Artificial neural network-based intelligent assessment and decision-making system for traumatic patient emergency treatment

Abstract

The invention relates to an artificial neural network-based traumatic patient first-aid intelligent evaluation and decision-making system, and relates to the technical field of artificial intelligence. Data of the first-aid trauma patient is acquired through the data acquisition device, the first-aid treatment difficulty of the first-aid trauma patient is evaluated through the evaluation device, and a treatment scheme of the first-aid trauma patient is decided through the decision-making device. Wherein the evaluation equipment evaluates the first-aid treatment difficulty by using a pre-trained evaluation model and realizes accurate and efficient evaluation based on an artificial neural network, and the decision-making equipment determines a treatment scheme of a first-aid trauma patient by using a pre-trained decision-making model and realizes accurate and efficient decision-making based on the artificial neural network. Therefore, according to the technical scheme, intelligent assessment and decision-making of trauma patient first aid based on the artificial neural network can be achieved, and then the treatment effect of a first-aid treatment doctor on the trauma patient can be improved in an auxiliary mode.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, and in particular to an intelligent assessment and decision-making system for emergency treatment of trauma patients based on artificial neural networks. Background Art

[0002] With the development of artificial intelligence technology, the application of artificial neural networks has also developed. Artificial neural networks are also used in medical scenarios. For example, artificial neural networks can be used to analyze medical data to obtain corresponding data analysis results.

[0003] In emergency medical scenarios, trauma patients often appear, and these patients can be called emergency trauma patients. Currently, some clinical databases can assist doctors in treating emergency trauma patients. Summary of the invention

[0004] The purpose of the present disclosure is to provide an intelligent assessment and decision-making system and method for first aid of trauma patients based on artificial neural networks.

[0005] In order to achieve the above-mentioned objectives, in a first aspect, the present disclosure provides an intelligent assessment and decision-making system for emergency treatment of trauma patients based on an artificial neural network, comprising: a data acquisition device, configured to obtain emergency treatment data of trauma patients; an assessment device, communicatively connected to the data acquisition device, configured to determine an assessment result of the emergency treatment of trauma patients based on the emergency treatment data of trauma patients, a clinical trauma database and a pre-trained assessment model, wherein the assessment result of the emergency treatment of trauma patients is used to characterize the difficulty of emergency treatment of trauma patients; a decision-making device, communicatively connected to the data acquisition device and the assessment device respectively, configured to obtain information of emergency treatment doctors, and determine an emergency treatment plan for trauma patients based on the assessment result of the emergency treatment of trauma patients, the information of emergency treatment doctors, a clinical trauma treatment database and a pre-trained decision-making model.

[0006] Optionally, the data acquisition device includes: a first acquisition device, configured to acquire physical description information of emergency trauma patients, wherein the physical description information is used to describe the physical reaction of the emergency trauma patients caused by trauma; a second acquisition device, configured to acquire trauma description information of emergency trauma patients, wherein the trauma description information is used to describe the wound condition of the emergency trauma patients; and a processing device, configured to integrate the physical description information and the trauma description information to obtain the emergency trauma patient data.

[0007] Optionally, the emergency trauma patient data includes physical description information and trauma description information of the emergency trauma patient, the clinical trauma database includes: multiple clinical trauma data, each clinical trauma data includes clinical physical description information and clinical trauma description information, and the evaluation device is further configured to: determine a first information weight based on the physical description information of the emergency trauma patient and the multiple clinical physical description information; determine a second information weight based on the trauma description information of the emergency trauma patient and the multiple clinical trauma description information; input the physical description information of the emergency trauma patient, the first information weight, the trauma description information of the emergency trauma patient and the second information weight into the pre-trained evaluation model to obtain the emergency trauma patient evaluation result output by the pre-trained evaluation model.

[0008] Optionally, the first information weight is used to characterize the complexity or rarity of the physical description information of the emergency trauma patient, and the second information weight is used to characterize the complexity or rarity of the trauma description information of the emergency trauma patient, and the evaluation device is further configured to: if target clinical physical description information matching the physical description information of the emergency trauma patient is determined from the multiple clinical physical description information, the first information weight is determined according to the complexity feature or rarity feature of the target clinical physical description information in the multiple clinical physical description information; if target clinical physical description information matching the physical description information of the emergency trauma patient is not determined from the multiple clinical physical description information, the first information weight is determined according to the complexity feature or rarity feature of the target clinical physical description information in the multiple clinical physical description information. The first information weight is determined according to the similarity between the body description information of the trauma patient and the multiple clinical body description information; and / or, if target clinical trauma description information matching the trauma description information of the emergency trauma patient is determined from the multiple clinical trauma description information, the second information weight is determined according to the complexity characteristics or rarity characteristics of the target clinical trauma description information in the multiple clinical trauma description information; if target clinical trauma description information matching the trauma description information of the emergency trauma patient is not determined from the multiple clinical trauma description information, the second information weight is determined according to the similarity between the trauma description information of the emergency trauma patient and the multiple clinical trauma description information.

[0009] Optionally, the evaluation device is further configured to: obtain a first training data set, the first training data set includes multiple first training samples, each of the first training samples includes: sample body description information, weight corresponding to the sample body description information, sample trauma description information, weight corresponding to the sample trauma description information and an evaluation label, and the evaluation label is used to characterize the difficulty of emergency treatment; determine a second training data set based on the first training data set, the second training data set includes multiple second training samples, each of the second training samples includes: sample body description information, sample trauma description information and an evaluation label; train the evaluation model to be trained based on the second training data set to obtain an initially trained evaluation model; optimize the initially trained evaluation model based on the first training data set to obtain the pre-trained evaluation model.

