Intelligent triage method suitable for children oral soft tissue injury

Abstract

The invention discloses an intelligent triage method suitable for oral soft tissue injury of children, which relates to the technical field of medical image processing, and comprises the following steps: collecting the age and gender of a patient through a patient information form, extracting the pain degree and injury time information from the symptom description of the patient by using a natural language processing technology, and calculating the pain degree and injury time information by using an X-ray machine. Collecting oral image data and extracting input features; based on children of different age groups, respectively setting threshold ranges of input features; and training a support vector machine model to realize automatic classification of oral emergency surgical injury levels. According to the intelligent triage method for oral emergency surgical injury, the triage efficiency and accuracy are remarkably improved, the method is particularly suitable for child patients, personal errors are reduced through an automatic process, it is ensured that emergency injury is rapidly responded, meanwhile, doctor-patient communication is enhanced through visual data display, and the medical experience is improved. And powerful support is provided for formulating a personalized treatment scheme.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and in particular to an intelligent triage method suitable for children's oral soft tissue injuries. Background Art

[0002] Oral emergency surgical injuries refer to a series of acute traumas that occur in the oral and maxillofacial regions. These injuries are often accompanied by severe pain and functional impairment, requiring timely and effective medical intervention. Such injuries include, but are not limited to, broken, dislocated or fallen teeth, which may be caused by external forces acting directly on the teeth or jaws. In addition, jaw fractures are also a common type of injury in oral emergency surgery, usually caused by traffic accidents, falls or other violent incidents, which may result in limited chewing, breathing and speech functions. Soft tissue injuries should not be ignored, such as lacerations of the gums, lips and tongue, which can cause bleeding, swelling and pain, and in severe cases may affect the patient's eating and speech. When dealing with these injuries, oral emergency surgeons will quickly assess the injuries and take necessary treatment measures, such as suturing wounds, fixing fracture sites, and controlling bleeding and preventing infection to ensure the patient's oral health and overall well-being.

[0003] In order to solve the problem of insufficient grading and diagnosis of oral emergency surgical injuries in children of different age groups, the existing technology mainly relies on the clinical experience and subjective judgment of medical staff to deal with it. However, this method often leads to misdiagnosis or inaccurate triage, especially for younger children with limited expressive ability. The specific circumstances and severity of their oral injuries may be difficult to accurately describe, which leads to untimely treatment or waste of medical resources. In actual operation, medical staff may need to infer the extent of injury based on the child’s crying, facial expression and simple physical examination. This method is not only inefficient, but also difficult to guarantee accuracy. In addition, differences in experience and skill levels among different medical staff may also lead to inconsistent triage results. Therefore, in order to improve the triage accuracy and efficiency of children’s oral emergency surgical injuries, an intelligent triage method suitable for children’s oral soft tissue injuries is proposed. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent triage method for children's oral soft tissue injuries to solve the problems raised in the above-mentioned background technology.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: an intelligent triage method for children's oral soft tissue injuries, comprising the following steps:

[0006] S1. Collect the patient's age and gender through the patient information form, use natural language processing technology to extract the pain level and injury time information from the patient's symptom description, and use ultrasonic imaging to collect oral soft tissue image data and extract input features;

[0007] S2. Based on children of different age groups, set the threshold range of input features respectively;

[0008] S3, training support vector machine model to achieve automatic classification of oral emergency surgery injury levels;

[0009] S4. Based on the classification results of oral emergency surgery injury levels, recommend corresponding treatment plans for target pediatric patients;

[0010] S5. Through communication protocols and technologies, data transmission between modules is realized, and visualization tools are used to display the patient's oral characteristic data and examination results in the form of charts.

[0011] A further improvement of the technical solution of the present invention is that in S1, the process of collecting the patient's age and gender through the patient information form and extracting the pain degree and injury time information from the patient's symptom description using natural language processing technology includes:

[0012] Medical staff manually enter a form containing the patient's basic personal information, collect the patient's age and gender, convert the patient's text description into a computer-processable data format, and use a text analysis algorithm to identify the patient's oral pain level and specific time of injury.

