Batch wounded transferring treatment strategy control system and method based on deep learning

Through a large-scale treatment strategy control system for batches of injured people transfers based on deep learning, the situation of injuries and medical institutions are analyzed, and the route of injured people transfers is optimized, which solves the problem of inflexible transfers of injured people in the existing technology, and improves the treatment efficiency and cure rate.

CN118969212APending Publication Date: 2024-11-15XUANWU HOSPITAL OF CAPITAL UNIV OF MEDICAL SCI
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410924254.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-11
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The existing methods of transferring injured people are inflexible and failing to consider a variety of situations in a comprehensive way, which may waste the best treatment time or delay treatment, reducing the success rate of cure.

Method used

Design a strategy control system for batches of injured people transferred to treatment based on deep learning. By obtaining detailed information of injured people, using trained deep learning models to analyze injuries, wait for treatment time, divide the injury class, prioritize, and calculate each medical treatment When transferring and treatment time of the institution, select the most suitable medical institution for transferring the injured.

Benefits of technology

It improves the efficiency and cure rate of the injured, ensures that the injured receive the most appropriate medical treatment and avoids wasted or delayed treatment time.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118969212A_ABST
    Figure CN118969212A_ABST
Patent Text Reader

Abstract

The invention discloses a batch wounded person transfer treatment strategy control system and method based on deep learning. The method comprises the steps that a wounded person transfer treatment strategy control center obtains injury condition information of each wounded person in batch wounded persons to be transferred; inputting the injury condition information of the wounded person into a trained deep learning model for learning to obtain the time for waiting for treatment of the injury condition of the wounded person; the wounded persons belonging to the same injury class are divided into corresponding wounded person injury class sets, and the injury class level of each wounded person in each wounded person injury class set is analyzed; the wounded persons in each wounded person injury class set are subjected to priority ranking according to the injury class level from high to low; the transfer time of each wounded person in the kth priority in the mth wounded person injury class set to each first medical institution corresponding to the injury class level of the wounded person and the medical institution treatment waiting time are calculated, the maximum value of the transfer time and the medical institution treatment waiting time is taken, and the wounded person with the maximum value smaller than the injury state wait treatment time of the wounded person is the first medical institution which can be delivered; one is selected as a target first medical institution to schedule transfer.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of control systems, and in particular to a control system and method for a batch wounded transfer and treatment strategy based on deep learning. Background Art

[0002] The existing transfer method of the wounded is generally to designate the hospital to which the wounded are transferred, and transfer the wounded to the designated hospital. Another method is the principle of nearby transfer, that is, to transfer the wounded to the nearest hospital. The existing transfer method is inflexible and does not comprehensively consider multiple situations. For example, if the wounded is shot and the nearest hospital is not capable of treating the gunshot wound, then sending the wounded to the nearest hospital will waste the best treatment time. If the designated hospital is far away from the wounded, sending the wounded to the designated hospital will delay the treatment of the wounded and reduce the success rate of the wounded's cure, etc. Based on this, the present invention has taken multiple considerations and designed a batch wounded transfer treatment strategy control system and method based on deep learning. Summary of the invention

[0003] In view of the problems and shortcomings of the prior art, the present invention provides a control system and method for batch wounded transfer and treatment strategy based on deep learning.

[0004] The present invention solves the above technical problems through the following technical solutions:

[0005] The present invention provides a batch wounded transfer and treatment strategy control system based on deep learning, which is characterized in that it includes a first medical institution management system corresponding to multiple first medical institutions corresponding to various types of injuries and a wounded transfer and treatment strategy control center;

[0006] The wounded transfer and treatment strategy control center is used to obtain the injury information of each wounded in the batch of wounded to be transferred, the injury information includes the wounded's identity information, gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury severity, complication information and preliminary treatment operations of the injury, and the complication information includes whether there are complications and the specific complications when there are complications;

[0007] The wounded transfer and treatment strategy control center is also used to input the wounded's gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury, complication information and preliminary treatment operations into the trained deep learning model for learning to obtain the wounded's injury waiting time for treatment, thereby obtaining the injury waiting time for treatment of each wounded to be transferred;

[0008] The patient transfer and treatment strategy control center is further used to classify the patients belonging to the same injury category into corresponding patient injury category sets based on the injury category, and a total of M patient injury category sets are divided. For each patient in each patient injury category set, the injury category level of the patient is analyzed based on the waiting time for treatment of the patient's injury, and the injury category levels include light injury category level, moderate injury category level, severe injury category level and critical injury category level;

[0009] The wounded transfer and treatment strategy control center is also used to prioritize the wounded in each wounded injury set in descending order of injury level. Wounded with the same injury level have the same priority, and a total of K priorities are sorted.

[0010] The wounded transfer treatment strategy control center is also used to calculate, for the kth priority level in the mth wounded injury category set, the transfer time of each wounded in the kth priority level in the mth wounded injury category set to each first medical institution corresponding to the injury category level of its injury category, and based on the current injury treatment situation of each first medical institution corresponding to the injury category level of its injury category, the medical institution treatment waiting time and transfer time of each first medical institution corresponding to the injury category level of its injury category are analyzed, and the maximum value of the medical institution treatment waiting time and transfer time of each first medical institution corresponding to the injury category level of its injury category is taken, and the maximum value is less than the injury condition of the wounded. The first medical institution with the waiting time for treatment is the first medical institution that can be delivered. One of the first medical institutions that can be delivered is selected as the target first medical institution, and the injured person is arranged to be transferred to the target first medical institution. The injured person's injury information, the waiting time for treatment and the transfer time are transmitted to the first medical institution management system corresponding to the target first medical institution, and a reminder message is sent to the first medical institution management system based on the transfer time so that the target first medical institution can arrange medical staff to receive the injured person, where m takes values ​​of 1-M, k takes values ​​of 1-K, and m, M, k and K are all positive integers.

