Patient information-based wish data processing system and method

Abstract

The invention discloses a wish data processing system and method based on patient information. The system comprises a wish data acquisition module, an answer matching classification module and a result statistics display module. The wish data acquisition module provides a question list for a target patient through wish listening software, collects reply contents of the target patient for the question list, and pre-processes and preliminarily analyzes the reply contents to obtain keywords of the reply contents; the answer matching classification module inputs the keywords into a wish classification model for classification to obtain wish labels corresponding to the keywords, and provides corresponding personalized services for the target patient according to the wish labels; and the result statistics display module uploads the wish label of the target patient to a medical staff center for display and storage through an encryption technology, and forms a visual instrument panel according to the wish labels of the plurality of patients for reference by medical staff. According to the invention, the wish data of the patient can be collected in time, and the patient is comprehensively humanistic.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and particularly to a wish data processing system and method based on patient information. Background Art

[0002] Hospice care is mainly a comprehensive and patient-centered care approach provided for patients with incurable diseases and limited life expectancy, aiming to improve the quality of life of these patients and their families. The wish messengers in hospice care are mainly volunteers or professionals who are committed to helping terminally ill patients realize their last wishes in the hospice care service, aiming to provide spiritual comfort and support to patients and improve their quality of life in the last stage of life by realizing their wishes. The work content of wish messengers includes in-depth communication with patients to understand their inner wishes, unfulfilled wishes, and any emotions or thoughts that patients want to express, and helping patients record important memories or information, such as writing letters to family members, recording video messages, etc., for leaving as a souvenir for relatives and friends. The role of wish messengers in hospice care is very important. It is not only to help realize specific wishes, but more importantly, to bring spiritual comfort to patients, so that patients can feel warmth and love at the last moment of their lives. This personalized care is crucial for improving the quality of life and end-of-life experience of patients.

[0003] However, in actual interpersonal relationships, some family members do not have the time and opportunity to listen to the inner thoughts of patients. Most of the wishes of hospice care patients are not understood and listened to early, missing the opportunity for medical staff, family members, and social workers to help patients realize their wishes and create memories early.

[0004] Therefore, the existing technology still needs to be improved and developed. Summary of the Invention

[0005] The main purpose of the present invention is to provide a wish data processing system and method based on patient information, aiming to solve the problem that most of the inner wishes of hospice care patients cannot be understood and listened to early in the prior art, and there are difficulties in communication between doctors and patients in the last stage of the patient's life, and it is impossible to collect the wish data of patients in time, resulting in the inability to provide comprehensive humanistic care for patients.

[0006] To achieve the above object, the present invention provides a wish data processing system based on patient information, and the wish data processing system based on patient information includes: a wish data collection module, an answer matching and classification module, and a result statistics and display module;

[0007] The wish data collection module, the answer matching and classification module, and the result statistics and display module are connected in sequence;

[0008] The wish data collection module is used to provide a list of questions to the target patient through the wish listening software, collect the response content of the target patient to the list of questions, preprocess and preliminarily analyze the response content, and obtain the keywords of the response content;

[0009] The answer matching and classification module is used to input the keywords into a pre-trained wish classification model for classification, obtain the wish labels corresponding to the keywords, and provide corresponding personalized services for the target patient;

[0010] The result statistics and display module is used to upload the wish labels of the target patient to the medical staff center for display and storage through encryption technology, and form a visual dashboard based on the wish labels of multiple patients for the reference of medical staff.

[0011] Optionally, for the wish data processing system based on patient information, the wish data collection module includes a question list sending unit, a response content collection unit, and a data preprocessing unit;

[0012] Among them, the question list sending unit, the response content collection unit, and the data preprocessing unit are connected in sequence;

[0013] The list sending unit is used to provide a list of questions to the target patient through the wish listening software after determining that the target patient is in the final stage of life;

[0014] The response content collection unit is used to collect text responses, voice responses, image responses, or file responses made by the target patient to the list of questions;

[0015] The data preprocessing unit is used to normalize the response content to obtain normalized data, use the Pandsa library or the Numpy library to clean the normalized data to obtain cleaned data, and use the open-source NIP library to extract keywords from the cleaned data to obtain the keywords of the response content.

[0016] Optionally, for the wish data processing system based on patient information, the wish data processing system based on patient information further includes: a model pre-training module, and the model pre-training module includes: a data collection unit, a data set division unit, and a model training unit;

[0017] Among them, the data collection unit, the data set division, and the model training unit are connected in sequence;

[0018] The data acquisition unit is used to construct a machine learning model for wish classification and obtain historical data, where the historical data includes historical keywords and wish labels corresponding to the historical keywords;

[0019] The dataset division unit is used to construct a dataset according to the historical keywords and wish labels corresponding to the historical keywords, and divide the dataset into a training set and a test set according to a preset ratio;

[0020] The model training unit is used to train the machine learning model using the training set and evaluate the machine learning model using the test set to obtain the wish classification model.

