection data of the current sample test object can be processed according to the preset processing rules in the following manner to obtain sample training data of the current sample test object that meets the requirements:

[0127] S3-1: According to the preset processing rules, detect whether there are positive labels among the preset number of first-level labels ranked first in the sample detection data of the current sample test object;

[0128] S3-2: When it is determined that there are no positive labels among the top-ranked primary labels in the sample detection data of the current sample test object, calculate the mean and variance of the current detection data based on the detection data of multiple test time points in the sample detection data of the current sample test object.

[0129] S3-3: Based on the mean and variance of the current detection data, and the detection data at multiple test time points, calculate the score values ​​of the detection data at multiple test time points for the current sample test object.

[0130] S3-4: Calculate the mean and variance of the current score based on the scores of the detection data at multiple test time points;

[0131] S3-5: Subtract the mean of the current score from the test data at multiple test time points to obtain the adjusted score of the test data at multiple test time points;

[0132] S3-6: Divide the adjusted scores of the detection data from multiple test time points by the variance of the current score to obtain the sample training data of the current sample test object that meets the requirements.

[0133] The preset quantity can be 2, or other quantities.

[0134] In practice, by detecting whether a positive label exists among the top-ranked, predetermined number of primary labels in the current sample test subject's sample detection data, it can be determined whether the current sample test subject already possessed a relatively clear target emotion at the beginning. If so, the key features of the blood cells of the current sample test subject have little reference value for subsequent model training in predicting the emotional health status of other subjects who did not originally possess the target emotion, and may even mislead the model, affecting its accuracy.

[0135] Therefore, if it is determined that at least one positive label exists among the top-ranked preset number of primary labels in the current test object's sample detection data, the sample detection data of that current test object can be removed and not used in subsequent model training. Conversely, if it is determined that no positive label exists among the top-ranked preset number of primary labels in the current test object's sample detection data, the sample detection data of that current test object can be further processed to obtain qualified sample training data for subsequent model training.

[0136] By processing the detection data of each sample test object in the above manner, the processed data can have relatively good statistical distribution characteristics as a whole, which is more suitable for subsequent model analysis of data change trajectories, mining and utilizing the data patterns, and obtaining sample training data with better results that are suitable for training the preset prediction model.

[0137] In some embodiments, see Figure 4 As shown, the above-mentioned sample training data is used to train a preset prediction model that meets the requirements by jointly using logistic regression analysis, generalized linear mixed effects model, Poisson regression analysis, and latent class mixed effects model. In specific implementation, it may include the following:

[0138] S4-1: Split the sample training data into a first dataset, a second dataset, and a third dataset;

[0139] S4-2: Based on auxiliary reference data, construct multiple initial models based on the latent class mixed effects model;

[0140] S4-3: Use the first dataset to train multiple initial models to obtain multiple intermediate models;

[0141] S4-4: Use the second dataset to test multiple intermediate models; and based on the model test results, select the target intermediate model that meets the requirements from the multiple intermediate models.

[0142] S4-5: Using the third dataset, logistic regression analysis, generalized linear mixed effects model, and Poisson regression analysis are used in combination to adjust the target intermediate model in multiple rounds to obtain a preset prediction model that meets the requirements.

[0143] In practice, the sample training data can be randomly divided into three groups according to the corresponding proportions, which are respectively used as the first dataset, the second dataset, and the third dataset.

[0144] In practice, auxiliary reference data can be used as covariates to construct multiple initial models using a latent class mixed effects model as the base model. These multiple initial models include at least one initial model that does not consider the auxiliary reference data, and an initial model that considers a class of related auxiliary reference data.

[0145] Before implementation, historical data can be acquired and used to perform correlation analysis on multiple auxiliary reference data to obtain correlation analysis results. Based on the correlation analysis results, auxiliary reference data that meet the correlation requirements are divided into one category, resulting in multiple auxiliary reference data groups. Each auxiliary reference data group corresponds to one category.

[0146] In practice, a classification model that uses a latent class mixed effects model as the base model can be directly constructed as the initial model without considering auxiliary reference data, denoted as initial model 0.

[0147] Furthermore, in the initial model 0, a model adjustment term based on a class of related auxiliary reference data can be introduced and added, and the corresponding model can be adjusted and modified accordingly to obtain an initial model that considers a class of related auxiliary reference data, such as initial model 1, initial model 2, initial model 3, etc.

[0148] Furthermore, based on the initial model that considers one type of related auxiliary reference data, and according to the correlation analysis results, adjustment terms can be added for other types of related auxiliary reference data whose correlation with the previously considered type of related auxiliary reference data is greater than a preset correlation threshold. The corresponding model can then be adjusted and modified accordingly to obtain an initial model that considers multiple types of related auxiliary reference data.

[0149] In this way, multiple initial models with comprehensive coverage can be obtained, taking into account various auxiliary parameter data and combinations of various auxiliary reference data.

