Triage system for outpatient and emergency treatment

Abstract

The invention belongs to the technical field of hospital triage systems, and provides an outpatient and emergency triage system, comprising: an information collection module collecting basic information and symptoms of a patient as original data, and preprocessing the data; the intelligent triage module classifies and predicts symptom data input by a patient through a medical knowledge graph and a machine learning algorithm; the triage result output module allocates departments to the patients according to the classification and prediction results of the intelligent triage module, and predicts the waiting time of the patients based on historical data; the interaction module displays and informs the patient according to the distribution information of the triage result output module and the predicted waiting time; through the intelligent triage module, the severity of the illness state of a patient can be rapidly judged, critical patients can be processed preferentially, so that distribution of medical resources is optimized, the waiting time of the patient in a hospital can be shortened through prediction and scheduling of waiting time of the patient, and the treatment efficiency is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of hospital triage systems, in particular to an outpatient and emergency triage system. Background Art

[0002] At present, the application of outpatient and emergency triage systems in hospitals has gradually become popular. The traditional triage method usually relies on manual judgment and experience. After entering the hospital, the patient consults the front desk nurse or doctor for a preliminary assessment;

[0003] The Chinese patent document with publication number CN212381238U discloses a contactless automatic triage system for patients visiting hospitals, including a server, a main switch, a floor switch, a triage device and a medical device. The medical device includes a pager, and the triage device includes a management device and a LoRa gateway connected to the management device, and also includes a LoRa terminal. The LoRa terminal includes a controller, a LoRa module connected to the controller, a power module and a prompt module. With this system, patients can hold the LoRa terminal in the rest area outside the hospital outpatient building or a leisure place outside the hospital to wait for medical treatment, so that people can wait in a dispersed manner, achieve contactless triage of patients, and avoid people waiting in the lobby;

[0004] However, the LoRa terminal configured by the invention is carried and used by individual users in the hospital. Since the users themselves may not comply with the epidemic prevention policy, there are still users who exceed the number specified by the epidemic prevention policy gathering in the same place, which results in the LoRa terminal still not being able to truly achieve uniform dispersion of the user population. And because triage work mainly relies on the experience and subjective judgment of medical staff, different nurses or doctors may have different assessments of the patient's condition. This inconsistency may cause patients who are in urgent need of medical services to be delayed in seeing a doctor, further affecting the safety and satisfaction of patients; at the same time, patients lack real-time understanding of their waiting time and progress while waiting for treatment, which increases unnecessary anxiety. The hospital also failed to effectively grasp the real-time dynamics of medical treatment and triage, resulting in management lags.

[0005] To this end, technicians in this field have proposed an outpatient and emergency triage system, which aims to achieve efficient processing and management of patient information, triage results, waiting time, etc. through intelligent means, which can improve patients' medical experience and improve the efficiency and quality of medical services. Summary of the invention

[0006] In order to solve the above technical problems, the present invention provides an outpatient and emergency triage system to solve the problems raised in the background technology.

[0007] An outpatient and emergency department triage system, comprising:

[0008] An information collection module is used to collect the patient's basic information and symptoms as raw data and pre-process the data;

[0009] Intelligent triage module, which is used to classify and predict symptom data input by patients through medical knowledge graph and machine learning algorithm;

[0010] A triage result output module, used to assign patients to departments according to the classification and prediction results of the intelligent triage module, and to predict the waiting time of patients based on historical data;

[0011] The interactive module is used to display and inform the patient of the allocation information and predicted waiting time of the triage result output module.

[0012] Preferably, the intelligent triage module infers the patient's condition by constructing a medical knowledge graph, and the construction of the medical knowledge graph includes the following steps:

[0013] Based on the preprocessed data, entity information is constructed and relationships are defined. The entities include diseases, symptoms, and drugs, and the relationships include causal relationships and correlation relationships. The graph model is used to represent:

[0014] G=(V,E)

[0015] Where G represents the knowledge graph, V represents a node set, including diseases, symptoms, and drugs, and E represents an edge set, which represents the relationship between entities. Each node v∈V is represented as a vector, and each edge e∈E is used to describe the relationship between nodes.