[0021] Optionally, in the wish data processing system based on patient information, the answer matching and classification module includes: a wish label matching unit, a rule engine processing unit, and a message sending unit;

[0022] Wherein, the wish label matching unit, the rule engine processing unit, and the message sending unit are connected in sequence;

[0023] The answer matching and classification module is used to input the keyword into a pre-trained wish classification model for classification to obtain the wish label corresponding to the keyword;

[0024] The rule engine processing unit is used to input the wish label into the rule engine for processing and output personalized suggestions or resource links for the wish label;

[0025] The message sending unit is used to send the personalized suggestions or the resource links to the target patient and record the receiving and browsing situations of the target patient.

[0026] Optionally, in the wish data processing system based on patient information, the answer matching and classification module further includes: an emotion recognition and adjustment unit;

[0027] The emotion recognition and adjustment unit is used to input the keyword into a pre-trained sentiment analysis model for recognition to obtain the emotional state of the target patient, and formulate the time and content for adjusting the personalized service according to the emotional state.

[0028] Optionally, in the wish data processing system based on patient information, the result statistics and display module includes: an encrypted upload unit, a similarity calculation unit, and a grouped statistics unit;

[0029] Wherein, the encrypted upload unit, the similarity calculation unit, and the grouped statistics unit are connected in sequence;

[0030] The encryption and upload unit is used to encrypt the wish label through the AES algorithm to obtain the encrypted wish label, upload the encrypted wish label to the medical staff center through SSL or TLS technology, decrypt the encrypted wish label to obtain the wish label, and display and save the wish label;

[0031] The similarity calculation unit is used to obtain the wish labels of multiple patients within the same time range, and calculate the similarity between the wish labels of multiple patients through the cosine similarity or JACCARD index algorithm;

[0032] The grouping and statistics unit is used to group and statistically analyze the wish labels of multiple patients according to the similarity to obtain a visual interface in the form of a chart.

[0033] In addition, to achieve the above object, the present invention also provides a method for processing wish data based on patient information for a wish data processing system based on patient information, wherein the method for processing wish data based on patient information includes:

[0034] Provide a question list to the target patient through the wish listening software, collect the reply content of the target patient for the question list, preprocess and preliminarily analyze the reply content to obtain the keywords of the reply content;

[0035] Input the keywords into a pre-trained wish classification model for classification to obtain the wish label corresponding to the keywords, and provide corresponding personalized services for the target patient according to the wish label;

[0036] Upload the wish label of the target patient to the medical staff center for display and storage through encryption technology, and form a visual dashboard based on the wish labels of multiple patients for the reference of medical staff.

[0037] Optionally, in the data processing method of the wish data processing system based on patient information, the types of the question list include: unfulfilled wishes and regrets, expressing emotions and relationship repair, farewell, review and sharing, spiritual comfort and inner peace, daily life and current experience, and special supplements.

[0038] Optionally, after providing corresponding personalized services for the target patient according to the wish label, it further includes:

[0039] Input the keywords into a pre-trained emotion analysis model for recognition to obtain the emotional state of the target patient, and formulate the time and content for adjusting the personalized service according to the emotional state;

[0040] The sentiment analysis model includes an input layer, an embedding layer, a convolutional layer, a bidirectional LSTM layer, an attention layer, and an output layer;

[0041] Among them, the input layer is used to receive the data to be processed and preprocess it to obtain a text sequence. The embedding layer is used to convert the text sequence into a vector representation. The convolutional layer is used to obtain the hidden layer representation of the vector representation. The bidirectional LSTM layer is used to capture long-distance context dependencies. The attention layer calculates the attention of a word to other words in the text sequence or words in other sequences, assigns different weights to each word in the sequence, and obtains a vector sequence with weight information. The output layer is used to perform a softmax function calculation on the vector sequence to obtain a scalar, and focus on key sentiment words according to the scalar to identify the sentiment trend of the target text.

[0042] Optionally, in the data processing method of the wish data processing system based on patient information, wherein the wish label of the target patient is uploaded to the medical staff center for display and storage through an encryption technology, and a visual dashboard is formed according to the wish labels of multiple patients, specifically including:

[0043] Encrypt the wish label through the AES algorithm to obtain the encrypted wish label, upload the encrypted wish label to the medical staff center through SSL or TLS technology, decrypt the encrypted wish label to obtain the wish label, and display and store the wish label;

[0044] Obtain the wish labels of multiple patients within the same time range, calculate the similarity between the wish labels of multiple patients through the cosine similarity or JACCARD index algorithm, group and count the wish labels of multiple patients according to the similarity, and obtain a visual interface in the form of a chart.

