Intelligent intervention evaluation system based on cancer patient symptom group analysis

Abstract

The invention belongs to the technical field of intelligent medical treatment, and particularly relates to an intelligent intervention evaluation system based on cancer patient symptom group analysis, which comprises a data acquisition module, a calculation analysis module, an identification module and an output module, the symptom data of the target object can be obtained, calculated, analyzed, identified and output an effective cognitive behavior intervention scheme. According to the intelligent intervention evaluation system based on the cancer patient symptom group analysis, on the basis of machine learning based on symptom data provided by the patient, an effective clinical intervention scheme can be analyzed, judged and matched to carry out xerotherapy on cognitive behaviors of the patient, so that medical resources can be effectively integrated to a certain extent, and the treatment efficiency is improved. The management efficiency and the implementation efficiency are improved, and the doctor seeing efficiency and the treatment expectation of the patient are improved.

Description

Technical Field

[0001] This application belongs to the field of intelligent medical technology. More specifically, it relates to an intelligent intervention evaluation system based on the analysis of symptom clusters of cancer patients. Background Art

[0002] Cancer, as a malignant disease that has been continuously concerned in the medical field, has been studied in different degrees and types worldwide. The cognitive-behavioral intervention program proposed based on psychological state and cognitive interference is of great significance for alleviating the psychological pain of patients, improving the quality of life of patients, and enhancing the treatment effect.

[0003] Traditional intervention evaluation methods mostly rely on the clinical work experience of doctors and nurses. A large amount of time is required in the early stage for clinical work accumulation and professional training. Moreover, there are differences in the intervention evaluation methods that may be adopted for the same symptom among different hospitals and different physicians, which reduces the utilization rate of medical resources and management efficiency to a certain extent. Summary of the Invention

[0004] The purpose of the embodiments of this application is to provide an intelligent intervention evaluation system based on the analysis of symptom clusters of cancer patients, so as to solve the technical problem of poor implementation efficiency of intervention evaluation and treatment for cancer patients in the existing technology.

[0005] To achieve the above purpose, the technical solution adopted in this application is: to provide an intelligent intervention evaluation system based on the analysis of symptom clusters of cancer patients, including:

[0006] A data acquisition module, configured to obtain symptom data of a target object;

[0007] A calculation and analysis module, configured to analyze the symptom data and obtain a first evaluation result;

[0008] An identification module, configured to identify the number of first symptom cluster categories in the first evaluation result and determine the intervention type of the target object;

[0009] An output module, configured to output the intervention type.

[0010] According to some embodiments of this application, the data acquisition module is further configured to obtain a second evaluation result of the target object, and the second evaluation result is the second symptom cluster category corresponding to the historical symptoms of the target object.

[0011] According to some embodiments of this application, the step of identifying the number of first symptom cluster categories in the first evaluation result and determining the intervention type of the target object includes:

[0012] When the number of first symptom cluster categories is 0, the identification result is to continue observation;

[0013] When the number of the first symptom group categories is 1, the recognition result is to output the first symptom group category and its corresponding intervention type;

[0014] When the number of the first symptom group categories is greater than 1, compare the first evaluation result with all the symptom thresholds of all the second evaluation results, and the recognition result is to output the two symptom group categories with the highest symptom thresholds and their corresponding intervention types.

[0015] According to some embodiments of the present application, the symptom group categories involved in the first symptom group category and the second symptom group category at least include a nerve-related symptom group, a cachexia-related symptom group, a respiratory disorder-related symptom group, and a psychology-related symptom group.

[0016] According to some embodiments of the present application, the symptoms of the nerve-related symptom group at least include pain, amnesia, restless sleep, numbness, and blurred vision; the symptoms of the cachexia-related symptom group at least include anorexia, nausea, and vomiting; the symptoms of the respiratory disorder-related symptom group at least include cough, dyspnea, and fatigue; the symptoms of the psychology-related symptom group at least include psychological depression and low mood.

