Intelligent clinical pathway and disease group pathway generation method and device based on multi-source data fusion, equipment and storage medium

Abstract

The invention provides an intelligent clinical pathway and disease group pathway generation method and device based on multi-source data fusion, equipment and a storage medium. The invention provides an innovative solution aiming at the structural defects of an existing clinical pathway, including the contradiction between a standardized process and personalized diagnosis and treatment, insufficient adaptability of a general template, too high requirements on medical staff and the problem of pathway stiffness caused by the conflict between supervision constraint and implementation requirements. The core technology breakthrough lies in that multi-source heterogeneous data of a patient is verified and analyzed based on a bottom-layer medical knowledge base, and a personalized clinical path or an adaptive disease group path is generated according to individual characteristics. The system takes a standard path as a benchmark, and realizes full-process monitoring and dynamic correction through a multi-source data fusion technology. The innovativeness of the method is mainly reflected in four dimensions: 1) intelligent decision support is enhanced to realize accurate path generation; 2) establishing a dynamic optimization closed loop to improve the adaptability of the path; 3) integrating a multidisciplinary diagnosis and treatment system to break professional barriers; and (4) man-machine interaction is optimized, and the operation complexity is reduced. Through a loop management mode of generation-execution-feedback-generation, the system effectively solves the problem of balance between clinical standardization and individualized diagnosis and treatment.

Description

TECHNICAL FIELD

[0001] The present application relates to a method and device for generating intelligent clinical pathway and disease group pathway based on multi-source data fusion, and a storage medium. BACKGROUND

[0002] As a key management tool for standardizing medical behavior and improving medical quality and efficiency, the development of clinical pathway has gradually transitioned from initial paper-based management to electronic pathway system, mainly relying on evidence-based medicine to establish standardized processes and time nodes. However, the current widely used clinical pathway has exposed significant structural limitations in practice. The core contradiction lies in the difficulty of preset standardized processes to flexibly adapt to individual differences of different patients, especially for patients with comorbidities or special constitution. The existing pathway lacks effective dynamic adjustment mechanism. At the same time, the execution process of the pathway highly depends on manual monitoring, and cannot integrate and respond to key clinical information such as the patient's continuously changing signs and test results in real time, resulting in deviation of the pathway in actual diagnosis and treatment and difficulty in being discovered and corrected in time. The deeper problem is that the rigid supervision of pathway execution sometimes forces medical staff to mechanically follow the established template, which may delay the best intervention opportunity for the individual condition of the patient, greatly reducing the practical value of the pathway.

[0003] The present application aims to fundamentally solve the above challenges through an innovative multi-source data fusion method. Its core advantage lies in the construction of a dynamic and intelligent pathway generation and optimization system. The system first integrates information from heterogeneous medical data sources such as electronic medical records, laboratory systems, and imaging systems to build a comprehensive health portrait of the patient; then through the underlying medical knowledge base and intelligent algorithms, it deeply checks and analyzes these data to customize individualized clinical pathways for individual patients or accurately include them in appropriate standardized disease group pathways according to their diagnostic characteristics. More importantly, this technology breaks through the static limitations in the previous pathway execution process, and can use multi-source data fusion technology to monitor and analyze the whole process of patient diagnosis and treatment in real time, realizing dynamic correction and optimization of the pathway. This dynamic adaptability significantly improves the usability and effectiveness of the pathway in actual complex medical scenarios. The breakthrough value of this technology lies in successfully reconciling the tension between standardized processes and individualized diagnosis and treatment. By constructing a closed-loop management mode of "generation-execution-feedback-generation", the system not only significantly improves pathway adherence and optimizes the allocation efficiency of medical resources, but also significantly enhances medical safety and patient diagnosis and treatment experience. In terms of diagnosis and treatment, the real-time data-driven pathway correction mechanism provides strong support for precision medicine; in terms of management, its intelligent supervision capability helps medical institutions achieve more refined cost control and performance management. Therefore, the present application represents an important development direction for the evolution of clinical pathway from a static management tool to an intelligent diagnosis and treatment infrastructure. SUMMARY

[0004] The application relates to a method, device and equipment for generating intelligent clinical paths and disease group paths based on multi-source data fusion and a storage medium.