[0016] Use natural language processing technology to convert unstructured text, such as doctors' medical notes, into structured data and integrate it into the knowledge graph;

[0017] The Bayesian network is used for reasoning, the dependency relationship between different symptoms and diseases is modeled, and the reasoning mechanism is established, which is expressed as:

[0018]

[0019] Where D represents the disease, S represents the symptom, and P(SD) represents the probability of the symptom occurring under the condition of the disease.

[0020] Preferably, the intelligent triage module classifies and predicts the symptom data input by the patient through a machine learning algorithm. When predicting a disease, the symptom feature vector x = [s1, s2, ..., s n ], the classification is expressed as:

[0021]

[0022] Among them, P(D iX) means that under the given symptom feature X, the patient has the disease D i The posterior probability of

[0023] For symptom data S = {s1, s2, ..., s n} to make predictions and construct label data Y = {y1, y2, ..., y n}, where y contains the disease type and urgency of each sample, and the prediction model is constructed, whose input feature vector is Where d is the feature dimension, the hidden layer of the prediction model is:

[0024] h1=f(W1X+b1)

[0025] h2=f(W2h1+b2)

[0026] Its output layer is expressed as: Where f is the activation function, W i 、b i are the weights and biases of each layer respectively;

[0027] The loss function is constructed as Where N is the number of samples, C is the number of categories, and y ij is the true label of sample i belonging to category j, is the predicted probability output by the model;

[0028] Get the prediction result as

[0029] Preferably, the intelligent triage module performs multi-task learning by predicting probability, and for the disease type, the output layer is expressed as:

[0030]

[0031] According to the urgency, the output layer is expressed as:

[0032]

[0033] The prediction of urgency is then expressed as:

[0034]

[0035] Among them, C d , C e are the number of categories for disease and urgency, respectively.

[0036] Preferably, the triage result output module is used to assign patients to departments according to the classification and prediction results of the intelligent triage module, by establishing a rule base for each department, including the professional field, common diseases and corresponding treatment methods of each department, and establishing a symptom matching model according to each symptom s∈S and each department c∈C to calculate the matching score:

[0037]

[0038] Among them, ω s is the importance weight of symptom s, ∏(s∈c) represents the indicator function, which takes 1 when symptom s matches the department, otherwise it takes 0;

[0039] Define the triage level according to the patient's severity_level (P):

[0040]

[0041] Assign patients to appropriate departments.

[0042] Preferably, the triage result output module also predicts the patient's waiting time based on historical data, and constructs a waiting time prediction model by extracting features from historical data, which is expressed as:

[0043] Y t =φ1Y t-1 +φ2Y t-2 +...+φ p Y t-p +θ1ε t-1 +θ2ε t-2 +...+θ q ε t-q +μ

[0044] Among them, Y t represents the number of patients arriving or the waiting time at time t, φ1, φ2, ..., φ p is the autoregressive coefficient, which indicates the influence of the data of the past p time points on the current time point; θ1,θ2,...,θ p is the sliding average coefficient, which indicates the impact of the errors of the past q time points on the current time point; ε i is a white noise process, represents the random error term; μ is a constant term.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] 1. The present invention can quickly determine the severity of the patient's condition through the intelligent triage module, give priority to critically ill patients, thereby optimizing the allocation of medical resources, and by predicting and scheduling the patient's waiting time, it can reduce the patient's waiting time in the hospital and improve the efficiency of medical treatment.

[0047] 2. The present invention significantly improves the patient's medical experience through faster triage and reduced waiting time, increasing the patient's trust and satisfaction with the hospital. In addition, real-time status updates and waiting time estimates allow patients to clearly understand the medical situation, thereby reducing anxiety.