[0045] In the present invention, a wish data processing system based on patient information includes: a wish data collection module, an answer matching and classification module, and a result statistics and display module; the wish data collection module provides a question list to target patients through wish listening software, collects the response content of the target patients to the question list, preprocesses and preliminarily analyzes the response content, and obtains the keywords of the response content; the answer matching and classification module inputs the keywords into a wish classification model for classification, obtains the wish labels corresponding to the keywords, and provides corresponding personalized services for the target patients according to the wish labels; the result statistics and display module uploads the wish labels of the target patients to the medical staff center for display and storage through encryption technology, and forms a visual dashboard according to the wish labels of multiple patients for the reference of medical staff. The present invention can timely understand and listen to the wishes of palliative care patients in the early stage, can timely collect the wish data of patients, and provide comprehensive humanistic care for patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is a structural diagram of a preferred embodiment of the wish data processing system based on patient information of the present invention;

[0047] Figure 2 is a flowchart of a preferred embodiment of the data processing method of the wish data processing system based on patient information of the present invention;

[0048] Figure 3 is a schematic diagram of the principle of the data processing method of the wish data processing system based on patient information of the present invention;

[0049] Figure 4 is a schematic diagram of the patient grouping interface in the data processing method of the wish data processing system based on patient information of the present invention;

[0050] Figure 5 is a schematic diagram of the patient wish distribution interface in the data processing method of the wish data processing system based on patient information of the present invention; DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] The present application provides a wish data processing system and method based on patient information. To make the purpose, technical solution and effect of the present application clearer and more definite, the following further describes the present application in detail with reference to the accompanying drawings and by way of examples. 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.

[0052] Those skilled in the art can understand that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as the general understanding of those of ordinary skill in the art to which this application belongs. It should also be understood that terms such as those defined in a general dictionary should be understood to have a meaning consistent with the meaning in the context of the prior art, and will not be interpreted in an idealized or overly formal sense unless specifically defined as here.

[0053] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, such descriptions of "first", "second", etc. are only for descriptive purposes and should not be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, features defined with "first", "second" may explicitly or implicitly include at least one such feature. In addition, the technical solutions between various embodiments can be combined with each other, but it must be based on the ability of those of ordinary skill in the art to implement. When the combination of technical solutions results in contradictions or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0054] The wish data processing system based on patient information according to a preferred embodiment of the present invention includes: a wish data collection module, an answer matching and classification module, and a result statistics and display module; the wish data collection module, the answer matching and classification module, and the result statistics and display module are connected in sequence.

[0055] Among them, the wish data collection module is used to provide a question list to the target patient through the wish listening software, collect the reply content of the target patient to the question list, preprocess and preliminarily analyze the reply content, and obtain the keywords of the reply content.

[0056] Specifically, the wish data collection module includes a question list sending unit, a reply content collection unit, and a data preprocessing unit; among them, the question list sending unit, the reply content collection unit, and the data preprocessing unit are connected in sequence.

[0057] The list sending unit is used to provide a question list to the target patient through the wish listening software after determining that the target patient is in the final stage of life; the reply content collection unit is used to collect text replies, voice replies, image replies, or file replies made by the target patient to the question list; the data preprocessing unit is used to normalize the reply content to obtain normalized data, perform data cleaning on the normalized data using the Pandsa library or the Numpy library to obtain cleaned data, and perform keyword extraction on the cleaned data using the open-source NIP library to obtain the keywords of the reply content.

[0058] It is understandable that the above list of questions for asking the wishes of palliative care patients aims to provide a safe space for palliative care patients, enabling them to have the opportunity to share their innermost feelings, memories, wishes, and regrets. At the same time, these questions also encourage family members to participate and provide them with a better way to understand and support their loved ones.

[0059] At the same time, palliative care patients can choose their own ways to answer the questions, including writing, typing, recording, taking pictures, shooting videos, etc. Therefore, the data obtained is in various formats. Database normalization is to eliminate data redundancy, improve data consistency, optimize data storage, and enhance the scalability of the database. Data redundancy can lead to data inconsistency, increasing storage costs and complexity. And data cleaning is to process the data to remove errors, incompleteness, inaccuracy, repetition, redundancy, etc. in the data, thereby improving data quality and accuracy. Finally, natural language processing techniques such as the TF-IDF (Term Frequency-Inverse Document Frequency) algorithm are used for keyword extraction. The TF-IDF algorithm is a commonly used keyword extraction method that evaluates the importance of a word in a document by calculating the term frequency (TF) and the inverse document frequency (IDF).

[0060] Furthermore, the wish data processing system based on patient information further includes: a model pre-training module, and the model pre-training module includes: a data collection unit, a data set division unit, and a model training unit; wherein, the data collection unit, the data set division unit, and the model training unit are connected in sequence.