[0011] The present invention also provides a method for controlling a batch wounded transfer and treatment strategy based on deep learning, which is characterized in that it includes the following steps:

[0012] S1. The wounded transfer and treatment strategy control center obtains the injury information of each wounded in the batch of wounded to be transferred. The injury information includes the wounded's identity information, gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury severity, complication information and preliminary treatment operations. The complication information includes whether there are complications and the specific complications if there are complications;

[0013] S2. The wounded transfer and treatment strategy control center inputs the wounded's gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury, complication information and preliminary treatment operations into the trained deep learning model for learning to obtain the wounded's injury waiting time for treatment, thereby obtaining the wounded's injury waiting time for treatment for each wounded to be transferred;

[0014] S3, the patient transfer and treatment strategy control center divides the patients of the same injury category into corresponding patient injury category sets based on the injury category, and divides a total of M patient injury category sets. For each patient in each patient injury category set, the injury category level of the patient is analyzed based on the waiting time for treatment of the patient's injury, and the injury category levels include light injury category level, moderate injury category level, severe injury category level and critical injury category level;

[0015] S4. The wounded transfer and treatment strategy control center prioritizes the wounded in each wounded injury set in descending order of injury level. Wounded with the same injury level have the same priority, and a total of K priorities are sorted.

[0016] S5. The patient transfer treatment strategy control center calculates the transfer time of each patient in the kth priority level in the mth patient injury category set to each first medical institution corresponding to the injury category level of the patient, and analyzes the medical institution treatment waiting time of each first medical institution corresponding to the current injury category treatment situation of each first medical institution corresponding to the injury category level of the patient based on the kth priority level in the mth patient injury category set. The maximum value of the medical institution treatment waiting time and transfer time of each first medical institution corresponding to the injury category level of the patient is taken, and the maximum value is less than the injury condition of the patient. The first medical institution with the waiting time for treatment is the first medical institution that can be delivered. One of the first medical institutions that can be delivered is selected as the target first medical institution, and the injured person is arranged to be transferred to the target first medical institution. The injured person's injury information, the waiting time for treatment and the transfer time are transmitted to the first medical institution management system corresponding to the target first medical institution, and a reminder message is sent to the first medical institution management system based on the transfer time so that the target first medical institution can arrange medical staff to receive the injured person, where m takes values ​​of 1-M, k takes values ​​of 1-K, and m, M, k and K are all positive integers.

[0017] The positive and progressive effects of the present invention are:

[0018] The present invention has taken multiple considerations into account, including the waiting time for treatment of the wounded, the corresponding appropriate medical institutions that can treat injuries of different injury levels, the transfer time to each first medical institution corresponding to the injury level of the wounded, the waiting time for treatment at each first medical institution, etc., so that relatively suitable first medical institutions can be obtained, which is conducive to more reasonable transfer of the wounded, effectively improving the transfer efficiency of batches of wounded, and improving the cure rate of the wounded. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a structural schematic diagram of a batch wounded transfer and treatment strategy control system based on deep learning in a preferred embodiment of the present invention.

[0020] Figure 2 This is a flowchart of a method for controlling a strategy for transferring and treating a batch of wounded based on deep learning according to a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0021] 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 making creative work are within the scope of protection of the present invention.

[0022] like Figure 1 As shown, this embodiment provides a batch wounded transfer treatment strategy control system based on deep learning, which includes a first medical institution management system 100 corresponding to multiple first medical institutions corresponding to various types of injuries, a wounded transfer treatment strategy control center 200, and a second medical institution management system 300 corresponding to multiple transferable second medical institutions corresponding to each injury type managed by each first medical institution management system 100.

[0023] The types of injuries here include gunshot wounds, explosion injuries, burns, etc. There are multiple medical institutions that can treat gunshot wounds, there are multiple medical institutions that can treat explosion injuries, and there are multiple medical institutions that can treat burns.

[0024] The wounded transfer and treatment strategy control center 200 is used to build different deep learning models and initialize the hyperparameters of each deep learning model; collect the injury information of multiple historical wounded patients, and professional doctors mark the waiting time for treatment of each historical wounded patient based on the injury information of each historical wounded patient, and build a sample data set based on the gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury, complication information and preliminary processing operations of the injury (input samples) and the corresponding waiting time for treatment of the injury (output samples) in the injury information of each historical wounded patient.

[0025] The wounded transfer and treatment strategy control center 200 is also used to use the sample training subset divided from the sample data set to iteratively train each deep learning model, continuously adjust the hyperparameters until the optimal solution is obtained, and use the sample test subset divided from the sample data set to test each trained deep learning model and obtain the accuracy. Thereafter, repeated random subsampling cross-validation technology is used to train and test each deep learning model multiple times, so that the injury information of each historical wounded person is used as model training data and model test data, thereby obtaining the accuracy of each deep learning model after multiple tests, calculating the average accuracy corresponding to each deep learning model, and taking the deep learning model corresponding to the largest average accuracy as the target deep learning model.