[0150] In some embodiments, see Figure 5As shown, the above-mentioned method utilizes a third dataset and combines logistic regression analysis, a generalized linear mixed-effects model, and Poisson regression analysis to perform multiple rounds of adjustments on the target intermediate model, thereby obtaining a preset prediction model that meets the requirements. In specific implementation, logistic regression analysis, a generalized linear mixed-effects model, and Poisson regression analysis can be combined to adjust the target intermediate model in the current round in the following manner:

[0151] S5-1: Obtain the target intermediate model from the previous round, and use the target intermediate model from the previous round to process the sample training data of the current round in the third dataset to obtain the prediction result of the current round.

[0152] S5-2: Based on the prediction results of the current round and the primary and secondary labels of the sample training data of the current round, check whether the target intermediate model of the previous round meets the requirements;

[0153] S5-3: If the target intermediate model of the previous round does not meet the requirements, the sample training data of the current round in the third dataset is processed by the first auxiliary model based on logistic regression analysis, the second auxiliary model based on generalized linear mixed effects model, and the third auxiliary model based on Poisson regression analysis, respectively, to obtain the corresponding first auxiliary result, second auxiliary result, and third auxiliary result of the current round.

[0154] S5-4: Based on the prediction results of the current round, the first auxiliary result of the current round, the second auxiliary result of the current round, and the third auxiliary result of the current round, determine the model adjustment rules for the current round for the target intermediate model of the previous round;

[0155] S5-5: Adjust the target intermediate model of the previous round according to the model adjustment rules of the current round to obtain the target intermediate model of the current round.

[0156] In practice, when the prediction results of the current round and the primary and secondary labels of the sample training data of the current round are used to check whether the target intermediate model of the previous round meets the requirements, and it is determined that the target intermediate model of the previous round meets the requirements, the model training can be terminated; and the target intermediate model of the previous round is determined as the preset prediction model that meets the requirements.

[0157] In practice, after adjusting and obtaining the target intermediate model for the current round, the next round of training can be carried out in the same manner as described above until the obtained target intermediate model meets the requirements.

[0158] In specific implementation, the above-mentioned use of the first auxiliary model based on logistic regression analysis to process the sample training data of the current round in the third dataset may include: using the first auxiliary model to process the sample training data of the current round and extracting the prediction reference information generated during the processing; wherein, the prediction reference information may include: odds ratio (OR) and / or 95% confidence interval (CI); based on the above prediction reference information, combined with the classification rules of the degree level of the target emotion and the time information of the test time point, the corresponding first auxiliary result of the current round is obtained.

[0159] Based on the above approach, by introducing and using the first auxiliary model, the advantages of logistic regression analysis can be fully utilized. By extracting and using the prediction reference information in the processing, the changing characteristics of the same sample test objects based on time updates can be effectively considered, thereby enabling more accurate analysis and processing of data patterns involving multivariate classification of individual objects.

[0160] Accordingly, by utilizing the first auxiliary result of the current round, the network structure involving multi-class classification processing in the target intermediate model of the previous round can be adjusted and optimized in a more targeted and effective manner.

[0161] In specific implementation, the above-mentioned use of the second auxiliary model based on the generalized linear mixed effects model to process the current round of sample training data in the third dataset may include:

[0162] S1: Based on the sample training data of the current round, determine the object identifier of the sample test object of the current round;

[0163] S2: Based on the object identifier of the sample test object in the current round, generate a corresponding random number as the random intercept term. For example, it can be represented as: ;

[0164] S3: Add the random intercept term to the sample training data of the current round to obtain the adjusted sample training data of the current round;

[0165] S4: Use the second auxiliary model to process the adjusted sample training data of the current round to obtain the corresponding second auxiliary result of the current round.

[0166] The object identifier of the aforementioned sample test object can be the identity ID of the sample test object, etc.

[0167] Specifically, for example, it can be configured according to the core formula model in the second auxiliary model:

[0168]

[0169] in, This represents the probability value of the predicted presence of the target emotion when the sample test subject i is numbered j. Represents the random intercept term. This indicates key characteristics of blood cells that influence their function. The key features of blood cells in test subject i, with sample number j. This represents the data value of the auxiliary reference data (or covariate) numbered k when the sample test object i is numbered j. This indicates the influence item of the auxiliary reference data, numbered k. denoted by j, the random response of the sample test object i is given by j; logit represents the logistic regression operation based on the generalized linear mixed effects model; and K is the total number of auxiliary reference data.

[0170] Based on the above approach, by introducing and using a second auxiliary model, the advantages of the generalized linear mixed-effects model can be fully utilized. By adding an additional random intercept term corresponding to the sample test objects, the differences between individuals in multiple tests of different sample test objects can be effectively considered, thereby enabling more accurate analysis and processing of data patterns involving binary classification of individual objects.

[0171] Correspondingly, by utilizing the second auxiliary results of the current round, the network structure involving binary classification processing in the target intermediate model of the previous round can be adjusted and optimized in a more targeted and effective manner.