[0005] In view of the structural defects of the existing clinical paths, including the contradiction between standardized processes and personalized diagnosis and treatment, insufficient adaptability of general templates, excessively high requirements for medical staff, and the conflict between supervision constraints and implementation requirements, the present application provides an innovative solution. The core technical breakthrough is that, relying on a bottom-layer medical knowledge base, multi-source heterogeneous data of a patient is checked and analyzed, and a personalized clinical path or an adaptive disease group path is generated according to individual characteristics. The system takes a standard path as a benchmark, and realizes full-process monitoring and dynamic correction through multi-source fusion technology. The innovation mainly lies in four dimensions: 1) strengthening intelligent decision support to realize accurate path generation; 2) establishing a dynamic optimization closed loop to improve path adaptability; 3) integrating a multi-disciplinary diagnosis and treatment system to break down professional barriers; and 4) optimizing human-computer interaction to reduce operation complexity. Through the loop management mode of "generation-execution-feedback-generation", the system effectively solves the balance problem between clinical standardization and individualized diagnosis and treatment.

[0006] As a further optimization, the multi-source data sources include network transmission, interface, memory copy, manual filling, paper graphics and electronic data, etc.

[0007] As a further optimization, the group analysis is based on the initial group recommendation or analysis obtained by the preliminary diagnosis of the patient.

[0008] As a further optimization, patients who cannot enter the group or patients who have not entered the group are treated with independent clinical paths.

[0009] As a further optimization, after the patient enters the group, the patient can be re-grouped according to the patient's condition.

[0010] As a further optimization, the knowledge graph is maintained and audited by artificial maintenance, and includes diseases, drugs, surgeries, policies, regulations, medical insurance rules, price rules, hospital rules, clinical paths, etc.

[0011] As a further optimization, the generated clinical path can be updated as a fixed template or only used for the patient, and other patients can be re-generated.

[0012] As a further optimization, the path execution monitoring can collect data to prove the execution according to the doctor's and nurse's operation or other data.

[0013] As a further optimization, after the path ends, information can be analyzed, and big data analysis capabilities such as path score, quantity, trend, range, etc. based on the path can be prompted.

[0014] The embodiment of the application has the following advantages: I. Dynamic adaptability enhancement The present technology builds a dynamic information network covering the entire diagnosis and treatment cycle of patients by integrating multi-source heterogeneous data in real time. Based on the underlying medical knowledge base and intelligent algorithms, the system can analyze key parameters such as patient condition evolution, test result fluctuations, and treatment response in real time, and automatically trigger path adjustment logic. This adaptive mechanism transforms standardized paths from rigid templates to flexible frameworks, maintaining the standardization of diagnosis and treatment while flexibly responding to individualized needs of complex cases, achieving precise response in both acute conditions and chronic disease management.

[0015] II. Precision of diagnosis and treatment By constructing a multi-dimensional patient health portrait, the system integrates medical data, combines evidence-based medical rules and clinical path logic in the knowledge base, and forms a dynamic decision support system. Intelligent algorithms can identify potential conflicts between individual patient characteristics and standardized paths. At the same time, by continuously tracking treatment effectiveness and patient feedback, the system dynamically optimizes the priority and timing of intervention measures, reducing reliance on human experience.

[0016] III. Optimization of medical resource allocation The closed-loop management mode realizes precise connection of diagnosis and treatment processes through intelligent prediction and dynamic scheduling. Based on historical path data and real-time progress, the system predicts resource demand at each stage (such as examination equipment, bed occupancy, and medical staff), and optimizes scheduling logic through algorithms to reduce waiting time and resource idleness. In addition, the dynamic path correction function can reduce repeated examinations or ineffective treatments caused by scheme deviation, reducing medical consumables and equipment wear and tear. This dynamic resource allocation capability not only improves the efficiency of single case handling, but also optimizes the operational efficiency of the overall medical system through large-scale application.