[0017] According to some embodiments of the present application, each of the symptoms is numerically collected and normalized using severity level I i for calculation, and its normalization calculation formula is:

[0018]

[0019] In the formula, x i is the symptom severity, specifically an evaluation value between 1 and 10;

[0020] μ i is the symptom historical mean, and σ i is the symptom historical standard deviation.

[0021] According to some embodiments of the present application, the symptom threshold is the ratio of the number of symptoms with severity level I i exceeding N (N is greater than 0, and the N values for different symptom group categories are set differently) to the total number of all symptoms in the symptom group.

[0022] According to some embodiments of the present application, before obtaining the symptom data of the target object, it further includes constructing an intelligent intervention evaluation model and performing model training.

[0023] According to some embodiments of the present application, the constructing of the intelligent intervention evaluation model and performing model training includes:

[0024] Collecting open-source symptom data and their corresponding treatment plans;

[0025] Open-source symptom data cleaning and construction of different categories of symptom groups based on the cleaned symptom text data;

[0026] Establish a response strategy between symptom groups and intervention types and conduct model training;

[0027] Make real-time treatment decisions and conduct verification and optimization.

[0028] According to some embodiments of the present application, the open-source symptom data cleaning includes:

[0029] Perform text conversion on the open-source symptom data and collect the obtained symptom text;

[0030] Preprocess the symptom text and add tags;

[0031] Utilize the forward propagation of the transformer attention mechanism and calculate the loss;

[0032] Take the symptom group with the highest probability as the classification label.

[0033] The beneficial effects of the intelligent intervention evaluation system based on the analysis of cancer patient symptom groups provided by the present application are as follows:

[0034] Compared with the prior art, the intelligent intervention evaluation system based on the analysis of cancer patient symptom groups provided by the present application can provide diagnosis and treatment services for patients in an intelligent manner. Specifically, based on machine learning of the symptom data provided by the patient, it can analyze, identify and match effective clinical intervention plans to intervene and treat the patient's cognitive behavior. Therefore, to a certain extent, it can effectively integrate medical resources, improve management efficiency and implementation efficiency, and is conducive to improving the patient's medical treatment efficiency and treatment expectations. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0036] Figure 1 It is a schematic diagram of the working logic of the intelligent intervention evaluation system based on the analysis of cancer patient symptom groups provided by the embodiments of the present application;

[0037] Figure 2 It is a schematic diagram of the construction steps of the intelligent intervention evaluation model provided by the embodiments of the present application;

[0038] Figure 3Schematic diagram of the steps for data cleaning provided by the embodiments of the present application;

[0039] Figure 4 Schematic diagram of the structural composition of the intelligent intervention evaluation system based on the analysis of cancer patient symptom clusters provided by the embodiments of the present application;

[0040] Figure 5 Schematic diagram of the environmental structure for the operation of the intelligent intervention evaluation program based on the analysis of cancer patient symptom clusters provided by the embodiments of the present application;

[0041] Figure 6 Logic diagram of the preliminary work for the collection of the symptom text of the target object. Detailed implementation manners

[0042] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present application more clear and understandable, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0043] It should be noted that when an element is referred to as being "fixed to" or "disposed on" another element, it can be directly on the other element or indirectly on the other element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or indirectly connected to the other element.

[0044] It should be understood that the orientation or positional relationship indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation to the present application.

[0045] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present application, "a plurality of" means two or more unless otherwise specifically defined.

[0046] Based on the background art, most of the existing cognitive-behavioral intervention programs for cancer patients are implemented manually, resulting in problems such as poor management efficiency and low utilization rate of medical resources. Therefore, this application proposes an intelligent intervention evaluation system based on the analysis of cancer patient symptom clusters, which provides intelligent symptom cluster analysis and guidance and assistance for the implementation of intervention types in the form of a computer or network, reconstructs the cognitive behavior of patients from the perspective of improving their cognitive behavior, can improve the quality of life and treatment expectations of patients during treatment by enhancing their treatment emotions, and is conducive to significantly reducing the evaluation time of intervention types by physicians and improving the utilization rate of medical resources and management efficiency.