[0172] In specific implementation, the above-mentioned use of a third auxiliary model based on Poisson regression analysis to process the current round's sample training data in the third dataset can include:

[0173] S1: Based on the sample training data of the current round, determine the time interval between two adjacent test time points; and based on the time interval, calculate the time offset term of the blood cell detection data between the two adjacent test time points, for example, it can be expressed as: log(Days).

[0174] S2: Add the time offset term to the sample training data of the current round to obtain the adjusted sample training data of the current round;

[0175] S3: Use the third auxiliary model to process the adjusted sample training data of the current round to obtain the corresponding third auxiliary result of the current round.

[0176] Specifically, for example, it can be configured according to the core formula model in the third auxiliary model:

[0177]

[0178] in, Indicates the number is The probability value of the target emotion predicted for the sample test subjects. Indicates a fixed intercept term. Indicating the influence coefficient of key characteristics of blood cells, Indicates the number is Key characteristics of blood cells in the sample test subjects. Indicates the number is The influence coefficient of auxiliary reference data, Indicates the number is The sample test object number is The data values ​​of the auxiliary reference data, Indicates the number is The time interval during sample test object detection.

[0179] Based on the above approach, by introducing and using a third auxiliary model, the advantages of Poisson regression analysis can be fully utilized. By adding an additional time offset term for blood cell detection data at two adjacent test time points, the influence of time interval on the detection data collected in two consecutive tests can be effectively considered, thereby enabling more accurate analysis and processing of data patterns based on time continuity.

[0180] Correspondingly, by utilizing the third auxiliary results of the current round, the network structure involving the processing of time continuity relationships in the target intermediate model of the previous round can be adjusted and optimized in a more targeted and effective manner.

[0181] In some embodiments, the model adjustment rules for the current round based on the prediction results of the current round, the first auxiliary results of the current round, the second auxiliary results of the current round, and the third auxiliary results of the current round are determined. In specific implementation, this may include the following:

[0182] S1: Calculate the first deviation value, second deviation value, and third deviation value between the prediction result of the current round and the first auxiliary result, the second auxiliary result, and the third auxiliary result of the current round, respectively;

[0183] S2: Based on the first deviation value, the second deviation value, and the third deviation value, determine the matching target model adjustment rule from the preset model adjustment rule library; wherein, the preset model adjustment rule library contains multiple preset model adjustment rules obtained by clustering based on historical model adjustment records, and each preset model adjustment rule corresponds to an interval combination of the first deviation value, the second deviation value, and the third deviation value;

[0184] S3: Calculate and adjust the target model adjustment rules based on the deviation between the first-level and second-level labels of the current round's prediction results and the current round's sample training data, to obtain the current round's model adjustment rules for the target intermediate model from the previous round.

[0185] Based on the above embodiments, the advantages and characteristics of different algorithm models can be fully and effectively utilized to make systematic and precise modifications and adjustments to the target intermediate model, so as to efficiently train and obtain a preset prediction model that meets the requirements.

[0186] In some embodiments, the preset prediction model may further include: a first auxiliary sub-model based on logistic regression analysis, a second auxiliary sub-model based on a generalized linear mixed-effects model, a third auxiliary sub-model based on Poisson regression analysis, and a joint prediction sub-model based on a latent class mixed-effects model connected in series with the first auxiliary sub-model, the second auxiliary sub-model, and the third auxiliary sub-model.

[0187] Accordingly, when the aforementioned determining module specifically uses a preset prediction model to determine the emotional health prediction result of the target object based on the key feature change trajectory of blood cells in the first time period of the target object, it may include: using the first auxiliary sub-model, the second auxiliary sub-model, and the third auxiliary sub-model in the preset prediction model to process the key feature change trajectory of blood cells in the first time period of the target object, respectively, to obtain the corresponding first auxiliary result, the second auxiliary result, and the third auxiliary result; splicing the first auxiliary result, the second auxiliary result, the third auxiliary result, and the key feature change trajectory of blood cells in the first time period of the target object to obtain an intermediate joint data set; and using the joint prediction sub-model to process the above intermediate joint data to obtain and output the corresponding emotional health prediction result.

[0188] In some embodiments, while processing the blood cell detection data of the target object for a first time period and determining the key characteristic change trajectory of the blood cells of the target object for the first time period, the acquisition module is also used to acquire auxiliary reference data of the target object; wherein, the auxiliary reference data may specifically include at least one of the following: age, gender, lifestyle habits, medical history, etc.

[0189] Correspondingly, the processing module can also be used to construct a joint feature data set of the target object based on the key feature change trajectory of blood cells in the first time period of the target object and auxiliary reference data;

[0190] Specifically, the determining module can also be used to determine the emotional health prediction result of the target object by processing the joint feature data group of the target object using a preset prediction model.

[0191] Based on the above embodiments, by introducing and using auxiliary reference data combined with the key feature change trajectory of blood cells of the target object in the first time period, it is possible to more accurately predict the emotional health status of the target object.