[0026] In this scheme, the sample data set is divided into a sample training subset and a sample test subset. The deep learning model corresponding to the maximum average accuracy is the relatively optimal deep learning model, which is used as the target deep learning model. After obtaining the target deep learning model, it can be used to analyze the waiting time for treatment of each injured person in the batch of transferred injured people.

[0027] The wounded transfer and treatment strategy control center 200 is also used to obtain the injury information of each wounded in the batch of wounded to be transferred. The injury information includes the wounded's identity information, gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury, complication information and preliminary treatment operations. The complication information includes whether there are complications and the specific complications when there are complications. When the wounded has multiple types of injuries, the most serious injury type is selected as the injury type in the injury information.

[0028] In this plan, if the injured person has both gunshot wounds and explosion injuries, and the gunshot wounds are more serious, the gunshot wounds will be used as the injury type of the injured person.

[0029] The wounded person transfer and treatment strategy control center 200 is also used to input the wounded person's gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury, complication information and preliminary treatment operations into the target deep learning model for learning to obtain the wounded person's injury waiting time for treatment, thereby obtaining the injury waiting time for treatment of each wounded person to be transferred.

[0030] The wounded transfer and treatment strategy control center 200 is also used to classify the wounded belonging to the same injury category into corresponding wounded injury category sets based on the injury category, and a total of M wounded injury category sets are divided. For each wounded in each wounded injury category set, the injury level of the wounded is analyzed based on the waiting time for treatment based on his injury. The injury levels include light injury level, moderate injury level, severe injury level and critical injury level.

[0031] In this scheme, the wounded with the same gunshot wound are classified into the wounded gunshot wound set, and the wounded in the wounded gunshot wound set have their injury level analyzed based on the waiting time for treatment of their injuries. The wounded with the same explosion wound are classified into the wounded explosion wound set, and the wounded in the explosion wound set have their injury level analyzed based on the waiting time for treatment of their injuries.

[0032] The wounded transfer treatment strategy control center 200 is also used to prioritize the wounded in each wounded injury set in descending order of injury level. Wounded with the same injury level have the same priority, and a total of K priorities are sorted.

[0033] For example: the wounded in the gunshot wound set are prioritized in descending order according to the injury level, and 2 gunshot wound serious injury levels and 2 gunshot wound minor injury levels are obtained; the wounded in the bomber gunshot wound set are prioritized in descending order according to the injury level, and 1 explosion injury critical injury level, 1 explosion injury serious injury level, 1 explosion injury moderate injury level and 1 explosion injury minor injury level are obtained.

[0034] The wounded transfer treatment strategy control center 200 is also used to calculate the transfer time for each wounded in the kth priority level in the mth wounded injury category set to reach each first medical institution corresponding to the injury category level of its injury category, and analyze the medical institution treatment waiting time corresponding to each first medical institution corresponding to the injury category level of its injury category based on the current injury treatment situation of each first medical institution corresponding to the injury category level of its injury category, take the maximum value of the medical institution treatment waiting time and transfer time of each first medical institution corresponding to the injury category level of its injury category, and determine the number of first medical institutions whose maximum value is less than the waiting time for treatment of the wounded's injury. When the number is zero, directly use the first medical institution with the shortest transfer time among the first medical institutions corresponding to the injury category level of its injury category as the target first medical institution, arrange for the wounded to be transferred to the target first medical institution, and transmit the wounded's injury information, the waiting time for treatment of the injury and the transfer time to the first medical institution management system 100 corresponding to the target first medical institution, and issue an emergency rescue reminder message based on the transfer time. The first medical institution management system 100 receives the injured person and sends a message to the first medical institution management system 100 so that the target first medical institution can arrange medical staff to receive the injured person and provide immediate emergency rescue. When the number is one, the first medical institution is directly used as the target first medical institution, and the injured person is arranged to be transferred to the target first medical institution. The injury information of the injured person, the waiting time for treatment and the transfer time are transmitted to the first medical institution management system 100 corresponding to the target first medical institution, and a reception reminder message is sent to the first medical institution management system 100 based on the transfer time so that the target first medical institution can arrange medical staff to receive the injured person. When the number is multiple, the first medical institution corresponding to the shortest transfer time is selected from the first medical institutions that can be delivered as the target first medical institution, and the injured person is arranged to be transferred to the target first medical institution. The injury information of the injured person, the waiting time for treatment and the transfer time are transmitted to the first medical institution management system 100 corresponding to the target first medical institution, and a reception reminder message is sent to the first medical institution management system 100 based on the transfer time so that the target first medical institution can arrange medical staff to receive the injured person.

[0035] Among them, m takes values ​​of 1-M in sequence, k takes values ​​of 1-K in sequence, and m, M, k and K are all positive integers.

