Graded diagnosis and treatment resource allocation system based on AI priority pre-auditing mechanism

Abstract

The invention relates to a hierarchical diagnosis and treatment resource allocation system based on an AI priority pre-auditing mechanism, and belongs to the technical field of medical resource allocation, and the system comprises a user input module which is used for receiving initial information of a patient based on an AI applet, and the initial information at least comprises a symptom description, a medical history record, examination data and a diagnosis and treatment stage; the AI intelligent pre-inquiry module is used for performing multiple rounds of questions and answers on the patient according to the initial information to extract an illness state index, inputting the illness state index into a preset risk layering model, and determining an illness state severity / urgency index of the patient; the resource distribution module comprises a special number source pool reserved for experts, and the resource distribution module is used for opening expert numbers for the patients whose illness condition severity / emergency index is greater than an illness condition threshold value; and the referral guidance module is used for recommending a replacement doctor seeing scheme for the patient who does not accord with the open expert number. According to the invention, the technical problem of insufficient fairness and efficiency of an illness state emergency degree registration distribution mechanism caused by structural mismatching of expert resources is solved.

Description

Technical Field

[0001] This invention relates to the field of medical resource allocation technology, and in particular to a hierarchical medical resource allocation system based on an AI priority pre-screening mechanism. Background Technology

[0002] The imbalance between supply and demand for medical resources is particularly pronounced in the area of ​​high-quality specialist resources, especially in the conflict of priority between patients with acute and critical illnesses such as cancer and those undergoing routine follow-up visits. Currently, appointments with renowned specialists are often snapped up in seconds after they are released, leading to an imbalance in resource allocation. Patients with stable conditions who are familiar with the registration rules and procedures occupy a large number of specialist appointments, while patients with newly diagnosed cancers or complex diseases who urgently require priority treatment are forced to delay their golden treatment period due to insufficient registration efficiency. The existing registration system lacks a tiered mechanism based on the urgency of the patient's condition, and only alleviates superficial congestion by increasing scheduling and waiting lists. This fails to fundamentally solve the problem of resource misallocation caused by "speed-based" rather than "condition-based" approaches, resulting in a structural squeeze on the treatment needs of patients with acute and critical illnesses.

[0003] The implementation of the tiered medical service policy faces the dual obstacles of weak primary care capabilities and an inefficient referral mechanism. The two-way referral system suffers from a lack of close collaboration, insufficient medical insurance incentives, and information sharing barriers, hindering the smooth flow of patients between primary and higher-level hospitals. High-quality medical resources are consistently occupied by non-urgent needs, failing to effectively reach truly needed critical and severe cases. Traditional appointment review and on-site registration models are insufficient to meet the demands for efficient and accurate priority tiering. Currently, manual review of appointment requests requires doctors to sift through a massive volume of requests for high-risk cases, which is inefficient and susceptible to subjective influences. On-site appointments result in high rates of wasted patient trips, and last-minute appointments may disrupt the specialist's regular appointment schedule. Existing technologies fail to achieve pre-screening and tiering of patient conditions through intelligent means, and cannot automatically identify critically ill patients and assign them priority for registration, leading to insufficient fairness and efficiency in the allocation of medical resources, and hindering the rapid response to the urgent treatment needs of critically ill patients.

[0004] Therefore, how to use artificial intelligence technology to build an intelligent tiered registration mechanism based on the urgency of the illness, solve the structural mismatch of high-quality expert resources, and promote the implementation of tiered medical services has become an urgent technical problem to be solved in the current medical service field. Summary of the Invention

[0005] In view of this, it is necessary to provide a hierarchical medical resource allocation system based on an AI priority pre-screening mechanism to solve the technical problem of insufficient fairness and efficiency in the existing technology's structural mismatch of expert resources, which leads to the allocation mechanism based on the urgency of the patient's condition.

[0006] To address the aforementioned problems, this invention provides a hierarchical medical resource allocation system based on an AI-based priority pre-screening mechanism, comprising:

[0007] The user input module is used to receive the patient's initial information based on an AI mini-program. The initial information includes at least a description of symptoms, medical history, examination data, and treatment stage.

[0008] The AI-powered intelligent pre-diagnosis module is used to extract disease indicators from patients through multiple rounds of questioning based on the initial information, and input the disease indicators into a preset risk stratification model to determine the severity / urgency index of the patient's condition.