[0192] In some embodiments, the above-described emotional health data processing system may further include a prompting module;

[0193] In specific implementation, if the target object is determined to have an emotional health risk based on the emotional health prediction results, the prompting module can be used to generate emotional health risk prompt information; and based on the key characteristic change trajectory of blood cells in the target object during the first time period, determine a target emotional health adjustment plan that matches the target object's current emotional health status.

[0194] Specifically, the aforementioned emotional health risk warning information can be used to indicate whether the target subject has negative target emotions (e.g., depression, anxiety, etc.) in the future or at present.

[0195] Specifically, based on the emotional health prediction results of the target object, when it is predicted that the target object will have negative target emotions in the future, it can be determined that the target object has emotional health risks.

[0196] In practice, based on the key feature change trajectory of blood cells in the target object during a first time period, a preset adjustment scheme library can be queried to find the preset emotional health adjustment scheme with the highest matching degree between the key feature change trajectory template of blood cells and the key feature change trajectory of blood cells in the target object during the first time period. This preset emotional health adjustment scheme is then used as the matching target emotional health adjustment scheme. The preset adjustment scheme library can store multiple preset emotional health adjustment schemes, and each preset emotional health adjustment scheme corresponds to at least one key feature change trajectory template of blood cells.

[0197] Based on the above-mentioned target emotional health adjustment program, targeted and appropriate emotional adjustment interventions can be carried out on the target subjects to alleviate negative target emotions and / or increase positive target emotions (e.g., happiness).

[0198] Before implementation, a large number of historical emotional health adjustment records can be collected. These records, showing expected adjustment effects, are then selected as sample emotional health adjustment records. Based on these records, corresponding sample emotional health adjustment schemes and key characteristic change trajectories of blood cells are determined. The sample emotional health adjustment schemes are clustered to obtain multiple clusters, each containing one or more identical or similar sample emotional health adjustment schemes. Based on these clusters, multiple pre-defined emotional health adjustment schemes are determined. Furthermore, the key characteristic change trajectories of blood cells corresponding to the sample emotional health adjustment schemes within each cluster are fused and organized to determine a template for the key characteristic change trajectory of blood cells corresponding to the pre-defined emotional health adjustment scheme for that cluster. The corresponding pre-defined emotional health adjustment schemes and key characteristic change trajectory templates of blood cells are then associated and stored in a corresponding database to obtain the aforementioned pre-defined adjustment scheme library.

[0199] In practice, the prompting module can add the aforementioned target emotional health adjustment plan to the corresponding emotional health risk warning information; then send the emotional health risk warning information to the target or other relevant users so as to intelligently and accurately assist the target in making targeted emotional adjustments to ensure that the target's emotions are as healthy and stable as possible.

[0200] In some embodiments, after determining a target emotional health adjustment plan that matches the current emotional health status of the target object, the acquisition module can also be used to acquire blood cell test data of the target object for a second time period; wherein, the second time period is the time period after the target emotional health adjustment plan is started;

[0201] The processing module can also be used to determine the key feature change trajectory of blood cells in the target object during the second time period based on the blood cell detection data of the target object during the second time period.

[0202] The determining module can also be used to determine risk change data about the emotional health of the target user in the second time period based on the key feature change trajectory of blood cells in the second time period of the target object and the key feature change trajectory of blood cells in the first time period of the target object.

[0203] The prompting module can also be used to update the target emotional health adjustment plan based on the risk change data of the second time period, the key feature change trajectory of the target object's blood cells in the second time period, and the key feature change trajectory of the target object's blood cells in the first time period.

[0204] In practice, during the second phase of implementing the target emotional health adjustment plan and adjusting the target individual's emotions, blood cell test data can be collected multiple times at different time points to obtain the target individual's blood cell test data for the second phase. Then, based on the target individual's blood cell test data for the second phase, by identifying and utilizing the key characteristic changes in blood cell counts during that phase, continuous tracking and analysis of the target individual's emotional health changes based on the target emotional health adjustment plan can be conducted.

[0205] Furthermore, based on the key feature change trajectory of blood cells in the second time period of the target object and the key feature change trajectory of blood cells in the first time period of the target object, a preset prediction model can be used to determine the risk change data of the target user's emotional health in the second time period; then, based on the risk change data, the implementation effect of the target emotional health adjustment plan can be evaluated; based on the evaluation results, it can be determined whether the currently used target emotional health adjustment plan needs to be modified or updated.

[0206] Based on the assessment results, if it is determined that the target individual's emotional health status has improved and the rate of improvement is in line with expectations, the current target emotional health adjustment plan can continue to be implemented for the target individual.

[0207] Conversely, based on the assessment results, if it is determined that the target individual's emotional health status has not improved, or has improved but at a slower pace than expected, the current emotional health adjustment method can be specifically modified and adjusted based on the risk change data of the second time period, the key characteristic change trajectory of the target individual's blood cells in the second time period, and the key characteristic change trajectory of the target individual's blood cells in the first time period. This updates the target emotional health adjustment plan. Subsequently, the updated plan can be implemented for the target individual to effectively adjust their emotions and ensure that their emotions are as healthy and stable as possible.