[0048] 3. The present invention can effectively improve the overall operational efficiency of the hospital by reducing the patient's stay time in the hospital and enhancing information flow. The outpatient and emergency triage system can continuously optimize medical processes such as triage, medical treatment, and referral based on data feedback, thereby improving the flexibility and adaptability of hospital operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is a block diagram of the outpatient and emergency triage system of the present invention. DETAILED DESCRIPTION

[0050] The following embodiments of the present invention are described in further detail in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0051] As attached Figure 1 As shown:

[0052] Embodiment: The present invention provides an outpatient and emergency department triage system, comprising:

[0053] The information collection module is used to collect the patient's basic information and symptoms as raw data and preprocess the data; the preprocessing includes cleaning, denoising and standardization, so as to ensure the accuracy and consistency of the data.

[0054] The intelligent triage module is used to classify and predict the symptom data input by the patient through the medical knowledge graph and machine learning algorithm; the intelligent triage module infers the patient's condition by constructing a medical knowledge graph. The construction of the medical knowledge graph includes the following steps:

[0055] Based on the preprocessed data, entity information is constructed and relationships are defined. Entities include diseases, symptoms, and drugs, and relationships include causal relationships and correlation relationships. The graph model is used to represent:

[0056] G=(V,E)

[0057] Among them, G represents the knowledge graph, V represents the node set, including diseases, symptoms and drugs, and E is the edge set, which represents the relationship between entities; each node v∈V is represented as a vector, and each edge e∈E is used to describe the relationship between nodes; the inference rules are constructed through the connection relationship between nodes.

[0058] Use natural language processing technology to convert unstructured text, such as doctors' medical notes, into structured data and integrate it into the knowledge graph;

[0059] The Bayesian network is used for reasoning, the dependency relationship between different symptoms and diseases is modeled, and the reasoning mechanism is established, which is expressed as:

[0060]

[0061] Among them, D represents the disease, S represents the symptom, and P(SD) represents the probability of the symptom occurring under the condition of the disease. At the same time, in the process of building the medical knowledge graph, close cooperation with clinicians is also required to ensure the accuracy and practicality of the knowledge graph. Through the effective integration and reasoning mechanism of data, the patient's condition can be effectively inferred and predicted. This method can greatly improve the efficiency and accuracy of medical services.

[0062] The intelligent triage module classifies and predicts the symptom data input by the patient through machine learning algorithms. When predicting the disease, the symptom feature vector x = [s1, s2, ..., s n ], the classification is expressed as:

[0063]

[0064] Among them, P(D i X) means that under the given symptom feature X, the patient has the disease D i The posterior probability of

[0065] For symptom data S = {s1, s2, ..., s n} to make predictions and construct label data Y = {y1, y2, ..., y n}, where y contains the disease type and urgency of each sample, and the prediction model is constructed, whose input feature vector is Where d is the feature dimension, the hidden layer of the prediction model is:

[0066] h1=f(W1X+b1)

[0067] h2=f(W2h1+b2)

[0068] Its output layer is expressed as: Where f is the activation function, W i 、b iare the weights and biases of each layer respectively;

[0069] The loss function is constructed as Where N is the number of samples, C is the number of categories, and y ij is the true label of sample i belonging to category j, is the predicted probability output by the model;

[0070] Get the prediction result as

[0071] The intelligent triage module performs multi-task learning by predicting probability. For the disease type, the output layer is expressed as:

[0072]

[0073] According to the urgency, the output layer is expressed as:

[0074]

[0075] The prediction of urgency is then expressed as:

[0076]

[0077] Among them, C d , C e are the number of categories of diseases and urgency, respectively. Through the training and prediction of deep learning models, the intelligent triage module can efficiently and accurately classify patients' symptom data, and then predict possible disease types and urgency, providing patients with better medical services.