[0010] Optionally, the clinical trauma treatment database includes: multiple clinical trauma treatment plans, and the decision-making device is further configured to: evaluate the emergency treatment doctor according to the emergency treatment doctor information, and determine the emergency treatment doctor evaluation result, wherein the emergency treatment doctor evaluation result is used to characterize the emergency treatment level of the emergency treatment doctor; input the multiple clinical treatment plans into the pre-trained decision model to obtain the clinical treatment difficulty labels and clinical doctor treatment level labels corresponding to each of the clinical trauma treatment plans output by the pre-trained decision model; determine the emergency trauma patient treatment plan according to the clinical treatment difficulty labels and clinical doctor treatment level labels corresponding to each of the clinical trauma treatment plans, the emergency trauma patient evaluation result and the emergency treatment doctor evaluation result.

[0011] Optionally, the pre-trained decision model includes a first module and a second module, and the decision device is further configured to: obtain a third training data set, the third training data set includes multiple third training samples, each of the third training samples includes: a sample trauma treatment plan, a sample trauma treatment plan feature, a treatment difficulty label, and a doctor's treatment level label; determine a fourth training data set and a fifth training data set based on the third training data set, the fourth training data set includes multiple fourth training samples, each of the fourth training samples includes a sample trauma treatment plan and a sample trauma treatment plan feature, and the fifth training data set includes multiple fifth training samples, each of the fifth training samples includes a sample trauma treatment plan, a treatment difficulty label, and a doctor's treatment level label; train the first module of the decision model to be trained according to the fourth training data set to obtain a decision model for initial training; train the second module of the decision model for initial training according to the fifth training data set to obtain the pre-trained decision model.

[0012] Optionally, the decision-making device is further configured to: determine a first clinical trauma treatment plan based on the emergency trauma patient assessment result and the clinical treatment difficulty labels corresponding to each of the clinical trauma treatment plans, the clinical treatment difficulty label corresponding to the first clinical trauma treatment plan being higher than the emergency treatment difficulty represented by the emergency trauma patient assessment result; determine a second clinical trauma treatment plan based on the emergency treatment doctor assessment result and the clinical doctor treatment level labels corresponding to each of the clinical trauma treatment plans, the clinical doctor treatment level label corresponding to the second clinical trauma treatment plan being lower than the emergency treatment level represented by the emergency treatment doctor assessment result; determine the emergency trauma patient treatment plan based on the first clinical trauma treatment plan and the second clinical trauma treatment plan.

[0013] Optionally, the intelligent assessment and decision-making system for emergency treatment of trauma patients also includes: a visualization device, which is communicatively connected to the data acquisition device, the assessment device and the decision-making device respectively, and is configured to: respectively display the emergency treatment data of trauma patients, the emergency treatment assessment results of trauma patients and the emergency treatment plan of trauma patients; in response to receiving an auxiliary treatment request initiated by the emergency treatment doctor, play a locally stored auxiliary treatment video recorded by a preset doctor.

[0014] In a second aspect, the present disclosure provides an intelligent assessment and decision-making method for emergency treatment of trauma patients based on an artificial neural network, comprising: obtaining emergency treatment data of trauma patients; determining an assessment result of the emergency treatment of trauma patients based on the emergency treatment data of trauma patients, a clinical trauma database and a pre-trained assessment model, wherein the assessment result of the emergency treatment of trauma patients is used to characterize the difficulty of emergency treatment of emergency treatment of trauma patients; obtaining emergency treatment doctor information; determining an emergency treatment plan for trauma patients based on the emergency treatment patient assessment result, the emergency treatment doctor information, a clinical trauma treatment database and a pre-trained decision-making model.

[0015] Through the above technical solution, the emergency trauma patient data is obtained through the data acquisition device, the emergency treatment difficulty of the emergency trauma patient is evaluated through the evaluation device, and the emergency treatment plan for the trauma patient is decided through the decision-making device. Among them, the evaluation device uses the pre-trained evaluation model to evaluate the difficulty of emergency treatment, and realizes accurate and efficient evaluation based on artificial neural networks. The decision-making device uses the pre-trained decision model to determine the emergency treatment plan for trauma patients, and realizes accurate and efficient decision-making based on artificial neural networks. Therefore, this technical solution can realize intelligent evaluation and decision-making of emergency treatment for trauma patients based on artificial neural networks, thereby assisting in improving the treatment effect of emergency treatment doctors on emergency trauma patients.

[0016] Other features and advantages of the present disclosure will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings are used to provide a further understanding of the present disclosure and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the present disclosure but do not constitute a limitation of the present disclosure. In the accompanying drawings: Figure 1 It is a structural block diagram of an intelligent assessment and decision-making system for emergency treatment of trauma patients based on an artificial neural network according to an exemplary embodiment.

[0018] Figure 2 It is a block diagram of a data acquisition device according to an exemplary embodiment.

[0019] Figure 3 The figure is a schematic diagram of a training process of an evaluation model according to an exemplary embodiment.

[0020] Figure 4 The figure is a flowchart of training and applying a decision model according to an exemplary embodiment.

[0021] Figure 5 It is a block diagram of another intelligent assessment and decision-making system for first aid of trauma patients based on artificial neural network according to an exemplary embodiment.

[0022] Figure 6 The present invention is a flowchart of an intelligent assessment and decision-making method for emergency treatment of trauma patients based on an artificial neural network according to an exemplary embodiment.

[0023] Figure 7 It is a block diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0024] The specific implementation of the present disclosure is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described herein is only used to illustrate and explain the present disclosure, and is not used to limit the present disclosure.

[0025] With the development of artificial intelligence technology, the application of artificial neural networks has also developed. Artificial neural networks are also used in medical scenarios. For example, artificial neural networks can be used to analyze medical data to obtain corresponding data analysis results.

