Intelligent evaluation and intervention system and method for psychological state of postoperative patient in hepatobiliary pancreatic surgery

Abstract

The invention relates to the technical field of medical information, in particular to an intelligent assessment and intervention system and method for psychological states of patients after hepatobiliary and pancreatic surgery. According to the method, technologies such as a BP neural network, speech recognition, natural language processing and an improved LSTM neural network are fused, and a psychological state evaluation model is constructed; the system comprises a psychological test module, a music intervention module and an evaluation module, and realizes data intercommunication of evaluation results. The psychological test module selects a proper scale for emotion evaluation according to an evaluation result; the music intervention module recommends playing music according to the evaluation result; the evaluation module analyzes the voice communication text, judges psychological problems by using a support vector machine algorithm, and constructs a postoperative health state data model; according to the system, through innovative technical means and overall design, the comprehensiveness of evaluation, the individuation of intervention and the continuity of service are improved, the defects of a traditional method are effectively overcome, and more accurate and comprehensive postoperative psychological evaluation and intervention service is provided for patients.

Description

Technical Field

[0001] The present invention relates to the field of medical information technology, and particularly to an intelligent evaluation and intervention system and method for the psychological state of patients after hepatobiliary and pancreatic surgery. Background Art

[0002] In recent years, with the progress of medical technology, the success rate of hepatobiliary and pancreatic surgery has increased significantly. However, the complexity and trauma of such surgeries still bring huge physical and mental pressures to patients. Postoperative patients generally face psychological problems such as anxiety, depression, and sleep disorders, which not only affect the quality of life of patients, but may also delay the recovery process and increase the risk of complications. Therefore, it becomes particularly important to conduct timely and effective psychological evaluation and intervention for patients after hepatobiliary and pancreatic surgery.

[0003] Traditional postoperative psychological evaluation mainly relies on regular manual questionnaires and face-to-face psychological counseling. Although this method can understand the psychological state of patients to a certain extent, it has many limitations. First, the evaluation frequency is limited, and it is often difficult to capture the dynamic changes of patients' psychological states in a timely manner. Second, standardized questionnaires are difficult to comprehensively reflect individual differences, and evaluation biases are likely to occur. Third, face-to-face counseling is limited by time and human resources, and it is difficult to provide continuous psychological support for patients.

[0004] In terms of intervention, common methods include psychological counseling, music therapy, etc. However, these methods often adopt unified standards and lack personalized customization. For example, music therapy usually uses a preset list of tracks without considering the patient's current emotional state and personal preferences. In addition, the evaluation of the effects of traditional intervention methods also lacks objective and quantitative criteria, making it difficult to make precise adjustments and optimizations.

[0005] With the development of artificial intelligence technology, some researchers have tried to apply machine learning algorithms to the field of mental health. However, most of these attempts are limited to single-dimensional data analysis, such as only analyzing voice or text data, and it is difficult to comprehensively grasp the psychological state of patients. At the same time, most existing intelligent systems stay at the evaluation stage and lack effective intervention measures and closed-loop management mechanisms.

[0006] In view of the deficiencies of the existing technology, there is an urgent need for an intelligent system that can achieve comprehensive evaluation, personalized intervention, real-time monitoring, and continuous optimization to meet the mental health needs of patients after hepatobiliary and pancreatic surgery. Summary of the Invention

[0007] The intelligent evaluation and intervention system and method for the psychological state of patients after hepatobiliary and pancreatic surgery of the present invention are precisely designed for the above problems. The system aims to achieve all-round and real-time evaluation and precise intervention of the psychological state of patients through multi-modal data fusion, deep learning algorithms, and personalized intervention strategies.

[0008] Specifically, the system of the present invention includes a mental state assessment model module, a psychological test module, a music intervention module, an evaluation module, and a remote psychological consultation module. This modular design enables the system to collect and analyze patient data from multiple dimensions, including clinical indicators, voice features, text content, body movements, etc., so as to construct a comprehensive mental state assessment model. At the same time, the system also integrates personalized music intervention and remote consultation functions to provide timely and targeted psychological support for patients.

[0009] The present invention proposes an intelligent assessment and intervention system for the mental state of patients after hepatobiliary and pancreatic surgery, including:

[0010] A mental state assessment model module, used for:

[0011] Based on pre-input clinical data, psychological test scale results, and emotion recognition data, train a BP neural network to construct a mental state assessment model;

[0012] Use speech recognition technology to convert the patient's speech into text, perform natural language processing, and extract words, phrases, and named entities in the text;

[0013] Adopt an improved LSTM neural network to perform emotion recognition and classification on the processed text;

[0014] A psychological test module, data-connected to the mental state assessment model module, used for:

[0015] Receive the evaluation results sent by the mental state assessment model module;

[0016] Based on the evaluation results, select an appropriate psychological test scale to evaluate the patient's emotions;

[0017] A music intervention module, data-connected to the mental state assessment model module and the psychological test module, used for:

[0018] Receive the evaluation results sent by the mental state assessment model module and the psychological test module;

[0019] According to the evaluation results, recommend and play different types of music to the patient;

[0020] An evaluation module, data-connected to the mental state assessment model module, the psychological test module, and the music intervention module, used for:

[0021] Receive the data sent by the mental state assessment model module, the psychological test module, and the music intervention module;

[0022] Analyze the text of the patient's voice communication with the nurse to identify the patient's voice features;

[0023] Use the support vector machine algorithm to classify speech features and determine whether the patient has psychological problems;

[0024] Construct a postoperative health status data model to evaluate the patient's postoperative health status.