[0047] In one embodiment of the present application, please refer to Figure 5 , the intelligent intervention evaluation system based on the analysis of cancer patient symptom clusters can be implemented as a mobile terminal device such as a PC, a smart phone, a tablet computer, or a portable computer. The intelligent intervention evaluation system may include: a processor 1001 (such as a CPU), a network interface 1004, a user interface 1003, a memory 1005, and a communication bus 1002. Among them, the communication bus 1002 is used to realize the connection and communication between these components; the user interface 1003 may include a display screen, an input unit such as a keyboard, a remote control, and an optional standard wired interface, a wireless interface, etc.; the network interface 1004 may be an optional standard finite interface, a wireless interface (such as a WI-FI interface); the memory 1005 may be an optional high-speed RAM memory, a disk memory, or other storage devices, and an intelligent intervention evaluation program based on the analysis of cancer patient symptom clusters is stored therein.

[0048] On this basis, the above-mentioned intelligent intervention evaluation system based on the analysis of cancer patient symptom clusters may further include a radio frequency circuit, an audio circuit, a wifi module, and so on. It can be understood that Figure 5 the intelligent intervention evaluation system shown in

[0049] Please refer to Figure 1 , in a further embodiment of the present application, the operation of the above-mentioned intelligent intervention evaluation program based on the analysis of cancer patient symptom clusters involves the following implementation steps or implementation methods:

[0050] Step S100: Obtain the symptom data of the target object;

[0051] Step S200: Analyze the symptom data and obtain a first evaluation result;

[0052] Step S300: Identify the number of first symptom group categories in the first evaluation result and determine the intervention type of the target object;

[0053] Step S400: Output the intervention type.

[0054] Obviously, the intelligent intervention evaluation system based on the analysis of cancer patient symptom clusters provided by the embodiments of the present application, on the basis of fully collecting symptom data and performing analysis, identification, and evaluation, matches effective intervention treatment plans, can be effectively implemented using mobile terminals such as computers and mobile phones, effectively reduces the offline treatment time of physicians, and provides patients with an efficient cognitive behavioral intervention treatment method for remote (such as home diagnosis and treatment), thereby achieving the purpose of fully improving their quality of life and treatment expectations.

[0055] Please further refer to Figure 6 , in a further embodiment of the present application, before obtaining the symptom data of the target object, it further includes:

[0056] Step S1001: Actively collect based on the symptom text provided by the target object;

[0057] Step S1002: Analyze the symptom data in the symptom text and preliminarily determine whether it belongs to a certain symptom cluster. If it belongs, access the intelligent intervention evaluation system for evaluation; if it does not belong, further perform manual correction through online consultation with experts or physicians, and access the intelligent intervention evaluation system for evaluation with the corrected symptom text.

[0058] Thus, the evaluation effectiveness and accuracy of the intelligent intervention evaluation system can be fully improved.

[0059] It can be understood that a certain symptom cluster described above reflects that the intelligent intervention evaluation system provided by the embodiments of the present application focuses on the normalization processing of symptom data, which can be further set to belong to multiple symptom cluster categories, or perform multi-category symptom cluster determination identification, and then access the intelligent intervention evaluation system for evaluation. On this premise, Step S1001 and Step S1002 can be set as a correction link before the intelligent intervention evaluation program, or a functional module or device external to the intelligent intervention evaluation program. In actual implementation, no matter which form is adopted, it should be regarded as an equivalent implementation manner of the embodiments of the present application when the principles or steps are the same.