[0009] The resource allocation module includes a dedicated pool of appointment slots reserved for specialists. The resource allocation module is used to open specialist appointments for patients whose disease severity / urgency index is greater than the disease threshold.

[0010] The referral guidance module is used to recommend alternative treatment plans for patients who do not qualify for open specialist appointments.

[0011] In one possible implementation, the AI ​​intelligent pre-diagnosis module includes a dynamic interaction unit and a risk stratification model unit;

[0012] The dynamic interaction unit is used to obtain clinical indicators such as symptom duration and tumor stage based on the patient's initial information and through preset question-and-answer logic;

[0013] The risk stratification model unit is used to classify the patient's clinical indicators based on a fully trained machine learning model and output a level number representing the severity / urgency of the condition, wherein the level number ranges from 1 to 5.

[0014] In one possible implementation, the resource allocation module includes a traffic splitting logic unit and a timeliness control unit;

[0015] The triage logic unit is used to allocate specialist appointments to patients with a level number greater than or equal to 4 based on the relationship between the severity / urgency index and the disease threshold.

[0016] The timeliness control unit is used to allocate a preset waiting confirmation period to patients after they have obtained the qualification to add an expert appointment. If payment is not made after the preset waiting period, the corresponding expert appointment will be automatically released to the dedicated appointment pool.

[0017] In one possible implementation, the referral guidance module includes a triage logic unit and a data connectivity unit;

[0018] The triage logic unit includes defining the severity / urgency index of the illness as level 1, level 2, and level 3, and is used to assign patients with a severity / urgency index of level 3 to the waiting queue or recommend further examination; and to guide patients with a severity / urgency index of less than level 2 to the general outpatient clinic or the Internet medical platform.

[0019] The data connectivity unit is used to push the patient's historical medical records to the corresponding destination based on the triage results of the triage logic unit.

[0020] In one possible implementation, the risk stratification model includes at least indication determination rules, risk assessment rules, and referral suggestion rules;

[0021] The indication determination rule is based on the tumor size, pathological type, and stage information in the initial information to determine whether it meets the indications for surgical or drug treatment;

[0022] The risk assessment rule is to generate a risk level score by weighting the patient's laboratory test results, history of complications, and treatment history.

[0023] The referral prompt rule is that when a preset critical value appears in the initial information, a referral request form containing the patient's basic information and key clinical data is automatically generated and pushed to the designated medical institution interface; the critical value includes at least tumor rupture, bleeding risk, and organ failure indicators.

[0024] In one possible implementation, a memory module is also included, which is used to record the patient's historical treatment stages and corresponding initial information. When the patient enters the follow-up visit stage, the initial information is automatically retrieved for comparison and analysis. If abnormal fluctuations in indicators are detected, a follow-up recommendation is triggered through the preset risk stratification model.

[0025] In one possible implementation, a risk notification module is also included, which is used to generate prompt information when a patient applies for an additional appointment. The prompt information includes notification of the appointment review time, notification of the risk of appointment failure, and notification of the risk of delay in the patient's condition due to appointment failure.

[0026] In one possible implementation, the referral guidance module further includes a multimodal interaction unit that supports outputting guidance instructions in the form of text, voice, and visual charts.

[0027] One possible implementation also includes an anti-abuse module for establishing a user credit blacklist.

[0028] In one possible implementation, the user input module further includes a data verification unit, which performs format verification and logical verification on the collected initial information. If the data is missing or contradictory, the unit sends a supplementary information prompt to the user and pauses the process switching until the data verification passes.

[0029] The beneficial effects of this invention are as follows: First, the initial information of the patient is obtained based on the user input module. Then, based on the AI ​​intelligent pre-diagnosis module, the patient undergoes multiple rounds of targeted questioning based on the symptom description, medical history, examination data, and treatment stage in the initial information to determine the patient's condition indicators. Furthermore, based on a preset risk stratification model, the severity / urgency index of the patient's condition is determined. A resource allocation module then prioritizes allocating specialist appointments to patients with urgent or severe conditions, achieving intelligent triage to screen and classify patient conditions, thus improving the efficiency of medical resource allocation. Finally, a referral guidance module recommends treatment plans for patients who do not meet the priority for specialist appointments, further improving the utilization rate of automated condition screening. Attached Figure Description

[0030] Figure 1 This is a system architecture diagram of an embodiment of the hierarchical medical resource allocation system based on an AI priority pre-screening mechanism provided by the present invention. Detailed Implementation

[0031] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.