[0026] In emergency medical scenarios, there are often trauma patients, who can be called emergency trauma patients. At present, some clinical databases can assist doctors in treating emergency trauma patients. Therefore, when doctors are inexperienced, they can help ensure the effectiveness of emergency treatment for patients.

[0027] However, the current emergency treatment programs based on clinical databases do not make good use of artificial neural networks. For example, they are only modeled based on clinical databases, but the actual application effect of the artificial neural network model is not very good.

[0028] Based on this, the present disclosure provides a technical solution, which fully integrates artificial neural networks into the evaluation and decision-making of emergency treatment scenarios for trauma patients, realizes intelligent evaluation and decision-making of emergency treatment for trauma patients based on artificial neural networks, and then assists in improving the treatment effect of emergency treatment doctors on emergency trauma patients.

[0029] It can be understood that the application scenario of the embodiment of the present disclosure is the first aid scenario for trauma patients. In this application scenario, on the one hand, the level of doctors performing first aid varies, and it is difficult to guarantee the treatment effect; on the other hand, the condition of the first aid patient is difficult to predict, and it is also difficult to guarantee the treatment effect. Therefore, if artificial neural networks can be used to perform more accurate and reliable data analysis, and then to make evaluations and decisions, it can play a good auxiliary role.

[0030] Figure 1 FIG. 1 is a structural block diagram of an intelligent assessment and decision system 10 for emergency treatment of trauma patients based on an artificial neural network according to an exemplary embodiment. Figure 1 As shown, the system includes: a data collection device 11, an evaluation device 12 and a decision-making device 13.

[0031] In some embodiments, the evaluation device 12 is communicatively connected to the data acquisition device 11, and the decision device 13 is communicatively connected to the data acquisition device 11 and the evaluation device 12 respectively, so that both the evaluation device 12 and the decision device 13 can obtain data from the data acquisition device 11, and the decision device 13 can obtain data from the evaluation device 12 and the data acquisition device 11 respectively.

[0032] Regarding the communication connection method of these devices, it can be a communication connection method based on the Internet of Things to ensure data transmission efficiency and accuracy, etc.

[0033] In some embodiments, the data collection device 11 is configured to obtain emergency trauma patient data.

[0034] In some embodiments, the emergency trauma patient data includes physical description information and trauma description information of the emergency trauma patient.

[0035] Physical description information is used to describe the physical reactions of emergency trauma patients caused by trauma, such as headache, dizziness, numbness, and fever, etc., which may include multiple descriptive words.

[0036] Trauma description information is used to describe the wound condition of trauma patients receiving emergency treatment, such as wound size, wound bleeding, and wound infection, etc. It can be presented in different forms such as pictures and descriptive words.

[0037] Figure 2 is a block diagram of a data acquisition device 11 according to an exemplary embodiment. Figure 2 As shown, the data acquisition device 11 includes: a first acquisition device 110 , a second acquisition device 112 and a processing device 114 .

[0038] The first acquisition device 110 is configured to acquire physical description information of the emergency trauma patient. As an example, the first acquisition device 110 can be a data input device, and the emergency trauma patient, doctor or nurse can input the physical description information on the device. The input form of information is not limited to text, voice and video.

[0039] The second acquisition device 112 is configured to acquire trauma description information of emergency trauma patients. As an example, the second acquisition device 112 may have an image acquisition function, which can acquire wound images, and an image processing function, which can analyze the wound conditions based on the acquired wound images. Of course, the second acquisition device 112 may also be a data input device, and emergency trauma patients, doctors or nurses, etc., can input trauma description information describing the wound conditions on the device.

[0040] Furthermore, the processing device 114 may integrate the data collected by the two collection devices to obtain the final emergency trauma patient data, for example, removing duplicate information, normalizing and standardizing the information, etc., which are not limited here.

[0041] The emergency trauma patient data collected by the data collection device 11 can be stored locally first, and can be obtained at any time when the evaluation device 12 or the decision-making device 13 needs it.

[0042] The evaluation device 12 may be a device integrated with a pre-trained evaluation model, which may have functions such as data analysis and processing.

[0043] In some embodiments, the assessment device 12 is configured to determine an emergency trauma patient assessment result based on emergency trauma patient data, a clinical trauma database, and a pre-trained assessment model, and the emergency trauma patient assessment result is used to characterize the emergency treatment difficulty of the emergency trauma patient.

[0044] In some embodiments, the clinical trauma database includes: a plurality of clinical trauma data, each of which includes clinical body description information and clinical trauma description information.

[0045] It can be understood that the clinical trauma database is a database constructed by collecting a large amount of clinical trauma data, and each piece of clinical trauma data involves clinical body description information and clinical trauma description information. The acquisition method of the clinical body description information and the clinical trauma description information may be different from or the same as the acquisition method of the emergency trauma patient data.

[0046] It is understandable that the conditions of clinical patients are different from those of emergency patients, and the conditions of clinical patients can be used as a reference for emergency patients. Also, the role of the pre-trained evaluation model is to evaluate, which can be combined with relevant reference information obtained from clinical patients for evaluation.

[0047] Thus, the evaluation process of the evaluation device 12 may include: determining a first information weight based on the physical description information of the emergency trauma patient and multiple clinical physical description information; determining a second information weight based on the trauma description information of the emergency trauma patient and multiple clinical trauma description information; inputting the physical description information of the emergency trauma patient, the first information weight, the trauma description information of the emergency trauma patient and the second information weight into a pre-trained evaluation model to obtain an emergency trauma patient evaluation result output by the pre-trained evaluation model.

[0048] In some embodiments, the first information weight is used to characterize the complexity or rarity of the physical description information of the first aid trauma patient, and the second information weight is used to characterize the complexity or rarity of the trauma description information of the first aid trauma patient.