[0013] A further improvement of the technical solution of the present invention is that in S1, the process of collecting oral soft tissue image data using ultrasonic imaging includes:

[0014] A coupling agent is applied in the patient's mouth. The piezoelectric crystal in the ultrasonic probe generates high-frequency ultrasonic pulses after being stimulated by electrical signals. The high-frequency ultrasonic pulses are transmitted to the patient's oral soft tissue through the coupling agent. When the ultrasonic wave encounters the interface of tissues with different densities, it is reflected. The probe receives the reflected ultrasonic signal, converts the reflected ultrasonic signal into an electrical signal, amplifies and digitizes it, and converts the digitized reflected signal into a two-dimensional image to show the structural characteristics of the patient's oral soft tissue.

[0015] A further improvement of the technical solution of the present invention is that in S1, the process of preprocessing the collected age, gender, pain level, injury time and oral soft tissue image data includes:

[0016] The actual age of the patient was directly used, and the gender information was converted into binary code, with male as 0 and female as 1. The pain level was divided into three levels: mild, moderate, and severe, and converted into numerical data, with mild as 1, moderate as 2, and severe as 3. The injury time was recorded in hours, and the digital image generated by the ultrasound imaging machine was standardized, the contrast and brightness of the image were adjusted, and the mean filtering algorithm was applied to remove noise in the image.

[0017] A further improvement of the technical solution of the present invention is that in S1, the process of extracting features from the pre-processed age, gender, pain level, injury time and oral soft tissue image data includes:

[0018] Extract the gray value of each pixel from the standardized image, calculate the average gray value and the standard deviation of the gray value, and define the gray level co-occurrence matrix Among them, I (m,n) It represents the gray value of the image at position (m,n), Δm and Δn represent the displacement in the row and column directions respectively, and the texture features of the image of size M×N are extracted by the gray level co-occurrence matrix. The texture features include angular second moment, contrast and correlation. The shape features in the image are extracted by morphological operations. The morphological features include area, perimeter and circularity.

[0019] A further improvement of the technical solution of the present invention is that in S2, the process of setting the average gray value and area threshold range based on children of different age groups includes:

[0020] Children aged 3 to 12 are divided into preschool children aged 3-6 and school-age children aged 6-12 respectively. The threshold ranges of the average grayscale value and area of ​​preschool children and school-age children are set. If the input feature data collected from children of different age groups exceeds the corresponding threshold range, the sample data is marked as abnormal.

[0021] A further improvement of the technical solution of the present invention is that in S3, the process of training the support vector machine model to realize automatic classification of the oral emergency surgery injury level includes:

[0022] Age, average gray value and area are used as input features, combined into a feature vector, and passed to the support vector machine model. The data set is divided into a training set, a data set and a test set, and each feature is converted into a zero mean and unit variance form;

[0023] Training the support vector machine model includes selecting a radial basis function and minimizing an objective function Among them, w is the weight vector, b is the bias term, C is the penalty coefficient, ξ i is the slack variable, and the problem is transformed into a dual problem by the Lagrange multiplier method. Among them, α i is the Lagrange multiplier, y i is the sample label, K(x i ,x j ) is the kernel function, and the optimal model parameters are determined through 5-fold cross validation. The accuracy of each validation is calculated and the average is taken as the evaluation index. The model parameter values ​​are adjusted until the optimal parameter combination is found. The model performance is evaluated on the test set, the accuracy is calculated, the confusion matrix is ​​generated, and the precision, recall and F1 score are introduced as evaluation indicators to evaluate the classification effect of the model.