[0032] A specific embodiment of the present invention discloses a hierarchical medical resource allocation system 1 based on an AI priority pre-screening mechanism. Please refer to [link / reference]. Figure 1 ,include:

[0033] User input module 11 is used to receive the patient’s initial information based on an AI mini-program. The initial information includes at least symptom description, medical history record, examination data and treatment stage.

[0034] It should be noted that the user input module is based on an AI mini-program, responding to user input requests and accepting the patient's initial information. This initial information includes symptom descriptions (such as pain intensity, duration, and accompanying symptoms), medical history records (past medical history, surgical history, and family history of genetic diseases), examination data (imaging reports, laboratory test results, and pathological diagnoses), demographic characteristics (age, gender, and occupation), medication history and allergy history, and the current stage of treatment (initial visit, follow-up visit, postoperative follow-up, etc.). The system dynamically collects the above information through structured forms and intelligent question answering, and automatically verifies the data integrity and logical consistency, providing multi-dimensional input basis for subsequent AI-based disease grading to support subsequent AI-based automated disease analysis and grading.

[0035] In one specific embodiment, the user input module further includes a data verification unit 111, which is used to perform format verification and logical verification on the collected initial information. If the data is missing or contradictory, a supplementary information prompt is sent to the user, and the process switching is paused until the data verification is passed.

[0036] In this embodiment, the data verification unit, through a closed-loop design of real-time verification → interception → correction, ensures the accuracy of system decisions, the smoothness of processes, and compliance and security. It is a fundamental support module for the medical AI system to achieve efficient resource allocation.

[0037] The AI ​​intelligent pre-diagnosis module 12 is used to extract disease indicators from the patient through multiple rounds of questioning based on the initial information, and input the disease indicators into a preset risk stratification model to determine the severity / urgency index of the patient's condition.

[0038] It should be noted that AI-based multi-turn question answering is an application of existing AI technology, and the question-and-answer dialogues are all pre-designed, which will not be described in detail here. Understandably, the purpose of AI question answering is to determine the severity and urgency of a patient's condition. Disease indicators refer to key clinical parameters extracted from the patient's initial information through multi-turn interactive consultations, which can objectively quantify the severity and urgency of the disease. These indicators are direct input variables for the risk stratification model.

[0039] It should be noted that the risk stratification model includes at least indication judgment rules, risk assessment rules, and referral suggestion rules;

[0040] The indication determination rule is based on the tumor size, pathological type, and stage information in the initial information to determine whether it meets the indications for surgical or drug treatment;

[0041] The risk assessment rule is to generate a risk level score by weighting the patient's laboratory test results, history of complications, and treatment history.

[0042] The referral prompt rule is that when a preset critical value appears in the initial information, a referral request form containing the patient's basic information and key clinical data is automatically generated and pushed to the designated medical institution interface; the critical value includes at least tumor rupture, bleeding risk, and organ failure indicators.

[0043] In one specific embodiment, the AI ​​intelligent pre-diagnosis module includes a dynamic interaction unit 121 and a risk stratification model unit 122;

[0044] The dynamic interaction unit 121 is used to obtain clinical indicators such as symptom duration and tumor stage based on the patient's initial information and through preset question-and-answer logic;

[0045] The risk stratification model unit 122 is used to classify the patient's clinical indicators based on a fully trained machine learning model and output a level number representing the severity / urgency index of the condition, wherein the level number is in the range of 1-5.

[0046] It should be noted that the machine learning model is trained using rules such as indication judgment rules, risk assessment rules, and referral prompt rules.

[0047] It should be noted that the AI-powered intelligent pre-diagnosis module can automatically analyze and assess a patient's condition using AI, determine the urgency of the patient's condition, and thus prioritize triage based on the urgency of the patient's condition.

[0048] The resource allocation module 13 includes a dedicated pool of appointment slots reserved for specialists. The resource allocation module is used to open specialist appointments for patients whose disease severity / urgency index is greater than the disease threshold.

[0049] It should be noted that the resource allocation module can be used to triage and treat patients with different degrees of urgency or severity of illness, or to add specialist appointments.