[0049] In this embodiment, complexity or rarity can be selected as the influence of the weight. For example, the higher the complexity or rarity of the body description information of the emergency trauma patient, the greater the weight of the first information; the higher the complexity or rarity of the trauma description information of the emergency trauma patient, the greater the weight of the second information.

[0050] Therefore, as an example, determining the first information weight may include: if target clinical physical description information that matches the physical description information of an emergency trauma patient is determined from multiple clinical physical description information, determining the first information weight based on the complexity characteristics or rarity characteristics of the target clinical physical description information in the multiple clinical physical description information; if target clinical physical description information that matches the physical description information of an emergency trauma patient is not determined from the multiple clinical physical description information, determining the first information weight based on the similarity between the physical description information of the emergency trauma patient and the multiple clinical physical description information.

[0051] In this embodiment, multiple clinical physical description information is compared with the physical description information of the first-aid trauma patient to determine whether the target clinical physical description information exists. The target clinical physical description information may be clinical physical description information whose similarity with the physical description information of the first-aid trauma patient is higher than a preset similarity. Therefore, the similarity between the multiple clinical physical description information and the physical description information of the first-aid trauma patient can be calculated respectively to determine whether the target clinical physical description information exists.

[0052] In some embodiments, the preset similarity can be a value between 95% and 100%, which can be set specifically according to the difference between the emergency scenario and the clinical scenario in different application scenarios. The greater the difference, the lower the preset similarity can be.

[0053] Furthermore, if the target clinical physical description information exists, the first information weight is determined by using the complexity feature or rarity feature of the target clinical physical description information in the multiple clinical physical description information. That is, the complexity or rarity of the physical description information of the emergency trauma patient is characterized by the complexity or rarity of the target clinical physical description information.

[0054] In some embodiments, the complexity feature or the rarity feature may be determined by frequency, similarity, etc. For example, the lower the frequency, the higher the rarity; the lower the average similarity with other clinical body description information, the higher the complexity, etc.

[0055] In some embodiments, the first information weight and complexity or rarity may have a corresponding conversion relationship. For example, the value range of the first information weight is 0~1, and the value range of complexity or rarity is 0~100%. Then, based on the positive proportional relationship between the first information weight and complexity or rarity, a corresponding conversion function may be configured to achieve conversion. For example, the conversion function is: y=0.01x, where x represents complexity or rarity, and y represents the first information weight.

[0056] In some embodiments, if the target clinical physical description information does not exist, the first information weight can be determined based on the average similarity between the physical description information of the emergency trauma patient and multiple clinical physical description information. The lower the average similarity, the lower the first information weight. Regarding the conversion relationship between the first information weight and the average similarity, reference can be made to the implementation method of the conversion between the first information weight and the complexity or rarity, which will not be repeated here.

[0057] As an example, determining the second information weight may include: if target clinical trauma description information that matches the trauma description information of an emergency trauma patient is determined from multiple clinical trauma description information, determining the second information weight based on the complexity characteristics or rarity characteristics of the target clinical trauma description information in the multiple clinical trauma description information; if target clinical trauma description information that matches the trauma description information of the emergency trauma patient is not determined from the multiple clinical trauma description information, determining the second information weight based on the similarity between the trauma description information of the emergency trauma patient and the multiple clinical trauma description information.

[0058] The determination of the second information weight is similar to the determination of the first information weight, and is also divided into the case where the target clinical trauma description information exists and the case where the target clinical trauma description information does not exist. Therefore, the detailed implementation of each step involved is referred to the detailed implementation of the first information weight, and will not be repeated here.

[0059] Furthermore, the body description information of the emergency trauma patient, the first information weight, the trauma description information of the emergency trauma patient and the second information weight are input into the pre-trained evaluation model to obtain the emergency trauma patient evaluation result output by the pre-trained evaluation model.

[0060] In some embodiments, the pre-trained evaluation model is an artificial neural network model, and its corresponding artificial neural network algorithm can be a natural language processing algorithm, etc.

[0061] In some embodiments, the body description information and the trauma description information of the emergency trauma patient can be converted into corresponding vectors respectively to facilitate model processing. For details, reference can be made to mature artificial neural network technology in the field.

[0062] In some embodiments, the training process of the pre-trained evaluation model may include: A first training data set is obtained, the first training data set includes multiple first training samples, each of which includes: sample body description information, weight corresponding to the sample body description information, sample trauma description information, weight corresponding to the sample trauma description information and an evaluation label, and the evaluation label is used to characterize the difficulty of emergency treatment; based on the first training data set, a second training data set is determined, the second training data set includes multiple second training samples, each of which includes: sample body description information, sample trauma description information and an evaluation label; an evaluation model to be trained is trained based on the second training data set to obtain an evaluation model for initial training; and the evaluation model for initial training is optimized and trained based on the first training data set to obtain a pre-trained evaluation model.

[0063] In some embodiments, the sample body description information and the sample trauma description information can be obtained from an existing relevant database regardless of clinical and emergency scenarios. Correspondingly, the weight corresponding to the sample body description information, the weight corresponding to the sample trauma description information, and the evaluation label can be manually configured labels, or the evaluation label can also be a label determined according to the actual processing difficulty.

[0064] In some embodiments, the difficulty of emergency treatment can be represented by a difficulty level, for example: level 1 to 5, the higher the level, the greater the difficulty.

[0065] In some embodiments, part of the data may be separated from the first training data set as a second training data set, and the information weight is not involved in the second training data set.

[0066] Furthermore, the evaluation model to be trained is first trained using the second training data set to obtain an evaluation model for initial training. The evaluation model does not require weight information and can also be evaluated.