[0036] For example: calculate the transfer time for two wounded persons in the first priority group in the first wounded injury group (the wounded gunshot wound group) (i.e., two wounded persons with a gunshot wound grade of severe injury, wounded person X and wounded person Y) to arrive at the two medical institutions corresponding to the gunshot wound grade of severe injury (i.e., medical institution A and medical institution B), and analyze the medical institution treatment waiting time required for medical institution A to treat wounded persons with a gunshot wound grade of severe injury and the medical institution treatment waiting time required for medical institution B to treat wounded persons with a gunshot wound grade of severe injury based on the current injury treatment situation of medical institution A and medical institution B. Take the maximum value of the medical institution treatment waiting time required by medical institution A and the transfer time to medical institution A, and take the maximum value of the medical institution treatment waiting time required by medical institution B and the transfer time to medical institution B. Determine the number of medical institutions whose maximum value of these two maximum values ​​is less than the waiting time for treatment of wounded person X. If medical institution A meets the conditions, directly select medical institution A as the first target medical institution , and transmit the injury information of the injured person X, the time that the injury can wait for treatment, and the transfer time to the A medical institution to the A medical institution management system corresponding to the A medical institution, and send a receiving reminder message to the A medical institution management system based on the transfer time so that the A medical institution arranges medical staff to receive the injured person X; judge the number of medical institutions whose maximum value among the two maximum values ​​is less than the time that the injury of the injured person Y can wait for treatment. If both the A medical institution and the B medical institution meet the conditions, select the medical institution corresponding to the shortest transfer time from the A medical institution and the B medical institution as the target first medical institution. If the transfer time to the B medical institution is the shortest, the B medical institution is selected as the target first medical institution, and the injury information of the injured person Y, the time that the injury can wait for treatment, and the transfer time to the B medical institution are transmitted to the B medical institution management system corresponding to the B medical institution, and send a receiving reminder message to the B medical institution management system based on the transfer time so that the B medical institution arranges medical staff to receive the Y injured person.

[0037] If the injured person cannot be fully treated in the first medical institution, it is necessary to move the injured person to a second medical institution of a higher level. Specifically, the first medical institution management system 100 is used to transmit the latest injury information of the injured person to the injured person transfer and treatment strategy control center 200 when the injured person needs to be transferred again. The latest injury information includes the information that the injury has been treated and the estimated time the injury can wait for treatment.

[0038] The wounded transfer and treatment strategy control center 200 is also used to calculate the transfer time of the re-transferred wounded to each of the multiple transferable second medical institutions corresponding to its injury type, and analyze the medical institution treatment waiting time of each second medical institution corresponding to its injury type based on the current injury treatment situation of each second medical institution corresponding to its injury type, take the maximum value of the medical institution treatment waiting time and transfer time of each second medical institution corresponding to its injury type, and judge the number of second medical institutions whose maximum value is less than the estimated time that the wounded's injury can wait for treatment. When the number is zero, directly use the second medical institution with the shortest transfer time among the second medical institutions corresponding to its injury type as the target second medical institution, arrange for the wounded to be transferred to the target second medical institution, and transmit the wounded's latest injury information, the estimated time that the injury can wait for treatment and the transfer time to the second medical institution management system 300 corresponding to the target second medical institution, and send an emergency rescue reminder message to the second medical institution management system 300 based on the transfer time so that the target second medical institution Arrange medical staff to receive the injured person and provide immediate emergency rescue; when the number is one, directly use this second medical institution as the target second medical institution, arrange for the injured person to be transferred to the target second medical institution, and transmit the injured person's latest injury information, estimated time for waiting for treatment and transfer time to the second medical institution management system 300 corresponding to the target second medical institution, and send a reception reminder message to the second medical institution management system 300 based on the transfer time so that the target second medical institution can arrange medical staff to receive the injured person; when the number is multiple, select the second medical institution with the shortest transfer time from the accessible second medical institutions as the target second medical institution, arrange for the injured person to be transferred to the target second medical institution, and transmit the injured person's latest injury information, estimated time for waiting for treatment and transfer time to the second medical institution management system 300 corresponding to the target second medical institution, and send a reception reminder message to the second medical institution management system 300 based on the transfer time so that the target second medical institution can arrange medical staff to receive the injured person.

[0039] like Figure 2 As shown, this embodiment also provides a method for controlling a batch wounded transfer and treatment strategy based on deep learning, which includes the following steps:

[0040] Preparation: The wounded transfer and treatment strategy control center builds different deep learning models and initializes the hyperparameters of each deep learning model; collects the injury information of multiple historical wounded, and professional doctors mark the waiting time for treatment of each historical wounded based on the injury information of each historical wounded, and builds a sample data set based on the gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury, complication information and initial treatment operations of the injury and the corresponding waiting time for treatment of the injury in each historical wounded. Use the sample training subset divided from the sample data set to train each deep learning model; The deep learning model is iteratively trained, and the hyperparameters are continuously adjusted until the optimal solution is obtained. The sample test subset divided from the sample data set is used to test each deep learning model after training and obtain the accuracy. Subsequently, each deep learning model is trained and tested multiple times using repeated random subsampling cross-validation technology, so that the injury information of each historical casualty is used as model training data and model test data, thereby obtaining the accuracy of each deep learning model after multiple tests, calculating the average accuracy corresponding to each deep learning model, and taking the deep learning model with the largest average accuracy as the target deep learning model.

[0041] S1. The control center for the transfer and treatment of wounded patients obtains the injury information of each wounded patient in the batch of wounded patients to be transferred. The injury information includes the wounded patient's identity information, gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury severity, complication information, and preliminary treatment operations. The complication information includes whether there are complications and the specific complications when there are complications. If there are multiple types of injuries for the wounded patient, the most serious injury type is selected as the injury type in the injury information.

