Remote severe case rapid response management platform

Abstract

The invention discloses a remote severe case rapid response management platform, which relates to the technical field of medical treatment and comprises a field information acquisition module, an illness state diagnosis module and a remote medical module. According to the invention, the AI voice packet related to the candidate disease can be intelligently retrieved from the AI voice library, voice conversation is carried out, the second audio and video information is further obtained, the target disease and the target disease level are accurately determined, emergency rescue guidance and remote medical guidance processes can be triggered according to the target disease and the disease level, and the emergency rescue guidance and remote medical guidance efficiency is improved. The disease diagnosis efficiency and accuracy are improved, and timely and effective medical services can be provided for patients.

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and in particular to a remote critical illness rapid response management platform. Background Art

[0002] In the current medical system, the treatment of critically ill patients faces many challenges. Time urgency and shortage of medical resources are the two most prominent issues. The traditional medical model relies on patients to go to the hospital on their own, but in emergency situations, such as when patients are in remote areas or in places with inconvenient transportation, they may not be able to reach the hospital in time for treatment. In addition, even if patients can reach the hospital in time, due to the limited medical resources, the hospital may not be able to immediately provide patients with sufficient medical resources, such as expert consultations, special drugs or advanced equipment.

[0003] During the diagnosis process, doctors often rely on patients' oral statements and paper medical records to understand the condition. However, this method has many shortcomings. On the one hand, the patient's oral statement may be inaccurate due to vague memory, language barriers or emotional tension; on the other hand, paper medical records may lack diagnostic information due to poor preservation, untimely information updates or unclear handwriting. These problems may affect the doctor's accurate judgment of the condition, thereby affecting the formulation and implementation of the treatment plan.

[0004] In order to solve these problems, telemedicine service platforms have emerged in recent years. These platforms use Internet technology to achieve remote communication and exchanges between doctors and patients. However, most of the existing telemedicine service platforms only provide simple online consultation or appointment registration services, and lack the ability to monitor patients' conditions in real time and respond quickly. When the patient's condition changes, the platform is often unable to obtain relevant information in a timely manner and cannot quickly collect measures based on the severity of the patient's symptoms. Summary of the invention

[0005] In view of the defects in the prior art, the present invention provides a remote critical illness rapid response management platform.

[0006] The technical solution of the present invention is: a remote critical care rapid response management platform, the platform comprising: A site information collection module, used to obtain the patient's condition information, the condition information including vital sign information and first audio and video information; A disease diagnosis module, for determining the candidate disease and disease level of the patient based on the disease information and in combination with the historical medical record information in the medical database, retrieving an AI voice package related to the candidate disease from the AI ​​voice library according to the disease level, acquiring second audio and video information according to the AI ​​voice package, determining a target disease and a target disease level according to the second audio and video information; and determining the disease level according to the target disease and the target disease level; The telemedicine module triggers emergency rescue guidance and telemedicine guidance according to the target disease and disease level.

[0007] Preferably, based on the condition information and in combination with historical medical record information in a medical database, the candidate conditions and condition levels of the patient are determined, including: Preprocessing the medical condition information, extracting vital signs and first audio and video features from the preprocessed medical condition information, extracting medical record features from historical medical record information in a medical database, The vital sign feature, the first audio and video feature and the medical record feature are input into a diagnosis model to obtain candidate symptoms and candidate symptom levels, and the symptom level is determined according to the candidate symptom level and the candidate symptom weight.

[0008] Preferably, retrieving an AI voice package related to the candidate disease from an AI voice library according to the disease level includes: Determining a voice packet fusion weight according to the candidate disease level; When the disease level is greater than a set value, the first AI voice sub-data packets corresponding to each candidate disease are retrieved from the first AI voice database for fusion according to the voice packet fusion weight; otherwise, the first AI voice sub-data packets corresponding to each candidate disease are retrieved from the second AI voice database for fusion to obtain an AI voice package.

[0009] Preferably, obtaining the second audio and video information according to the AI ​​voice package, and determining the target symptom and the target symptom level according to the second audio and video information include: Conduct a voice conversation with the site according to the AI ​​voice package to obtain second audio and video information; Preprocessing the second audio and video information, and extracting second audio and video features from the preprocessed second audio and video information; The second audio and video features are input into a diagnostic model to obtain a target symptom and a target symptom level.