[0067] Next, the first training data set is used to optimize the initially trained evaluation model to train the generalization ability of the evaluation model. For example, the first training data set can be used as a test data set and input into the evaluation model to obtain a test result. The test result is used to determine the accuracy of the evaluation model with weight information. Then, according to the accuracy, the training data including the weight information is selected. For example, the higher the accuracy, the less training data is selected, so as to perform optimization training.

[0068] It can be understood that if the model training is performed directly based on the first training data set, the evaluation model cannot be evaluated based on the unweighted situation. However, if the training method disclosed in the present invention is adopted, the evaluation model can be evaluated based on the unweighted situation, and the unweighted situation can correspond to the situation where the accuracy of the first information weight and the second information weight is poor. Therefore, the accuracy of the evaluation model can be guaranteed in various situations.

[0069] Figure 3 is a schematic diagram of a training process of an evaluation model according to an exemplary embodiment. Figure 3 As shown, first configure the original training data set (i.e., the first training data set), select part of the original training data set as a sub-training data set (i.e., the second training data set) based on the original training data set, first use the sub-training data set for training, then use the original training data set to test the accuracy, and then select a new sub-training data set for optimization training based on the test results.

[0070] Furthermore, the pre-trained evaluation model can eventually evaluate the difficulty of emergency treatment for trauma patients and obtain the evaluation results of emergency treatment for trauma patients.

[0071] In some embodiments, the decision device 13 may be a device integrated with a pre-trained decision model, which may have functions such as data analysis and processing.

[0072] In some embodiments, the evaluation device 12 and the decision device 13 may be integrated into the same device, but their functions are different.

[0073] In some embodiments, the decision device 13 is configured to obtain emergency treatment physician information, and determine an emergency trauma patient treatment plan based on emergency trauma patient assessment results, emergency treatment physician information, a clinical trauma treatment database, and a pre-trained decision model.

[0074] In some embodiments, the clinical trauma treatment database includes: multiple clinical trauma treatment plans. These multiple clinical trauma treatment plans are clinical trauma treatment plans that are actually used in clinical trauma scenarios, and all of them are treatment plans that meet the treatment effect standards. By constructing a clinical trauma treatment database, it can be used to decide on treatment plans.

[0075] In some embodiments, the clinical trauma treatment program may include a treatment method, and more specifically, may also involve various treatment steps. For example, if the treatment method is conservative treatment, the corresponding treatment steps may include: first cleaning the wound, then bandaging the wound, then observing the patient's condition, and finally choosing whether to hospitalize the patient according to the patient's condition. For another example, if the treatment method is one-time treatment, the corresponding treatment steps may include: cleaning the wound, bandaging the wound, and prescribing a wound recovery prescription. It is understandable that more treatment plans may be involved, and they are not listed one by one here.

[0076] Furthermore, different treatment plans may require matching different levels of doctors. For example, conservative treatment has higher requirements for doctors' level, and it is necessary to judge whether the patient needs further treatment; one-time treatment has lower requirements for doctors' level, and only requires bandaging ability.

[0077] Therefore, as an optional implementation, the process of determining a decision plan includes: evaluating the emergency treatment physician based on the emergency treatment physician information to determine the evaluation result of the emergency treatment physician, and the evaluation result of the emergency treatment physician is used to characterize the emergency treatment level of the emergency treatment physician; inputting multiple clinical treatment plans into a pre-trained decision model to obtain clinical treatment difficulty labels and clinical physician treatment level labels corresponding to each clinical trauma treatment plan output by the pre-trained decision model; determining the emergency trauma patient treatment plan based on the clinical treatment difficulty labels and clinical physician treatment level labels corresponding to each clinical trauma treatment plan, the emergency trauma patient evaluation result and the emergency treatment physician evaluation result.

[0078] In this embodiment, the emergency treatment physician is first evaluated, and then a decision model is used to determine the clinical treatment plan that matches the clinical treatment difficulty label and the clinician treatment level label, and then a plan is selected.

[0079] It is understandable that multiple clinical treatment plans can also be implemented by pre-configuring clinical treatment difficulty labels and clinical doctor treatment level labels. However, since multiple clinical treatment plans may be updated at any time, the labels configured in this way are less accurate. Using models for real-time decision-making is more accurate.

[0080] In some embodiments, the emergency treatment doctor information may include: the doctor's age, the doctor's years of practice, the doctor's grade, the number of emergency treatments performed by the doctor, the number of clinical treatments performed by the doctor, etc. Based on this information, the emergency treatment doctor is evaluated accordingly to determine the evaluation result of the emergency treatment doctor. For example, the longer the doctor's years of practice, the higher the doctor's grade, the more emergency treatments performed by the doctor, and the more clinical treatments performed by the doctor, the higher the emergency treatment level.

[0081] In some embodiments, the pre-trained decision model includes a first module and a second module, wherein the first module can be used for feature extraction and the second module can be used for label determination.

[0082] The training process of the decision model may include: obtaining a third training data set, the third training data set includes multiple third training samples, each of which includes: a sample trauma treatment plan, a sample trauma treatment plan feature, a treatment difficulty label, and a doctor's treatment level label; according to the third training data set, determining a fourth training data set and a fifth training data set, the fourth training data set includes multiple fourth training samples, each of which includes a sample trauma treatment plan and a sample trauma treatment plan feature, and the fifth training data set includes multiple fifth training samples, each of which includes a sample trauma treatment plan, a treatment difficulty label, and a doctor's treatment level label; training a first module of the decision model to be trained according to the fourth training data set to obtain a decision model for initial training; training a second module of the decision model for initial training according to the fifth training data set to obtain a pre-trained decision model.