[0025] Preferably, it further includes a remote psychological consultation module, which is data-connected to the evaluation module and is used for:

[0026] Receive the patient status evaluation information sent by the evaluation module;

[0027] Extract the text information of the conversation between the patient and the doctor through speech recognition and natural language processing technologies;

[0028] According to the extracted text information, perform problem keyword extraction and medical literature retrieval;

[0029] Provide the remote doctor with the patient status information and relevant medical suggestions to achieve online consultation and diagnosis.

[0030] Preferably, the psychological state evaluation model module includes:

[0031] A psychological state evaluation model training unit, which is used for:

[0032] Take the pre-input data as the training set, validation set, and test set with probabilities of 50%, 70%, and 90% respectively;

[0033] Preprocess the training set and the test set;

[0034] Use the preprocessed data to train a BP neural network and verify the evaluation accuracy;

[0035] A speech text processing unit, which is used for:

[0036] Complete Chinese text annotation using the CRF++-1.3.6 version of Python;

[0037] Use the conditional random field model to extract effective features from the text corpus and train the model;

[0038] Associate the word library with the part of speech and emotional tendency;

[0039] An emotion analysis unit, which is used for:

[0040] Calculate the emotion score of each word in the corpus;

[0041] Based on the BP neural network, use the emotion score of the corpus as the target vector to construct an emotion analysis model;

[0042] An improved LSTM model unit, which is used for:

[0043] Based on the text emotion score, train the text using an improved LSTM model;

[0044] Utilize deep learning technology for final emotion recognition and classification.

[0045] Preferably, the psychological test module includes:

[0046] Scale selection unit, for:

[0047] According to the psychological state assessment result, select an appropriate scale from a preset psychological test scale library;

[0048] The psychological test scale library includes the HAMA anxiety scale, Hamilton depression scale, self-rating depression scale, self-rating anxiety scale, symptom checklist, etc.;

[0049] Test execution unit, for:

[0050] Present the selected psychological test scale to the patient;

[0051] Collect the patient's answer results;

[0052] Result analysis unit, for:

[0053] Calculate the scores of each scale according to the patient's answer results;

[0054] Generate a psychological test report.

[0055] Preferably, the music intervention module includes:

[0056] Music recommendation unit, for:

[0057] Based on the psychological state assessment result and psychological test result, select music suitable for the patient's current emotional state from a preset music library;

[0058] Music playback unit, for:

[0059] Play the recommended music through the system interface or an external device connected thereto;

[0060] Effect feedback unit, for:

[0061] Record the patient's reaction during the music listening process;

[0062] Collect the patient's subjective evaluation of the music intervention effect.

[0063] Preferably, the evaluation module includes:

[0064] Voice analysis unit, for:

[0065] Analyze the formants, MFCCs, gain adjustments and other features of the patient's speech using acoustic analysis techniques; classify the extracted speech features using the support vector machine algorithm;

[0066] A health status modeling unit for:

[0067] Collect the patient's postoperative rehabilitation data;

[0068] Construct a postoperative health status data model;

[0069] A comprehensive evaluation unit for:

[0070] Combine the speech analysis results, health status model and psychological state evaluation results;

[0071] Generate a comprehensive evaluation report for the patient.

[0072] Preferably, the remote psychological consultation module includes:

[0073] A dialogue processing unit for:

[0074] Convert the conversation between the patient and the doctor into text through speech recognition technology;

[0075] Perform preprocessing such as word segmentation and dependency syntactic analysis on the text;

[0076] A problem extraction unit for:

[0077] Extract keywords from the preprocessed text;

[0078] Conduct medical literature retrieval based on the extracted keywords;

[0079] A consultation support unit for:

[0080] Provide the evaluation results of the patient and relevant medical literature to the remote doctor;

[0081] Support the doctor to conduct online diagnosis and treatment suggestions.

[0082] Preferably, it further includes:

[0083] A knowledge base module, which is data-connected to the remote psychological consultation module, for:

[0084] Store articles, books, and image materials related to the psychological state after hepatobiliary and pancreatic surgery;

[0085] Provide knowledge support for the remote psychological consultation module;

[0086] The psychological state evaluation model module further includes:

[0087] A body movement analysis unit for:

[0088] Use the VOT algorithm to perform motion estimation and tracking on the patient's limb movements;

[0089] Extract the motion features of limb movements, such as joint angular displacement;

[0090] Based on the extracted motion features, construct a limb behavior feature model.