[0061] Specifically, the data collection unit is used to construct a machine learning model for wish classification and obtain historical data, and the historical data includes historical keywords and wish labels corresponding to the historical keywords; the data set division unit is used to construct a data set according to the historical keywords and wish labels corresponding to the historical keywords, and divide the data set into a training set and a test set according to a preset ratio; the model training unit is used to train the machine learning model using the training set and evaluate the machine learning model using the test set to obtain the wish classification model.

[0062] It is understandable that the model pre-training module is independently connected to the answer matching and classification module. When the keyword is input into the answer matching and classification module, the answer matching and classification module will call the wish classification model that has been pre-trained in the model pre-training module.

[0063] The answer matching and classification module is used to input the keywords into a pre-trained wish classification model for classification, obtain the wish label corresponding to the keywords, and provide corresponding personalized services for the target patient according to the wish label.

[0064] Specifically, the answer matching and classification module includes: a wish label matching unit, a rule engine processing unit, and a message sending unit; among them, the wish label matching unit, the rule engine processing unit, and the message sending unit are connected in sequence.

[0065] The answer matching and classification module is used to input the keywords into a pre-trained wish classification model for classification, and obtain the wish label corresponding to the keywords; the rule engine processing unit is used to input the wish label into the rule engine for processing, and output personalized suggestions or resource links for the wish label; the message sending unit is used to send the personalized suggestions or the resource links to the target patient, and record the receiving and browsing situations of the target patient.

[0066] Furthermore, the answer matching and classification module further includes: an emotion recognition and adjustment unit; the emotion recognition and adjustment unit is used to input the keywords into a pre-trained sentiment analysis model for recognition, obtain the emotional state of the target patient, and formulate the time and content for adjusting the personalized service according to the emotional state.

[0067] By paying attention to and responding to the emotional needs of patients, the present invention can not only improve the physiological health level of patients, but also significantly improve their overall medical experience and increase their satisfaction with medical services.

[0068] The result statistics and display module is used to upload the wish label of the target patient to the medical staff center for display and storage through encryption technology, and form a visual dashboard according to the wish labels of multiple patients for the reference of medical staff.

[0069] Specifically, the result statistics and display module includes: an encrypted upload unit, a similarity calculation unit, and a grouped statistics unit; among them, the encrypted upload unit, the similarity calculation unit, and the grouped statistics unit are connected in sequence.

[0070] The encrypted upload unit is used to encrypt the wish label through the AES algorithm to obtain the encrypted wish label, upload the encrypted wish label to the medical staff center through SSL or TLS technology, decrypt the encrypted wish label to obtain the wish label, and display and save the wish label; the similarity calculation unit is used to obtain the wish labels of multiple patients within the same time range, and calculate the similarity between the wish labels of multiple patients through the cosine similarity or JACCARD index algorithm; the grouping and statistics unit is used to group and statistically analyze the wish labels of multiple patients according to the similarity to obtain a visual interface in the form of a chart.

[0071] The method for processing wish data based on patient information of the wish data processing system based on patient information in the preferred embodiment of the present invention is as Figure 2 and Figure 3 shown, the method for processing wish data based on patient information includes the following steps:

[0072] Step S10: Provide a question list to the target patient through the wish listening software, collect the reply content of the target patient for the question list, preprocess and preliminarily analyze the reply content to obtain the keywords of the reply content.

[0073] Specifically, when it is determined that the target patient is in the final stage of life, a question list is provided to the target patient through the wish listening software; wherein, the types of the question list include: unfulfilled wishes and regrets, expressing emotions and relationship repair, farewell, review and sharing, spiritual comfort and inner peace, daily life and current experience, and special supplements.

[0074] As an example, the questions of unfulfilled wishes and regrets include: "So far, what wishes do you still have unfulfilled", "What are the things you always wanted to do but didn't", "What wishes do you still want to fulfill in this life", and "What regrets do you have in this life", etc.

[0075] Expressing emotions and relationship repair includes: "In your life journey, which people have given you great help or support and how do you want to express your gratitude to them", "For the people you cherish, what unexpressed love and care do you want to tell them", "Looking back, is there anything that makes you feel the need to apologize to someone or seek forgiveness and how do you hope to convey your apology", and "At this special moment, do you have anything to say to someone or hope they know", etc.

[0076] Farewell-related questions include: "Facing the upcoming farewell, are there any people you hope to see in the last moments?", "Are there any people you hope to spend more time with?", "How do you hope your family and friends will accompany you?", "Do you have any special expectations or wishes for your family and friends?", "In what way do you hope to say goodbye to important relatives or friends?", and "Are there any specific arrangements or things you hope we can help you with?", etc.

[0077] Review and sharing-related questions include: "In your memory, what are the particularly wonderful moments that made you feel happy?", "What was your most cherished family gathering or friends' get-together like?", "In your life, which moments made you feel the most accomplished?", and "What is your proudest experience and why is it so important to you?", etc.

[0078] Spiritual comfort and inner peace-related questions include: "Do you have any special thoughts or needs regarding your faith or spiritual life now?", "For you, what can bring you inner peace", and "Are there specific prayers, meditations, or other spiritual activities that can comfort you?", etc.