[0083] In this embodiment, a third training data set is first obtained, and the third training data set involves sample trauma treatment plans, sample trauma treatment plan features, treatment difficulty labels, and doctor treatment level labels. The sample trauma treatment plan may be a real treatment plan recorded in a database. The sample trauma treatment plan features may be obtained by extracting features from the plan, which may include: plan type features (such as the aforementioned one-time type, conservative type, etc.), plan keyword features (such as the aforementioned cleaning, bandaging, hospitalization, etc.).

[0084] In some embodiments, the treatment difficulty label and the doctor's treatment level label can be labels determined based on actual recorded situations, or can be manually configured labels.

[0085] In some embodiments, the third training data is split to obtain a fourth training data set for feature extraction training and a fifth training data set for label determination training.

[0086] Furthermore, the first module of the decision model to be trained is trained according to the fourth training data set to obtain the initially trained decision model; and the second module of the initially trained decision model is trained according to the fifth training data set to obtain the pre-trained decision model.

[0087] It can be understood that through this separate training method, the accuracy of different functional modules can be guaranteed, thereby improving the accuracy of the entire model.

[0088] Figure 4 is a schematic diagram of a training and application process of a decision model according to an exemplary embodiment. Figure 4 As shown, during model training, the first module and the second module are trained separately. When the model is applied, the input of the first module is the input of the entire model, the output of the first model is the input of the second module, and the output of the second model is the output of the entire model.

[0089] Furthermore, determining the emergency treatment plan for trauma patients based on the clinical treatment difficulty labels and clinical doctor treatment level labels corresponding to each clinical trauma treatment plan, the emergency trauma patient assessment results and the emergency treatment doctor assessment results may include: determining a first clinical trauma treatment plan based on the emergency trauma patient assessment results and the clinical treatment difficulty labels corresponding to each clinical trauma treatment plan, the clinical treatment difficulty label corresponding to the first clinical trauma treatment plan being higher than the emergency treatment difficulty represented by the emergency trauma patient assessment results; determining a second clinical trauma treatment plan based on the emergency treatment doctor assessment results and the clinical doctor treatment level labels corresponding to each clinical trauma treatment plan, the clinical doctor treatment level label corresponding to the second clinical trauma treatment plan being lower than the emergency treatment level represented by the emergency treatment doctor assessment results; determining the emergency treatment plan for trauma patients based on the first clinical trauma treatment plan and the second clinical trauma treatment plan.

[0090] In some embodiments, the first clinical trauma treatment plan and the second clinical trauma treatment plan are treatment plans determined based on different label conditions.

[0091] In some embodiments, if the first clinical trauma treatment plan and the second clinical trauma treatment plan have the same treatment plan, a treatment plan is randomly selected from the same treatment plan. If the first clinical trauma treatment plan and the second clinical trauma treatment plan are not repeated, a treatment plan is selected from the clinical trauma treatment plans involving a larger number.

[0092] In some embodiments, if the first clinical wound treatment plan and the second clinical wound treatment plan do not exist, the treatment plan with a label closer to the current situation is determined as the final treatment plan.

[0093] Furthermore, after the decision-making device 13 determines a plan, feedback can be provided and the doctor can make a choice according to needs.

[0094] Figure 5 is a block diagram of another artificial neural network-based intelligent assessment and decision system 10 for emergency treatment of trauma patients according to an exemplary embodiment. Figure 5 As shown, the system further includes: a visualization device 14 .

[0095] The visualization device 14 is respectively connected to the data acquisition device 11, the evaluation device 12 and the decision device 13. The visualization device 14 has visualization and interaction functions, and therefore, it may have an input and output module, a display module and the like.

[0096] In some embodiments, the visualization device 14 is configured to display emergency trauma patient data, emergency trauma patient assessment results, and emergency trauma patient treatment plans, respectively. It is understood that it can display various data involved in the system.

[0097] In some embodiments, the visualization device 14 is configured to: in response to receiving an auxiliary treatment request initiated by an emergency treatment physician, play a locally stored auxiliary treatment video recorded by a preset physician.

[0098] In this embodiment, the emergency treatment doctor can initiate an auxiliary treatment request through the interactive function of the visualization device 14. For example, the interface displayed by the visualization device 14 includes an auxiliary treatment control. When the emergency treatment doctor triggers the control, the visualization device 14 is deemed to have received the auxiliary treatment request.

[0099] In some embodiments, the visualization device 14 stores an auxiliary treatment video recorded by a preset doctor, and if auxiliary treatment is needed, the auxiliary treatment video can be played directly.

[0100] In some embodiments, the preset doctor may be a doctor with strong expertise in emergency treatment, and the auxiliary treatment video recorded by the doctor may include: first aid precautions, key first aid treatment steps and other first aid reference information to assist first aid doctors, nurses, etc. in providing more professional treatment.

[0101] Through this implementation, a comprehensive auxiliary role in emergency treatment can be played, further improving the effect of emergency treatment.

[0102] Figure 6 is a flow chart of an intelligent assessment and decision-making method for emergency treatment of trauma patients based on an artificial neural network according to an exemplary embodiment. Figure 6 As shown, the method includes: Step S61, obtaining emergency trauma patient data.

[0103] Step S62, determining the emergency trauma patient assessment result based on the emergency trauma patient data, the clinical trauma database and the pre-trained assessment model, and the emergency trauma patient assessment result is used to characterize the emergency treatment difficulty of the emergency trauma patient.

[0104] Step S63, obtaining emergency treatment doctor information.

[0105] Step S64, determining the emergency treatment plan for trauma patients based on the emergency trauma patient assessment results, emergency treatment physician information, clinical trauma treatment database and pre-trained decision-making model.

[0106] Regarding the method in the above embodiment, the specific manner of each step has been described in detail in the embodiment of the system and will not be elaborated here.