[0091] Preferably, the evaluation module further includes:

[0092] An intervention effect evaluation unit for:

[0093] Use an improved HCAHPS scoring system to evaluate the patient's intervention effect;

[0094] Generate an intervention effect evaluation report;

[0095] According to the evaluation results, provide feedback to the music intervention module and the remote psychological consultation module to adjust the intervention strategy.

[0096] An intelligent evaluation and intervention method for the psychological state of patients after hepatobiliary and pancreatic surgery, including the following steps:

[0097] S1. Obtain the patient's clinical data, psychological test results, and emotion recognition data, input them into the psychological state evaluation model for evaluation, and obtain the patient's psychological state evaluation result;

[0098] S2. Based on the psychological state evaluation result, select an appropriate psychological test scale to conduct further psychological tests on the patient;

[0099] S3. According to the psychological state evaluation result and the psychological test result, match the corresponding music intervention plan, recommend and play music to the patient;

[0100] S4. Use the support vector machine algorithm to analyze the patient's speech features to determine whether the patient has psychological problems;

[0101] S5. Construct a postoperative health status data model of the patient to evaluate the patient's postoperative health status;

[0102] S6. Based on speech recognition and natural language processing technologies, extract the text information of the patient's remote conversation with the doctor;

[0103] S7. Perform keyword extraction and medical literature retrieval on the extracted text information;

[0104] S8. Provide the patient's status information and relevant medical advice to the remote doctor to support online consultation and diagnosis;

[0105] S9. Use the VOT algorithm to analyze the patient's limb movements, extract motion features, and construct a limb behavior feature model;

[0106] S10. Use the improved HCAHPS scoring system to evaluate the intervention effect and adjust the intervention strategy according to the evaluation results.

[0107] The beneficial effects of the present invention are mainly reflected in the following aspects:

[0108] First of all, in terms of the comprehensiveness and accuracy of evaluation, the present invention is significantly superior to traditional methods. By integrating multi-source data and applying advanced machine learning algorithms, the system can capture subtle psychological changes that are difficult to detect by traditional questionnaires. For example, voice analysis can infer the emotional state of patients from their intonation and speech rate, while body movement analysis can detect potential anxiety or depression symptoms. This multi-dimensional evaluation greatly improves the accuracy of psychological state judgment.

[0109] Secondly, in terms of the personalization and timeliness of intervention, the present invention has achieved a qualitative leap. The system can not only automatically adjust the intervention strategy according to the real-time evaluation results, but also learn from the feedback of patients and continuously optimize the recommendation algorithm. This means that each patient can obtain customized psychological support instead of a one-size-fits-all standardized treatment. Especially in the aspect of music intervention, the system can accurately recommend the most effective music works according to the patient's current emotional state and personal preferences, greatly improving the intervention effect.

[0110] Furthermore, the system of the present invention realizes all-weather monitoring and support. Different from traditional regular evaluations, this system can continuously collect and analyze patient data and timely detect potential psychological problems. This real-time monitoring can not only give intervention at the early stage of the problem, but also provide a report on the dynamic changes of the patient's psychological state for medical staff to assist in clinical decision-making.

[0111] In addition, the remote psychological counseling consultation module of the present invention breaks through geographical restrictions, enabling patients to obtain professional psychological counseling services at any time. The system can also provide a comprehensive evaluation report of the patient and relevant medical literature for consulting doctors, improving the efficiency and quality of consultation. This remote service model not only facilitates patients, but also alleviates the problem of uneven distribution of medical resources.

[0112] Finally, the system of the present invention has the ability of self-learning and optimization. By continuously collecting patient data and feedback on intervention effects, the system can continuously improve its evaluation model and intervention strategy. This dynamic optimization mechanism ensures the continuous improvement of system performance and can adapt to changing clinical needs.

[0113] In summary, the intelligent psychological state assessment and intervention system and method for patients after hepatobiliary and pancreatic surgery of the present invention effectively solve the deficiencies of traditional methods in terms of comprehensive assessment, personalized intervention, service continuity, etc. through innovative technical means and overall design. It can not only significantly improve the mental health status of patients, but also accelerate the postoperative rehabilitation process, improve the medical effect and patient satisfaction. This intelligent, personalized, and whole-process mental health management model provides a new solution for the comprehensive rehabilitation of patients after hepatobiliary and pancreatic surgery, and has important clinical application value and broad promotion prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0114] Figure 1 is the overall logic flow chart of the system of the present invention;

[0115] Figure 2 is the detailed structure diagram of the psychological state assessment model module of the present invention;

[0116] Figure 3 is the working flow chart of the psychological test module of the present invention;

[0117] Figure 4 is the functional logic diagram of the music intervention module of the present invention;

[0118] Figure 5 is the structure and function diagram of the assessment module of the present invention;

[0119] Figure 6 is the operation flow chart of the remote psychological consultation and consultation module of the present invention; DETAILED DESCRIPTION OF THE INVENTION

[0120] Please refer to the attached Figure 1-6 , the present invention provides an intelligent psychological state assessment and intervention system and method for patients after hepatobiliary and pancreatic surgery. The system aims to solve the common psychological problems of patients after hepatobiliary and pancreatic surgery, and improve the mental health level and quality of life of patients through intelligent assessment and personalized intervention.

