Painless surgical anesthesia target-controlled drug delivery system

Abstract

This invention provides a target-controlled assisted drug delivery system for painless surgery anesthesia, belonging to the field of medical technology. It includes: an initial protocol generation module for determining the initial drug delivery protocol based on the type of painless surgery anesthesia and the patient's basic physiological information; a trial drug delivery control module for administering trial drugs to the patient based on the initial drug delivery protocol and recording the changes in the patient's electroencephalogram (EEG) signals and physiological signs during the trial drug delivery process; a resistance analysis module for analyzing the patient's resistance to the anesthetic drug in real time based on the EEG signal changes and physiological sign changes; and a drug delivery adjustment module for adjusting the initial drug delivery protocol in real time based on the resistance level and continuing drug delivery to the patient based on the adjusted protocol and medical intervention instructions until the drug delivery is completed. This improves the target control accuracy of anesthetic drug dosage and delivery rate during painless surgery.

Description

Technical Field

[0001] This invention relates to the field of medical technology, and in particular to a target-controlled assisted drug delivery system for painless surgical anesthesia. Background Technology

[0002] With the continuous development of medical technology, painless surgical anesthesia is applied in various painless examinations, treatments, and surgical procedures, such as painless childbirth, painless gastroscopy and colonoscopy, painless dentistry, bronchoscopy, and cosmetic procedures. Current painless surgical anesthesia aims to control the patient's anesthesia to the required depth with the minimum amount of medication while ensuring safety. To reduce human target control errors during anesthesia, some drug delivery target control systems or devices exist to assist in the anesthesia process. Most existing drug delivery target control systems or devices for assisting in the anesthesia process use pre-built analytical calculation models to analyze the patient's drug delivery parameters (such as dosage and delivery rate) for medical staff reference.

[0003] However, the pre-built analytical calculation model considers a limited number of information items that affect the anesthesia process, which limits the information available for determining anesthesia parameters. Furthermore, the pre-built analytical calculation model cannot take into account all individual physiological differences of each patient. Both of these reasons may directly lead to inaccurate determination of drug administration parameters.

[0004] Therefore, this invention proposes a target-controlled assisted drug delivery system for painless surgical anesthesia. Summary of the Invention

[0005] This invention provides a target-controlled assisted drug delivery system for painless surgical anesthesia. It determines the initial drug delivery regimen based on the type of painless surgical anesthesia and the patient's basic physiological information. Based on the patient's response data (EEG and physiological signs) during administration of the initial regimen, it analyzes the patient's resistance to the anesthetic drug and adjusts the initial drug delivery regimen accordingly. By incorporating medical intervention instructions, it completes a precise and safe drug delivery process. This system achieves target-controlled assisted delivery of the anesthetic drug dosage and rate during painless surgical anesthesia, improving the accuracy of target control. By analyzing the patient's physiological response during the trial drug delivery phase, this system compensates for individual physiological differences that may not have been considered when determining the initial drug delivery regimen, thus improving the accuracy of target control and the safety of anesthesia during the drug delivery process.

[0006] This invention provides a target-controlled drug delivery system for painless surgical anesthesia, comprising: The initial protocol generation module is used to determine the patient's initial medication regimen based on the type of painless surgical anesthesia and the patient's basic physiological information. The trial drug administration control module is used to administer trial drugs to patients based on the initial drug administration protocol and record the changes in the patient's electroencephalogram and physiological signs during the trial drug administration process. The resistance analysis module is used to analyze the patient's resistance to anesthetic drugs in real time based on changes in electroencephalogram (EEG) signals and physiological signs. The dosing assistance adjustment module is used to adjust the initial dosing regimen in real time based on the level of resistance, and to continue administering medication to the patient based on the adjusted dosing regimen and medical intervention instructions until the dosing is completed.

[0007] Preferably, the initial scheme generation module includes: The type determination submodule is used to determine the type of painless surgical anesthesia based on the patient's physical examination report and historical medical records. The protocol generation submodule is used to determine the safe dosage and safe administration rate for patients based on the type of painless surgical anesthesia and the patient's basic physiological information, and to generate an initial dosing protocol based on the safe dosage and safe administration rate.

[0008] Preferably, the category determination submodule includes: The text extraction unit is used to extract text from the patient's physical examination report and historical medical records to obtain the decision reference text. The information determination unit is used to determine all anesthesia type decision items and the information source for each anesthesia type decision item in the decision reference text; The reliability determination unit is used to determine the reliability of the anesthesia type determination information based on the information source of the anesthesia type determination information; The first single-item decision value determination unit is used to determine the decision value of the corresponding anesthesia type decision item based on the preset decision method of each anesthesia type decision item information; The main decision value determination unit is used to take the sum of the products of the decision values ​​of all anesthesia type decision items in the decision reference text, the reliability of the corresponding information source, and the decision weight of the corresponding anesthesia type decision item as the main decision value of the patient's anesthesia type. The secondary decision value determination unit is used to determine the secondary decision value of the patient's anesthesia type based on potential influencing factors contained in the patient's physical examination report and historical medical records. The anesthesia type determination unit is used to take the sum of the patient's primary anesthesia type determination value and secondary anesthesia type determination value as the anesthesia type determination value, and determine the patient's painless surgical anesthesia type based on the anesthesia type determination value and the preset determination value range corresponding to each painless surgical anesthesia type.