[0060] Please refer to Figure 1 and Figure 4 , in another embodiment of the present application, there is also provided an intelligent intervention evaluation system based on the analysis of cancer patient symptom clusters, and the evaluation system includes:

[0061] A data collection module 10, configured to obtain the symptom data of the target object;

[0062] The calculation and analysis module 20 is used to analyze the symptom data and obtain the first evaluation result;

[0063] The identification module 30 is used to identify the number of the first symptom group categories in the first evaluation result and determine the intervention type of the target object;

[0064] The output module 40 is used to output the intervention type.

[0065] According to the intelligent intervention evaluation system based on the analysis of cancer patients' symptom clusters provided by the embodiments of the present application, on the basis of collecting the symptom data of the target object, through the data collection module 10, the calculation and analysis module 20, the identification module 30 and the output module 40, the collection, analysis, identification and classification output of the symptom data can be realized in sequence, and then an effective clinical intervention plan can be analyzed and identified to intervene and treat the cognitive behavior of the patient. Therefore, the intelligent intervention evaluation system based on the analysis of cancer patients' symptom clusters provided by the embodiments of the present application can effectively integrate medical resources to a certain extent, improve the management efficiency and implementation efficiency, and is beneficial to improving the patient's medical treatment efficiency and treatment expectation.

[0066] Please refer to again Figure 1 In a further embodiment of the present application, the data collection module 10 is further used to obtain the second evaluation result of the target object, and the second evaluation result is the second symptom group category corresponding to the historical symptoms of the target object. It can be understood that in the case of first applying the intelligent intervention evaluation system to the patient, the second evaluation result can use the data of multiple patients with the same cancer as a reference. Thus, the evaluation accuracy of the entire evaluation system can be improved by making great use of the prior symptom group classification knowledge.

[0067] Please refer to further Figure 1 In a further embodiment of the present application, the method for identifying the number of the first symptom group categories in the first evaluation result and determining the intervention type of the target object includes:

[0068] When the number of the first symptom group categories is 0, the identification result is to continue to observe;

[0069] When the number of the first symptom group categories is 1, the identification result is to output the first symptom group category and its corresponding intervention type;

[0070] When the number of the first symptom group categories is greater than 1, compare the first evaluation result with all the symptom thresholds of all the second evaluation results, and the identification result is to output the two symptom group categories with the highest symptom thresholds and their corresponding intervention types.

[0071] Obviously, when the number of the first symptom group categories is greater than 1, the intervention types corresponding to the multiple existing symptom group categories may have similarities or commonalities, such as cognitive teaching of pathological factors, psychological counseling, etc. Therefore, when determining the intervention type, the common intervention part of the intervention type can be considered for pre-implementation or synchronous implementation to quickly relieve the psychological and emotional distress of the patient.

[0072] In a further embodiment of the present application, the symptom group categories involved in the above-mentioned first symptom group category and second symptom group category at least include a nerve-related symptom group, a cachexia-related symptom group, a respiratory disorder-related symptom group, and a psychological-related symptom group.

[0073] Obviously, the symptom group categories here are only exemplary descriptions based on existing clinical experience, and do not limit that the first symptom group category and the second symptom group category only include or must be the symptom group categories here. Based on subsequent optimization and differential implementation, the number of symptom group categories can be increased or decreased, and their names, types, etc. can also be adjusted. No matter what name and form are used for definition, in the case of consistent implementation steps, it should be regarded as equivalent to the implementation method of the embodiment of the present application.

[0074] In a further embodiment of the present application, the symptoms of the above-mentioned nerve-related symptom group at least include pain, amnesia, restless sleep, numbness, and blurred vision; the symptoms of the above-mentioned cachexia-related symptom group at least include anorexia, nausea, and vomiting; the symptoms of the above-mentioned respiratory disorder-related symptom group at least include cough, dyspnea, and fatigue; the symptoms of the above-mentioned psychological-related symptom group at least include psychological depression and low mood.