[0042] S2. The wounded person transfer and treatment strategy control center inputs the wounded person's gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury, complications information and preliminary treatment operations into the target deep learning model for learning to obtain the wounded person's injury waiting time for treatment, thereby obtaining the injury waiting time for treatment of each wounded person to be transferred.

[0043] S3. The wounded transfer and treatment strategy control center divides the wounded of the same injury category into corresponding injury category sets based on the injury type, and divides M injury category sets in total. For each wounded in each injury category set, the injury level of the wounded is analyzed based on the waiting time for treatment due to his / her injury. The injury levels include light injury level, moderate injury level, severe injury level and critical injury level.

[0044] S4. The control center of the strategy for transferring and treating the wounded prioritizes the wounded in each wounded injury set in descending order of injury level. The wounded of the same injury level have the same priority, and a total of K priorities are sorted.

[0045] S5. The wounded transfer and treatment strategy control center calculates the transfer time for each wounded in the kth priority level in the mth wounded injury category set to each first medical institution corresponding to the injury category level of its injury category, and analyzes the medical institution treatment waiting time corresponding to each first medical institution corresponding to the injury category level of its injury category based on the current injury treatment situation of each first medical institution corresponding to the injury category level of its injury category, takes the maximum value of the medical institution treatment waiting time and transfer time of each first medical institution corresponding to the injury category level of its injury category, and determines the number of first medical institutions whose maximum value is less than the waiting time for treatment of the wounded's injury. When the number is zero, directly use the first medical institution with the shortest transfer time among the first medical institutions corresponding to the injury category level of its injury category as the target first medical institution, arrange for the wounded to be transferred to the target first medical institution, and transmit the wounded's injury information, the waiting time for treatment of the injury and the transfer time to the first medical institution management system corresponding to the target first medical institution, and issue an emergency rescue based on the transfer time. A reminder message is sent to the first medical institution management system so that the target first medical institution can arrange medical staff to receive the injured person and provide immediate emergency rescue; when the number is one, the first medical institution is directly used as the target first medical institution, and the injured person is arranged to be transferred to the target first medical institution, and the injury information of the injured person, the waiting time for treatment and the transfer time are transmitted to the first medical institution management system corresponding to the target first medical institution, and a reception reminder message is sent to the first medical institution management system based on the transfer time so that the target first medical institution can arrange medical staff to receive the injured person; when the number is multiple, the first medical institution corresponding to the shortest transfer time is selected from the first medical institutions that can be delivered as the target first medical institution, and the injured person is arranged to be transferred to the target first medical institution, and the injury information of the injured person, the waiting time for treatment and the transfer time are transmitted to the first medical institution management system corresponding to the target first medical institution, and a reception reminder message is sent to the first medical institution management system based on the transfer time so that the target first medical institution can arrange medical staff to receive the injured person.

[0046] Among them, m takes values ​​of 1-M in sequence, k takes values ​​of 1-K in sequence, and m, M, k and K are all positive integers.

[0047] The following steps are included after step S5: when a certain injured person needs to be transferred again, the first medical institution management system transmits the latest injury information of the injured person to the injured person transfer treatment strategy control center, and the latest injury information includes the information that the injury has been treated and the estimated time that the injury can wait for treatment.

[0048] The wounded transfer and treatment strategy control center calculates the transfer time of the wounded to arrive at each of the multiple transferable second medical institutions corresponding to his injury type, and analyzes the medical institution treatment waiting time corresponding to each second medical institution based on the current injury treatment situation of each second medical institution corresponding to his injury type. The medical institution treatment waiting time and transfer time of each second medical institution corresponding to his injury type are taken as the maximum value, and the number of second medical institutions whose maximum value is less than the estimated time that the wounded's injury can wait for treatment is determined. When the number is zero, the second medical institution with the shortest transfer time among the second medical institutions corresponding to his injury type is directly used as the target second medical institution, and the wounded is arranged to be transferred to the target second medical institution, and the latest injury information of the wounded, the estimated time that the injury can wait for treatment and the transfer time are transmitted to the second medical institution management system corresponding to the target second medical institution, and an emergency rescue reminder message is sent to the second medical institution management system based on the transfer time so that the target second medical institution Arrange medical staff to receive the injured person and provide immediate emergency rescue; when the number is one, directly use this second medical institution as the target second medical institution, arrange for the injured person to be transferred to the target second medical institution, and transmit the injured person's latest injury information, estimated time for waiting for treatment and transfer time to the second medical institution management system corresponding to the target second medical institution, and send a reception reminder message to the second medical institution management system based on the transfer time so that the target second medical institution can arrange medical staff to receive the injured person; when the number is multiple, select the second medical institution with the shortest transfer time from the accessible second medical institutions as the target second medical institution, arrange for the injured person to be transferred to the target second medical institution, and transmit the injured person's latest injury information, estimated time for waiting for treatment and transfer time to the second medical institution management system corresponding to the target second medical institution, and send a reception reminder message to the second medical institution management system based on the transfer time so that the target second medical institution can arrange medical staff to receive the injured person.

[0049] Although the specific embodiments of the present invention are described above, those skilled in the art should understand that these are only examples, and the protection scope of the present invention is defined by the appended claims. Those skilled in the art may make various changes or modifications to these embodiments without departing from the principles and essence of the present invention, but these changes and modifications all fall within the protection scope of the present invention.