[0009] Preferably, the reliability determination unit includes: The information consistency verification subunit is used to compare the information of the current anesthesia type determination with the comparison information extracted from at least another independent medical record of the same patient that describes the same medical facts. When the content is consistent, assign a first numerical factor to the anesthesia type determination information currently being evaluated; When there is a content conflict, a second numerical factor is assigned to the anesthesia type determination information currently being evaluated; The first numerical factor is greater than the second numerical factor; The data source quality assessment subunit is used to identify the original data record format that generated the anesthesia type determination information for the current assessment; When the original data record is in the form of a structured electronic medical record data field, a third numerical factor is assigned to that information; When the original data record is information extracted from unstructured text through natural language processing technology, a fourth numerical factor is assigned to that information. The third numerical factor is greater than the fourth numerical factor; The reliability calculation subunit is used to output the reliability value of the anesthesia type determination information under the current assessment, based on the numerical factors allocated by the information consistency verification subunit and the numerical factors allocated by the data source quality assessment subunit, and through preset calculation rules.

[0010] Preferably, the secondary determination value determining unit includes: The potential mining subunit is used to mine the mineable text items in the patient's physical examination report and historical medical records, and to perform deep correlation mining on the mineable text items to obtain at least one potentially influential text item. The second single-item decision value determination subunit is used to determine the decision value of each potential impact text information based on the preset decision method of all potential impact text information; The secondary decision value determination subunit is used to take the sum of the products of the decision values ​​of all potentially influential text information in the mineable text item information, the reliability of the corresponding information source, and the decision weight of the corresponding potentially influential text information item as the secondary decision value of the patient's anesthesia type.

[0011] Preferably, the solution generation submodule includes: The decision information extraction unit is used to extract medication decision information from the patient's basic physiological information; The drug administration parameter determination unit is used to determine the patient's safe drug dosage within the safe drug dosage range corresponding to the type of painless surgical anesthesia based on the drug administration decision information, and to determine the patient's safe drug administration rate within the safe drug administration rate range corresponding to the type of painless surgical anesthesia. The initial dosing protocol generation unit is used to generate an initial dosing protocol based on a safe dosing dose and a safe dosing rate.

[0012] Preferably, the physiological sign change data includes: heart rate change data, blood oxygen change data, blood pressure change data, consciousness index change data, and injury index change data.

[0013] Preferably, the resistance analysis module includes: The first test time determination submodule is used to determine the time when the EEG signal reaches the EEG signal marker value based on the EEG signal change map, and use it as the EEG signal test time. The second test time determination submodule is used to determine the time when the corresponding physiological sign change data of the patient reaches the physiological sign marker value in each change data in the physiological sign change data, and use it as the physiological sign test time of the corresponding physiological sign item. The resistance level calculation submodule is used to calculate the patient's resistance to anesthetic drugs based on the EEG signal test time, the physiological sign test time of all physiological signs, and the reference EEG signal test time and the reference physiological sign test time of all physiological signs.

[0014] Preferably, the drug administration assistance adjustment module includes: The regimen adjustment submodule is used to adjust the initial dosing regimen in real time based on the level of resistance, so as to obtain the target dosage and target dosing rate for the patient. The drug administration auxiliary control submodule is used to generate auxiliary target control prompts based on the target dosage and target drug administration rate, and transmit the auxiliary target control prompts to the medical staff terminal. After the auxiliary target control prompts are received, the submodule receives medical intervention instructions input by the medical staff terminal, and continues to administer drugs to the patient based on the medical intervention instructions until the drug administration requirements contained in the medical intervention instructions are met, at which point the drug administration is terminated.

[0015] Preferably, the scheme adjustment submodule includes: The rate of increase determination unit is used to determine the rate of increase in dosage and the rate of increase in administration rate based on the degree of resistance. The dosing target determination unit is used to calculate the target dosing dose and target dosing rate for the patient based on the dosing dose increase rate and dosing rate increase rate, as well as the patient's current dosing dose and dosing rate.

[0016] The advantages of this invention over the prior art are as follows: By combining in-depth analysis of patients' historical medical records (physical examination reports, medical history) to determine the initial type and protocol of anesthesia, and utilizing real-time physiological feedback (EEG, vital signs) during the trial drug administration phase to dynamically assess drug resistance, the system achieves a shift from static model estimation to dynamic individual response adaptation, significantly improving the accuracy of matching drug dosage and rate with the patient's actual physiological state. Through continuous monitoring and analysis of key physiological indicators (such as consciousness index and injury index) during the trial drug administration phase, the system can identify early abnormal reactions or potential risks to anesthetic drugs and make adjustments through a human-machine collaborative model of auxiliary prompts and medical intervention, providing dual protection for anesthesia safety and helping to avoid adverse events such as excessive anesthesia and intraoperative awareness. By automatically extracting text, assessing information reliability, and conducting deep correlation mining on unstructured physical examination reports and historical medical records, scattered and potentially overlooked medical information (such as potential complication risks and the impact of special medication history) is transformed into effective supporting data for anesthesia decision-making, assisting physicians in making more comprehensive judgments. The system of this invention provides a closed-loop auxiliary decision support from information integration, plan generation, real-time monitoring to dynamic adjustment, which reduces the cognitive load of anesthesiologists in complex information processing and rapid judgment, allowing doctors to focus more on the overall monitoring of patients and key decisions, thereby improving work efficiency and decision quality.