[0017] IV. Strengthening of whole-cycle quality control capability The system builds a three-level quality control system through embedded quality control nodes throughout the diagnosis and treatment process, including pre-incident warning, in-incident monitoring, and post-incident tracing. Before path execution, the system automatically checks patient data and path applicability, intercepts mismatched intervention measures; during the diagnosis and treatment process, it monitors key indicators deviating from thresholds in real time, triggers multi-level alerts and recommends correction schemes; after completion, the system automatically generates quality evaluation reports, analyzes path deviation causes and improvement directions.

[0018] V. Collaborative innovation of clinical research Dynamic pathway data sedimentation forms a multi-dimensional real-world research database, including standardized pathway execution records, individualized adjustment logs, patient outcome associated data, etc., to provide traceable standardized analysis units for clinical research. Researchers can verify the effectiveness of new treatment strategies by comparing the implementation effects of different pathway versions, or find potential risk factors and improvement directions by analyzing large-scale pathway deviation cases. This closed-loop interaction mechanism between research and clinical practice promotes the rapid iteration of medical knowledge, while supporting guideline updates and policy making with real-world evidence. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 is a flowchart; DETAILED DESCRIPTION

[0020] The flowchart, see Figure 1 , includes the following steps: S1, collate multi-source patient data from hospital HIS, PACS, LIS, electronic medical record, hand numbness system, etc., including generated clinical pathways or stored creative pathway library. At the same time, support storage device copy, page entry, etc. S2, based on the obtained patient data, analyze and judge whether the patient is suitable for group, and execute group path for the enrolled patients, and execute clinical path for the patients not enrolled; S3, based on patient data, analyze each piece of information to determine the corresponding clinical pathway for each piece of information. S4, based on each individual clinical pathway, integration is carried out, and the interaction and restriction between each clinical pathway are analyzed during the integration process, and finally a complete clinical pathway suitable for the current user is generated. In this process, a clinical pathway in the clinical pathway library that is more suitable may have been adopted, and only content verification or supplement is needed; S5, monitor the generated clinical pathway, and monitor whether each clinical pathway item is executed, has been changed or added, and the influence on other clinical pathway items. The execution of the clinical pathway item can be known by means of multi-source patient information grabbing or manual annotation. The generated complete clinical pathway is also used as multi-source patient data for analysis.

Claims

1. The present application relates to a method, device and storage medium for generating intelligent clinical pathway and disease group pathway based on multi-source data fusion. In view of the structural defects of existing clinical pathway, including the contradiction between standardized process and personalized diagnosis and treatment, insufficient adaptability of general template, excessive requirements for medical staff, and the conflict between regulatory constraints and implementation needs leading to rigid pathway, the present application proposes an innovative solution. The core technology breakthrough lies in: relying on the underlying medical knowledge base, checking and analyzing the multi-source data of patients, and generating personalized clinical pathway or adaptive disease group pathway according to individual characteristics. The system takes the standard pathway as the benchmark, and realizes full-process monitoring and dynamic correction through multi-source data fusion technology.

2. The underlying medical knowledge base of claim 1, wherein, The knowledge base is manually entered, maintained, audited and can form a new knowledge base automatically according to the operation of the clinical pathway.

3. The multi-source data of claim 1, wherein, The data comes from hospital HIS, PACS, LIS, electronic medical record, hand anesthesia system, etc., including the generated clinical pathway or stored creative pathway library. At the same time, it supports storage device copying, page entry and other methods.

4. The data fusion and whole process monitoring according to claim 1, characterized in that, The newly acquired multi-source patient data is used to judge the execution result of the clinical pathway or manually mark whether the clinical pathway item is executed and the execution result.

5. The dynamic correction of claim 1 is characterized by reanalyzing all items of the clinical pathway based on each piece of information obtained from the multi-source patient data. The data analysis between each clinical pathway item, and the analysis result is synchronized as a data source to participate in the generation of the clinical pathway.

Images