[0208] As can be seen from the above, based on the emotional health data processing system provided in the embodiments of this specification, before specific implementation, continuous sample detection data of sample test subjects over a relatively long sample test period can be collected and utilized. By jointly using logistic regression analysis, generalized linear mixed effects model, Poisson regression analysis, and latent class mixed effects model, a preset prediction model suitable for emotional health prediction can be trained. In specific implementation, the acquisition module and processing module first acquire and determine the key feature change trajectory of the target object's blood cells in the first time period based on the blood cell detection data of the target object; then, the determination module uses the preset prediction model to determine the emotional health prediction result of the target object based on the key feature change trajectory of the target object's blood cells in the first time period. Thus, the data information carried by the target object's blood cell detection data can be effectively mined and fully utilized, and the emotional health status of the target object can be automatically determined efficiently and intelligently.

[0209] See Figure 6 As shown in the embodiments of this specification, an emotional health data processing method is also provided, which may include the following in specific implementation:

[0210] S601: Obtain blood cell test data of the target object for the first time period; wherein, the blood cell test data for the first time period includes blood cell test data at multiple time points;

[0211] S602: Process the blood cell detection data of the target object in the first time period to determine the key characteristic change trajectory of the blood cells of the target object in the first time period;

[0212] S603: Using a preset prediction model, determine the emotional health prediction result of the target object based on the key characteristic change trajectory of blood cells in the first time period.

[0213] In some embodiments, the preset prediction model is trained in the following manner: acquiring sample detection data of the sample test objects based on the sample test time period; processing the sample detection data according to preset processing rules to obtain sample training data that meets the requirements; and using the sample training data, training the preset prediction model that meets the requirements by jointly using logistic regression analysis, generalized linear mixed effects model, Poisson regression analysis and latent class mixed effects model.

[0214] In some embodiments, the step of using the sample training data to train a predetermined prediction model that meets the requirements by jointly using logistic regression analysis, a generalized linear mixed-effects model, Poisson regression analysis, and a latent class mixed-effects model may specifically include: splitting the sample training data into a first dataset, a second dataset, and a third dataset; constructing multiple initial models based on the latent class mixed-effects model according to auxiliary reference data; training the multiple initial models using the first dataset to obtain multiple intermediate models; testing the multiple intermediate models using the second dataset; and selecting a target intermediate model that meets the requirements from the multiple intermediate models based on the model test results; and using the third dataset to jointly use logistic regression analysis, a generalized linear mixed-effects model, and Poisson regression analysis to perform multiple rounds of adjustments on the target intermediate model to obtain a predetermined prediction model that meets the requirements.

[0215] In some embodiments, the above-mentioned use of a third dataset, combined with logistic regression analysis, a generalized linear mixed-effects model, and Poisson regression analysis, is used to adjust the target intermediate model in multiple rounds to obtain a preset prediction model that meets the requirements. In specific implementation, logistic regression analysis, a generalized linear mixed-effects model, and Poisson regression analysis can be combined to adjust the target intermediate model in the current round in the following manner: obtain the target intermediate model of the previous round, and use the target intermediate model of the previous round to process the sample training data of the current round in the third dataset to obtain the prediction result of the current round; based on the prediction result of the current round and the first-level and second-level labels of the sample training data of the current round, detect whether the target intermediate model of the previous round meets the requirements; If the target intermediate model from the previous round is determined to be unsuitable, the training data of the current round in the third dataset is processed using a first auxiliary model based on logistic regression analysis, a second auxiliary model based on a generalized linear mixed-effects model, and a third auxiliary model based on Poisson regression analysis, respectively, to obtain the corresponding first auxiliary result, second auxiliary result, and third auxiliary result for the current round. Based on the prediction results, first auxiliary result, second auxiliary result, and third auxiliary result for the current round, the model adjustment rules for the target intermediate model from the previous round are determined. The target intermediate model from the previous round is adjusted according to the model adjustment rules for the current round to obtain the target intermediate model for the current round.

[0216] In some embodiments, while processing the blood cell detection data of the target object for a first time period and determining the key characteristic change trajectory of the blood cells of the target object for the first time period, the method may further include: acquiring auxiliary reference data of the target object; wherein, the auxiliary reference data includes at least one of the following: age, gender, lifestyle habits, and medical history;

[0217] Accordingly, based on the key feature change trajectory of blood cells of the target object in the first time period and auxiliary reference data, a joint feature data set of the target object is constructed; by processing the joint feature data set of the target object using a preset prediction model, the emotional health prediction result of the target object is determined.

[0218] In some embodiments, the key characteristics of blood cells may specifically include: key characteristics of red blood cells, key characteristics of white blood cells, etc.

[0219] The key characteristics of red blood cells may include at least one of the following: red blood cell count, red blood cell distribution width (SD), red blood cell distribution width (CV), hematocrit, mean corpuscular hemoglobin concentration, mean corpuscular hemoglobin content, mean corpuscular volume, hemoglobin concentration, etc.