[0017] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in this application.

[0018] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0019] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of the target-controlled assisted drug delivery system for painless surgical anesthesia in an embodiment of the present invention; Figure 2 This is a schematic diagram of the initial scheme generation module in an embodiment of the present invention. Detailed Implementation

[0020] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention. Example 1:

[0021] This invention provides a target-controlled assisted drug delivery system for painless surgical anesthesia, with reference to... Figure 1 ,include: The initial protocol generation module is used to determine the patient's initial medication regimen based on the type of painless surgical anesthesia (the choice of anesthesia type often depends on the patient's general condition and the difficulty of the procedure; anesthesia types include moderate sedation, deep sedation, and general anesthesia; moderate sedation is only suitable for patients in good physical condition and for simple endoscopic procedures; deep sedation requires close monitoring of vital signs, airway management, and preparation for endotracheal intubation if necessary; for patients with special positions, difficult procedures, or multiple comorbidities, general anesthesia with endotracheal intubation remains the safest and most effective anesthesia regimen) and the patient's basic physiological information (i.e., the physiological information that plays a decisive role in the patient's initial medication regimen, such as gender, age, and weight). This initial medication regimen includes at least a safe dosage and a safe administration rate, and may also include other information such as the name of the medication. The trial drug control module is used to administer trial drugs to patients based on the initial drug administration protocol (i.e., administer trial drugs to patients based on the drug administration parameters in the initial drug administration protocol), and record the changes in the patient's electroencephalogram (EEG) signal (i.e., an image containing the changes in the patient's EEG signal during the trial drug administration) and the changes in physiological signs (i.e., data containing the changes in the patient's physiological signs during the trial drug administration). The resistance analysis module is used to analyze the patient's resistance to anesthetic drugs in real time based on EEG signal change maps and physiological sign change data (that is, how much greater the patient's resistance to anesthetic drugs is compared to the resistance of patients to anesthetic drugs in the reference anesthesia example). The drug administration adjustment module is used to adjust the initial drug administration regimen in real time based on the resistance level, and continue to administer drugs to the patient based on the adjusted drug administration regimen and medical intervention instructions (i.e., control instructions input by medical staff after receiving the adjusted drug administration regimen to intervene and adjust the drug administration dosage and rate of administration) until the drug administration ends (the drug administration is controlled to end when the drug administration dosage reaches the medical intervention instruction).

[0022] Based on the type of painless surgical anesthesia and the patient's basic physiological information, an initial dosing regimen is determined. The patient's resistance to the anesthetic is analyzed based on EEG signal data and physiological sign changes during administration of the initial regimen. The initial dosing regimen is then adjusted based on the level of resistance, and medical intervention instructions are incorporated to complete a precise and safe drug administration process. This approach achieves assisted target control of anesthetic dosage and administration rate during painless surgical anesthesia, improving the accuracy of target control. By analyzing the patient's physiological response during the trial administration phase, this method compensates for individual physiological differences that may not have been considered when determining the initial dosing regimen, thus improving the accuracy of assisted target control and anesthesia safety. Example 2:

[0023] Based on Example 1, the initial scheme generation module, refer to Figure 2 ,include: The category determination submodule is used to determine the type of painless surgical anesthesia based on the patient's physical examination report (i.e., a report that at least includes information obtained after the patient has undergone routine physical examinations) and historical medical records (the historical medical records here can be records of the patient's visits to all hospitals obtained through the hospital information linkage system, and the medical records at least include the patient's medical needs and diagnosis results). The protocol generation submodule is used to determine the patient's safe dosage (i.e., the maximum dosage that can be achieved while ensuring patient safety) and safe administration rate (i.e., the maximum administration rate that can be achieved while ensuring patient safety) based on the type of painless surgical anesthesia and the patient's basic physiological information. Based on the safe dosage and safe administration rate, an initial dosing protocol is generated (other information such as drug names can be added here as needed to obtain the initial dosing protocol).

[0024] The above process enables the determination of the appropriate type of painless surgical anesthesia for the patient based on the patient's physical examination report and historical medical records, and further combines the patient's basic physiological information to determine an initial dosing regimen that includes at least a safe dosage and a safe dosing rate. Example 3:

[0025] Based on Example 2, the category determination submodule includes: The text extraction unit is used to extract text from the patient's physical examination report and historical medical records (based on a preset list of decision text items, which contains information items that determine the type of anesthesia for painless surgery, such as special medication history, special disease history, etc.) to obtain decision reference text (that is, text extracted from the patient's physical examination report and historical medical records that summarizes the specific information of the information items that determine the type of anesthesia for painless surgery, such as extracting that the patient has acute myocardial infarction or asthma). The information determination unit is used to determine all anesthesia type decision items and the information source for each anesthesia type decision item in the decision reference text; The reliability determination unit is used to determine the reliability of the anesthesia type determination information based on the information source of the anesthesia type determination information (i.e., the information source of the information item that determines the type of anesthesia for painless surgery) (i.e., the issuing institution of the physical examination report and the hospital of the historical medical record). The first single-item decision value determination unit is used to determine the decision value of the corresponding anesthesia type decision item based on a preset decision method for each anesthesia type decision item information (i.e., a preset method for determining the decision value of the corresponding anesthesia type decision item for the patient based on a single anesthesia type decision item information, such as summing the preset decision values ​​corresponding to each disease in the patient's special medical history, or multiplying the preset decision value corresponding to the medication type in the special medication history with the product of the multiple corresponding to the duration of medication use, as the decision value of the corresponding anesthesia type decision item). This determines the decision value of the corresponding anesthesia type decision item (i.e., a numerical value determined based on the patient's information on a certain anesthesia type decision item, which also represents its contribution to the patient's final anesthesia type decision value without considering the reliability of the information source and the decision weight of the anesthesia type decision item information). The main decision value determination unit is used to take the sum of the products of the decision values ​​of all anesthesia type decision items in the decision reference text, the reliability of the corresponding information source, and the decision weight of the corresponding anesthesia type decision item (i.e., the contribution ratio of the absolute value of the corresponding anesthesia type decision item to the patient's main anesthesia type decision value) as the patient's main anesthesia type decision value (i.e., the main reference value determined based on the anesthesia type decision item information contained in the patient's physical examination report and historical medical records, used to determine the patient's painless surgery anesthesia type). The secondary decision value determination unit is used to determine the secondary decision value of the patient's anesthesia type (i.e., another secondary reference value determined based on the potential influencing factors contained in the patient's physical examination report and historical medical records, excluding the information on the anesthesia type determination item) to determine the patient's anesthesia type secondary decision value (i.e., another secondary reference value determined based on the potential influencing factors contained in the patient's physical examination report and historical medical records to determine the patient's anesthesia type for painless surgery). The anesthesia type determination unit is used to take the sum of the patient's primary anesthesia type determination value and secondary anesthesia type determination value as the anesthesia type determination value (that is, the reference value used to determine the patient's painless surgery anesthesia type by comprehensively considering the primary anesthesia type determination value and the secondary anesthesia type determination value). Based on the anesthesia type determination value and the preset determination value range corresponding to each painless surgery anesthesia type (that is, the value range of different anesthesia type determination values ​​corresponding to each painless surgery anesthesia type), the unit determines the patient's painless surgery anesthesia type (taking the painless surgery anesthesia type corresponding to the preset determination value range in which the patient's anesthesia type determination value is located as the patient's painless surgery anesthesia type).

[0026] The above process determines the primary and secondary decision values ​​for deciding the type of anesthesia for a patient's painless surgery from two perspectives: information on anesthesia type determination items obtained directly from the patient's physical examination report and historical medical records, and potential influencing factors that are mined. This achieves comprehensive extraction and consideration of information in the patient's physical examination report and historical medical records that may influence the determination of the type of anesthesia, ensuring the accuracy of the final determined type of anesthesia for painless surgery. Secondly, when determining the primary decision value, it is not only based on the preset decision method for each anesthesia type determination item, but also comprehensively considers the reliability and decision weight of the anesthesia type determination item information, further improving the accuracy of the determined type of anesthesia for painless surgery based on the aforementioned steps. Example 4:

[0027] Based on Example 3, the reliability determination unit includes: The information consistency verification subunit is used to compare the information of the current anesthesia type determination with the comparison information extracted from at least another independent medical record of the same patient that describes the same medical facts. When the content is consistent, assign a first numerical factor to the anesthesia type determination information currently being evaluated; When there is a content conflict, a second numerical factor is assigned to the anesthesia type determination information currently being evaluated; The first numerical factor is greater than the second numerical factor; The data source quality assessment subunit is used to identify the original data record format that generated the anesthesia type determination information for the current assessment; When the original data record is in the form of a structured electronic medical record data field, a third numerical factor is assigned to that information; When the original data record is information extracted from unstructured text through natural language processing technology, a fourth numerical factor is assigned to that information. The third numerical factor is greater than the fourth numerical factor; The reliability calculation subunit is used to output the reliability value of the anesthesia type determination information under the current assessment, based on the numerical factors allocated by the information consistency verification subunit and the numerical factors allocated by the data source quality assessment subunit, and through preset calculation rules.

[0028] In this embodiment, the core content that needs to be explained includes: information on the type of anesthesia, another independent medical record, the same medical facts, content consistency and conflict, structured electronic medical record data fields, unstructured text, numerical factors, and the calculation rules for reliability values.