[0107] Figure 7 FIG. 7 is a block diagram of an electronic device 700 according to an exemplary embodiment. Figure 7 As shown, the electronic device 700 may include: a processor 701 , a memory 702 . The electronic device 700 may also include one or more of a multimedia component 703 , an input / output (I / O) interface 704 , and a communication component 705 .

[0108] It can be understood that the electronic device 700 can be used as part or all of the aforementioned data acquisition device, decision-making device, evaluation device and visualization device.

[0109] The processor 701 is used to control the overall operation of the electronic device 700 to complete all or part of the steps in the above-mentioned artificial neural network-based trauma patient emergency intelligent assessment decision method. The memory 702 is used to store various types of data to support the operation of the electronic device 700, and these data may include, for example, instructions for any application or method used to operate on the electronic device 700, and application-related data, such as contact data, sent and received messages, pictures, audio, video, etc. The memory 702 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read-Only Memory, referred to as EPROM), programmable read-only memory (Programmable Read-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, magnetic disk or optical disk. The multimedia component 703 may include a screen and an audio component. The screen may be, for example, a touch screen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone, which is used to receive external audio signals. The received audio signal may be further stored in the memory 702 or sent through the communication component 705. The audio component also includes at least one speaker for outputting audio signals. The I / O interface 704 provides an interface between the processor 701 and other interface modules, and the above-mentioned other interface modules may be a keyboard, a mouse, buttons, etc. These buttons may be virtual buttons or physical buttons. The communication component 705 is used for wired or wireless communication between the electronic device 700 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G or 4G, or a combination of one or more of them, so the corresponding communication component 705 may include: Wi-Fi module, Bluetooth module, NFC module.

[0110] In an exemplary embodiment, the electronic device 700 can be implemented by one or more application specific integrated circuits (ASIC), digital signal processors (DSP), digital signal processing devices (DSPD), programmable logic devices (PLD), field programmable gate arrays (FPGA), controllers, microcontrollers, microprocessors or other electronic components to execute part or all of the steps of the above-mentioned artificial neural network-based intelligent assessment and decision-making method for first aid of trauma patients.

[0111] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided, and when the program instructions are executed by a processor, some or all of the steps of the above-mentioned method for intelligent assessment and decision-making of first aid for trauma patients based on artificial neural networks are implemented. For example, the computer-readable storage medium can be the above-mentioned memory 702 including program instructions, and the above-mentioned program instructions can be executed by the processor 701 of the electronic device 700 to complete some or all of the steps of the above-mentioned method for intelligent assessment and decision-making of first aid for trauma patients based on artificial neural networks.

[0112] In another exemplary embodiment, a computer program product is also provided, which includes a computer program that can be executed by a processor, and when the computer program is executed by the processor, it implements part or all of the steps of the above-mentioned artificial neural network-based intelligent assessment and decision-making method for first aid of trauma patients.

[0113] The preferred embodiments of the present disclosure are described in detail above in conjunction with the accompanying drawings; however, the present disclosure is not limited to the specific details in the above embodiments. Within the technical concept of the present disclosure, a variety of simple modifications can be made to the technical solution of the present disclosure, and these simple modifications all fall within the protection scope of the present disclosure.

[0114] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the present disclosure will not further describe various possible combinations.

[0115] In addition, various embodiments of the present disclosure may be arbitrarily combined, and as long as they do not violate the concept of the present disclosure, they should also be regarded as the contents disclosed by the present disclosure.

Claims

1. An intelligent assessment and decision-making system for emergency treatment of trauma patients based on artificial neural network, characterized in that: include: a data acquisition device configured to acquire emergency trauma patient data; An evaluation device, which is in communication with the data acquisition device and is configured to determine an evaluation result of an emergency trauma patient based on the emergency trauma patient data, a clinical trauma database and a pre-trained evaluation model, wherein the evaluation result of the emergency trauma patient is used to characterize the difficulty of emergency treatment of the emergency trauma patient; The decision-making device is respectively communicatively connected to the data acquisition device and the evaluation device, and is configured to obtain the emergency treatment doctor information, and determine the emergency treatment plan for the trauma patient according to the emergency trauma patient evaluation result, the emergency treatment doctor information, the clinical trauma treatment database and the pre-trained decision-making model.

2. The intelligent assessment and decision-making system for emergency treatment of trauma patients according to claim 1, characterized in that: The data acquisition device comprises: A first collection device is configured to collect physical description information of the emergency trauma patient, wherein the physical description information is used to describe the physical reaction of the emergency trauma patient caused by the trauma; The second collection device is configured to collect trauma description information of the emergency trauma patient, wherein the trauma description information is used to describe the wound condition of the emergency trauma patient; The processing device is configured to integrate the body description information and the trauma description information to obtain the emergency trauma patient data.

3. The intelligent assessment and decision-making system for emergency treatment of trauma patients according to claim 1, characterized in that: The emergency trauma patient data includes body description information and trauma description information of the emergency trauma patient, the clinical trauma database includes: a plurality of clinical trauma data, each clinical trauma data includes clinical body description information and clinical trauma description information, and the evaluation device is further configured to: Determining a first information weight according to the physical description information of the emergency trauma patient and the plurality of clinical physical description information; determining a second information weight according to the trauma description information of the emergency trauma patient and the plurality of clinical trauma description information; The body description information of the emergency trauma patient, the first information weight, the trauma description information of the emergency trauma patient and the second information weight are input into the pre-trained evaluation model to obtain the emergency trauma patient evaluation result output by the pre-trained evaluation model.