[0079] Daily life and current experience-related questions include: "How do you want to spend today?", "What are your favorite foods, music, books, movies, or other activities?", "Are there any special habits or activities you want to continue to maintain?", and "In the coming days, is there anything that can make you feel more comfortable?", etc.

[0080] Special supplement-related questions include: "Are there any other important memories or information you want to record?" and "How do you hope people will remember you in the future?", etc.

[0081] In this embodiment, the above question list for asking about the wishes of palliative care patients aims to provide a safe space for palliative care patients, enabling them to have the opportunity to share their innermost feelings, memories, wishes, and regrets. At the same time, these questions also encourage family members to participate and provide them with a better way to understand and support their loved ones.

[0082] Furthermore, collect the text responses, voice responses, image responses, or file responses made by the target patients to the question list.

[0083] It is understandable that after the target patient logs in to the software system with their real name, they can complete their personal information; are allowed to call the shooting and recording system for shooting and recording; are allowed to push the recorded content to the medical staff port; there is an answer area set under the questions of the software system, and the patient can choose the way to answer the questions by themselves, including writing, typing, recording, taking pictures, shooting videos, etc.; for the convenience of the patient's operation and recording, any text, recording, photo and video recorded by the patient do not need to be clicked to save manually, and the system will automatically save the recorded content; when new content needs to be recorded, a new segment of questions and conversations will be started.

[0084] Furthermore, preprocess and preliminarily analyze the reply content to obtain the keywords of the reply content, specifically including:

[0085] Normalize the reply content to obtain normalized data, and use the Pandsa library (a tool based on NumPy) or the Numpy library (an open-source numerical computing extension of Python) to clean the normalized data to obtain the cleaned data; use the open-source NIP (Natural Language Processing) library to extract keywords from the cleaned data to obtain the keywords of the reply content.

[0086] [[ID=**9**]]It is understandable that the patient can choose the way to answer the questions by themselves, including writing, typing, recording, taking pictures, shooting videos, etc., so the data formats obtained are various. Database normalization is to eliminate data redundancy, improve data consistency, optimize data storage, and improve the scalability of the database. Data redundancy will lead to data inconsistency, increase storage costs and complexity. And data cleaning is to process the data to remove problems such as errors, incompleteness, inaccuracy, repetition, redundancy, etc. in the data, so as to improve data quality and accuracy. Finally, natural language processing techniques such as the TF-IDF (Term Frequency-Inverse Document Frequency) algorithm are used for keyword extraction. The TF-IDF algorithm is a commonly used keyword extraction method, which evaluates the importance of a word in a document by calculating the term frequency (TF) and the inverse document frequency (IDF).

[0087] Step S20: Input the keywords into a pre-trained wish classification model for classification to obtain the wish labels corresponding to the keywords, and provide corresponding personalized services for the target patient according to the wish labels.

[0088] Before inputting the keywords into the pre-trained wish classification model for classification, it further includes:

[0089] Build a machine learning model for wish classification, obtain historical data (the historical data is obtained from a wish label library, and the wish label library provides a set of standardized labels or terms for various types of wishes, enabling similar wishes expressed by different patients to be consistently identified and processed), the historical data includes historical keywords and the wish labels corresponding to the historical keywords; construct a data set according to the historical keywords and the wish labels corresponding to the historical keywords, and divide the data set into a training set and a test set according to a preset ratio; use the training set to train the machine learning model, and use the test set to evaluate the machine learning model to obtain the wish classification model.

[0090] By pre - defining wish labels, it can help machine learning models or rule - based systems classify patients' wishes more accurately. These labels serve as the basis for classification and help reduce the occurrence of misclassification.

[0091] Further, input the keyword into a pre - trained wish classification model for classification to obtain the wish label corresponding to the keyword.

[0092] In this embodiment, input the keyword into a pre - trained wish classification model for classification. The wish classification model can automatically classify the input keyword and output the wish label corresponding to the keyword. For example, wish keywords such as "hoping for family company" and "wanting to see friends from afar" can be classified under the wish label of "family relationships and social connections".

[0093] It can be understood that as the system is used, new wish types may appear. At this time, the classification system can be extended by adding new labels to the wish label library to ensure the adaptability and flexibility of the classification system. The wish label library is not only the basis for intelligent classification, but also plays an important role in improving classification accuracy, supporting personalized services, and promoting data analysis. Its relationship with classification is interdependent, and a good wish label library can significantly improve the efficiency of the entire system.

[0094] Further, provide corresponding personalized services for the target patient according to the wish label, specifically including:

[0095] Input the wish label into a rule engine for processing, and output personalized suggestions or resource links for the wish label;

[0096] Send the personalized suggestions or the resource links to the target patient, and record the receiving and browsing situations of the target patient.