[0121] The system of the present invention includes a psychological state assessment model module 1, a psychological test module 2, a music intervention module 3, an assessment module 4, and a remote psychological consultation and consultation module 5. These modules are connected by data to form a complete assessment and intervention closed loop.

[0122] The psychological state assessment model module 1 is the core component of the system. Based on pre-input clinical data, psychological test scale results, and emotion recognition data, this module trains a BP neural network to construct a psychological state assessment model. Preferably, the pre-input data includes clinical data such as the patient's age, gender, type of surgery, postoperative recovery, etc., as well as the scores of psychological test scales such as anxiety and depression. In one embodiment of the present invention, the number of hidden layer nodes of the BP neural network is set to 1.5 times the number of input layer nodes, which is a better choice based on empirical values and can achieve a balance between model complexity and computational efficiency.

[0123] The psychological state assessment model module 1 also uses speech recognition technology to convert the patient's speech into text and performs natural language processing. In this process, the system uses the jieba word segmentation tool to segment the text and extract keywords and phrases. At the same time, named entity recognition is performed through the Stanford NER tool to identify entities such as person names, place names, and organization names in the text. These processes help the system understand the patient's language expression more accurately, thereby better evaluating their psychological state.

[0124] Next, the psychological state assessment model module 1 uses an improved LSTM neural network to perform emotion recognition and classification on the processed text. The present invention improves the traditional LSTM network by introducing an attention mechanism, enabling the model to better capture key emotional information in long texts. Specifically, the improved LSTM model can be expressed as:

[0125] f t =σ(W f ·[h t-1 ,x t +b f ),

[0126] i t =σ(W i ·[h t-1 ,x t +b i ),

[0127]

[0128] o t =σ(W o ·[h t-1 ,x t +b o ),

[0129] h t =o t *tanh(C t ),

[0130] where ft and i t and o t are the forget gate, input gate, and output gate respectively. C t is the cell state, h t is the hidden state, and W f , W i , W C , W o are weight matrices, and b f , b i , b C , b o are bias vectors, σ is the sigmoid activation function, and tanh is the hyperbolic tangent activation function.

[0131] The psychological test module 2 is data-connected to the psychological state assessment model module 1 and receives the assessment results sent by the psychological state assessment model module 1. Based on these results, the psychological test module 2 selects an appropriate psychological test scale to further evaluate the patient's emotions. In the system of the present invention, the psychological test scale library includes, but is not limited to, the HAMA anxiety scale, Hamilton depression scale, self-rating depression scale, self-rating anxiety scale, and symptom checklist scale, etc. The system will select the most suitable scale for the patient's current state according to the preliminary assessment results. For example, if the preliminary assessment shows that the patient may have a depressive tendency, the system will preferentially select the Hamilton depression scale or the self-rating depression scale for testing.

[0132] The music intervention module 3 is data-connected to the psychological state assessment model module 1 and the psychological test module 2. This module receives the assessment results sent by the psychological state assessment model module 1 and the psychological test module 2, and recommends and plays different types of music to the patient according to these results. In one embodiment of the present invention, the music library is classified according to emotion types and intensities, including multiple categories such as relaxation music, uplifting music, and soothing music. The system will select the most suitable music type for recommendation according to the patient's psychological state. For example, for patients with a higher level of anxiety shown in the assessment results, the system may recommend classical music with a slow rhythm and beautiful melody, such as Mozart's piano concertos, to help the patient relax.

[0133] This personalized music recommendation and intervention method based on psychological state can effectively relieve the patient's negative emotions and improve their mental health level. Research shows that appropriate music intervention can reduce the patient's anxiety level, improve sleep quality, and even possibly accelerate the postoperative recovery process.

[0134] The evaluation module 4 of the present invention is connected to the psychological state evaluation model module 1, the psychological test module 2, and the music intervention module 3 for data connection. The evaluation module 4 receives the data sent by these modules and conducts comprehensive analysis and evaluation. A key function of this module is to analyze the text of the voice communication between the patient and the nurse and identify the voice characteristics of the patient. This process utilizes advanced acoustic analysis techniques to extract multiple characteristic parameters of the patient's voice, including but not limited to formants, Mel Frequency Cepstral Coefficients (MFCC), and gain adjustment, etc.

