Intelligent assessment and auxiliary decision-making system for perioperative period of high-risk surgical patient

Abstract

The invention discloses an intelligent assessment and auxiliary decision-making system for a perioperative period of a high-risk surgical patient, and belongs to the technical field of intelligent medical treatment. Comprising a set definition module used for defining a perioperative period state set; the logic set binding module is used for presetting a state binding logic set; the data migration module is used for inputting the multi-source heterogeneous perioperative period data into the state migration engine; the logic set activation module is used for dynamically adjusting the monitoring frequency, the early warning threshold and the decision support focus; the suggestion generation module is used for generating instant early warning and intervention suggestions; and the data recording module is used for feeding back and recording. According to the method, the dynamic state transition mechanism is constructed, and the corresponding monitoring and decision logic is bound for each state, so that the problems of early warning lag, untimely intervention and easy interruption of an intelligent management process caused by disconnection of system response logic and a clinical actual stage in the prior art are solved.

Description

Technical Field

[0001] This invention relates to the field of intelligent medical technology, and in particular to an intelligent perioperative assessment and decision support system for high-risk surgical patients. Background Technology

[0002] With the deep integration of medical informatics and artificial intelligence technologies, intelligent decision support systems are being used more and more widely in clinical medicine. In the perioperative management of high-risk surgical patients, how to achieve continuous and accurate risk assessment and timely clinical decision support through technological means has become a key research direction for improving surgical safety and patient prognosis. Existing intelligent systems mostly process patient data based on preset fixed rules or single models, aiming to provide medical staff with risk warnings and intervention references.

[0003] The above-disclosed technical solutions have at least the following technical problems: Traditional methods typically employ a uniform, static logic to process monitoring data across different surgical stages. This lack of adaptability to dynamic changes in the clinical setting leads to a mismatch between the monitoring data stream and the decision-making logic in terms of timing. Key risk information is not captured and responded to in a timely manner, ultimately resulting in delayed warnings and missed optimal intervention windows, significantly reducing the closed-loop management effectiveness of the intelligent system. To address these issues, this invention proposes a solution. Summary of the Invention

[0004] This application provides an intelligent perioperative assessment and auxiliary decision-making system for high-risk surgical patients, which solves the problem of delayed risk warning response in the preoperative, intraoperative, and postoperative stages caused by rigid assessment logic in the prior art. It realizes dynamic adjustment of monitoring and decision focus based on the real-time status of the patient, thereby significantly improving the timeliness of risk warning, the accuracy of intervention measures, and the coherence of closed-loop management throughout the entire process.

[0005] This application provides a perioperative intelligent assessment and auxiliary decision-making system for high-risk surgical patients, including: a set definition module: used to predefine a perioperative state set based on the perioperative clinical pathway; Logical set binding module: used to preset a logical set for binding each state in the perioperative state set; Data migration module: used to acquire multi-source heterogeneous perioperative data of patients in real time, input the multi-source heterogeneous perioperative data into the state migration engine, and determine the current perioperative status of patients in real time; Logic set activation module: used to determine the current perioperative state, activate and load the state binding logic set corresponding to the current perioperative state, and dynamically adjust the monitoring frequency, early warning threshold and decision support focus; Suggestion generation module: used to generate real-time warnings and intervention suggestions that are strongly correlated with the current state based on the loaded state binding logic set; Data recording module: used to record the current perioperative status, generated real-time warnings and intervention suggestions, and medical staff operation feedback.

[0006] Furthermore, the step of pre-defining a set of perioperative states includes: Based on clinical pathways and physiological stages, the perioperative period is divided into a series of discrete state nodes with clear clinical significance; The state nodes include at least: the preoperative stable period, which represents the patient's state of having completed preoperative preparation and having stable vital signs; Anesthesia induction period, which describes the state from the start of anesthesia to the establishment of the surgical incision; The intraoperative stability maintenance period represents the state in which vital signs fluctuate within the expected range during the main surgical steps. The high-risk period during surgery is characterized by a synergistic deterioration of one or more key physiological parameters or the occurrence of serious adverse events. The anesthesia recovery period refers to the state from the end of surgery until the patient's consciousness and spontaneous breathing are restored; the postoperative monitoring period refers to the state of the patient returning to the intensive care unit and being under close observation.