Claims

1. A deep learning-based control system for batch casualty transfer and treatment strategy, characterized in that: It includes a first medical institution management system corresponding to multiple first medical institutions corresponding to various types of injuries and a wounded transfer and treatment strategy control center; The wounded transfer and treatment strategy control center is used to obtain the injury information of each wounded in the batch of wounded to be transferred, the injury information includes the wounded's identity information, gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury severity, complication information and preliminary treatment operations of the injury, and the complication information includes whether there are complications and the specific complications when there are complications; The wounded transfer and treatment strategy control center is also used to input the wounded's gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury, complication information and preliminary treatment operations into the trained deep learning model for learning to obtain the wounded's injury waiting time for treatment, thereby obtaining the injury waiting time for treatment of each wounded to be transferred; The patient transfer and treatment strategy control center is further used to classify the patients belonging to the same injury category into corresponding patient injury category sets based on the injury category, and a total of M patient injury category sets are divided. For each patient in each patient injury category set, the injury category level of the patient is analyzed based on the waiting time for treatment of the patient's injury, and the injury category levels include light injury category level, moderate injury category level, severe injury category level and critical injury category level; The wounded transfer and treatment strategy control center is also used to prioritize the wounded in each wounded injury set in descending order of injury level. Wounded with the same injury level have the same priority, and a total of K priorities are sorted. The wounded transfer treatment strategy control center is also used to calculate, for the kth priority level in the mth wounded injury category set, the transfer time of each wounded in the kth priority level in the mth wounded injury category set to each first medical institution corresponding to the injury category level of its injury category, and based on the current injury treatment situation of each first medical institution corresponding to the injury category level of its injury category, the medical institution treatment waiting time and transfer time of each first medical institution corresponding to the injury category level of its injury category are analyzed, and the maximum value of the medical institution treatment waiting time and transfer time of each first medical institution corresponding to the injury category level of its injury category is taken, and the maximum value is less than the injury condition of the wounded. The first medical institution with the waiting time for treatment is the first medical institution that can be delivered. One of the first medical institutions that can be delivered is selected as the target first medical institution, and the injured person is arranged to be transferred to the target first medical institution. The injured person's injury information, the waiting time for treatment and the transfer time are transmitted to the first medical institution management system corresponding to the target first medical institution, and a reminder message is sent to the first medical institution management system based on the transfer time so that the target first medical institution can arrange medical staff to receive the injured person, where m takes values ​​of 1-M, k takes values ​​of 1-K, and m, M, k and K are all positive integers.

2. The deep learning-based batch wounded transfer and treatment strategy control system according to claim 1, characterized in that: The wounded transfer and treatment strategy control center is also used to build different deep learning models and initialize the hyperparameters of each deep learning model; The wounded transfer treatment strategy control center is also used to collect the injury information of multiple historical wounded, and professional doctors mark the waiting time for treatment of each historical wounded based on the injury information of each historical wounded, and construct a sample data set based on the gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury, complication information and preliminary treatment operations of the injury and the corresponding waiting time for treatment of the injury in each injury information of each historical wounded; The wounded transfer and treatment strategy control center is also used to use the sample training subset divided from the sample data set to iteratively train each deep learning model, continuously adjust the hyperparameters until the optimal solution is obtained, and use the sample test subset divided from the sample data set to test each trained deep learning model and obtain the accuracy, and then use repeated random subsampling cross-validation technology to train and test each deep learning model multiple times, so that the injury information of each historical wounded person is used as model training data and model test data, so as to obtain the accuracy of each deep learning model after multiple tests, calculate the average accuracy corresponding to each deep learning model, and use the deep learning model corresponding to the maximum average accuracy as the target deep learning model; The wounded person transfer and treatment strategy control center is also used to input the wounded person's gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury, complication information and preliminary injury processing operations into the target deep learning model for learning to obtain the wounded person's injury waiting time for treatment, thereby obtaining the injury waiting time for treatment of each wounded person to be transferred.

3. The deep learning-based batch wounded transfer and treatment strategy control system according to claim 1, characterized in that: The wounded transfer and treatment strategy control center is also used to determine, for each wounded in the kth priority level in the mth wounded injury type set, the number of first medical institutions whose maximum value is less than the time that the wounded can wait for treatment. When the number is zero, directly use the first medical institution with the shortest transfer time among the first medical institutions corresponding to the injury type level of the wounded as the target first medical institution, arrange for the wounded to be transferred to the target first medical institution, transmit the wounded's injury information, the time that can wait for treatment and the transfer time to the first medical institution management system corresponding to the target first medical institution, and send an emergency rescue reminder message to the first medical institution management system based on the transfer time so that the target first medical institution can arrange medical staff to receive the wounded and provide immediate emergency rescue; When the number is one, this first medical institution is directly used as the target first medical institution, and the injured person is arranged to be transferred to the target first medical institution, and the injury information, the waiting time for treatment and the transfer time of the injured person are transmitted to the first medical institution management system corresponding to the target first medical institution, and a reception reminder message is sent to the first medical institution management system based on the transfer time so that the target first medical institution can arrange medical staff to receive the injured person; when the number is multiple, the first medical institution corresponding to the shortest transfer time is selected from the first medical institutions that can be delivered as the target first medical institution, and the injured person is arranged to be transferred to the target first medical institution, and the injury information, the waiting time for treatment and the transfer time of the injured person are transmitted to the first medical institution management system corresponding to the target first medical institution, and a reception reminder message is sent to the first medical institution management system based on the transfer time so that the target first medical institution can arrange medical staff to receive the injured person.