[0010] Preferably, the diagnostic model is an improved LSTM model, whose input layer includes two extension layers, which are respectively used to receive medical record features and second audio and video features. A switch is provided before the extension layer, and the switch is used to execute an extension layer skip instruction when there is no extension layer input.

[0011] Preferably, remote medical guidance is triggered according to the target disease and the disease level, including: Determine a set of optional doctors according to the target disease and the disease level; The most matching doctor is selected from the set of available doctors as the target doctor for telemedicine guidance.

[0012] Preferably, selecting the most matching doctor from the optional doctor set as the target doctor for remote medical guidance includes: Obtain the medical resource margin of the doctor's hospital and the distance between the patient and the hospital; The medical resource margin of the hospital is corrected according to the medical resource margin of the adjacent hospitals and the distance between hospitals to obtain the target medical resource margin; A doctor matching score is determined according to the target medical resource margin and the distance, and the doctor with the highest matching score is selected as the target doctor for remote medical guidance.

[0013] Preferably, the medical resource margin of a hospital is corrected according to the medical resource margin of adjacent hospitals and the distance between hospitals to obtain a target medical resource margin, including: Determining whether the adjacent hospital is one of the best matching hospitals, the best matching hospital being a hospital within a preset distance from the patient; If yes, then calculating the ratio between the number of best matching hospitals among the adjacent hospitals and the total number of best matching hospitals, and determining a correction coefficient according to the ratio, the difference between the medical resource margin of the adjacent hospital and the corresponding threshold value of the medical resource margin of the adjacent hospital, and the distance between the hospitals; If not, a correction coefficient is determined based on the difference between the medical resource margin of the adjacent hospital and the corresponding margin threshold of the adjacent hospital and the distance between the hospitals; The medical resource margin is corrected according to the correction coefficient to obtain the target medical resource margin.

[0014] Preferably, the medical resource margin of the hospital is corrected according to the medical resource margin of the adjacent hospitals and the distance between the hospitals, and the formula for obtaining the target medical resource margin is: in, is the original medical resource margin of the doctor’s hospital, is the number of neighboring hospitals, For the The weight of the neighboring hospital, For the The correction coefficient of the neighboring hospitals on the medical resource margin of the doctor's hospital; For the The indicator variable of whether the neighboring hospital is the best matching hospital. If not, then P i =1, if so, then P i A value less than one is determined by the ratio of the number of neighboring hospitals that are the best matching hospitals to the total number of the best matching hospitals and the location dispersion of the best matching hospitals, indicating its support potential. For the The medical resource adequacy of the neighboring hospitals, For the The medical resource margin threshold of adjacent hospitals, For the The distance between the neighboring hospital and the doctor's hospital, is a distance attenuation factor.

[0015] The beneficial effects of the present invention are mainly reflected in: the present invention provides a remote critical rapid response management platform, the platform includes a field information acquisition module for obtaining the patient's condition information, the condition information includes vital sign information and the first audio and video information, a module for determining the patient's candidate disease and disease level based on the condition information and in combination with the historical medical record information in the medical database, and retrieving the AI ​​voice package related to the candidate disease from the AI ​​voice library according to the disease level, obtaining the second audio and video information according to the AI ​​voice package, and determining the target disease and the target disease level according to the second audio and video information; a disease diagnosis module for determining the disease level according to the target disease and the target disease level, and a telemedicine module for triggering emergency rescue guidance and telemedicine guidance according to the target disease and the disease level. The present invention can intelligently retrieve the AI ​​voice package related to the candidate disease from the AI ​​voice library, conduct voice dialogue, and further obtain the second audio and video information to accurately determine the target disease and the target disease level. According to the target disease and the disease level, the emergency rescue guidance and telemedicine guidance process can be triggered. The present invention improves the efficiency and accuracy of disease diagnosis and can provide patients with timely and effective medical services. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the description of the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0017] Figure 1 A schematic diagram of the structure of a remote critical care rapid response management platform provided in an embodiment of the present invention. DETAILED DESCRIPTION The following embodiments of the technical solution of the present invention are described in detail in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and are therefore only used as examples, and cannot be used to limit the protection scope of the present invention. It should be noted that, unless otherwise specified, the technical terms or scientific terms used in this application should have the common meanings understood by those skilled in the art to which the present invention belongs.