[0007] Furthermore, the step of presetting a state binding logic set for each state in the perioperative state set includes: Configure a separate set of monitoring, early warning, and decision support logic for each state; The monitoring logic includes presetting a specific data monitoring frequency and data source priority for this state; The early warning logic includes setting a dynamic early warning threshold for this state; The decision support logic includes a highly focused decision knowledge base and suggestion templates preset for this state.

[0008] Further steps for determining the patient's current perioperative status in real time include: When the patient is in a stable maintenance phase during surgery, calculate the multi-parameter synergistic deterioration risk value. ,when Not less than the preset migration threshold At this time, it triggers the transition from the intraoperative stable maintenance period to the intraoperative high-risk period; The method for calculating the risk value of multi-parameter synergistic deterioration is as follows: right Key physiological parameters Real-time data stream Perform normalization processing; Calculate the first derivative of each parameter and second derivative ; Calculated using the following formula : ; In the formula, It is the first Normalized physiological parameters over time The value, and They are The first and second derivatives, and It is preset and targeted at the first Risk weights for the rate of change and acceleration of change of each parameter. and It is a non-linear activation function.

[0009] Furthermore, the step of activating and loading the state binding logic set corresponding to the current perioperative state includes: After the state transition engine determines that the patient's state has changed, it sends a state update command to the logic set activation module. After receiving the instruction, the logic set activation module unloads all logic sets bound to the previous state, and retrieves and loads the complete state-bound logic set corresponding to the new current perioperative state from the logic set binding module.

[0010] Furthermore, the steps for generating immediate warnings and intervention recommendations that are strongly correlated with the current state include: The suggestion generation module analyzes the real-time data stream under the guidance of the loaded state binding logic set; When the warning conditions of the current state are met, the knowledge base that matches the preset decision support focus of that state is invoked to generate structured intervention suggestions.

[0011] Further steps to dynamically adjust monitoring frequency, early warning thresholds, and decision support focus include: When the current perioperative status switches to the anesthesia induction phase or the current perioperative status is the intraoperative stable maintenance phase, a conventional early warning threshold is applied. ; When the current perioperative status switches to the anesthesia induction period or the intraoperative high-risk period, a more sensitive dynamic early warning threshold is applied. .

[0012] Furthermore, the dynamic early warning threshold The calculation method is as follows: ; In the formula, This is a basic early warning baseline value under this high-risk state. Key physiological parameters In the past period of time Volatility within, It is a positive sensitivity coefficient. .

[0013] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: By dividing the perioperative period into multiple distinct states and pre-setting unique monitoring and early warning rules for each state, the focus of attention can be intelligently adjusted according to the patient's actual stage. This effectively solves the response delay problem caused by the mismatch between monitoring data flow and decision-making logic, and improves the timeliness of risk warnings.

[0014] Furthermore, in achieving rapid response, the generated intervention recommendations are closely related to the current surgical stage because a highly matched decision knowledge base can be immediately loaded when the state changes. This highly contextualized decision support ensures that the measures provided when risks are detected are not only timely but also extremely accurate and actionable, effectively avoiding the failure of interventions due to generalization or detachment from the context, and reducing the risk of patient complications.