[0075] Similarly, the specific symptom types and quantities involved in different symptom group categories here can also be increased or decreased during actual optimization and differential implementation, and the symptom names, types, etc. can be adjusted according to requirements. No matter what name and form are used for definition, in the case of consistent implementation steps, it should be regarded as equivalent to the implementation method of the embodiment of the present application.

[0076] On the basis of the above embodiment, an exemplary description of the processing of symptom data is further given here. In a further embodiment of the present application, after each symptom data is collected, it is necessary to use severity I i for numerical acquisition and normalization calculation, and its normalization calculation formula is:

[0077]

[0078] In the formula, x i is the symptom severity, specifically an evaluation value between 1 and 10;

[0079] μ i is the symptom historical mean, and σi is the standard deviation of the symptom history.

[0080] The σ in this calculation formula i is regarded as the variance factor, which can effectively eliminate outliers and merge similar symptoms.

[0081] In this way, the symptom data can be refined into symptom features in a digital form, and the normalization calculation is mainly used to normalize and reduce the noise of the symptom severity, so as to improve the prediction accuracy of the intervention type corresponding to the symptom features.

[0082] In a further embodiment of the present application, the symptom threshold is severity I i The ratio of the number of symptoms exceeding N (N>0, and the N values for different symptom group categories are set differently) to the total number of symptoms in the symptom group. In this way, the degree of attention or importance of different symptom group categories can be differentiated to further improve the accuracy of the system prediction and evaluation.

[0083] Please refer further to Figure 2 , the intelligent intervention evaluation system based on the analysis of cancer patient symptom groups involved in all embodiments of the present application needs to rely on a prediction model based on artificial intelligence or machine learning for implementation. Specifically, the prediction model adopted in the present application is based on multivariate Logistic regression analysis, and uses a random forest model to predict the preferred intervention type corresponding to a specific symptom group (the intervention type in the present application is equivalent to the intervention treatment plan), and it can use a neural network to mine the non-linear relationship between the symptom group formed by multiple symptom combinations and the intervention type. The input variables of the prediction model are mainly symptom group characteristics, and tumor stage characteristics (different time sequence elements of early, middle and late stages of cancer) and / or molecular typing can also be added during optimization or differential implementation; while the output variables are mainly intervention types, and the remission efficiency, survival period extension rate, and quality of life improvement rate can also be added on the basis of optimization or differential implementation to further improve the comprehensiveness and comprehensiveness of the prediction model.

[0084] Specifically, in some embodiments of the present application, before putting the intelligent intervention evaluation system based on the analysis of cancer patient symptom groups into use, model construction and model training are also required. In other words, before obtaining the symptom data of the target object mentioned above, it also includes constructing an intelligent intervention evaluation model and performing model training.

[0085] Please refer further to Figure 2 , in a further embodiment of the present application, constructing an intelligent intervention evaluation model and performing model training includes:

[0086] Collect open-source symptom data and their corresponding treatment plans;

[0087] Open-source symptom data cleaning and construction of different categories of symptom groups based on the cleaned symptom text data;

[0088] Establish a response strategy between symptom groups and intervention types and conduct model training;

[0089] Make real-time treatment decisions and conduct verification and optimization.

[0090] Regarding the model construction and its training provided by the embodiments of the present application, when collecting open-source symptom data, clustering analysis (such as K-means, hierarchical clustering) can be used to classify the natural symptoms therein to extract the internal associations existing between symptoms. In this context, the present application considers using the Aprior algorithm to identify the high-frequency symptom combinations therein to form multiple preliminary symptom group categories.

[0091] On this basis, further verification is carried out through clinical expert review and biological verification (such as case-control and the generation or annihilation of molecular markers involved in the comparison process) (for example, the inflammatory factor IL-6 caused by pain has a high correlation with the fatigue symptom in cancer symptoms). In this way, the evaluation accuracy of the open-source symptom data and its corresponding treatment plan can be fully guaranteed.