[0220] The key characteristics of the white blood cells may include at least one of the following: white blood cell count, neutrophil count, neutrophil percentage, lymphocyte count, lymphocyte percentage, eosinophil count, eosinophil percentage, basophil count, basophil percentage, monocyte count, monocyte percentage, T lymphocyte CD3 count, T lymphocyte CD3 percentage, T lymphocyte CD4 count, T lymphocyte CD4 percentage, T lymphocyte CD8 count, T lymphocyte CD8 percentage, T lymphocyte CD4 / CD8 ratio, etc.

[0221] In some embodiments, the key characteristics of the blood cells may also include: key characteristics of platelets and / or composite characteristics among multiple types of blood cells, etc.

[0222] The key characteristics of platelets may include at least: platelet count, etc.

[0223] The composite characteristics among the various types of blood cells may specifically include at least one of the following: the ratio of monocytes to lymphocytes, the ratio of neutrophils to lymphocytes, the ratio of platelets to lymphocytes, etc.

[0224] In some embodiments, the method may further include: generating emotional health risk warning information when it is determined that the target object has an emotional health risk based on the emotional health prediction result of the target object; and determining a target emotional health adjustment plan that matches the current emotional health status of the target object based on the key characteristic change trajectory of blood cells of the target object in a first time period.

[0225] In some embodiments, after determining a target emotional health adjustment plan that matches the current emotional health status of the target object, the method further includes: acquiring blood cell test data of the target object over a second time period; wherein the second time period is the period after the target emotional health adjustment plan is started.

[0226] Accordingly, based on the blood cell detection data of the target object in the second time period, the key feature change trajectory of the target object's blood cells in the second time period can be determined; based on the key feature change trajectory of the target object's blood cells in the second time period and the key feature change trajectory of the target object's blood cells in the first time period, the risk change data of the target user's emotional health in the second time period can be determined; based on the risk change data of the second time period, the key feature change trajectory of the target object's blood cells in the second time period and the key feature change trajectory of the target object's blood cells in the first time period, the target emotional health adjustment plan can be updated.

[0227] As can be seen from the above, the emotional health data processing method provided in the embodiments of this specification first obtains and determines the key characteristic change trajectory of the target object's blood cells during the first time period by acquiring and analyzing the blood cell detection data of the target object during that first time period; then, it uses a preset prediction model to determine the target object's emotional health prediction result based on the key characteristic change trajectory of the target object's blood cells during the first time period. Thus, the emotional health status of the target object can be determined efficiently and intelligently and automatically.

[0228] This specification provides an embodiment of a computer device, see below. Figure 7 As shown. The computer device includes a network communication port 701, a processor 702, and a memory 703. These structures are connected by internal cables so that they can perform specific data interaction.

[0229] Specifically, the network communication port 701 can be used to acquire blood cell detection data of the target object in a first time period; wherein the blood cell detection data in the first time period includes blood cell detection data at multiple time points.

[0230] The processor 702 can specifically be used to process the blood cell detection data of the target object in the first time period, determine the key feature change trajectory of the blood cells of the target object in the first time period, and use a preset prediction model to determine the emotional health prediction result of the target object based on the key feature change trajectory of the blood cells of the target object in the first time period.

[0231] The memory 703 can be used to store the corresponding instruction program, as well as preset prediction models and other related data.

[0232] Based on the above methods, the relevant structural performance of computer equipment can be effectively utilized to improve the data processing speed of electronic devices and efficiently realize emotional health data processing.

[0233] In this embodiment, the network communication port 701 can be a virtual port bound to different communication protocols, thereby enabling the sending or receiving of different data. For example, the network communication port can be a port responsible for web data communication, a port responsible for FTP data communication, or a port responsible for email data communication. Furthermore, the network communication port can also be a physical communication interface or communication chip. For example, it can be a wireless mobile network communication chip, such as GSM or CDMA; it can also be a Wi-Fi chip; or it can be a Bluetooth chip.

[0234] In this embodiment, the processor 702 can be implemented in any suitable manner. For example, the processor can take the form of a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers, etc. This specification is not limiting.

[0235] In this embodiment, the memory 703 may include multiple layers. In a digital system, anything that can store binary data can be a memory. In an integrated circuit, a circuit with storage function but no physical form is also called a memory, such as RAM, FIFO, etc. In a system, a storage device with a physical form is also called a memory, such as a memory stick, TF card, etc.

[0236] This specification also provides a computer-readable storage medium based on the above-described emotional health data processing method. The computer-readable storage medium stores computer program instructions that, when executed, implement the following: acquiring blood cell detection data of a target object over a first time period; wherein the blood cell detection data over the first time period includes blood cell detection data from multiple time points; processing the blood cell detection data of the target object over the first time period to determine the key characteristic change trajectory of blood cells in the target object over the first time period; and using a preset prediction model to determine the emotional health prediction result of the target object based on the key characteristic change trajectory of blood cells in the target object over the first time period.