[0024] A further improvement of the technical solution of the present invention is that in S3, the process of using the support vector machine model to predict the patient's oral injury level includes:

[0025] The feature vector of the new patient including age, average gray value and area was standardized and input into the trained support vector machine model. According to the output values ​​of each decision function corresponding to the radial basis function in the support vector machine model, the category with the highest score was selected as the final prediction result, and the corresponding oral injury level was output. The oral injury levels included mild, moderate and severe. When the output value of the decision function for preschool children was >0.7, -0.3<output value ≤0.7 and output value ≤-0.3, and the output value of the decision function for school-age children was >0.5, -0.5<output value ≤0.5 and output value ≤-0.5, they corresponded to mild, moderate and severe injury levels respectively.

[0026] A further improvement of the technical solution of the present invention is that in S4, the process of recommending a corresponding treatment plan for the target child patient based on the classification result of the oral emergency surgery injury level includes:

[0027] If the classification result is a mild injury, the target child patient's wound will be cleaned, disinfectant and antibiotic ointment will be used, and the parents will closely observe and follow up regularly. If the classification result is a moderate injury, the target child patient's wound will be sutured under local anesthesia, antibiotics will be used, and the doctor will arrange a follow-up examination to ensure good healing. If the classification result is a severe injury, the target child patient will undergo surgery to repair the damaged tissue, antibiotics will be used, and the child will be hospitalized for observation to ensure there are no complications, and psychological support will be provided.

[0028] A further improvement of the technical solution of the present invention is that in S5, the process of realizing data transmission through communication protocols and technologies and using visualization tools to display the patient's oral characteristic data and examination results in the form of charts includes:

[0029] Convert the preprocessed feature data into JSON format, use HTTPS protocol for data transmission, and send the collected data to the central processing unit through HTTP request;

[0030] According to the patient's specific situation, use bar charts to show the distribution of different feature values, use line graphs to show the changing trend of time series data, use heat maps to show the grayscale value distribution and texture features to assist doctors in identifying abnormal areas, use pie charts to show the proportion of different injury levels, and show the generated charts and graphs to doctors and patients' families to help them understand the patient's examination results and treatment plans.

[0031] Due to the adoption of the above technical solution, the present invention has the following technical advances compared with the prior art:

[0032] 1. The present invention provides an intelligent triage method for children's oral soft tissue injuries. The method can comprehensively and accurately obtain the patient's age, gender, pain level, injury time, oral soft tissue image data and other key information by comprehensively using patient information forms, natural language processing technology, ultrasonic imaging image acquisition and analysis, etc., laying a solid foundation for subsequent accurate triage.

[0033] 2. The present invention provides an intelligent triage method for oral soft tissue injuries in children. It sets specific input feature threshold ranges for children of different age groups and trains a support vector machine model for automatic classification, which effectively improves the accuracy and pertinence of triage. It is especially suitable for child patients with limited expressive ability and reduces the risk of misdiagnosis and missed diagnosis.

[0034] 3. The present invention provides an intelligent triage method for children's oral soft tissue injuries. Through the application of data transmission and visualization tools, it can display the patient's oral characteristic data and examination results in real time, providing medical staff with an intuitive and clear diagnostic basis, which helps to quickly formulate and implement effective treatment plans, thereby improving the efficiency and quality of emergency surgical injury treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0036] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] Examples, such as Figure 1 As shown, the present invention provides an intelligent triage method for children's oral soft tissue injuries, comprising the following steps:

[0039] S1. Collect the patient's age and gender through the patient information form. Use natural language processing technology to extract the pain level and injury time information from the patient's symptom description. Use an ultrasonic imaging machine to collect oral soft tissue image data and extract input features. Medical staff manually enter the form containing the patient's basic personal information, collect the patient's age and gender, convert the patient's text description into a computer-processable data format, and use the text analysis algorithm to identify the patient's oral pain level and specific injury time. Apply coupling agent to the patient's mouth. The piezoelectric crystal in the ultrasonic probe generates high-frequency ultrasonic pulses after being stimulated by electrical signals. The high-frequency ultrasonic pulses are transmitted to the patient's oral soft tissues through the coupling agent. When the ultrasonic wave encounters the interface of tissues with different densities, it is reflected and the probe receives the reflected The ultrasonic signal emitted by the ultrasonic imaging machine is converted into an electrical signal, and the reflected ultrasonic signal is converted into an electrical signal, which is amplified and digitized. The digitized reflected signal is converted into a two-dimensional image to show the structural characteristics of the patient's oral soft tissue. The actual age value of the patient is directly used, and the gender information is converted into a binary code, with males being 0 and females being 1. The pain level is divided into three levels: mild, moderate, and severe, and converted into numerical data, with mild being 1, moderate being 2, and severe being 3. The injury time is recorded in hours. The digital image generated by the ultrasonic imaging machine is standardized, the contrast and brightness of the image are adjusted, and the mean filtering algorithm is applied to remove noise in the image. The grayscale value of each pixel is extracted from the standardized image, and the average grayscale value and the standard deviation of the grayscale value are calculated. The grayscale co-occurrence matrix is ​​defined. Among them, I (m,n) Represents the gray value of the image at position (m, n), Δm and Δn represent the displacement in the row and column directions respectively, and extracts the texture features of the image of size M×N through the gray level co-occurrence matrix. The texture features include angular second moment, contrast and correlation. The shape features in the image are extracted by morphological operations. The morphological features include area, perimeter and circularity;

[0040] S2. Set the threshold range of input features based on children of different age groups:

[0041] 3-6 years old preschool age: 65-75, 18-22, 0.65-0.75, 0.28-0.32, 0.55-0.65, 650-750mm 2 , 130-150mm, 0.55-0.65;

[0042] 6-12 years old school age: 75-85, 23-27, 0.55-0.65, 0.38-0.42, 0.45-0.55, 750-850mm 2 , 150-170mm, 0.5-0.6;

[0043] If the input feature data collected from children of different age groups exceeds the corresponding threshold range, the sample data is marked as abnormal;

[0044] S3, training the support vector machine model to achieve automatic classification of the level of oral emergency surgery injuries, taking age, average gray value and area as input features, combining them into a feature vector, passing them to the support vector machine model, dividing the data set into a training set, a data set and a test set, and converting each feature into a zero mean and unit variance form;

[0045] Training the support vector machine model includes selecting a radial basis function and minimizing an objective function Among them, w is the weight vector, b is the bias term, C is the penalty coefficient, ξ i is the slack variable, and the problem is transformed into a dual problem by the Lagrange multiplier method. Among them, α i is the Lagrange multiplier, y i is the sample label, K(x i ,x j ) is the kernel function, the optimal model parameters are determined by 5-fold cross validation, the accuracy of each validation is calculated and the average is taken as the evaluation index, the model parameter values ​​are adjusted until the optimal parameter combination is found, the model performance is evaluated on the test set, the accuracy is calculated, the confusion matrix is ​​generated, the precision, recall and F1 scores are introduced as evaluation indicators to evaluate the model classification effect, the feature vector of the new patient including age, average gray value and area is standardized and input into the trained support vector machine model, according to the output values ​​of each decision function corresponding to the radial basis function in the support vector machine model, the category with the highest score is selected as the final prediction result, and the corresponding oral injury level is output, the oral injury level includes mild, moderate and severe, when the output value of the decision function of preschool children is>0.7, -0.3<output value≤0.7 and output value≤-0.3, the output value of the decision function of school-age children is>0.5, -0.5<output value≤0.5 and output value≤-0.5, it corresponds to mild, moderate and severe injury levels respectively;

[0046] S4. Based on the classification results of the oral emergency surgery injury level, recommend corresponding treatment plans for the target child patients. If the classification result is a mild injury, the target child patients will have their wounds cleaned, disinfectants and antibiotic ointments used, and parents will closely observe and conduct regular follow-up visits. If the classification result is a moderate injury, the target child patients will have their wounds sutured under local anesthesia and antibiotics used. The doctor will arrange a follow-up to ensure good healing. If the classification result is a severe injury, the target child patients will undergo surgery to repair the damaged tissue, use antibiotics, and be hospitalized for observation to ensure there are no complications, and psychological support will be provided.