[0015] Furthermore, by recording the status and operations throughout the entire process, a complete intelligent management closed loop is formed. This allows for seamless integration of preoperative, intraoperative, and postoperative care without the need for frequent switching between automatic and manual modes. It fundamentally prevents the disruption of closed-loop management due to system obsolescence, ensuring the overall improvement and stability of perioperative nursing quality. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the structure of a perioperative intelligent assessment and auxiliary decision-making system for high-risk surgical patients provided in an embodiment of this application. Detailed Implementation

[0017] This application provides an intelligent perioperative assessment and decision support system for high-risk surgical patients, which solves the problem of delayed early warning response caused by the asynchronous monitoring data flow and decision logic in the prior art. By establishing a state transition model covering the perioperative period and dynamically loading a matching set of monitoring, early warning and decision logic when the state changes, the system achieves early detection and precise intervention of patient risks, effectively improving the timeliness of early warning, the accuracy of measures and the continuity of the whole process management.

[0018] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0019] like Figure 1 As shown, this application provides a perioperative intelligent assessment and auxiliary decision-making system for high-risk surgical patients, including: a set definition module: used to predefine a discrete set of perioperative states covering preoperative, intraoperative and postoperative periods based on the perioperative clinical pathway; Logical set binding module: used to pre-set an independent state binding logical set for each state in the perioperative state set; Data migration module: used to acquire multi-source heterogeneous perioperative data of patients in real time, including vital signs data, medical event data and surgical progress data; input the multi-source heterogeneous perioperative data into the state migration engine, which performs fusion judgment on the data based on the state migration rule base to determine the patient's current perioperative status in real time; Logic set activation module: used to determine the current perioperative state, activate and load the state binding logic set corresponding to the current perioperative state, and dynamically adjust the monitoring frequency, early warning threshold and decision support focus of the logic set; Suggestion generation module: used to perform context-dependent intelligent evaluation and auxiliary decision-making on the multi-source heterogeneous perioperative data based on the loaded state binding logic set, and generate real-time early warning and intervention suggestions that are strongly correlated with the current state; Data recording module: used to record the current perioperative status, generated real-time warnings and intervention suggestions, and medical staff operation feedback to form closed-loop data, which is used for iterative optimization of the state transition rule base.

[0020] Furthermore, the step of pre-defining a set of perioperative states includes: Based on clinical pathways and physiological stages, the perioperative period is divided into a series of discrete state nodes with clear clinical significance; The state nodes include at least: the preoperative stable period, which represents the patient's state of having completed preoperative preparation and having stable vital signs; Anesthesia induction period, which describes the state from the start of anesthesia to the establishment of the surgical incision; The intraoperative stability maintenance period represents the state in which vital signs fluctuate within the expected range during the main surgical steps. The high-risk period during surgery is characterized by a synergistic deterioration of one or more key physiological parameters or the occurrence of serious adverse events. The anesthesia recovery period refers to the state from the end of surgery until the patient's consciousness and spontaneous breathing are restored; the postoperative monitoring period refers to the state of the patient returning to the intensive care unit and being under close observation. Each state node is interconnected through a logical sequence and transition conditions, together forming a state network that covers the entire perioperative process.

[0021] Furthermore, the step of presetting a state binding logic set for each state in the perioperative state set includes: Configure a separate set of monitoring, early warning, and decision support logic for each state; The monitoring logic includes presetting specific data monitoring frequencies and data source priorities for this state. For example, during high-risk periods during surgery, the system automatically increases the sampling and analysis frequency of hemodynamic parameters and oxygenation indicators. The early warning logic includes a preset dynamic early warning threshold for the state, which can be adaptively adjusted based on the characteristics of the state itself and the fluctuations of the patient's recent physiological data. The decision support logic includes a highly focused decision knowledge base and suggestion templates preset for this state, ensuring that the generated suggestions are strongly relevant to the current clinical scenario. For example, during the anesthesia induction period, the decision support focus is on airway management and blood flow stability maintenance strategies.