[0135] Preferably, the present invention uses the Support Vector Machine (SVM) algorithm to classify the extracted voice characteristics to determine whether the patient has psychological problems. The core idea of the SVM algorithm is to construct an optimal separating hyperplane in the high-dimensional feature space to maximize the margin between two types of samples. In the embodiment of the present invention, the decision function of the SVM can be expressed as:

[0136]

[0137] x is the input feature vector, x i is the support vector, y i is the class label, α i is the Lagrange multiplier, K(x i , x) is the kernel function, and b is the bias term. The present invention preferably uses the Radial Basis Function (RBF) as the kernel function, and its expression is:

[0138] K(x i , x) = exp(-γ||x i - x||),

[0139] where γ is the kernel function parameter, which is used to control the radial action range of the function. In practical applications, the value of γ usually ranges from 0.01 to 1. The present invention determines the optimal value of γ through the cross-validation method to balance the generalization ability of the model and the risk of overfitting.

[0140] The evaluation module 4 is also responsible for constructing a postoperative health status data model to evaluate the postoperative health status of the patient. This model comprehensively considers various physiological indicators of the patient, the recovery progress, and the results of the psychological state evaluation to form a comprehensive health status evaluation. In the preferred embodiment of the present invention, a Multi-Layer Perceptron (MLP) neural network is used to construct this model, and its structure includes an input layer, two hidden layers, and an output layer. The ReLU activation function is used in the hidden layer, and the Sigmoid function is used in the output layer to obtain a health status score between 0 and 1.

[0141] The remote psychological consultation module 5 is connected to the evaluation module 4 for data, and receives the patient status evaluation information sent by the evaluation module 4. The design of this module aims to achieve efficient remote psychological consultation and diagnosis. Through speech recognition and natural language processing technologies, the system can extract the text information of the conversation between the patient and the doctor in real time. During the text processing, the present invention adopts a pre-trained model based on BERT (Bidirectional Encoder Representations from Transformers) to better understand the semantic content of the conversation.

[0142] Another important function of the remote psychological consultation module 5 is to extract problem keywords and retrieve medical literature. Here, the present invention uses an improved TextRank algorithm for keyword extraction. Compared with the traditional TextRank, the improved version takes into account the position information and part-of-speech information of words, and can more accurately identify the key terms in medical conversations. The extracted keywords are then used to retrieve relevant medical literature to provide decision-making support for remote doctors.

[0143] The system of the present invention also includes a knowledge base module, which is connected to the remote psychological consultation module 5 for data. This knowledge base module stores a large number of articles, books and image materials related to the psychological state after hepatobiliary and pancreatic surgery. Through this module, the system can provide rich knowledge support for remote psychological consultation, help doctors understand the patient's situation more comprehensively, and formulate more accurate diagnosis and treatment plans.

[0144] Preferably, the knowledge base is constructed using graph database technology, and Neo4j is used as the backend storage engine. This structure allows the system to quickly retrieve relevant knowledge and establish complex associations between knowledge. For example, the system can quickly find out the common psychological problems related to a specific surgical type and the corresponding effective intervention methods.

[0145] Generally speaking, the intelligent evaluation and intervention system for the psychological state of patients after hepatobiliary and pancreatic surgery of the present invention realizes the full-process intelligent management from evaluation to intervention through the collaborative work of multiple functional modules. The innovation of the system lies not only in the advanced algorithms and technologies of each module, but also in their organic integration, forming a closed-loop evaluation-intervention-feedback mechanism. This design can provide continuous and personalized mental health support for patients after hepatobiliary and pancreatic surgery, and is expected to significantly improve the postoperative rehabilitation quality and quality of life of patients.

[0146] The mental state assessment model module 1 of the present invention further includes an innovative body movement analysis unit. This unit uses the advanced VOT (Visual Object Tracking) algorithm to perform motion estimation and tracking on the patient's body movements. The core idea of the VOT algorithm is to predict the motion trajectory of the target object through the position changes of the target object in consecutive video frames. In the application scenario of the present invention, this technology is used to analyze the patient's body language, which is an important supplementary indicator for evaluating the mental state.

[0147] Preferably, the VOT algorithm used in the present invention is based on the principle of the correlation filter. Its mathematical expression can be summarized as:

[0148] f(x) = w T φ(x),

[0149] where x is the input image, φ(x) is the feature extraction function, and w is the filter weight. The training objective of the filter is to minimize the following loss function:

[0150]

[0151] y i is the expected output, and λ is the regularization parameter.

[0152] Through this method, the system of the present invention can accurately extract the body movement characteristics of the patient, such as joint angular displacement. These characteristics are then used to construct a body behavior feature model, providing an additional dimension for mental state assessment. For example, frequent restlessness movements may indicate that the patient is in an anxious state, while slow and sluggish movements may be manifestations of depression.

[0153] The evaluation module 4 of the present invention further includes an intervention effect evaluation unit. This unit uses an improved HCAHPS (Hospital Consumer Assessment of Healthcare Providers and Systems) scoring system to evaluate the intervention effect of the patient. HCAHPS is a widely used tool for evaluating the quality of medical services. The present invention has made targeted improvements to better meet the special needs of patients after hepatobiliary and pancreatic surgery.