4. The deep learning-based batch wounded transfer and treatment strategy control system according to claim 1, characterized in that: The system further includes a second medical institution management system corresponding to a plurality of transferable second medical institutions corresponding to each injury type managed by each first medical institution management system; The first medical institution management system is used to transmit the latest injury information of a wounded person who needs to be transferred to the wounded transfer treatment strategy control center when the wounded person needs to be transferred again, and the latest injury information includes the information that the injury has been treated and the estimated time that the injury can wait for treatment; The wounded transfer and treatment strategy control center is also used to calculate the transfer time of the re-transferred wounded to each of the multiple transferable second medical institutions corresponding to the injury type, and analyze the medical institution treatment waiting time of each second medical institution corresponding to the injury type based on the current injury treatment situation of each second medical institution corresponding to the injury type, take the maximum value of the medical institution treatment waiting time and transfer time of each second medical institution corresponding to the injury type, and judge the number of second medical institutions whose maximum value is less than the estimated time that the wounded's injury can wait for treatment. When the number is zero, directly use the second medical institution with the shortest transfer time among the second medical institutions corresponding to the injury type as the target second medical institution, arrange the transfer of the wounded to the target second medical institution, and transmit the latest injury information of the wounded, the estimated time that the injury can wait for treatment and the transfer time to the second medical institution management system corresponding to the target second medical institution, and send an emergency rescue reminder message to the second medical institution management system based on the transfer time so that the target second medical institution can arrange medical staff to receive the wounded and provide immediate emergency rescue; When the number is one, this second medical institution is directly used as the target second medical institution, and the injured person is arranged to be transferred to the target second medical institution, and the latest injury information of the injured person, the estimated time for waiting for treatment and the transfer time are transmitted to the second medical institution management system corresponding to the target second medical institution, and a reception reminder message is sent to the second medical institution management system based on the transfer time so that the target second medical institution can arrange medical staff to receive the injured person; when the number is multiple, the second medical institution corresponding to the shortest transfer time is selected from the accessible second medical institutions as the target second medical institution, and the injured person is arranged to be transferred to the target second medical institution, and the latest injury information of the injured person, the estimated time for waiting for treatment and the transfer time are transmitted to the second medical institution management system corresponding to the target second medical institution, and a reception reminder message is sent to the second medical institution management system based on the transfer time so that the target second medical institution can arrange medical staff to receive the injured person.

5. The deep learning-based batch wounded transfer and treatment strategy control system according to claim 1, characterized in that: When the injured person has multiple types of injuries, the most serious injury type is selected as the injury type in the injury information.

6. A method for controlling the strategy of transferring and treating batches of wounded based on deep learning, characterized in that: It includes the following steps: S1. The wounded transfer and treatment strategy control center obtains the injury information of each wounded in the batch of wounded to be transferred. The injury information includes the wounded's identity information, gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury severity, complication information and preliminary treatment operations. The complication information includes whether there are complications and the specific complications if there are complications; S2. The wounded transfer and treatment strategy control center inputs the wounded's gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury, complication information and preliminary treatment operations into the trained deep learning model for learning to obtain the wounded's injury waiting time for treatment, thereby obtaining the wounded's injury waiting time for treatment for each wounded to be transferred; S3, the patient transfer and treatment strategy control center divides the patients of the same injury category into corresponding patient injury category sets based on the injury category, and divides a total of M patient injury category sets. For each patient in each patient injury category set, the injury category level of the patient is analyzed based on the waiting time for treatment of the patient's injury, and the injury category levels include light injury category level, moderate injury category level, severe injury category level and critical injury category level; S4. The wounded transfer and treatment strategy control center prioritizes the wounded in each wounded injury set in descending order of injury level. Wounded with the same injury level have the same priority, and a total of K priorities are sorted. S5. The patient transfer treatment strategy control center calculates the transfer time of each patient in the kth priority level in the mth patient injury category set to each first medical institution corresponding to the injury category level of the patient, and analyzes the medical institution treatment waiting time of each first medical institution corresponding to the current injury category treatment situation of each first medical institution corresponding to the injury category level of the patient based on the kth priority level in the mth patient injury category set. The maximum value of the medical institution treatment waiting time and transfer time of each first medical institution corresponding to the injury category level of the patient is taken, and the maximum value is less than the injury condition of the patient. The first medical institution with the waiting time for treatment is the first medical institution that can be delivered. One of the first medical institutions that can be delivered is selected as the target first medical institution, and the injured person is arranged to be transferred to the target first medical institution. The injured person's injury information, the waiting time for treatment and the transfer time are transmitted to the first medical institution management system corresponding to the target first medical institution, and a reminder message is sent to the first medical institution management system based on the transfer time so that the target first medical institution can arrange medical staff to receive the injured person, where m takes values ​​of 1-M, k takes values ​​of 1-K, and m, M, k and K are all positive integers.