[0097] In this embodiment, after obtaining the wish label, the wish label is input into a rule engine or a recommendation system for processing, and the output result is personalized suggestions or resource links for the wish label. The patient can directly click on the resource link to enter it. For example, if it is detected that the patient's wish label is "spiritual needs", resource links such as relevant meditation courses can be recommended.

[0098] Furthermore, after providing the corresponding personalized service for the target patient according to the wish label, it further includes:

[0099] Input the keyword into a pre-trained sentiment analysis model for recognition to obtain the emotional state of the target patient;

[0100] Adjust the time and content of the personalized service according to the emotional state.

[0101] In this embodiment, the sentiment analysis model includes an input layer, an embedding layer, a convolutional layer, a bidirectional LSTM layer, an attention layer, and an output layer. Among them, the input layer is used to receive the data to be processed and preprocess it, and the embedding layer converts the text sequence into a vector representation; the convolutional layer is used to obtain the hidden layer expression of the vector obtained from the embedding layer; the bidirectional LSTM layer is used to capture long-distance context dependencies, and the attention layer calculates the attention of a word to other words in the text sequence or words in other sequences, assigns different weights to each word in the sequence, so as to obtain a vector sequence with weight information; the output layer performs a softmax function calculation on the vector sequence passing through the attention layer to obtain a scalar, focuses on the key sentiment words, and thus identifies the emotional trend of the target text.

[0102] The embedding layer uses two methods, word embedding and character embedding, to convert the text sequence, increasing the dual-channel input at the character level and the word level. Among them: the word embedding layer performs two-dimensional convolution on the character sequence in each word, performs max pooling on the convolution result, and obtains the character-level vector corresponding to the word; the word embedding layer converts each word into a vector through pre-trained Glove word vectors, and finally concatenates it with the character-level vector to obtain the text sequence vector.

[0103] It can be understood that sentiment analysis is used to identify the emotional state of the patient (such as anxiety, depression, optimism, fear, etc.), and adjust the time and content of the personalized service according to the emotional state. For example, for patients showing a high level of anxiety or fear, more psychological counseling services can be arranged in the original push content, or additional attention and support can be given, more gentle and encouraging language can be selected for communication, and the personalized service time can be advanced. This helps to create a more user-friendly and caring medical environment.

[0104] The functions of sentiment analysis are mainly reflected in the following aspects: Improving the quality of personalized care: Once the emotional state of the patient is identified, medical staff can provide more personalized support and services based on this information. Improving communication strategies and optimizing communication methods: Understanding the emotional tendencies of patients helps to adjust the communication methods with patients. For example, if it is found that the patient is in a low mood, medical staff may choose to communicate in a more gentle and encouraging language to enhance the patient's comfort. Assisting in clinical decision-making: The emotional state may affect the patient's acceptance and response to treatment plans. When formulating or adjusting treatment plans, it is very important to consider the patient's current emotional state. Mental health monitoring: Conducting sentiment analysis can help medical staff track changes in the patient's mental health status, detect potential psychological problems in a timely manner, and take corresponding intervention measures. Enhancing patient satisfaction: By paying attention to and responding to the emotional needs of patients, not only can the patient's physical health level be improved, but also their overall medical experience can be significantly improved, increasing their satisfaction with medical services.

[0105] Step S30: Upload the wish label of the target patient to the medical staff center for display and storage through encryption technology, and form a visual dashboard based on the wish labels of multiple patients for medical staff to refer to.

[0106] Specifically, encrypt the wish label through the AES algorithm to obtain the encrypted wish label, upload the encrypted wish label to the medical staff center through SSL or TLS technology, decrypt the encrypted wish label to obtain the wish label, and display and store the wish label.

[0107] The AES algorithm (Advanced Encryption Standard) is a symmetric encryption algorithm that uses the same key for encryption and decryption operations, including: Block division: Divide the plaintext into blocks of 128 bits (16 bytes). Initial round key addition: Perform an exclusive OR operation on the plaintext block and the first round key. Main round transformation (repeat 10 / 12 / 14 rounds, depending on the key length): Byte substitution: Replace each byte with an S-box. Row shift: Circularly shift the bytes in each row to the left. Column confusion: Perform a matrix transformation on each column (omitted in the last round). Round key addition: Exclusive OR with the current round key. Final round: Perform byte substitution, row shift, and round key addition, skipping column confusion. Output ciphertext: Obtain a 128-bit ciphertext block (encrypted wish label).

[0108] Upload the encrypted wish label to the medical staff center through SSL (Secure Socket Layer) or TLS (Transport Layer Security) technology to protect the security of the patient's personal information.

[0109] Further, obtain the wish tags of multiple patients within the same time range, calculate the similarity between the wish tags of multiple patients through the cosine similarity or JACCARD index algorithm, group and count the wish tags of multiple patients according to the similarity, and obtain a visual interface in the form of a chart.