[0154] In the preferred embodiment of the present invention, the improved HCAHPS scoring system includes the following main dimensions:

[0155] 1. Degree of improvement in mental state: Evaluate the alleviation of negative emotions such as anxiety and depression in the patient.

[0156] 2. Improvement in physical symptoms: Evaluate the improvement of physical symptoms such as pain and sleep quality in the patient.

[0157] 3. Satisfaction with intervention methods: Evaluate the satisfaction of patients with intervention methods such as music intervention and remote consultation.

[0158] 4. Overall rehabilitation progress: Evaluate the overall rehabilitation speed and quality of patients.

[0159] The scoring for each dimension uses a quantitative scale of 1 - 10, where 1 represents "very dissatisfied" or "no improvement", and 10 represents "very satisfied" or "significant improvement". The final evaluation score is the weighted average of the scores of each dimension, and the weights are determined based on expert opinions and statistical analysis.

[0160] Based on this evaluation method, the intervention effect evaluation unit can generate a detailed intervention effect evaluation report. This report not only reflects the overall effect of the intervention but also points out which specific aspects need improvement. According to the evaluation results, the system will automatically provide feedback to the music intervention module 3 and the remote psychological consultation module 5 to adjust the intervention strategy. For example, if the evaluation shows that a certain type of music has a particularly good effect on relieving patients' anxiety, the system will increase the weight of this type of music in the recommendation list.

[0161] Finally, the present invention also provides a method for intelligent evaluation and intervention of the psychological state of patients after hepatobiliary and pancreatic surgery corresponding to the above system. This method includes a series of steps starting from obtaining patient data, through evaluation, testing, intervention, until effect evaluation and strategy adjustment, forming a complete closed loop.

[0162] Specifically, the method first obtains the clinical data, psychological test results, and emotion recognition data of the patient, inputs these data into the psychological state evaluation model for evaluation, and obtains the psychological state evaluation result of the patient. Based on this preliminary result, the method will select an appropriate psychological test scale to conduct a more in-depth psychological test on the patient.

[0163] Then, according to the evaluation results and test results, the system will match the corresponding music intervention plan, recommend and play suitable music for the patient. At the same time, the system will also use the support vector machine algorithm to analyze the voice characteristics of the patient to further determine whether the patient has psychological problems.

[0164] During this process, the method of the present invention will also construct a postoperative health status data model of the patient to comprehensively evaluate the postoperative health status of the patient. In addition, through speech recognition and natural language processing technologies, the system can extract the text information of the remote conversation between the patient and the doctor, and perform keyword extraction and medical literature retrieval on this information to provide decision-making support for remote doctors.

[0165] Innovatively, this method also uses the VOT algorithm to analyze the patient's limb movements, extract motion features, and construct a limb behavior feature model. This provides an additional dimension for psychological state assessment, making the assessment results more comprehensive and accurate.

[0166] Finally, this method uses an improved HCAHPS scoring system to evaluate the intervention effect and dynamically adjusts the intervention strategy according to the evaluation results. This closed-loop design ensures that the intervention plan can be continuously optimized to best meet the personalized needs of patients.

[0167] Generally speaking, the intelligent psychological state assessment and intervention system and method for patients after hepatobiliary and pancreatic surgery provided by the present invention provide comprehensive and continuous mental health support for patients after hepatobiliary and pancreatic surgery through multi-dimensional data analysis, application of advanced algorithms, and personalized intervention strategies. This not only helps improve the patient's psychological state, but may also accelerate the physical rehabilitation process, improve the overall treatment effect and patient satisfaction.

[0168] To verify the effectiveness and superiority of the intelligent psychological state assessment and intervention system and method for patients after hepatobiliary and pancreatic surgery of the present invention, a 3-month clinical trial was conducted. The trial was carried out in the hepatobiliary and pancreatic surgery ward of a certain tertiary hospital, and a total of 150 patients who had just completed hepatobiliary and pancreatic surgery were recruited. These patients were randomly divided into three groups, with 50 patients in each group, and different postoperative psychological assessment and intervention programs were adopted.

[0169] Example 1 (the present invention's solution): The intelligent assessment and intervention system and method of the present invention were adopted.

[0170] Comparative Example 1: Traditional manual psychological assessment and face-to-face psychological counseling methods were adopted.

[0171] Comparative Example 2: Simple questionnaire assessment and standardized music therapy methods were adopted.

[0172] During the trial, all three groups of patients received routine postoperative care. The patients in the Example 1 group used the system of the present invention for daily psychological state assessment and personalized intervention; the patients in the Comparative Example 1 group received face-to-face assessment and counseling by a psychologist once a week; the patients in the Comparative Example 2 group filled out a standardized mental health questionnaire once a week and listened to fixed relaxation music every day.