[0018] like Figure 1 As shown, a remote critical care rapid response management platform includes: The on-site information collection module is used to obtain the patient's condition information, which includes vital sign information and first audio and video information.

[0019] Specifically, the on-site information collection module is equipped with a number of vital sign sensors, high-definition cameras, microphones and other collection devices to comprehensively collect patient information, among which vital sign information includes body temperature, heart rate, blood pressure, etc.; the first audio and video information includes the patient's self-introduction, auditory symptoms, visual symptoms, etc.

[0020] The condition diagnosis module is used to determine the patient's candidate conditions and condition levels based on the condition information and in combination with historical medical record information in a medical database, retrieve an AI voice package related to the candidate condition from an AI voice library according to the condition level, obtain second audio and video information according to the AI ​​voice package, determine a target condition and a target condition level according to the second audio and video information; and determine the condition level according to the target condition and the target condition level.

[0021] The disease diagnosis module is an intelligent platform that integrates data analysis, information retrieval and interactive feedback functions, and is mainly used to assist medical workers in assessing disease conditions.

[0022] First, the module will match the historical medical records information. The module will compare the medical condition information with the historical medical records information in the medical database. The medical database is a huge knowledge base that stores a large amount of historical data of cases, including disease descriptions, diagnosis results, treatment plans, etc. Then, based on the disease information and the historical medical records information in the medical database, the candidate diseases and disease levels of the patient are determined.

[0023] Specifically, based on the medical condition information and combined with the historical medical record information in the medical database, the candidate diseases and disease levels of the patient are determined, including: preprocessing the medical condition information, extracting vital signs and first audio and video features from the preprocessed medical condition information; extracting medical record features from the historical medical record information in the medical database, inputting the vital sign features, the first audio and video features, and the medical record features into a diagnostic model to obtain candidate diseases and candidate disease levels, and determining the disease level according to the candidate disease levels and candidate disease weights.

[0024] Specifically, the diagnostic model is an improved LSTM model (Fast-LSTM), and its input layer includes two extension layers, which are used to receive medical record features and second audio and video features respectively. A switch is set before the extension layer, and the switch is used to execute the extension layer skip instruction when there is no extension layer input.

[0025] More specifically, first, an additional input node is introduced into the input layer of Fast-LSTM to receive the preprocessed medical record features. The medical record features are converted into a fixed-length vector through the preprocessing step and used as the input of the extended input layer. Then, the information received by the extended input layer is fused with the previous hidden state to generate an extended context vector. The extended context vector is used to calculate the output of the gated unit of Fast-LSTM, including the input gate, forget gate, output gate, and candidate memory unit. A dynamic gating mechanism is introduced to dynamically adjust the output of the gated unit according to the current input and the previous hidden state to better capture the temporal dependency and nonlinear relationship of the disease information. Then, the output of Fast-LSTM is used as a feature vector and input into a classifier (such as a softmax classifier) ​​to predict candidate diseases. According to the output of the classifier and the preset disease level classification standard, the candidate disease and disease level of the patient are determined.

[0026] The Fast-LSTM provided in the embodiment of the present invention is expanded on the basis of the original model, which reduces the time of model training and improves the training efficiency.

[0027] In an embodiment of the present invention, an AI voice package related to the candidate disease is retrieved from an AI voice database according to the disease level, second audio and video information is obtained according to the AI ​​voice package, and a target disease and a target disease level are determined according to the second audio and video information, including: determining a voice package fusion weight according to the candidate disease level; when the disease level is greater than a set value, retrieving the first AI voice sub-data package corresponding to each candidate disease from the first AI voice database according to the voice package fusion weight for fusion, otherwise, retrieving the first AI voice sub-data package corresponding to each candidate disease from the second AI voice database for fusion to obtain an AI voice package; conducting a voice conversation with the scene according to the AI ​​voice package to obtain second audio and video information; preprocessing the second audio and video information, and extracting second audio and video features from the preprocessed second audio and video information; inputting the second audio and video features into a diagnostic model to obtain a target disease and a target disease level.