[0237] In this embodiment, the storage medium includes, but is not limited to, Random Access Memory (RAM), Read-Only Memory (ROM), Cache, Hard Disk Drive (HDD), or Memory Card. The memory can be used to store computer program instructions. The network communication unit can be an interface configured according to standards specified in the communication protocol for network connection communication.

[0238] In this embodiment, the specific functions and effects implemented by the program instructions stored in the computer-readable storage medium can be explained in comparison with other embodiments, and will not be repeated here.

[0239] This specification also provides a computer program product, comprising at least a computer program, which, when executed by a processor, implements the following method steps: acquiring blood cell detection data of a target object over a first time period; wherein the blood cell detection data over the first time period includes blood cell detection data at multiple time points; processing the blood cell detection data of the target object over the first time period to determine the key feature change trajectory of blood cells in the target object over the first time period; and using a preset prediction model to determine the emotional health prediction result of the target object based on the key feature change trajectory of blood cells in the target object over the first time period.

[0240] It should be noted that the units, devices, or modules described in the above embodiments can be implemented by computer chips or physical entities, or by products with certain functions. For ease of description, the above devices are described by dividing them into various modules according to their functions. Of course, in implementing this specification, the functions of each module can be implemented in one or more software and / or hardware, or the module that implements the same function can be implemented by a combination of multiple sub-modules or sub-units, etc. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection between the devices or units shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0241] While this specification provides the steps of operation for the methods described in the embodiments or flowcharts, more or fewer steps may be included based on conventional or non-inventive means. The order of steps listed in the embodiments is merely one possible order of execution among many steps and does not represent the only possible order. In actual device or client product execution, the methods shown in the embodiments or drawings may be executed sequentially or in parallel (e.g., in a parallel processor or multi-threaded processing environment, or even a distributed data processing environment). The terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, product, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, product, or apparatus. Without further limitations, the presence of other identical or equivalent elements in a process, method, product, or apparatus that includes said elements is not excluded. The terms "first," "second," etc., are used to denote names and do not indicate any particular order.

[0242] Those skilled in the art will also know that, besides implementing the controller using purely computer-readable program code, the same functions can be achieved by logically programming the method steps, making the controller function as logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers (PLCs), and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the devices within it used to implement various functions can also be considered structures within that hardware component. Alternatively, the devices used to implement various functions can be considered as both software modules implementing the method and structures within a hardware component.

[0243] This specification can be described in the general context of computer-executable instructions that are executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, classes, etc., that perform a specific task or implement a specific abstract data type. This specification can also be practiced in distributed computing environments, where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer-readable storage media, including storage devices.

[0244] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this specification can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solutions of this specification can essentially be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, mobile terminal, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments of this specification.

[0245] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on its differences from other embodiments. This specification can be used in numerous general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable electronic devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices, etc.

[0246] Although this specification has been described by way of examples, those skilled in the art will recognize that many variations and modifications are possible without departing from the spirit of this specification, and it is intended that the appended claims cover such variations and modifications without departing from the spirit of this specification.

Claims

1. An emotional health data processing system, characterized in that, include: The acquisition module is used to acquire blood cell test data of the target object in a first time period; wherein, the blood cell test data in the first time period includes blood cell test data at multiple time points; The processing module is used to process the blood cell detection data of the target object in the first time period and determine the key feature change trajectory of the blood cells of the target object in the first time period. The determination module is used to determine the emotional health prediction result of the target object based on the key feature change trajectory of blood cells of the target object in the first time period using a preset prediction model. The preset prediction model is trained as follows: the sample training data is split into a first dataset, a second dataset, and a third dataset; multiple initial models based on a latent class mixed effects model are constructed according to auxiliary reference data; multiple initial models are trained using the first dataset to obtain multiple intermediate models; multiple intermediate models are tested using the second dataset; and a target intermediate model that meets the requirements is selected from the multiple intermediate models based on the model test results; using the third dataset, logistic regression analysis, a generalized linear mixed effects model, and Poisson regression analysis are used in combination to adjust the target intermediate model in multiple rounds to obtain a preset prediction model that meets the requirements.

2. The system according to claim 1, characterized in that, The sample training data was obtained in the following manner: Obtain sample detection data of the sample test objects based on the sample test time period; The sample detection data is processed according to preset processing rules to obtain sample training data that meets the requirements.