[0173] During the trial, the researchers regularly evaluated the patients, mainly focusing on the following indicators:

[0174] 1. Anxiety level: Evaluated using the HAMA (Hamilton Anxiety Scale)

[0175] 2. Depression degree: Evaluated using the HAMD (Hamilton Depression Scale)

[0176] 3. Sleep quality: Assessed using the PSQI (Pittsburgh Sleep Quality Index)

[0177] 4. Quality of life: Assessed using the SF-36 (Short Form 36 Health Survey)

[0178] 5. Patient satisfaction: Assessed using the improved HCAHPS questionnaire

[0179] 6. Recovery speed: The doctor evaluated the patient's recovery and gave a score from 1 to 10

[0180] At the end of the trial, the researchers conducted a final evaluation of the various indicators of the three groups of patients. The results are shown in the following table:

[0181]

[0182]

[0183] As can be seen from the above results, the intelligent evaluation and intervention system of the present invention is significantly superior to the traditional method and the simple standardized method in all indicators. The specific analysis is as follows:

[0184] 1. Improvement of anxiety and depression: Through real-time monitoring and timely intervention, the system of the present invention has better alleviated the anxiety and depression symptoms of patients. This may be because the system can give targeted intervention when the symptoms first appear, rather than waiting until the regular evaluation to find problems.

[0185] 2. Sleep quality: The personalized music recommendation function of the present invention has significantly improved the sleep quality of patients. The system can select the most suitable music according to the real-time state of patients, which is more effective than the fixed music therapy program.

[0186] 3. Quality of life: The improvement of the overall psychological state is directly reflected in the quality of life of patients. The all-round intervention of the present invention enables patients to feel more positive emotions in their daily lives.

[0187] 4. Patient satisfaction: The satisfaction rate of up to 92% indicates that patients highly recognize the intelligent system of the present invention. This may be because the system provides 24-hour uninterrupted support, making patients feel continuously concerned and cared for.

[0188] 5. Recovery speed: The recovery speed of the patients in the group of the present invention is significantly faster than that of the other two groups. This shows that a good psychological state can indeed promote physical recovery, verifying the clinical value of the present invention.

[0189] It should be noted that although traditional face-to-face psychological counseling (Comparative Example 1) has also achieved good results, the degree of improvement is less than that of the system of the present invention. This may be because the frequency of face-to-face counseling is limited and it is difficult to provide continuous support. The system of the present invention can monitor the patient's state all day long and give timely intervention.

[0190] In addition, although the simple standardized method (Comparative Example 2) is better than no intervention at all, the effect is significantly inferior to personalized intelligent intervention. This highlights the advantage of the present invention in terms of personalized customization.

[0191] In summary, the intelligent psychological state assessment and intervention system and method for patients after hepatobiliary and pancreatic surgery of the present invention have shown remarkable superiority in practical applications. It can not only effectively improve the psychological state of patients, improve the quality of life, but also accelerate the rehabilitation process of patients, improve the medical effect and patient satisfaction. This intelligent, personalized and all-day mental health management mode provides a new solution for the comprehensive rehabilitation of patients after hepatobiliary and pancreatic surgery and has broad clinical application prospects.

[0192] It should be noted that the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. An intelligent assessment and intervention system for the psychological state of patients after hepatobiliary and pancreatic surgery, characterized by: include: Mental state assessment model module, used to: Based on the pre-input clinical data, psychological test scale results and emotion recognition data, the BP neural network is trained to build a psychological state assessment model; Use speech recognition technology to convert patient speech into text and perform natural language processing to extract words, phrases and named entities from the text; Use the improved LSTM neural network to perform emotion recognition and classification on the processed text; The psychological test module is connected to the psychological state assessment model module data and is used to: Receiving the evaluation result sent by the mental state evaluation model module; Based on the assessment results, select an appropriate psychological test scale to conduct an emotional assessment on the patient; The music intervention module is data-connected with the mental state assessment model module and the mental test module, and is used to: Receiving the evaluation results sent by the psychological state evaluation model module and the psychological test module; recommending and playing different types of music to the patient based on the results of said assessment; An evaluation module, connected with the psychological state evaluation model module, the psychological test module and the music intervention module, is used to: Receiving data sent by the psychological state assessment model module, the psychological test module and the music intervention module; Analyze the text of the voice communication between patients and nurses to identify the patient's voice characteristics; Use the support vector machine algorithm to classify speech features and determine whether the patient has psychological problems; Construct a postoperative health status data model to evaluate the patient's postoperative health status.

2. The system according to claim 1, characterized in that It also includes a remote psychological consultation module, which is connected to the evaluation module data and is used to: Receiving patient status assessment information sent by the assessment module; Extract text information of conversations between patients and doctors through speech recognition and natural language processing technology; Based on the extracted text information, question keywords are extracted and medical literature is searched; Provide remote doctors with patient status information and relevant medical advice, and realize online consultation and diagnosis.