[0022] Further steps for determining the patient's current perioperative status in real time include: When the patient is in a stable maintenance phase during surgery, calculate the multi-parameter synergistic deterioration risk value. ,when Not less than the preset migration threshold At this time, it triggers the transition from the intraoperative stable maintenance period to the intraoperative high-risk period; The method for calculating the risk value of multi-parameter synergistic deterioration is as follows: right Key physiological parameters (in Real-time data stream Perform normalization processing; Calculate the first derivative (rate of change) of each parameter. And the second derivative (changing acceleration) ; Calculated using the following formula : ; In the formula, It is the first A normalized physiological parameter (such as blood pressure, heart rate) over time The value, and They are The first and second derivatives are used to capture the changing trends of the parameters. and It is preset and targeted at the first The risk weights for the rate of change and acceleration of change of each parameter reflect the contribution of the deterioration of that parameter to the overall risk. and It is a non-linear activation function used to amplify deteriorating trends; its form is defined as follows: and ,in The preset gain coefficient, To correct the linear unit, this design causes the negative changes (deterioration) of the parameters to be nonlinearly amplified; The formula quantifies the synergistic deterioration by integrating the dynamic changing trends of multiple parameters (rate and acceleration), thereby achieving accurate capture of the "trend deterioration" pattern.

[0023] Furthermore, the step of activating and loading the state binding logic set corresponding to the current perioperative state includes: After the state transition engine determines that the patient's state has changed, it sends a state update command to the logic set activation module. After receiving the instruction, the logic set activation module unloads all logic sets bound to the previous state, and retrieves and loads the complete state-bound logic set corresponding to the new current perioperative state from the logic set binding module. The loading process is atomic, ensuring that the system executes only the logic most relevant to the current state at any given time, thereby enabling seamless and instantaneous switching between monitoring strategies, early warning sensitivity, and decision focus.

[0024] Furthermore, the steps for generating immediate warnings and intervention recommendations that are strongly correlated with the current state include: The suggestion generation module analyzes the real-time data stream under the guidance of the loaded state binding logic set; When the warning conditions of the current state are met, the knowledge base that matches the preset decision support focus of the current state is invoked to generate structured intervention suggestions; The content and level of detail of the intervention recommendations are closely related to the current situation. For example, during the high-risk period during surgery, the system generates a crisis management process that includes emergency diagnosis and identification, preferred drug dosage and administration dosage; while during the postoperative monitoring period, it generates guidance on adjusting the analgesia regimen and rehabilitation exercises.

[0025] Further steps to dynamically adjust monitoring frequency, early warning thresholds, and decision support focus include: When the current perioperative status switches to the anesthesia induction phase or the current perioperative status is the intraoperative stable maintenance phase, a conventional early warning threshold is applied. ; When the current perioperative status switches to the anesthesia induction period or the intraoperative high-risk period, a more sensitive dynamic early warning threshold is applied. .

[0026] Furthermore, the dynamic early warning threshold The calculation method is as follows: ; In the formula, This is a basic early warning baseline value under this high-risk state. Key physiological parameters In the past period of time The volatility within that period is calculated by... Standard deviation To quantify, It is a positive sensitivity coefficient. ; When the patient's recent physiological parameters fluctuate When the value increases, the warning threshold will be automatically lowered. (Making it more sensitive), because high volatility itself indicates increased instability, requiring earlier intervention.

[0027] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0028] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0029] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0030] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0031] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0032] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A perioperative intelligent assessment and auxiliary decision-making system for high-risk surgical patients, characterized in that, include: Set definition module: used to predefine a set of perioperative states based on the perioperative clinical pathway; Logical set binding module: used to preset a logical set for binding each state in the perioperative state set; Data migration module: used to acquire multi-source heterogeneous perioperative data of patients in real time, input the multi-source heterogeneous perioperative data into the state migration engine, and determine the current perioperative status of patients in real time; Logic set activation module: used to determine the current perioperative state, activate and load the state binding logic set corresponding to the current perioperative state, and dynamically adjust the monitoring frequency, early warning threshold and decision support focus; Suggestion generation module: used to generate real-time warnings and intervention suggestions that are strongly correlated with the current state based on the loaded state binding logic set; Data recording module: used to record the current perioperative status, generated real-time warnings and intervention suggestions, and medical staff operation feedback.