4. The intelligent assessment and decision-making system for emergency treatment of trauma patients according to claim 3, characterized in that: The first information weight is used to characterize the complexity or rarity of the body description information of the emergency trauma patient, the second information weight is used to characterize the complexity or rarity of the trauma description information of the emergency trauma patient, and the evaluation device is further configured to: If target clinical physical description information matching the physical description information of the emergency trauma patient is determined from the multiple clinical physical description information, the first information weight is determined according to the complexity feature or rarity feature of the target clinical physical description information in the multiple clinical physical description information; if target clinical physical description information matching the physical description information of the emergency trauma patient is not determined from the multiple clinical physical description information, the first information weight is determined according to the similarity between the physical description information of the emergency trauma patient and the multiple clinical physical description information; and / or, If target clinical trauma description information that matches the trauma description information of the emergency trauma patient is determined from the multiple clinical trauma description information, the second information weight is determined based on the complexity characteristics or rarity characteristics of the target clinical trauma description information in the multiple clinical trauma description information; if target clinical trauma description information that matches the trauma description information of the emergency trauma patient is not determined from the multiple clinical trauma description information, the second information weight is determined based on the similarity between the trauma description information of the emergency trauma patient and the multiple clinical trauma description information.

5. The intelligent assessment and decision-making system for emergency treatment of trauma patients according to claim 3 or 4, characterized in that: The evaluation device is further configured to: Acquire a first training data set, the first training data set comprising a plurality of first training samples, each of the first training samples comprising: sample body description information, a weight corresponding to the sample body description information, sample trauma description information, a weight corresponding to the sample trauma description information, and an evaluation label, the evaluation label being used to characterize the difficulty of emergency treatment; Determine a second training data set according to the first training data set, wherein the second training data set includes a plurality of second training samples, each of the second training samples includes: sample body description information, sample trauma description information and an evaluation label; Training the evaluation model to be trained according to the second training data set to obtain an evaluation model for initial training; The initially trained evaluation model is optimized and trained according to the first training data set to obtain the pre-trained evaluation model.

6. The intelligent assessment and decision-making system for emergency treatment of trauma patients according to claim 1, characterized in that: The clinical trauma treatment database includes: a plurality of clinical trauma treatment plans, and the decision-making device is further configured to: Evaluate the emergency treatment physician according to the emergency treatment physician information to determine an evaluation result of the emergency treatment physician, wherein the evaluation result of the emergency treatment physician is used to characterize the emergency treatment level of the emergency treatment physician; Inputting a plurality of clinical treatment plans into the pre-trained decision model, and obtaining clinical treatment difficulty labels and clinical doctor treatment level labels corresponding to each of the clinical trauma treatment plans output by the pre-trained decision model; The emergency trauma patient treatment plan is determined according to the clinical treatment difficulty labels and clinical doctor treatment level labels corresponding to each of the clinical trauma treatment plans, the emergency trauma patient assessment results and the emergency treatment doctor assessment results.

7. The intelligent assessment and decision-making system for emergency treatment of trauma patients according to claim 6, characterized in that: The pre-trained decision model includes a first module and a second module, and the decision device is further configured to: Acquire a third training data set, the third training data set comprising a plurality of third training samples, each of the third training samples comprising: a sample trauma treatment plan, a sample trauma treatment plan feature, a treatment difficulty label, and a doctor treatment level label; Determine a fourth training data set and a fifth training data set according to the third training data set, wherein the fourth training data set includes a plurality of fourth training samples, each of which includes a sample trauma treatment plan and a sample trauma treatment plan feature, and the fifth training data set includes a plurality of fifth training samples, each of which includes a sample trauma treatment plan, a treatment difficulty label, and a doctor's treatment level label; Training the first module of the decision model to be trained according to the fourth training data set to obtain an initially trained decision model; The second module of the initially trained decision model is trained according to the fifth training data set to obtain the pre-trained decision model.

8. The intelligent assessment and decision-making system for emergency treatment of trauma patients according to claim 6 or 7, characterized in that: The decision making device is further configured to: Determining a first clinical trauma treatment plan according to the emergency trauma patient assessment result and the clinical treatment difficulty labels corresponding to each of the clinical trauma treatment plans, wherein the clinical treatment difficulty label corresponding to the first clinical trauma treatment plan is higher than the emergency treatment difficulty represented by the emergency trauma patient assessment result; Determining a second clinical trauma treatment plan according to the emergency treatment doctor's evaluation result and the clinician's treatment level labels corresponding to each of the clinical trauma treatment plans, wherein the clinician's treatment level label corresponding to the second clinical trauma treatment plan is lower than the emergency treatment level represented by the emergency treatment doctor's evaluation result; The emergency trauma patient treatment plan is determined based on the first clinical trauma treatment plan and the second clinical trauma treatment plan.

9. The intelligent assessment and decision-making system for emergency treatment of trauma patients according to claim 1, characterized in that: The trauma patient emergency intelligent assessment and decision-making system also includes: A visualization device is communicatively connected with the data acquisition device, the evaluation device and the decision-making device, and is configured as follows: The emergency trauma patient data, the emergency trauma patient assessment results and the emergency trauma patient treatment plan are displayed respectively; In response to receiving the auxiliary treatment request initiated by the emergency treatment doctor, playing the locally stored auxiliary treatment video recorded by the preset doctor.

10. An intelligent assessment and decision-making method for emergency treatment of trauma patients based on artificial neural network, characterized in that: include: Obtain emergency trauma patient data; Determine an emergency trauma patient assessment result based on the emergency trauma patient data, the clinical trauma database and the pre-trained assessment model, wherein the emergency trauma patient assessment result is used to characterize the emergency treatment difficulty of the emergency trauma patient; Get emergency treatment physician information; The emergency trauma patient treatment plan is determined based on the emergency trauma patient assessment results, the emergency treatment doctor information, the clinical trauma treatment database and the pre-trained decision model.

Images