[0050] In one specific embodiment, the resource allocation module includes a traffic splitting logic unit 131 and a timeliness control unit 132;

[0051] The triage logic unit 131 is used to allocate expert appointments to patients with a level number greater than or equal to 4 based on the relationship between the severity / urgency index and the disease threshold.

[0052] The timeliness control unit 132 is used to allocate a preset waiting confirmation time to the patient after the patient obtains the qualification for expert appointment. If payment is not made after the preset waiting time has expired, the corresponding expert appointment will be automatically released to the dedicated appointment pool.

[0053] In this embodiment, the triage logic unit ensures a strict positive correlation between the number of specialist appointments and the urgency of the patient's condition; and the timeliness control unit automates management to reduce resource idleness.

[0054] The referral guidance module 14 is used to recommend alternative treatment plans for patients who do not meet the requirements for open specialist appointments. The referral guidance module 14 includes a triage logic unit 141 and a data connectivity unit 142;

[0055] The triage logic unit 141 includes defining the severity / urgency index of the illness as level 1, level 2 and level 3, and is used to assign patients with a severity / urgency index of level 3 to the waiting queue or recommend further examination; and to guide patients with a severity / urgency index of less than level 2 to the general outpatient clinic or the Internet medical platform.

[0056] The data connectivity unit 142 is used to push the patient's historical medical records to the corresponding destination based on the triage results of the triage logic unit.

[0057] The referral guidance module also includes a multimodal interaction unit 143, which supports outputting guidance instructions in the form of text, voice and visual charts.

[0058] The referral guidance module not only enables the automated AI platform to add patients with severe conditions to the list of patients, but also takes into account other patients and recommends appropriate treatment methods based on the severity of their conditions, thereby improving patient access efficiency and optimizing the allocation of medical resources.

[0059] In this embodiment, the initial patient information is first obtained through a user input module. Then, an AI-powered pre-diagnosis module conducts multiple rounds of targeted questioning based on symptom descriptions, medical history, examination data, and treatment stages from the initial information to determine the patient's condition indicators. A pre-defined risk stratification model is used to determine the severity / urgency index of the patient's condition. A resource allocation module then prioritizes allocating specialist appointments to patients with urgent or severe conditions, achieving intelligent triage to screen and categorize patient conditions and improve the efficiency of medical resource allocation. Finally, a referral guidance module recommends treatment plans for patients who do not meet the priority for specialist appointments, further enhancing the utilization rate of automated condition screening.

[0060] In some embodiments of the present invention, a memory module 15 is also included. The memory module is used to record the patient's historical treatment stages and corresponding initial information. When the patient enters the follow-up visit stage, the initial information is automatically retrieved for comparison and analysis. If abnormal fluctuations in indicators are detected, a follow-up recommendation is triggered through the preset risk stratification model.

[0061] In this embodiment, by setting up a memory module, the patient's condition can be quickly analyzed to improve the efficiency of patient triage and treatment.

[0062] In some embodiments of the present invention, a risk notification module 16 is also included. The risk notification module is used to generate prompt information when a patient applies for an additional appointment. The prompt information includes notification of the appointment review time, notification of the risk of appointment failure, and notification of the risk of delay in the patient's condition due to appointment failure.

[0063] In this embodiment, by setting up a risk disclosure module, the responsibility of automatic priority allocation can be reduced, patients can be informed, their right to know can be protected, the transparency of the priority triage system can be improved, the risk of medical disputes can be reduced, and patients' right to make independent decisions can be enhanced.

[0064] In some embodiments of the present invention, an anti-abuse module 17 is also included for establishing a user credit blacklist.

[0065] In this embodiment, a user behavior blacklist is established through the anti-abuse module, and accounts are disabled for the following behaviors:

[0066] The same WeChat account has been granted "plus" status 3 times within 30 days without paying;

[0067] The same device ID is used by three or more different patient accounts to submit additional appointment requests;

[0068] Using image manipulation tools to forge medical records or image reports.