[0029] In this embodiment, the anesthesia type determination information refers to specific medical information entries extracted from the patient's physical examination report or historical medical records, used to determine the specific type of painless surgical anesthesia (such as moderate sedation, deep sedation, or general anesthesia). Examples include a clear disease diagnosis (such as severe bronchial asthma), long-term medication records (such as daily warfarin 5mg), or significant abnormal test results (such as a left ventricular ejection fraction of 35%). The other independent medical record refers to a medical document that differs from the current information source in terms of time, institution, or record type. For example, if the current information comes from a physical examination report from Hospital A in 2023, the comparison information could be taken from a health record from Community Hospital B in 2022 or an earlier medical record from Clinic C, thus ensuring the independence of the comparison. The same medical facts refer to information describing the same medical entity or condition. For example, if the current assessment information is a 10-year history of diabetes, the comparison information should be the same description of the diabetes diagnosis or medical history (such as a history of diabetes or poor glycemic control), rather than other unrelated conditions. Content consistency refers to two pieces of information describing the same medical fact in a medical sense that are identical or mutually supportive, such as both confirming that a patient has hypertension. Content conflict refers to contradictory or conflicting descriptions, such as one record stating no history of drug allergies, while another records a penicillin allergy. Structured electronic medical record data fields refer to specific columns in the hospital information system that store data according to standardized formats, such as the blood pressure field in the vital signs table or the ICD-10 code field in the diagnosis list; their content is standardized and can be directly read by a computer. Unstructured text refers to medical record paragraphs, doctor's notes, or the conclusions of examination reports written in freehand style; its content requires parsing using natural language processing technology to extract key information. The first, second, third, and fourth numerical factors are preset weight values ​​for achieving reliable quantification; for example, they can be set to 0.9, 0.3, 0.8, and 0.5 respectively. Their relative magnitudes reflect the positive impact of information consistency and the degree of source structuring on reliability. The reliability score is calculated using a pre-defined logic that combines the aforementioned factors into a single reliability index. For example, a multiplication rule can be used: if a piece of information matches the comparison (first factor 0.9) and originates from a structured field (third factor 0.8), its final reliability score is calculated to be 0.72; if the information conflicts during comparison (second factor 0.3) and originates from unstructured text (fourth factor 0.5), its reliability score is 0.15. This rule ensures that the output is a quantifiable evaluation value that comprehensively reflects the inherent consistency and source quality of the information.

[0030] This embodiment assesses information reliability through objective technical means: the system automatically cross-validates multiple medical records of patients, identifies the consistency between information and the degree of structure of the source data, and quantifies reliability accordingly. This method abandons reliance on non-technical rules such as hospital level, and directly improves information quality based on data processing, thereby enhancing the scientific, safe, and compliant nature of anesthesia protocol decisions. Example 5:

[0031] Based on Embodiment 3, the secondary determination value determination unit includes: The potential mining subunit is used to mine the mineable text item information in the patient's physical examination report and historical medical records (i.e., the specific information of the mineable text items included in the preset list of mineable text items in the patient's physical examination report and historical medical records) based on a preset list of mineable text items, and to perform deep association mining on the mineable text item information (i.e., based on the preset mining principle of the mineable text items, inferring new potential impact text information from the existing mineable text item information; the preset mining principle is, for example, based on the values ​​in mineable text item A and mineable text item B, it is concluded that the patient may have a special medication history of drug C, and the special medication history of drug C is the potential impact text information mined by deep association mining), and to obtain at least one potential impact text information (i.e., information that plays a decisive role in the type of painless surgery anesthesia for the patient based on the deep association mining of the mineable text item information). The second single-item decision value determination subunit is used to determine the decision value of each potential impact text information based on the preset decision method of all potential impact text information (that is, the numerical value determined based on all potential impact text information of the patient to determine the final anesthesia type decision value, and also represents its contribution to the patient's final anesthesia type decision value without considering the reliability of the information source and decision weight of all potential impact text information). The secondary decision value determination subunit is used to take the sum of the products of the decision values ​​of all potentially influential text information in the mineable text item information, the reliability of the corresponding information source (the reliability of the information source is determined in the same way as the reliability of the information source of the anesthesia type decision item information), and the decision weight of the corresponding potentially influential text information item (which is also preset) as the secondary decision value of the patient's anesthesia type.

[0032] In this embodiment, one specific implementation method for deep association mining is as follows: The deep correlation mining function in the potential mining sub-unit can be implemented through the following steps: Building / Invoking a Medical Knowledge Graph: The system pre-installs or accesses a medical knowledge graph, which contains medical entities such as diseases, symptoms, signs, drugs, and laboratory indicators, as well as the relationships between entities (e.g., Disease A - Contraindications - Drug B, Symptom C - Possible Indication - Disease D, Drug E - Common Side Effects - F).

[0033] Text information entityification: Extract the patient's mineable text information (such as high creatinine level, long-term aspirin use) into standard medical entities (such as renal insufficiency, aspirin) through natural language processing technology.

[0034] Association path discovery: The extracted entities are queried and traversed within a medical knowledge graph. For example, an influence-metabolism relationship was found between the entity of renal insufficiency and the slowed metabolism of certain anesthetic drugs (such as propofol); a cause-risk relationship was found between the entity of aspirin and increased bleeding risk, and bleeding risk is associated with the risk of certain anesthetic procedures (such as spinal anesthesia).

[0035] Generate textual information on potential impacts: Based on the discovered association paths, generate natural language descriptions of potential impacts. For example: A patient's elevated creatinine level suggests renal insufficiency, which may affect the clearance rate of anesthetic drugs metabolized by the kidneys, requiring dose adjustment or the selection of alternative drugs. Or, a patient taking aspirin long-term may have a bleeding tendency, requiring assessment of the risk of spinal anesthesia.