3. The system according to claim 1, characterized in that, The method of utilizing a third dataset and combining logistic regression analysis, a generalized linear mixed-effects model, and Poisson regression analysis to perform multiple rounds of adjustments on the target intermediate model to obtain a preset prediction model that meets the requirements includes: The target intermediate model for the current round is adjusted by combining logistic regression analysis, generalized linear mixed effects model, and Poisson regression analysis in the following manner: Obtain the target intermediate model from the previous round, and use the target intermediate model from the previous round to process the sample training data of the current round in the third dataset to obtain the prediction result of the current round. Based on the prediction results of the current round and the primary and secondary labels of the sample training data of the current round, check whether the target intermediate model of the previous round meets the requirements; If the target intermediate model of the previous round is determined to be unsuitable, the sample training data of the current round in the third dataset is processed by the first auxiliary model based on logistic regression analysis, the second auxiliary model based on generalized linear mixed effects model, and the third auxiliary model based on Poisson regression analysis, respectively, to obtain the corresponding first auxiliary result, second auxiliary result, and third auxiliary result of the current round. Based on the prediction results of the current round, the first auxiliary result of the current round, the second auxiliary result of the current round, and the third auxiliary result of the current round, the model adjustment rules for the current round for the target intermediate model of the previous round are determined; Based on the model adjustment rules of the current round, adjust the target intermediate model of the previous round to obtain the target intermediate model of the current round.

4. The system according to claim 1, characterized in that, While processing the blood cell detection data of the target object in the first time period and determining the key characteristic change trajectory of the blood cells of the target object in the first time period, the acquisition module is also used to acquire auxiliary reference data of the target object; wherein, the auxiliary reference data includes at least one of the following: age, gender, lifestyle habits, and medical history; The processing module is also used to construct a joint feature data set of the target object based on the key feature change trajectory of blood cells in the first time period of the target object and auxiliary reference data; The determining module is also used to determine the emotional health prediction result of the target object by processing the joint feature data group of the target object using a preset prediction model.

5. The system according to claim 1, characterized in that, The key characteristics of blood cells include: key characteristics of red blood cells and key characteristics of white blood cells; The key characteristics of the red blood cells include at least one of the following: red blood cell count, red blood cell distribution width (SD), red blood cell distribution width (CV), hematocrit, mean corpuscular hemoglobin concentration, mean corpuscular hemoglobin content, mean corpuscular volume, and hemoglobin concentration. The key characteristics of the white blood cells include at least one of the following: white blood cell count, neutrophil count, neutrophil percentage, lymphocyte count, lymphocyte percentage, eosinophil count, eosinophil percentage, basophil count, basophil percentage, monocyte count, monocyte percentage, T lymphocyte CD3 count, T lymphocyte CD3 percentage, T lymphocyte CD4 count, T lymphocyte CD4 percentage, T lymphocyte CD8 count, T lymphocyte CD8 percentage, and T lymphocyte CD4 / CD8 ratio.

6. The system according to claim 5, characterized in that, The key characteristics of blood cells also include: key characteristics of platelets and / or complex characteristics among multiple types of blood cells; The key characteristics of platelets include at least: platelet count; The composite characteristics among the various blood cell types include at least one of the following: the ratio of monocytes to lymphocytes, the ratio of neutrophils to lymphocytes, and the ratio of platelets to lymphocytes.

7. The system according to claim 1, characterized in that, It also includes a prompt module; If, based on the emotional health prediction results of the target object, it is determined that the target object has an emotional health risk, the prompting module is used to generate emotional health risk prompt information; Based on the key characteristic changes in blood cells of the target object during the first time period, a target emotional health adjustment plan that matches the target object's current emotional health status is determined. Furthermore, after determining a target emotional health adjustment plan that matches the current emotional health status of the target object, the acquisition module is also used to acquire blood cell test data of the target object for a second time period; wherein, the second time period is the period after the target emotional health adjustment plan is started. The processing module is also used to determine the key feature change trajectory of blood cells in the target object during the second time period based on the blood cell detection data of the target object during the second time period. The determining module is also used to determine risk change data about the emotional health of the target user in the second time period based on the key feature change trajectory of blood cells in the second time period of the target object and the key feature change trajectory of blood cells in the first time period of the target object; The prompting module is also used to update the target emotional health adjustment plan based on the risk change data of the second time period, the key feature change trajectory of the target object's blood cells in the second time period, and the key feature change trajectory of the target object's blood cells in the first time period.

8. A method for processing emotional health data, characterized in that, include: Obtain blood cell test data of the target object for a first time period; wherein, the blood cell test data for the first time period includes blood cell test data at multiple time points; Process the blood cell detection data of the target object in the first time period to determine the key characteristic change trajectory of the blood cells of the target object in the first time period; Using a pre-set prediction model, the emotional health prediction result of the target object is determined based on the key feature change trajectory of blood cells in the first time period. The preset prediction model is trained as follows: the sample training data is split into a first dataset, a second dataset, and a third dataset; multiple initial models based on a latent class mixed effects model are constructed according to auxiliary reference data; multiple initial models are trained using the first dataset to obtain multiple intermediate models; multiple intermediate models are tested using the second dataset; and a target intermediate model that meets the requirements is selected from the multiple intermediate models based on the model test results; using the third dataset, logistic regression analysis, a generalized linear mixed effects model, and Poisson regression analysis are used in combination to adjust the target intermediate model in multiple rounds to obtain a preset prediction model that meets the requirements.

9. A computer device, characterized in that, It includes a processor and a memory for storing processor-executable instructions, wherein the processor, when executing the instructions, implements the steps of the method of claim 8.

Images