[0078] The triage result output module is used to assign patients to departments according to the classification and prediction results of the intelligent triage module, and predict the waiting time of patients based on historical data; the triage result output module is used to assign patients to departments according to the classification and prediction results of the intelligent triage module, and establish a rule base for each department, including the professional field, common diseases and corresponding treatment methods of each department, and establish a symptom matching model according to each symptom s∈S and each department c∈C to calculate the matching score:

[0079]

[0080] Among them, ω s is the importance weight of symptom s, Π(s∈c) represents the indicator function, which takes 1 when symptom s matches the department, otherwise it takes 0;

[0081] Define the triage level according to the patient's severity_level (P):

[0082]

[0083] Assign patients to the corresponding departments. The system assigns patients to appropriate departments based on their symptoms and diseases, combined with the hospital's departmental division rules; and combined with intelligent triage and consultation recommendations, the system matches the doctor's expertise profile data, and recommends the doctor who is best at the field to the patient based on the doctor's expertise and the patient's needs, which helps improve the doctor-patient match and ensures that the patient receives accurate treatment.

[0084] The system will assess the severity of the patient's condition based on the patient's symptoms, signs and other information, and divide the patient into different levels based on the assessment results, such as emergency patients, non-emergency patients, etc.;

[0085] For emergency patients, the system will give priority to their visits to ensure that they receive timely treatment;

[0086] For non-emergency patients, the system will reasonably arrange the patient's consultation time based on the busyness of the department and the doctor's schedule.

[0087] The triage result output module also predicts the patient's waiting time based on historical data. By extracting features from historical data, a waiting time prediction model is constructed, which is expressed as:

[0088] Y t =φ1Y t-1 +φ2Y t-2 +...+φ p Y t-p +θ1ε t-1 +θ2ε t-2 +...+θ q ε t-q +μ

[0089] Among them, Y t represents the number of patients arriving or the waiting time at time t, φ1, φ2, ..., φ p is the autoregressive coefficient, which indicates the influence of the data of the past p time points on the current time point; θ1,θ2,...,θ p is the sliding average coefficient, which indicates the impact of the errors of the past q time points on the current time point; ε i is a white noise process, represents the random error term; μ is a constant term.

[0090] The interactive module is used to display and inform patients of the allocation information and predicted waiting time of the triage result output module. This module is designed to improve the user experience, allowing patients and doctors to easily get started and quickly complete related operations.

[0091] As can be seen from the above, the intelligent triage system can quickly determine the severity of the patient's condition and give priority to critically ill patients, thereby optimizing the allocation of medical resources; by predicting and scheduling the patient's waiting time, the patient's waiting time in the hospital can be reduced and the efficiency of medical treatment can be improved; intelligent allocation of doctors based on the patient's triage results can make the doctor's workload more balanced, while ensuring that patients in urgent need of medical services receive timely care. And through data sharing and information circulation, the system can promote collaboration between different departments and improve the overall level of patient care.

[0092] Importantly, it should be noted that the construction and arrangement of the present application shown in a plurality of different exemplary embodiments are only exemplary. Although only a few embodiments are described in detail in this disclosure, it should be readily understood by those who refer to this disclosure that many modifications are possible without substantially departing from the novel teachings and advantages of the subject matter described in this application. Without departing from the scope of the present invention, other replacements, improvements, changes and omissions may be made in the design, operating conditions and arrangement of the exemplary embodiments. Therefore, the present invention is not limited to specific embodiments, but extends to a variety of modifications still falling within the scope of the appended claims.

[0093] Additionally, in order to provide a concise description of exemplary embodiments, all features of an actual embodiment (ie, those features that are not relevant to the best mode presently contemplated for carrying out the invention or those that are not relevant to implementing the invention) may not be described.

[0094] It should be understood that in the development of any actual implementation, as in any engineering or design project, numerous implementation-specific decisions may be made. Such a development effort may be complex and time-consuming, but for those of ordinary skill having the benefit of this disclosure, without undue experimentation, the development effort will be a routine task of design, fabrication, and production.