[0047] S5. Through communication protocols and technologies, data transmission between modules is realized. Visual tools are used to display the patient's oral feature data and examination results in the form of charts. The pre-processed feature data is converted into JSON format. The HTTPS protocol is used for data transmission, and the collected data is sent to the central processing unit through HTTP requests. According to the specific situation of the patient, a bar chart is used to display the distribution of different feature values, a line chart is used to display the changing trend of time series data, and a heat map is used to display the gray value distribution and texture features to assist doctors in identifying abnormal areas. A pie chart is used to display the proportion of different injury levels. The generated charts and graphs are displayed to doctors and patients' families to help them understand the patient's examination results and treatment plans.

[0048] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An intelligent triage method for children's oral soft tissue injuries, characterized in that: The following steps are involved: S1. Collect the patient's age and gender through the patient information form, use natural language processing technology to extract the pain level and injury time information from the patient's symptom description, and use ultrasonic imaging to collect oral soft tissue image data and extract input features; S2. Based on children of different age groups, set the threshold range of input features respectively; S3, training support vector machine model to achieve automatic classification of oral emergency surgery injury levels; S4. Based on the classification results of oral emergency surgery injury levels, recommend corresponding treatment plans for target pediatric patients; S5. Through communication protocols and technologies, data transmission between modules is realized, and visualization tools are used to display the patient's oral characteristic data and examination results in the form of charts.

2. The intelligent triage method for children's oral soft tissue injuries according to claim 1, characterized in that: In S1, the process of collecting the patient's age and gender through the patient information form and extracting the pain degree and injury time information from the patient's symptom description using natural language processing technology includes: Medical staff manually enter a form containing the patient's basic personal information, collect the patient's age and gender, convert the patient's text description into a computer-processable data format, and use a text analysis algorithm to identify the patient's oral pain level and specific time of injury.

3. The intelligent triage method for children's oral soft tissue injuries according to claim 2, characterized in that: In S1, the process of collecting oral soft tissue image data using ultrasonic imaging includes: A coupling agent is applied in the patient's mouth. The piezoelectric crystal in the ultrasonic probe generates high-frequency ultrasonic pulses after being stimulated by electrical signals. The high-frequency ultrasonic pulses are transmitted to the patient's oral soft tissue through the coupling agent. When the ultrasonic wave encounters the interface of tissues with different densities, it is reflected. The probe receives the reflected ultrasonic signal, converts the reflected ultrasonic signal into an electrical signal, amplifies and digitizes it, and converts the digitized reflected signal into a two-dimensional image to show the structural characteristics of the patient's oral soft tissue.

4. The intelligent triage method for children's oral soft tissue injuries according to claim 3, characterized in that: In S1, the process of preprocessing the collected age, gender, pain level, injury time and oral soft tissue image data includes: The actual age of the patient was directly used, and the gender information was converted into binary code, with male as 0 and female as 1. The pain level was divided into three levels: mild, moderate, and severe, and converted into numerical data, with mild as 1, moderate as 2, and severe as 3. The injury time was recorded in hours, and the digital image generated by the ultrasound imaging machine was standardized, the contrast and brightness of the image were adjusted, and the mean filtering algorithm was applied to remove noise in the image.

5. The intelligent triage method for children's oral soft tissue injuries according to claim 4, characterized in that: In S1, the process of extracting features from the pre-processed age, gender, pain level, injury time, and oral soft tissue image data includes: Extract the gray value of each pixel from the standardized image, calculate the average gray value and the standard deviation of the gray value, and define the gray level co-occurrence matrix Among them, I (m,n) It represents the gray value of the image at position (m,n), Δm and Δn represent the displacement in the row and column directions respectively, and the texture features of the image of size M×N are extracted by the gray level co-occurrence matrix. The texture features include angular second moment, contrast and correlation. The shape features in the image are extracted by morphological operations. The morphological features include area, perimeter and circularity.