[0110] Cosine similarity, also known as cosine similarity, is used to evaluate the similarity between two vectors by calculating the cosine value of the angle between the two vectors. The JACCARD index algorithm, that is, the Jaccard similarity coefficient, is used to compare the similarity and difference between finite sample sets.

[0111] In this embodiment, the similarity between the wish tags of multiple patients is calculated through the cosine similarity or JACCARD index algorithm. Among them, cosine similarity is suitable for vector data and focuses on the consistency of direction. The Jaccard index is suitable for set data and focuses on the overlapping ratio of elements, and can be selected according to the actual situation. Further, wish tags from different patients are received in real time, and their similarity is calculated. The wish tags of multiple patients are grouped and counted according to the similarity (as Figure 4 shown), and a visual interface in the form of a chart that changes in real time is obtained (as Figure 5 shown). Finally, a downloadable and printable wish analysis report can also be generated based on the above statistical data and provided to medical staff, which helps to promote communication between doctors and patients.

[0112] It can be seen that the present invention provides a convenient, free and safe wish processing method for patients in palliative care. This method asks, listens to, and records the innermost feelings, memories, wishes and regrets of patients through a wish question list, analyzes and processes, matches and classifies, and dynamically groups the content uploaded by patients, gives patients personalized suggestions and resource links, and finally generates a downloadable and printable wish analysis report. This method can not only help patients find inner peace, but also leave precious memories for their families, enabling patients to feel love and care in the last stage of their lives. The results of the wish analysis report help to promote communication between doctors and patients, ensure that the voices of patients are heard, and their wishes are respected and supported, and can become an important tool for enhancing the humanistic care ability of hospitals.

[0113] In summary, the present invention proposes a wish data processing system and method based on patient information. The wish data processing system based on patient information includes: a wish data collection module, an answer matching and classification module, and a result statistics and display module. The wish data collection module provides a question list to the target patient through the wish listening software, collects the response content of the target patient to the question list, preprocesses and preliminarily analyzes the response content, and obtains the keywords of the response content. The answer matching and classification module inputs the keywords into the wish classification model for classification, obtains the wish labels corresponding to the keywords, and provides corresponding personalized services for the target patient according to the wish labels. The result statistics and display module uploads the wish labels of the target patient to the medical staff center for display and storage through encryption technology, and forms a visual dashboard based on the wish labels of multiple patients for the reference of medical staff. The present invention can timely understand and listen to the wishes of palliative care patients in the early stage, can timely collect the wish data of patients, and provide comprehensive humanistic care for patients.

[0114] It should be noted that in this article, the term "including", "comprising" or any other variant thereof is intended to cover a non-exclusive inclusion, so that a process, method, article or terminal including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or terminal. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, article or terminal including that element.

[0115] Of course, those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided by the present invention can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0116] It should be understood that the application of the present invention is not limited to the above examples. For those of ordinary skill in the art, improvements or transformations can be made according to the above description. All such improvements and transformations should fall within the protection scope of the appended claims of the present invention.

Claims

1. A wish data processing system based on patient information, characterized in that, The described wish data processing system based on patient information includes: a wish data collection module, an answer matching and classification module, and a result statistics and display module; The wish data collection module, the answer matching and classification module, and the result statistics and display module are connected in sequence; The wish data collection module is used to provide a question list to the target patient through wish listening software, collect the response content of the target patient to the question list, preprocess and preliminarily analyze the response content, and obtain the keywords of the response content; The answer matching and classification module is used to input the keywords into a pre-trained wish classification model for classification, obtain the wish labels corresponding to the keywords, and provide corresponding personalized services for the target patient according to the wish labels; The result statistics and display module is used to upload the wish labels of the target patient to the medical staff center for display and storage through encryption technology, and form a visual dashboard based on the wish labels of multiple patients for medical staff to refer to.

2. The wish data processing system based on patient information according to claim 1, wherein The wish data collection module includes a question list sending unit, a response content collection unit, and a data preprocessing unit; Among them, the question list sending unit, the response content collection unit, and the data preprocessing unit are connected in sequence; The list sending unit is used to provide a question list to the target patient through wish listening software after determining that the target patient is in the final stage of life; The response content collection unit is used to collect text responses, voice responses, image responses, or file responses made by the target patient to the question list; The data preprocessing unit is used to normalize the response content to obtain normalized data, clean the normalized data using the Pandsa library or the Numpy library to obtain cleaned data, and extract keywords from the cleaned data using the open-source NIP library to obtain the keywords of the response content.