[0028] The disease level is an important indicator for assessing the urgency of a patient's condition. In this process, the higher the disease level, the more urgent the patient's condition is, and diagnosis and treatment are needed as soon as possible. On the contrary, the lower the disease level, the less urgent the patient's condition is, and more time is available to obtain detailed and accurate disease information. According to the disease level, the platform will retrieve AI voice packages related to candidate diseases from the AI ​​voice library. The AI ​​voice packages here contain voice dialogue content related to different diseases, which are used for voice interaction with patients.

[0029] When selecting an AI voice package, the platform will also determine the voice package fusion weight based on the level of the disease. This weight is used to determine the importance of different AI voice sub-packets in the fusion process. If the disease level is greater than the preset maximum value (that is, the patient's condition is urgent), the platform will select a relatively streamlined AI voice package for fusion. This is because in an emergency, the pre-diagnosis time needs to be shortened so that follow-up treatment can be carried out as soon as possible. The streamlined AI voice package contains the necessary conversation content, but removes redundant and unnecessary parts. If the disease level is less than or equal to the preset maximum value (that is, the patient's condition is relatively non-urgent), the platform will select a relatively accurate and detailed AI voice package for fusion. This is to ensure the effectiveness of the pre-diagnosis and obtain the patient's condition information through more detailed and accurate conversation content.

[0030] Using the selected AI voice package, the platform will have a voice conversation with the patient. This conversation process will generate second audio and video information, including the patient's voice response and possible video images (if the platform supports video conversations). The acquired second audio and video information will be preprocessed, including denoising, sound quality enhancement and other steps to improve the accuracy of subsequent processing. The preprocessed second audio and video information will be used to extract second audio and video features. These features contain information related to the patient's condition, such as the tone, speed, volume of the voice, and possible actions and expressions in the video. The extracted second audio and video features will be input into the diagnostic model. This model has been trained to determine the target disease and the target disease level based on the input audio and video features.

[0031] Through the output of the diagnostic model, the platform can determine the patient's specific symptoms and symptom level, and provide guidance for subsequent treatment. The voice packet fusion weight is used to determine the number of calls for the first AI voice sub-data packet, and a random call method is used during scheduling.

[0032] Of course, in order to further improve the efficiency of pre-processing, when the disease level is greater than the preset maximum value, the process of triggering the second audio and video acquisition is stopped, and the emergency medical rescue process is directly triggered.

[0033] The telemedicine module is used to trigger emergency rescue guidance and telemedicine guidance according to the target disease and disease level.

[0034] Specifically, for patients with higher levels of illness, the telemedicine module will immediately trigger the emergency rescue guidance and telemedicine guidance process. This may include sending a distress signal to the emergency center, providing emergency self-rescue guidance (such as cardiopulmonary resuscitation, hemostasis, etc.), and coordinating nearby medical resources (such as ambulances, emergency personnel, etc.). During the emergency rescue process, the telemedicine module can maintain real-time communication with the patient and provide continuous guidance and support. At the same time, it can also monitor the patient's vital signs so that the rescue strategy can be adjusted in time.

[0035] Specifically, for patients with a low level of illness or in a stable period, the telemedicine module only triggers the telemedicine process and provides detailed disease management suggestions, which may include drug treatment, lifestyle adjustments, dietary suggestions, etc. At the same time, remote consultation and remote consultation services are provided to patients. Doctors can provide more professional and personalized medical advice based on the patient's specific situation.

[0036] Specifically, in order to ensure that patients can be quickly matched with doctors for treatment, the invention embodiment triggers remote medical guidance for patients with higher levels of illness, including: determining a set of optional doctors based on the target disease and the level of illness; selecting the most matching doctor from the set of optional doctors as the target doctor for remote medical guidance.

[0037] In order to facilitate subsequent treatment in the hospital, the present invention considers the distance and hospital medical resources to match the patient with a suitable doctor to achieve remote medical guidance.

[0038] Specifically, selecting the most matching doctor from the set of optional doctors as the target doctor for remote medical guidance includes: obtaining the medical resource margin of the doctor's hospital and the distance between the patient and the hospital; correcting the hospital's medical resource margin according to the medical resource margin of adjacent hospitals and the distance between hospitals to obtain the target medical resource margin; determining the doctor matching score according to the target medical resource margin and the distance, and selecting the doctor with the highest matching score as the target doctor for remote medical guidance.