[0092] Furthermore, when cleaning and constructing different categories of symptom groups based on open-source symptom data, the model selection is not fixed. For example, based on multivariate Logistic regression analysis, after initially evaluating the impact of the intervention type on the remission rate of symptom groups, the random forest model is further used as the main machine learning model to establish a response strategy between symptom groups and intervention types and conduct model training. Then, in the context of improving the prediction accuracy, real-time treatment decisions are combined and verified and optimized.

[0093] Furthermore, considering that there are various optimization methods here, tumor stage features (different time sequence elements in the early, middle, and late stages of cancer) and / or molecular typing can be added at the input end as mentioned above, and the remission efficiency, survival period extension rate, and quality of life improvement rate can be added at the output end for optimization. For example, a two-arm controlled study (comparison between the model-guided group and the conventional treatment group) can be carried out. An ethical consideration module can also be added. For example, an informed consent mechanism for patients can be established to explain the auxiliary decision-making nature of this prediction model, or an expert supervision committee can be set up to strengthen the artificial review mechanism when the intervention types obtained by prediction conflict with the actual work guidelines too much.

[0094] Under this premise, it is also possible to consider further constructing a national or even global symptom group database to improve the effectiveness of the foundation for constructing the prediction model with a large multi-disciplinary and multi-field database.

[0095] In a further embodiment of the present application, the preliminary evaluation reference formula for the intervention type based on multivariate Logistic regression analysis is:

[0096]

[0097] In the formula, β0 is the response probability in the absence of any symptoms; β i is the regression coefficient of the symptom corresponding to the intervention type, and x i is the aforementioned symptom severity.

[0098] When establishing a prediction model for symptom clusters and intervention types with a random forest model as the main machine learning model, the design of its hidden layer can refer to the empirical formula:

[0099]

[0100] to fully balance the expression ability and overfitting risk of the entire prediction model. For example, when the importance of a certain symptom in the prediction of the intervention type decision is 1.9 times that of another symptom, its priority is increased to facilitate improving the efficacy of the intervention type.

[0101] Please further refer to Figure 3 , in a further embodiment of the present application, the open-source symptom data cleaning includes:

[0102] Performing text conversion on the open-source symptom data and collecting the obtained symptom text;

[0103] Preprocessing the symptom text and adding tags;

[0104] Using the forward propagation of the transformer attention mechanism and performing loss calculation;

[0105] Taking the symptom cluster with the highest probability as the classification label.

[0106] It can be understood that when performing cluster analysis on the open-source symptom data collection here, K-means can be used for cluster analysis. Specifically, it is to divide based on the Euclidean distance of the refined matrix of symptom severity, thereby performing natural exploration and analysis of symptom cluster classification without prior knowledge.

[0107] On this premise, the clinical work guidelines (such as chemotherapy toxicity profiles) can be further combined to achieve symptom cluster marking between a small number of expert annotations and a large number of unannotated numbers, so as to fully improve the accuracy and effectiveness of data preprocessing.

[0108] On this basis, the forward propagation of the Transformer attention mechanism and the process of loss calculation involve cross-entropy loss calculation and weighted calculation when the symptom group categories are imbalanced. To optimize and improve the processing efficiency, the number of layers and heads can also be adjusted to fully adapt to long symptom texts, minimizing the cross-entropy loss and encouraging the model to distinguish key symptom words (such as the strong signal of fatigue for the symptom group related to respiratory disorders).

[0109] Obviously, the open-source symptom data cleaning method provided by the embodiments of the present application is beneficial for the prediction model to efficiently capture symptom semantics, and can achieve an end-to-end classification mode by combining supervision and fine-tuning, which is fully applicable to the complex, multiple, and fine-grained medical text analysis requirements of cancer symptoms.

[0110] In summary of the specific content of the above embodiments, the intelligent intervention evaluation system provided by the present application based on the analysis of cancer patient symptom groups can maximize the value of symptom data on the premise of effectively ensuring clinical rationality, laying a solid foundation for subsequent symptom group analysis and research (for example, on this basis, it can be further combined with visualization tools and real-time data update to optimize the system).