3. The wish data processing system based on patient information according to claim 1, wherein The wish data processing system based on patient information further includes: a model pre-training module, and the model pre-training module includes: a data collection unit, a data set division unit, and a model training unit; Among them, the data collection unit, the data set division, and the model training unit are connected in sequence; The data collection unit is used to construct a machine learning model for wish classification and obtain historical data, where the historical data includes historical keywords and the wish labels corresponding to the historical keywords; The data set division unit is used to construct a data set according to the historical keywords and the wish labels corresponding to the historical keywords, and divide the data set into a training set and a test set according to a preset ratio; The model training unit is used to train the machine learning model using the training set and evaluate the machine learning model using the test set to obtain the wish classification model.

4. The wish data processing system based on patient information according to claim 1, characterized in that, The answer matching and classification module includes: a wish label matching unit, a rule engine processing unit, and a message sending unit; Among them, the wish label matching unit, the rule engine processing unit, and the message sending unit are connected in sequence; The answer matching and classification module is used to input the keywords into a pre-trained wish classification model for classification, and obtain the wish label corresponding to the keywords; The rule engine processing unit is used to input the wish label into the rule engine for processing, and output personalized suggestions or resource links for the wish label; The message sending unit is used to send the personalized suggestions or the resource links to the target patient, and record the receiving and browsing situations of the target patient.

5. The wish data processing system based on patient information according to claim 4, wherein The answer matching and classification module further includes: an emotion recognition and adjustment unit; The emotion recognition and adjustment unit is used to input the keywords into a pre-trained sentiment analysis model for recognition, obtain the emotional state of the target patient, and formulate the time and content for adjusting the personalized service according to the emotional state.

6. The wish data processing system based on patient information according to claim 1, wherein, The result statistics and display module includes: an encrypted upload unit, a similarity calculation unit, and a grouped statistics unit; Among them, the encrypted upload unit, the similarity calculation unit, and the grouped statistics unit are connected in sequence; The encrypted upload unit is used to encrypt the wish label through the AES algorithm to obtain the encrypted wish label, upload the encrypted wish label to the medical staff center through SSL or TLS technology, decrypt the encrypted wish label to obtain the wish label, and display and save the wish label; The similarity calculation unit is used to obtain the wish labels of multiple patients within the same time range, and calculate the similarity between the wish labels of multiple patients through the cosine similarity or JACCARD index algorithm; The grouped statistics unit is used to group and statistically analyze the wish labels of multiple patients according to the similarity, and obtain a visual interface in the form of a chart.

7. A method for processing wish data based on patient information of a wish data processing system based on patient information according to any one of claims 1-6, characterized in that, The wish data processing method based on patient information includes: Providing a question list to the target patient through a wish listening software, collecting the reply content of the target patient for the question list, preprocessing and preliminarily analyzing the reply content, and obtaining the keywords of the reply content; Inputting the keywords into a pre-trained wish classification model for classification, obtaining the wish label corresponding to the keywords, and providing corresponding personalized services for the target patient according to the wish label; Uploading the wish label of the target patient to the medical staff center for display and storage through encryption technology, and forming a visual dashboard according to the wish labels of multiple patients for the reference of medical staff.

8. The method for processing wish data based on patient information according to claim 7, wherein The types of the question list include: unfulfilled wishes and regrets, expressing emotions and relationship repair, farewell, review and sharing, spiritual comfort and inner peace, daily life and current experience, and special supplements.

9. The method for processing wish data based on patient information according to claim 7, wherein After providing corresponding personalized services for the target patient according to the wish label, it further includes: Inputting the keywords into a pre-trained sentiment analysis model for recognition, obtaining the emotional state of the target patient, and formulating the time and content for adjusting the personalized service according to the emotional state; The sentiment analysis model includes an input layer, an embedding layer, a convolutional layer, a bidirectional LSTM layer, an attention layer, and an output layer; Among them, the input layer is used to receive the data to be processed and preprocess it to obtain a text sequence. The embedding layer is used to convert the text sequence into a vector representation. The convolutional layer is used to obtain the hidden layer expression of the vector representation. The bidirectional LSTM layer is used to capture long-distance context dependencies. The attention layer calculates the attention of a word to other words in the text sequence or words in other sequences, assigns different weights to each word in the sequence, and obtains a vector sequence with weight information. The output layer is used to perform a softmax function calculation on the vector sequence to obtain a scalar, and focus on key sentiment words according to the scalar to identify the sentiment trend of the target text.

10. The method for processing wish data based on patient information according to claim 7, wherein The wish label of the target patient is uploaded to the medical staff center for display and storage through encryption technology, and a visual dashboard is formed based on the wish labels of multiple patients. Specifically, it includes: Encrypt the wish label through the AES algorithm to obtain the encrypted wish label. Upload the encrypted wish label to the medical staff center through SSL or TLS technology, decrypt the encrypted wish label to obtain the wish label, and display and store the wish label; Obtain the wish labels of multiple patients within the same time range, calculate the similarity between the wish labels of multiple patients through the cosine similarity or JACCARD index algorithm, group and count the wish labels of multiple patients according to the similarity, and obtain a visual interface in the form of a chart.

Images