[0036] In this embodiment, another specific implementation of deep association mining is as follows: The deep correlation mining function can also be achieved through statistical models trained on a large amount of historical medical record data: Model Training: Using a massive amount of anonymized electronic medical record data as the training set, an association analysis model (such as the Apriori association rule mining algorithm or a neural network model) is trained. The input to this model is structured patient information (such as diagnosis, medication, and test result codes), and the output is potential risk factors or contraindications that are not explicitly recorded in the medical records but have statistical significance.

[0037] Model application: Transform the current patient's mineable text information into the same structured input format.

[0038] Model Inference: The structured input is fed into the trained model. Based on patterns learned from historical data, the model outputs one or more potential impact terms and their confidence levels. For example, the model might output with high confidence an increased risk of potential cardiovascular events, requiring enhanced intraoperative monitoring, based on the combination of age >65 years and ST segment changes on ECG.

[0039] Result generation: Transform the potential impact terms output by the model into readable textual information about potential impacts.

[0040] The above process determines the secondary decision value of the patient's anesthesia type based on the patient's physical examination report and the mineable text item information in the patient's historical medical records. It not only relies on the preset decision method of the mineable text item information, but also takes into account the reliability and decision weight of the mineable text item information. On the basis of the aforementioned steps, it further improves the accuracy of the determined painless surgical anesthesia type. Example 6:

[0041] Based on Example 2, the scheme generation submodule includes: The decision information extraction unit is used to extract dosing decision information from the patient's basic physiological information (that is, information contained in the patient's basic physiological information that determines the dosage and rate of administration, such as weight, age, and gender). The drug administration parameter determination unit is used to determine the patient's safe drug dose within the safe drug dose range corresponding to the type of painless surgical anesthesia (i.e., the preset maximum drug dose range that ensures patient safety as much as possible under the current type of anesthetic under the painless surgical anesthesia) based on the drug administration decision information, and to determine the patient's safe drug administration rate within the safe drug administration rate range corresponding to the type of painless surgical anesthesia (i.e., the preset maximum drug administration rate range that ensures patient safety as much as possible under the current type of anesthetic under the painless surgical anesthesia); The initial dosing protocol generation unit is used to generate an initial dosing protocol based on a safe dosing dose and a safe dosing rate.

[0042] The above process enables the determination of the safe dosage and safe administration rate for patients within the safe dosage and safe administration rate range corresponding to the type of painless surgical anesthesia. This provides a double guarantee for the determined safe dosage and safe administration rate, and greatly improves the safety of the determined safe dosage and safe administration rate. Example 7:

[0043] Based on Example 1, the physiological sign change data included: heart rate change data, blood oxygen change data, blood pressure change data, consciousness index change data, and injury index change data (respectively representing the changes in heart rate, blood oxygen, blood pressure, consciousness index IOC1, and injury index IOC2 during the trial drug administration process. The injury index IOC2 can not only evaluate the degree of analgesia in the unconscious state of the anesthetized patient, but also, based on the relationship between the injury index IOC2 and the consciousness index IOC1, can effectively avoid the phenomenon of excessive anesthesia and intraoperative awareness in surgical patients. The value range of IOC1 is 0~99, and 40~60 is the appropriate range of sedation during general anesthesia; the value range of IOC2 is 0~100, IOC2>50 indicates incomplete analgesia, IOC2<30 indicates excessive analgesia, and 30~50 indicates the appropriate depth of pain for surgery).

[0044] The above plan specifies the specific items of physiological changes in patients that need to be monitored during the trial drug administration process. Example 8:

[0045] Based on Example 1, the resistance analysis module includes: The first test time determination submodule is used to determine the time when the EEG signal reaches the EEG signal marker value (i.e., the preset marker value in the EEG signal, such as the peak or trough of the EEG signal) based on the EEG signal change graph, and use it as the EEG signal test time (i.e., the time required for the EEG signal to reach the EEG signal marker value as shown in the EEG signal change graph). The second test time determination submodule is used to determine the time when the corresponding physiological sign change data of the patient reaches the physiological sign marker value (e.g., the time when the heart rate first drops to 65 bpm, or the time when the consciousness index first drops to 60) in each physiological sign change data (e.g., heart rate change data, blood oxygen change data, blood pressure change data, consciousness index change data, injury index change data) in each physiological sign change data (e.g., the time required for the change data of the physiological sign item to show that the corresponding physiological sign item reaches the corresponding physiological sign marker value). The resistance level calculation submodule is used to calculate the patient's resistance to anesthetic drugs based on the EEG signal test time and the physiological sign test time of all physiological signs, as well as the reference EEG signal test time (i.e., the pre-prepared EEG signal test time that can characterize the majority of people under the same anesthesia conditions as the current patient) and the reference physiological sign test time of all physiological signs (i.e., the pre-prepared physiological sign test time that can characterize the majority of people under the same anesthesia conditions as the current patient). The resistance level is calculated as follows: the ratio of the difference between the EEG signal test time and the reference EEG signal test time to the reference EEG signal test time is taken as the first resistance level; the average of the ratios of the differences between the physiological sign test times of all physiological signs and the corresponding reference physiological sign test times to the corresponding reference physiological sign test times is taken as the second resistance level; and the maximum value between the first and second resistance levels is taken as the patient's resistance to anesthetic drugs.

[0046] The resistance level in this embodiment may be positive or negative, but it is positive in most cases.