6. The intelligent triage method for children's oral soft tissue injuries according to claim 5, characterized in that: In S2, the process of setting the average gray value and area threshold range based on children of different age groups includes: Children aged 3 to 12 are divided into preschool children aged 3-6 and school-age children aged 6-12 respectively. The threshold ranges of the average grayscale value and area of ​​preschool children and school-age children are set. If the input feature data collected from children of different age groups exceeds the corresponding threshold range, the sample data is marked as abnormal.

7. The intelligent triage method for children's oral soft tissue injuries according to claim 6, characterized in that: In S3, the process of training the support vector machine model to realize automatic classification of oral emergency surgery injury levels includes: Age, average gray value and area are used as input features, combined into a feature vector, and passed to the support vector machine model. The data set is divided into a training set, a data set and a test set, and each feature is converted into a zero mean and unit variance form; Training the support vector machine model includes selecting a radial basis function and minimizing an objective function Among them, w is the weight vector, b is the bias term, C is the penalty coefficient, ξ i is the slack variable, and the problem is transformed into a dual problem by the Lagrange multiplier method. Among them, α i is the Lagrange multiplier, y i is the sample label, K(x i ,x j ) is the kernel function, and the optimal model parameters are determined through 5-fold cross validation. The accuracy of each validation is calculated and the average is taken as the evaluation index. The model parameter values ​​are adjusted until the optimal parameter combination is found. The model performance is evaluated on the test set, the accuracy is calculated, the confusion matrix is ​​generated, and the precision, recall and F1 score are introduced as evaluation indicators to evaluate the classification effect of the model.

8. The intelligent triage method for children's oral soft tissue injuries according to claim 7, characterized in that: In S3, the process of using the support vector machine model to predict the patient's oral injury level includes: The feature vector of the new patient including age, average gray value and area was standardized and input into the trained support vector machine model. According to the output values ​​of each decision function corresponding to the radial basis function in the support vector machine model, the category with the highest score was selected as the final prediction result, and the corresponding oral injury level was output. The oral injury levels included mild, moderate and severe. When the output value of the decision function for preschool children was >0.7, -0.3<output value ≤0.7 and output value ≤-0.3, and the output value of the decision function for school-age children was >0.5, -0.5<output value ≤0.5 and output value ≤-0.5, they corresponded to mild, moderate and severe injury levels respectively.

9. The intelligent triage method for children's oral soft tissue injuries according to claim 8, characterized in that: In S4, based on the classification results of oral emergency surgery injury levels, the process of recommending corresponding treatment plans for target pediatric patients includes: If the classification result is a mild injury, the target child patient's wound will be cleaned, disinfectant and antibiotic ointment will be used, and the parents will closely observe and follow up regularly. If the classification result is a moderate injury, the target child patient's wound will be sutured under local anesthesia, antibiotics will be used, and the doctor will arrange a follow-up examination to ensure good healing. If the classification result is a severe injury, the target child patient will undergo surgery to repair the damaged tissue, antibiotics will be used, and the child will be hospitalized for observation to ensure there are no complications, and psychological support will be provided.

10. The intelligent triage method for children's oral soft tissue injuries according to claim 9, characterized in that: In S5, the process of realizing data transmission through communication protocols and technologies and using visualization tools to display the patient's oral characteristic data and examination results in the form of charts includes: Convert the preprocessed feature data into JSON format, use HTTPS protocol for data transmission, and send the collected data to the central processing unit through HTTP request; According to the patient's specific situation, use bar charts to show the distribution of different feature values, use line graphs to show the changing trend of time series data, use heat maps to show the grayscale value distribution and texture features to assist doctors in identifying abnormal areas, use pie charts to show the proportion of different injury levels, and show the generated charts and graphs to doctors and patients' families to help them understand the patient's examination results and treatment plans.

Images