[0095] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An outpatient and emergency department triage system, characterized in that: include: An information collection module is used to collect the patient's basic information and symptoms as raw data and pre-process the data; Intelligent triage module, which is used to classify and predict symptom data input by patients through medical knowledge graph and machine learning algorithm; A triage result output module, used to assign patients to departments according to the classification and prediction results of the intelligent triage module, and to predict the waiting time of patients based on historical data; The interactive module is used to display and inform the patient of the allocation information and predicted waiting time of the triage result output module.

2. An outpatient and emergency triage system as claimed in claim 1, characterized in that: The intelligent triage module infers the patient's condition by constructing a medical knowledge graph, and the construction of the medical knowledge graph includes the following steps: Based on the preprocessed data, entity information is constructed and relationships are defined. The entities include diseases, symptoms, and drugs, and the relationships include causal relationships and correlation relationships. The graph model is used to represent: G=(V,E) Where G represents the knowledge graph, V represents a node set, including diseases, symptoms, and drugs, and E represents an edge set, which represents the relationship between entities. Each node v∈V is represented as a vector, and each edge e∈E is used to describe the relationship between nodes. Use natural language processing technology to convert unstructured text, such as doctors’ medical notes, into structured data and integrate it into the knowledge graph; The Bayesian network is used for reasoning, the dependency relationship between different symptoms and diseases is modeled, and the reasoning mechanism is established, which is expressed as: Where D represents the disease, S represents the symptom, and P(SD) represents the probability of the symptom occurring under the condition of the disease.

3. An outpatient and emergency triage system as claimed in claim 2, characterized in that: The intelligent triage module classifies and predicts the symptom data input by the patient through a machine learning algorithm. When predicting a disease, the symptom feature vector x = [s1, s2, ..., s n ], the classification is expressed as: Among them, P(D i X) means that under the given symptom feature X, the patient has the disease D i The posterior probability of For symptom data S = {s1, s2, ..., s n } to make predictions and construct label data Y = {y1, y2, ..., y n }, where y contains the disease type and urgency of each sample, and the disease prediction model is constructed, and its input feature vector is Where d is the feature dimension, the hidden layer of the prediction model is: h1=f(W1X+b1) h2=f(W2h1+b2) Its output layer is expressed as: Where f is the activation function, W i , b i are the weights and biases of each layer respectively; The loss function is constructed as Where N is the number of samples, C is the number of categories, and y ij is the true label of sample i belonging to category j, is the predicted probability output by the model; Get the prediction result as 4. An outpatient and emergency triage system as claimed in claim 3, characterized in that: The intelligent triage module performs multi-task learning by predicting probability. For the disease type, the output layer is expressed as: According to the urgency, the output layer is expressed as: The prediction of urgency is then expressed as: Among them, C d , C e are the number of categories for disease and urgency, respectively.

5. The outpatient and emergency triage system according to claim 1, characterized in that: The triage result output module is used to assign patients to departments according to the classification and prediction results of the intelligent triage module. By establishing a rule base for each department, including the professional field, common diseases and corresponding treatment methods of each department, a symptom matching model is established according to each symptom s∈S and each department c∈C to calculate the matching score: Among them, ω s is the importance weight of symptom s, ∏(s∈c) represents the indicator function, which takes 1 when symptom s matches the department, otherwise it takes 0; Define the triage level according to the patient's severity_level (P): Assign patients to appropriate departments.

6. An outpatient and emergency triage system as claimed in claim 5, characterized in that: The triage result output module also predicts the patient's waiting time based on historical data. By extracting features from historical data, a waiting time prediction model is constructed, which is expressed as: Y t =φ1Y t-1 +φ2Y t-2 +...+φ p Y t-p +θ1ε t-1 +θ2ε t-2 +...+θ q e t-q +m Among them, Y t represents the number of patients arriving or the waiting time at time t, φ1, φ2, ..., φ p is the autoregressive coefficient, which indicates the influence of the data of the past p time points on the current time point; θ1,θ2,...,θ p is the sliding average coefficient, which indicates the impact of the errors of the past q time points on the current time point; ε i is a white noise process, represents the random error term; μ is a constant term.

Images