[0039] In an embodiment of the present invention, the medical resource margin of a hospital is corrected according to the medical resource margin of adjacent hospitals and the distance between hospitals to obtain a target medical resource margin, including: determining whether the adjacent hospital is one of the best matching hospitals, the best matching hospital being a hospital whose distance to the patient is within a preset range; if so, calculating the ratio between the number of best matching hospitals in the adjacent hospitals and the total number of best matching hospitals, and determining a correction coefficient according to the ratio, the difference between the medical resource margin of the adjacent hospital and the corresponding medical resource margin threshold of the adjacent hospital, and the distance between the hospitals; if not, determining a correction coefficient according to the difference between the medical resource margin of the adjacent hospital and the corresponding medical resource margin threshold of the adjacent hospital, and the distance between the hospitals; and correcting the medical resource margin according to the correction coefficient to obtain a target medical resource margin.

[0040] Specifically, the medical resource margin of the hospital is corrected according to the medical resource margin of the adjacent hospitals and the distance between hospitals, and the formula for obtaining the target medical resource margin is: in, is the original medical resource margin of the doctor’s hospital, is the number of neighboring hospitals, For the The weight of the neighboring hospital, For the The correction coefficient of the neighboring hospitals on the medical resource margin of the doctor's hospital; For the The indicator variable of whether the neighboring hospital is the best matching hospital. If not, then P i =1, if so, then P i A value less than one is determined by the ratio of the number of neighboring hospitals that are the best matching hospitals to the total number of the best matching hospitals and the location dispersion of the best matching hospitals, indicating its support potential. For the The medical resource adequacy of the neighboring hospitals, For the The medical resource margin threshold of adjacent hospitals is a reasonable medical resource margin standard. For the The distance between the neighboring hospital and the doctor's hospital, It is a distance attenuation factor, which is used to reflect the impact of distance on the call of medical resources.

[0041] When the adjacent hospital is the best matching hospital, the medical resources of the adjacent hospital are called upon. At this time, patients adjacent to the patient will go to the hospital selected by the patient, resulting in a decrease in the medical resource margin of the hospital selected by the patient. Therefore, the embodiment of the present invention studies whether the adjacent hospital is the best matching hospital. If it is the best matching hospital, a value less than one determined according to the ratio of the number of best matching hospitals to the total number of best matching hospitals and the location dispersion of the best matching hospitals is used as the hospital's support capacity.

[0042] It should be noted that the specific formula for determining a value less than one as the hospital's support capacity based on the ratio of the number of best matching hospitals to the total number of best matching hospitals and the dispersion of the locations of the best matching hospitals can be to determine a value based on the ratio and the dispersion, normalize the value, and then subtract the value from 1.

[0043] It should also be noted that the hospital weight is preset. In one embodiment, the formula for setting the weight is: Among them, α, β, and γ are weight coefficients of distance, medical quality, and complementarity of medical resources, respectively. These coefficients can be adjusted according to actual conditions. is the medical quality score of the i-th adjacent hospital. This score can be obtained based on factors such as the hospital's medical level, the doctor's professional ability, and patient satisfaction. is the complementarity score between the i-th adjacent hospital and the doctor’s hospital in terms of medical resources. This score can be comprehensively derived based on factors such as the type, quantity, and quality of the medical resources of the two hospitals.

[0044] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A remote critical care rapid response management platform, characterized in that: include: A site information collection module, used to obtain the patient's condition information, the condition information including vital sign information and first audio and video information; A disease diagnosis module, for determining the candidate disease and disease level of the patient based on the disease information and in combination with the historical medical record information in the medical database, retrieving an AI voice package related to the candidate disease from the AI ​​voice library according to the disease level, acquiring second audio and video information according to the AI ​​voice package, determining a target disease and a target disease level according to the second audio and video information; and determining the disease level according to the target disease and the target disease level; The telemedicine module is used to trigger emergency rescue guidance and telemedicine guidance according to the target disease and disease level.

2. A remote critical care rapid response management platform according to claim 1, characterized in that: Based on the condition information, combined with the historical medical record information in the medical database, the candidate conditions and condition levels of the patient are determined, including: Preprocessing the medical condition information, extracting vital sign features and first audio and video features from the preprocessed medical condition information, and extracting medical record features from historical medical record information in a medical database; The vital sign feature, the first audio and video feature and the medical record feature are input into a diagnosis model to obtain candidate symptoms and candidate symptom levels, and the symptom level is determined according to the candidate symptoms and the candidate symptom levels.