[0069] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A hierarchical medical resource allocation system based on an AI-priority pre-screening mechanism, characterized in that, include: The user input module is used to receive the patient's initial information based on an AI mini-program. The initial information includes at least a description of symptoms, medical history, examination data, and treatment stage. The AI-powered intelligent pre-diagnosis module is used to extract disease indicators from patients through multiple rounds of questioning based on the initial information, and input the disease indicators into a preset risk stratification model to determine the severity / urgency index of the patient's condition. The resource allocation module includes a dedicated pool of appointment slots reserved for specialists. The resource allocation module is used to open specialist appointments for patients whose disease severity / urgency index is greater than the disease threshold. The referral guidance module is used to recommend alternative treatment plans for patients who do not qualify for open specialist appointments.

2. The hierarchical medical resource allocation system based on an AI priority pre-screening mechanism according to claim 1, characterized in that, The AI-powered intelligent pre-diagnosis module includes a dynamic interaction unit and a risk stratification model unit; The dynamic interaction unit is used to obtain clinical indicators such as symptom duration and tumor stage based on the patient's initial information and through preset question-and-answer logic; The risk stratification model unit is used to classify the patient's clinical indicators based on a fully trained machine learning model and output a level number representing the severity / urgency of the condition, wherein the level number ranges from 1 to 5.

3. The hierarchical medical resource allocation system based on an AI priority pre-screening mechanism according to claim 2, characterized in that, The resource allocation module includes a traffic distribution logic unit and a timeliness control unit; The triage logic unit is used to allocate specialist appointments to patients with a level number greater than or equal to 4 based on the relationship between the severity / urgency index and the disease threshold. The timeliness control unit is used to allocate a preset waiting confirmation period to patients after they have obtained the qualification to add an expert appointment. If payment is not made after the preset waiting period, the corresponding expert appointment will be automatically released to the dedicated appointment pool.

4. The hierarchical medical resource allocation system based on an AI priority pre-screening mechanism according to claim 2, characterized in that, The referral guidance module includes a triage logic unit and a data connectivity unit; The triage logic unit includes defining the severity / urgency index of the illness as level 1, level 2, and level 3, and is used to assign patients with a severity / urgency index of level 3 to the waiting queue or recommend further examination; and to guide patients with a severity / urgency index of less than level 2 to the general outpatient clinic or the Internet medical platform. The data connectivity unit is used to push the patient's historical medical records to the corresponding destination based on the triage results of the triage logic unit.

5. The hierarchical medical resource allocation system based on an AI priority pre-screening mechanism according to claim 1, characterized in that, The risk stratification model includes at least indication determination rules, risk assessment rules, and referral prompting rules; The indication determination rule is based on the tumor size, pathological type, and stage information in the initial information to determine whether it meets the indications for surgical or drug treatment; The risk assessment rule is to generate a risk level score by weighting the patient's laboratory test results, history of complications, and treatment history. The referral prompt rule is that when a preset critical value appears in the initial information, a referral request form containing the patient's basic information and key clinical data is automatically generated and pushed to the designated medical institution interface; the critical value includes at least tumor rupture, bleeding risk, and organ failure indicators.

6. The hierarchical medical resource allocation system based on an AI priority pre-screening mechanism according to claim 1, characterized in that, It also includes a memory module, which is used to record the patient's historical treatment stages and corresponding initial information. When the patient enters the follow-up visit stage, the initial information is automatically retrieved for comparison and analysis. If abnormal fluctuations in indicators are detected, a follow-up recommendation is triggered through the preset risk stratification model.

7. The hierarchical medical resource allocation system based on an AI priority pre-screening mechanism according to claim 1, characterized in that, It also includes a risk notification module, which generates prompt information when a patient applies for an additional appointment. The prompt information includes notification of the appointment review time, notification of the risk of appointment failure, and notification of the risk of delay in treatment due to appointment failure.

8. The hierarchical medical resource allocation system based on an AI priority pre-screening mechanism according to claim 1, characterized in that, The referral guidance module also includes a multimodal interaction unit, which supports outputting guidance instructions in the form of text, voice and visual charts.

9. The hierarchical medical resource allocation system based on an AI priority pre-screening mechanism according to claim 1, characterized in that, It also includes an anti-abuse module for establishing a user credit blacklist.

10. The hierarchical medical resource allocation system based on an AI priority pre-screening mechanism according to claim 1, characterized in that, The user input module also includes a data verification unit, which performs format verification and logical verification on the collected initial information. If the data is missing or contradictory, it sends a supplementary information prompt to the user and pauses the process switching until the data verification passes.

Images