7. The method for controlling the strategy of transferring and treating the wounded in batches based on deep learning as claimed in claim 6, characterized in that: Before step S1, the following steps are included: the wounded transfer and treatment strategy control center constructs different deep learning models and initializes the hyperparameters of each deep learning model; The wounded transfer treatment strategy control center collects the injury information of multiple historical wounded, and professional doctors mark the waiting time for treatment of each historical wounded based on the injury information of each historical wounded, and constructs a sample data set based on the gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury, complication information and preliminary treatment operations of the injury and the corresponding waiting time for treatment of the injury in each injury information of each historical wounded; The wounded transfer and treatment strategy control center uses the sample training subset divided from the sample data set to iteratively train each deep learning model, continuously adjusts the hyperparameters until the optimal solution is obtained, and uses the sample test subset divided from the sample data set to test each trained deep learning model and obtain the accuracy, and then uses repeated random subsampling cross-validation technology to train and test each deep learning model multiple times, so that the injury information of each historical wounded person is used as model training data and model test data, thereby obtaining the accuracy of each deep learning model after multiple tests, calculating the average accuracy corresponding to each deep learning model, and taking the deep learning model corresponding to the maximum average accuracy as the target deep learning model; In step S2, the wounded person transfer and treatment strategy control center inputs the wounded person's gender, age, body temperature, blood pressure, blood oxygen, heart rate, injury time, injury type, injury part, injury type, injury, complication information and preliminary processed injury operations into the target deep learning model for learning to obtain the wounded person's injury waiting time for treatment, thereby obtaining the injury waiting time for treatment of each wounded person to be transferred.

8. The method for controlling the strategy of transferring and treating the wounded in batches based on deep learning as claimed in claim 6, characterized in that: In step S5, the wounded transfer treatment strategy control center determines the number of first medical institutions whose maximum value is less than the waiting time for treatment of the wounded in the mth wounded injury category set for each wounded in the kth priority category. When the number is zero, the first medical institution corresponding to the injury category level of the wounded with the shortest transfer time is directly used as the target first medical institution, and the wounded is transferred to the target first medical institution. The wounded's injury information, the waiting time for treatment and the transfer time are transmitted to the first medical institution management system corresponding to the target first medical institution, and an emergency rescue reminder message is sent to the first medical institution management system based on the transfer time so that the target first medical institution can arrange medical staff to receive the wounded and provide immediate emergency rescue. When the number is one, this first medical institution is directly used as the target first medical institution, and the injured person is arranged to be transferred to the target first medical institution, and the injury information, the waiting time for treatment and the transfer time of the injured person are transmitted to the first medical institution management system corresponding to the target first medical institution, and a reception reminder message is sent to the first medical institution management system based on the transfer time so that the target first medical institution can arrange medical staff to receive the injured person; when the number is multiple, the first medical institution corresponding to the shortest transfer time is selected from the first medical institutions that can be delivered as the target first medical institution, and the injured person is arranged to be transferred to the target first medical institution, and the injury information, the waiting time for treatment and the transfer time of the injured person are transmitted to the first medical institution management system corresponding to the target first medical institution, and a reception reminder message is sent to the first medical institution management system based on the transfer time so that the target first medical institution can arrange medical staff to receive the injured person.

9. The method for controlling the strategy of transferring and treating the wounded in batches based on deep learning as claimed in claim 6, characterized in that: The following steps are included after step S5: when a certain injured person needs to be transferred again, the first medical institution management system transmits the latest injury information of the injured person to the injured person transfer treatment strategy control center, wherein the latest injury information includes the information that the injury has been treated and the estimated time that the injury can wait for treatment; The wounded person transfer and treatment strategy control center calculates the transfer time of the re-transferred wounded person to each of the multiple transferable second medical institutions corresponding to his / her injury type, and analyzes the medical institution treatment waiting time of each second medical institution corresponding to his / her injury type based on the current injury treatment situation of each second medical institution corresponding to his / her injury type. The medical institution treatment waiting time and transfer time of each second medical institution corresponding to his / her injury type are taken as the maximum value, and the number of second medical institutions whose maximum value is less than the estimated time that the wounded person's injury can wait for treatment is determined. When the number is zero, the second medical institution with the shortest transfer time among the second medical institutions corresponding to his / her injury type is directly used as the target second medical institution, and the wounded person is arranged to be transferred to the target second medical institution. The latest injury information of the wounded person, the estimated time that the injury can wait for treatment, and the transfer time are transmitted to the second medical institution management system corresponding to the target second medical institution, and an emergency rescue reminder message is sent to the second medical institution management system based on the transfer time so that the target second medical institution The second medical institution arranges medical personnel to receive the injured person and provides immediate emergency rescue; when the number is one, the second medical institution is directly used as the target second medical institution, and the injured person is arranged to be transferred to the target second medical institution, and the latest injury information of the injured person, the estimated time for waiting for treatment and the transfer time are transmitted to the second medical institution management system corresponding to the target second medical institution, and a reception reminder message is sent to the second medical institution management system based on the transfer time so that the target second medical institution arranges medical personnel to receive the injured person; when the number is multiple, the second medical institution with the shortest transfer time is selected from the accessible second medical institutions as the target second medical institution, and the injured person is arranged to be transferred to the target second medical institution, and the latest injury information of the injured person, the estimated time for waiting for treatment and the transfer time are transmitted to the second medical institution management system corresponding to the target second medical institution, and a reception reminder message is sent to the second medical institution management system based on the transfer time so that the target second medical institution arranges medical personnel to receive the injured person.

10. The method for controlling the strategy of transferring and treating the wounded in batches based on deep learning as claimed in claim 6, characterized in that: In step S1, when the injured person has multiple types of injuries, the most serious injury is selected as the injury type in the injury information.