2. The perioperative intelligent assessment and auxiliary decision-making system for high-risk surgical patients as described in claim 1, characterized in that, The steps to predefine a set of perioperative states include: Based on clinical pathways and physiological stages, the perioperative period is divided into a series of discrete state nodes with clear clinical significance; The state nodes include at least: the preoperative stable period, which represents the patient's state of having completed preoperative preparation and having stable vital signs; Anesthesia induction period, which describes the state from the start of anesthesia to the establishment of the surgical incision; The intraoperative stability maintenance period represents the state in which vital signs fluctuate within the expected range during the main surgical steps. The high-risk period during surgery is characterized by a synergistic deterioration of one or more key physiological parameters or the occurrence of serious adverse events. The anesthesia recovery period refers to the state from the end of surgery until the patient's consciousness and spontaneous breathing are restored; the postoperative monitoring period refers to the state of the patient returning to the intensive care unit and being under close observation.

3. The perioperative intelligent assessment and auxiliary decision-making system for high-risk surgical patients as described in claim 1, characterized in that, The steps for presetting a state binding logic set for each state in the perioperative state set include: Configure a separate set of monitoring, early warning, and decision support logic for each state; The monitoring logic includes presetting a specific data monitoring frequency and data source priority for this state; The early warning logic includes setting a dynamic early warning threshold for this state; The decision support logic includes a highly focused decision knowledge base and suggestion templates preset for this state.

4. The perioperative intelligent assessment and auxiliary decision-making system for high-risk surgical patients as described in claim 1, characterized in that, The steps for determining a patient's current perioperative status in real time include: When the patient is in a stable maintenance phase during surgery, calculate the multi-parameter synergistic deterioration risk value. ,when Not less than the preset migration threshold At this time, it triggers the transition from the intraoperative stable maintenance period to the intraoperative high-risk period; The method for calculating the risk value of multi-parameter synergistic deterioration is as follows: right Key physiological parameters Real-time data stream Perform normalization processing; Calculate the first derivative of each parameter and second derivative ; Calculated using the following formula : ； In the formula, It is the first Normalized physiological parameters over time The value, and They are The first and second derivatives, and It is preset and targeted at the first Risk weights for the rate of change and acceleration of change of each parameter. and It is a non-linear activation function.

5. The perioperative intelligent assessment and auxiliary decision-making system for high-risk surgical patients as described in claim 1, characterized in that, The steps of activating and loading the state binding logic set corresponding to the current perioperative state include: After the state transition engine determines that the patient's state has changed, it sends a state update command to the logic set activation module. After receiving the instruction, the logic set activation module unloads all logic sets bound to the previous state, and retrieves and loads the complete state-bound logic set corresponding to the new current perioperative state from the logic set binding module.

6. The perioperative intelligent assessment and auxiliary decision-making system for high-risk surgical patients as described in claim 1, characterized in that, The steps to generate immediate warnings and intervention recommendations that are strongly correlated with the current state include: The suggestion generation module analyzes the real-time data stream under the guidance of the loaded state binding logic set; When the warning conditions of the current state are met, the knowledge base that matches the preset decision support focus of the current state is invoked to generate structured intervention suggestions.

7. The perioperative intelligent assessment and auxiliary decision-making system for high-risk surgical patients as described in claim 1, characterized in that, The steps for dynamically adjusting monitoring frequency, early warning thresholds, and decision support focus include: When the current perioperative status switches to the anesthesia induction phase or the current perioperative status is the intraoperative stable maintenance phase, a conventional early warning threshold is applied. ; When the current perioperative status switches to the anesthesia induction period or the intraoperative high-risk period, a more sensitive dynamic early warning threshold is applied. .

8. The perioperative intelligent assessment and auxiliary decision-making system for high-risk surgical patients as described in claim 7, characterized in that, The dynamic early warning threshold The calculation method is as follows: ； In the formula, This is a basic early warning baseline value under this high-risk state. Key physiological parameters In the past period of time Volatility within, It is a positive sensitivity coefficient. .

Images