3. A remote critical care rapid response management platform according to claim 2, characterized in that: According to the disease level, an AI voice package related to the candidate disease is retrieved from the AI ​​voice library, including: Determining a voice packet fusion weight according to the candidate disease level; When the disease level is greater than a set value, the first AI voice sub-data packets corresponding to each candidate disease are retrieved from the first AI voice database for fusion according to the voice packet fusion weight; otherwise, the first AI voice sub-data packets corresponding to each candidate disease are retrieved from the second AI voice database for fusion to obtain an AI voice package.

4. A remote critical care rapid response management platform according to claim 3, characterized in that: Acquiring second audio and video information according to the AI ​​voice package, and determining a target symptom and a target symptom level according to the second audio and video information, including: Conduct a voice conversation with the site according to the AI ​​voice package to obtain second audio and video information; Preprocessing the second audio and video information, and extracting second audio and video features from the preprocessed second audio and video information; The second audio and video features are input into a diagnostic model to obtain a target symptom and a target symptom level.

5. A remote critical care rapid response management platform according to claim 4, characterized in that: The diagnostic model is an improved LSTM model, and its input layer includes two extension layers, which are respectively used to receive medical record features and second audio and video features. A switch is set before the extension layer, and the switch is used to execute the extension layer skip instruction when there is no extension layer input.

6. A remote critical care rapid response management platform according to claim 1, characterized in that: Trigger telemedicine guidance based on target conditions and severity, including: Determine a set of optional doctors according to the target disease and the disease level; The most matching doctor is selected from the set of available doctors as the target doctor for telemedicine guidance.

7. A remote critical care rapid response management platform according to claim 6, characterized in that: Screening the most suitable doctor from the available doctor set as the target doctor for telemedicine guidance includes: Obtain the medical resource margin of the doctor's hospital and the distance between the patient and the hospital; The medical resource margin of the hospital is corrected according to the medical resource margin of the adjacent hospitals and the distance between the hospitals to obtain the target medical resource margin; A doctor matching score is determined according to the target medical resource margin and the distance, and the doctor with the highest matching score is selected as the target doctor for remote medical guidance.

8. A remote critical care rapid response management platform according to claim 1, characterized in that: The medical resource margin of the hospital is corrected according to the medical resource margin of the adjacent hospitals and the distance between the hospitals to obtain the target medical resource margin, including: Determining whether the adjacent hospital is one of the best matching hospitals, the best matching hospital being a hospital within a preset distance from the patient; If yes, then calculating the ratio between the number of best matching hospitals among the adjacent hospitals and the total number of best matching hospitals, and determining a correction coefficient according to the ratio, the difference between the medical resource margin of the adjacent hospital and the corresponding threshold value of the medical resource margin of the adjacent hospital, and the distance between the hospitals; If not, a correction coefficient is determined based on the difference between the medical resource margin of the adjacent hospital and the corresponding medical resource margin threshold of the adjacent hospital and the distance between the hospitals; The medical resource margin is corrected according to the correction coefficient to obtain the target medical resource margin.

9. A remote critical care rapid response management platform according to claim 1, characterized in that: The medical resource margin of the hospital is corrected according to the medical resource margin of the adjacent hospitals and the distance between the hospitals, and the formula for the target medical resource margin is obtained as follows: in, is the original medical resource margin of the doctor’s hospital, is the number of neighboring hospitals, For the The weight of the neighboring hospital, For the The correction coefficient of the neighboring hospitals on the medical resource margin of the doctor's hospital; For the The indicator variable of whether the neighboring hospital is the best matching hospital. If not, then P i =1, if so, then P i A value less than one is determined by the ratio of the number of neighboring hospitals that are the best matching hospitals to the total number of the best matching hospitals and the location dispersion of the best matching hospitals, indicating its support potential. For the The medical resource adequacy of the neighboring hospitals, For the The medical resource margin threshold of adjacent hospitals, For the The distance between the neighboring hospital and the doctor's hospital, is a distance attenuation factor.

Images