3. The system according to claim 1, characterized in that The mental state assessment model module includes: Mental state assessment model training unit, used for: The pre-input data is used as the training set, validation set, and test set with a probability of 50%, 70%, and 90% respectively; Preprocessing the training set and the test set; Use the preprocessed data to train the BP neural network and verify the evaluation accuracy; Speech and text processing unit, used for: Use Python's CRF++-1.3.6 version to complete Chinese text annotation; Use the conditional random field model to extract effective features from the text corpus and train the model; Associate the vocabulary with parts of speech and emotional tendencies; Sentiment analysis unit, used to: Calculate the sentiment score of each word in the corpus; Based on BP neural network, the sentiment score of the corpus is used as the target vector to build a sentiment analysis model; Improved LSTM model unit for: Based on the text sentiment score, the text is trained using an improved LSTM model; Deep learning technology is used for final emotion recognition and classification.

4. The system according to claim 1, characterized in that The psychological test module includes: Scale selection unit for: According to the results of psychological status assessment, select an appropriate scale from the preset psychological test scale library; The psychological test scale library includes HAMA Anxiety Scale, Hamilton Depression Rating Scale, Self-Rating Depression Scale, Self-Rating Anxiety Scale, Symptom Self-Rating Scale, etc. Test execution unit, used to: Present the patient with the selected psychological test scale; Collect the patients' answers; Result analysis unit, used for: According to the patient's answers, the scores of each scale were calculated; Generate psychological test reports.

5. The system according to claim 1, characterized in that The music intervention module includes: Music recommendation unit, used for: Based on the results of psychological status assessment and psychological test, music suitable for the patient's current emotional state is selected from the preset music library; Music playback unit, used for: Play the recommended music through the system interface or external devices connected thereto; Effects feedback unit for: Record the patient's response during the music listening process; Collect patients' subjective evaluation of the effect of music intervention.

6. The system according to claim 1, characterized in that The evaluation module includes: Speech analysis unit for: Use acoustic analysis technology to analyze the formants, MFCC, gain adjustment and other characteristics of the patient's voice; The extracted speech features are classified using the support vector machine algorithm; Health status modeling unit for: Collect patients’ postoperative recovery data; Construct a postoperative health status data model; Comprehensive Assessment Unit for: Combining speech analysis results, health status models, and psychological status assessment results; Generate a comprehensive patient assessment report.

7. The system according to claim 1, characterized in that The remote psychological consultation module includes: A dialogue processing unit for: Convert patient-doctor conversations into text using speech recognition technology; Preprocess the text by performing word segmentation, dependency syntax analysis, etc. Question extraction unit, used to: Extract keywords from the preprocessed text; Medical literature search based on extracted keywords; Consulting support unit for: Provide the remote physician with the patient's assessment results and relevant medical literature; Support doctors to provide online diagnosis and treatment advice.

8. The system according to claim 1, characterized in that Also includes: The knowledge base module is connected to the remote psychological consultation module data and is used to: Store articles, books, and imaging materials related to psychological status after hepatobiliary and pancreatic surgery; Provide knowledge support for remote psychological counseling consultation modules; The mental state assessment model module also includes: Body movement analysis unit for: The VOT algorithm is used to estimate and track the patient's limb movements; Extract motion features of limb movements, such as joint angular displacement; Based on the extracted motion features, a limb behavior feature model is constructed.

9. The system according to claim 1, characterized in that The evaluation module also includes: Intervention effectiveness evaluation unit, used to: The improved HCAHPS scoring system was used to evaluate the intervention effect on patients; Generate intervention effectiveness evaluation reports; Based on the evaluation results, feedback is provided to the music intervention module and the remote psychological counseling consultation module to adjust the intervention strategy.

10. An intelligent assessment and intervention method for the psychological state of patients after hepatobiliary and pancreatic surgery, characterized in that: The following steps are involved: S1. Obtain the patient's clinical data, psychological test results, and emotion recognition data, input them into the psychological state assessment model for assessment, and obtain the patient's psychological state assessment results; S2. Based on the psychological status assessment results, select an appropriate psychological test scale and conduct further psychological tests on the patient; S3. According to the psychological state assessment results and psychological test results, matching the corresponding music intervention program, and recommending and playing music to the patient; S4, using support vector machine algorithm to analyze the patient's voice characteristics and determine whether the patient has psychological problems; S5. Construct a data model of the patient's postoperative health status to evaluate the patient's postoperative health status; S6. Extract text information of remote conversations between patients and doctors based on speech recognition and natural language processing technology; S7, performing keyword extraction and medical literature retrieval on the extracted text information; S8. Provide patient status information and relevant medical advice to remote doctors, and support online consultation and diagnosis; S9, using VOT algorithm to analyze the patient's limb movements, extract movement features, and build a limb behavior feature model; S10. Use the improved HCAHPS scoring system to evaluate the intervention effect and adjust the intervention strategy based on the evaluation results.

Images