[0111] The above are only the preferred embodiments of the present application, and are not intended to limit the present application. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. An intelligent intervention evaluation system based on symptom group analysis of cancer patients, characterized in that: include: A data collection module (10) is used to obtain symptom data of a target subject; A calculation and analysis module (20), used for analyzing the symptom data and obtaining a first evaluation result; An identification module (30) is used to identify the number of first symptom group categories in the first assessment result and determine the intervention type of the target object; An output module (40) is used to output the intervention type.

2. The intelligent intervention evaluation system based on symptom group analysis of cancer patients according to claim 1, characterized in that: The data collection module (10) is also used to obtain a second evaluation result of the target object, where the second evaluation result is a second symptom group category corresponding to the historical symptoms of the target object.

3. The intelligent intervention evaluation system based on symptom group analysis of cancer patients according to claim 2, characterized in that: The method for determining the number of the first symptom group categories in the first assessment result and determining the intervention type of the target object includes: When the number of the first symptom group categories is 0, the judgment result is to continue observation; When the number of the first symptom group categories is 1, the identification result is outputting the first symptom group category and its corresponding intervention type; When the number of the first symptom group categories is greater than 1, the first assessment result is compared with all symptom thresholds of all the second assessment results, and the identification result is to output the two symptom group categories with the highest symptom thresholds and their corresponding intervention types.

4. The intelligent intervention evaluation system based on symptom group analysis of cancer patients according to claim 3, characterized in that: The symptom group categories involved in the first symptom group category and the second symptom group category include at least a neurological-related symptom group, a cachexia-related symptom group, a respiratory disorder-related symptom group and a psychological-related symptom group.

5. The intelligent intervention evaluation system based on symptom group analysis of cancer patients according to claim 4, characterized in that: The symptoms of the neurological-related symptom group include at least pain, forgetfulness, disturbed sleep, numbness and blurred vision; the symptoms of the cachexia-related symptom group include at least anorexia, nausea and vomiting; the symptoms of the respiratory disorder-related symptom group include at least cough, dyspnea and fatigue; the symptoms of the psychological-related symptom group include at least psychological depression and low mood.

6. The intelligent intervention evaluation system based on symptom group analysis of cancer patients according to claim 5, characterized in that: Each symptom is rated on a severity scale of I i Perform numerical collection and normalization calculation, and the normalization calculation formula is: In the formula, x i The severity of the symptoms is a numerical evaluation between 1 and 10; μ i is the historical mean of symptoms, σ i is the standard deviation of symptom history.

7. The intelligent intervention evaluation system based on symptom group analysis of cancer patients according to claim 6, characterized in that: The symptom threshold is the severity level I i The ratio of the number of symptoms exceeding N (N is greater than 0, and the N value setting for different symptom group categories) to the total number of symptoms in the symptom group.

8. The intelligent intervention evaluation system based on symptom group analysis of cancer patients according to any one of claims 1 to 7, characterized in that: Before obtaining the symptom data of the target object, it also includes building an intelligent intervention assessment model and performing model training.

9. The intelligent intervention evaluation system based on symptom group analysis of cancer patients according to claim 8, characterized in that: The constructing of the intelligent intervention assessment model and performing model training include: Collect open source symptom data and their corresponding treatment plans; Clean the open source symptom data and build different categories of symptom groups based on the cleaned symptom text data; Establish response strategies for symptom clusters and intervention types and conduct model training; Real-time treatment decision making and validation optimization.

10. The intelligent intervention evaluation system based on symptom group analysis of cancer patients according to claim 9, characterized in that: The open source symptom data cleaning includes: Performing text conversion on the open source symptom data and collecting the obtained symptom text; Preprocessing and marking the symptom text; Use the transformer attention mechanism to propagate forward and calculate the loss; Take the symptom group with the highest probability as the classification label.

Images