[0047] The above process uses the EEG signal testing time required for the patient's EEG signal to reach the EEG signal marker value and the physiological sign testing time required for the corresponding physiological sign change data to reach the physiological sign marker value as significant determinant features in the patient's EEG signal change map and physiological sign change data. This feature is then compared with reference features (i.e., reference EEG signal testing time and reference physiological sign testing time) to accurately determine the enhancement ratio (i.e., resistance degree) of the patient's resistance to anesthetic drugs relative to the reference situation. Example 9:

[0048] Based on Example 1, the drug administration assistance adjustment module includes: The protocol adjustment submodule is used to adjust the initial dosing protocol in real time based on the resistance level to obtain the patient's target dosing dose and target dosing rate (i.e., the dosing dose and dosing rate included in the adjusted dosing protocol, which are also the parameters included in the target-controlled auxiliary prompt information finally transmitted to the medical staff, i.e., the optimal dosing dose and optimal dosing rate currently adapted to the patient determined by the target-controlled auxiliary dosing system in this embodiment). The drug administration support control submodule is used to generate auxiliary target control prompts based on the target dosage and target administration rate (i.e., prompts containing the target dosage and target administration rate, for transmission to healthcare personnel for reference), and transmit the auxiliary target control prompts to the healthcare personnel's end. After receiving the auxiliary target control prompts, it receives healthcare intervention instructions input by the healthcare personnel (i.e., control intervention instructions containing the target dosage and target administration rate contained in the auxiliary target control prompts) regarding the final dosage and administration rate to the patient. Based on the healthcare intervention instructions, it continues to administer medication to the patient (i.e., administer medication to the patient based on the target dosage and target administration rate contained in the healthcare intervention instructions) until the medication requirements contained in the healthcare intervention instructions are met (i.e., when the dosage already administered to the patient is equal to the dosage that the healthcare personnel want to administer to the patient as contained in the healthcare intervention instructions), and then terminates the medication administration to the patient.

[0049] The above process adjusts the initial dosing regimen based on the level of resistance and generates auxiliary target control prompts based on the adjusted regimen, including new dosage and dosing rate, for reference by medical staff. Finally, the intervention command input by medical staff after receiving the auxiliary target control prompts enables the regulation and intervention of the patient's drug administration. This combines automatic and manual decision-making to further improve the accuracy and safety of anesthesia and drug administration to patients. Example 10:

[0050] Based on Example 9, the scheme adjusts the sub-modules, including: The increase rate determination unit is used to determine the dosage increase rate (i.e., the product of the resistance level and the preset dosage increase rate conversion coefficient is used as the dosage increase rate, which is the proportion of the dosage that needs to be increased based on the current dosage) and the administration rate increase rate (i.e., the product of the resistance level and the preset administration rate increase rate conversion coefficient is used as the administration rate increase rate, which is the proportion of the administration rate that needs to be increased based on the current administration rate, generally following the rule that the greater the resistance level, the greater the dosage and administration rate should be). The dosing target determination unit is used to calculate the target dosing dose and target dosing rate for the patient based on the dosing dose increase rate and the dosing rate increase rate, as well as the patient's current dosing dose and dosing rate (i.e., the product of 1 and the sum of the dosing dose increase rate and the current dosing dose is taken as the target dosing dose, and the product of 1 and the sum of the dosing rate increase rate and the current dosing rate is taken as the target dosing rate).

[0051] The above process determines the dose increase rate and administration rate increase rate based on the degree of resistance, and calculates a reasonable target dose and target administration rate based on the current dose and administration rate.

[0052] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A target-controlled assisted drug delivery system for painless surgical anesthesia, characterized in that, include: The initial protocol generation module is used to determine the patient's initial medication regimen based on the type of painless surgical anesthesia and the patient's basic physiological information. The trial drug administration control module is used to administer trial drugs to patients based on the initial drug administration protocol and record the changes in the patient's electroencephalogram and physiological signs during the trial drug administration process. The resistance analysis module is used to analyze the patient's resistance to anesthetic drugs in real time based on changes in electroencephalogram (EEG) signals and physiological signs. The dosing assistance adjustment module is used to adjust the initial dosing regimen in real time based on the level of resistance, and to continue administering medication to the patient based on the adjusted dosing regimen and medical intervention instructions until the dosing is completed.

2. The painless surgical anesthesia target-controlled drug delivery system according to claim 1, characterized in that, The initial scheme generation module includes: The type determination submodule is used to determine the type of painless surgical anesthesia based on the patient's physical examination report and historical medical records. The protocol generation submodule is used to determine the safe dosage and safe administration rate for patients based on the type of painless surgical anesthesia and the patient's basic physiological information, and to generate an initial dosing protocol based on the safe dosage and safe administration rate.

3. The painless surgical anesthesia target-controlled drug delivery system according to claim 2, characterized in that, The category determination submodule includes: The text extraction unit is used to extract text from the patient's physical examination report and historical medical records to obtain the decision reference text. The information determination unit is used to determine all anesthesia type decision items and the information source for each anesthesia type decision item in the decision reference text; The reliability determination unit is used to determine the reliability of the anesthesia type determination information based on the information source of the anesthesia type determination information; The first single-item decision value determination unit is used to determine the decision value of the corresponding anesthesia type decision item based on the preset decision method of each anesthesia type decision item information; The main decision value determination unit is used to take the sum of the products of the decision values ​​of all anesthesia type decision items in the decision reference text, the reliability of the corresponding information source, and the decision weight of the corresponding anesthesia type decision item as the main decision value of the patient's anesthesia type. The secondary decision value determination unit is used to determine the secondary decision value of the patient's anesthesia type based on potential influencing factors contained in the patient's physical examination report and historical medical records. The anesthesia type determination unit is used to take the sum of the patient's primary anesthesia type determination value and secondary anesthesia type determination value as the anesthesia type determination value, and determine the patient's painless surgical anesthesia type based on the anesthesia type determination value and the preset determination value range corresponding to each painless surgical anesthesia type.

4. The painless surgical anesthesia target-controlled drug delivery system according to claim 3, characterized in that, The reliability determination unit includes: The information consistency verification subunit is used to compare the information of the current anesthesia type determination with the comparison information extracted from at least another independent medical record of the same patient that describes the same medical facts. When the content is consistent, assign a first numerical factor to the anesthesia type determination information currently being evaluated; When there is a content conflict, a second numerical factor is assigned to the anesthesia type determination information currently being evaluated; The first numerical factor is greater than the second numerical factor; The data source quality assessment subunit is used to identify the original data record format that generated the anesthesia type determination information for the current assessment; When the original data record is in the form of a structured electronic medical record data field, a third numerical factor is assigned to that information; When the original data record is information extracted from unstructured text through natural language processing technology, a fourth numerical factor is assigned to that information. The third numerical factor is greater than the fourth numerical factor; The reliability calculation subunit is used to output the reliability value of the anesthesia type determination information being evaluated, based on the numerical factors allocated by the information consistency verification subunit and the numerical factors allocated by the data source quality assessment subunit, and through preset calculation rules.

5. The painless surgical anesthesia target-controlled drug delivery system according to claim 3, characterized in that, Secondary decision value determination unit, including: The potential mining subunit is used to mine the mineable text items in the patient's physical examination report and historical medical records, and to perform deep correlation mining on the mineable text items to obtain at least one potentially influential text item. The second single-item decision value determination subunit is used to determine the decision value of each potential impact text information based on the preset decision method of all potential impact text information; The secondary decision value determination subunit is used to take the sum of the products of the decision values ​​of all potentially influential text information in the mineable text item information, the reliability of the corresponding information source, and the decision weight of the corresponding potentially influential text information item as the secondary decision value of the patient's anesthesia type.

6. The painless surgical anesthesia target-controlled drug delivery system according to claim 2, characterized in that, The solution generation submodule includes: The decision information extraction unit is used to extract medication decision information from the patient's basic physiological information; The drug administration parameter determination unit is used to determine the patient's safe drug dosage within the safe drug dosage range corresponding to the type of painless surgical anesthesia based on the drug administration decision information, and to determine the patient's safe drug administration rate within the safe drug administration rate range corresponding to the type of painless surgical anesthesia. The initial dosing protocol generation unit is used to generate an initial dosing protocol based on a safe dosing dose and a safe dosing rate.

7. The painless surgical anesthesia target-controlled drug delivery system according to claim 1, characterized in that, Physiological change data include: heart rate change data, blood oxygen change data, blood pressure change data, consciousness index change data, and injury index change data.

8. The painless surgical anesthesia target-controlled drug delivery system according to claim 1, characterized in that, The resistance level analysis module includes: The first test time determination submodule is used to determine the time when the EEG signal reaches the EEG signal marker value based on the EEG signal change map, and use it as the EEG signal test time. The second test time determination submodule is used to determine the time when the corresponding physiological sign change data of the patient reaches the physiological sign marker value in each change data in the physiological sign change data, and use it as the physiological sign test time of the corresponding physiological sign item. The resistance level calculation submodule is used to calculate the patient's resistance to anesthetic drugs based on the EEG signal test time, the physiological sign test time of all physiological signs, and the reference EEG signal test time and the reference physiological sign test time of all physiological signs.

9. The painless surgical anesthesia target-controlled drug delivery system according to claim 1, characterized in that, The drug administration adjustment module includes: The regimen adjustment submodule is used to adjust the initial dosing regimen in real time based on the level of resistance, so as to obtain the target dosage and target dosing rate for the patient. The drug administration auxiliary control submodule is used to generate auxiliary target control prompts based on the target dosage and target drug administration rate, and transmit the auxiliary target control prompts to the medical staff terminal. After the auxiliary target control prompts are received, the submodule receives medical intervention instructions input by the medical staff terminal, and continues to administer drugs to the patient based on the medical intervention instructions until the drug administration requirements contained in the medical intervention instructions are met, at which point the drug administration is terminated.

10. The painless surgical anesthesia target-controlled drug delivery system according to claim 9, characterized in that, The solution adjustment submodule includes: The rate of increase determination unit is used to determine the rate of increase in dosage and the rate of increase in administration rate based on the degree of resistance. The dosing target determination unit is used to calculate the target dosing dose and target dosing rate for the patient based on the dosing dose increase rate and dosing rate increase rate, as well as the patient's current dosing dose and dosing rate.

Images