AI-based oral sedation treatment auxiliary monitoring method, system and medium

Abstract

The present invention discloses an AI-based oral sedation treatment auxiliary monitoring method, system and medium, which relates to the field of artificial intelligence. The method includes: when the patient is in a sedated state, receiving the test voice recognition result sent by the voice recognition module; based on the test voice recognition result, sending a test action recognition instruction to the image recognition module, so that the image recognition module obtains the patient's character image within a preset time period; receiving the test action recognition result sent by the image recognition module; determining the patient's sedation test reaction result according to the test voice recognition result and the test action recognition result; according to the sedation test reaction result, adjusting the sedative drug dose currently received by the patient. The present invention improves the safety and accuracy of oral disease patients in sedation treatment through AI technology, and improves the treatment efficiency and patient experience, providing a strong guarantee for the health and safety of patients and the smooth progress of treatment.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to an AI-based oral sedation treatment auxiliary monitoring method, system and medium. Background Art

[0002] Before performing surgery on a patient, a dentist usually needs to take sedative measures to calm the patient in order to reduce the patient's fear, anxiety and other emotions, and relieve the pain during the surgery, which is conducive to the smooth progress of the oral surgery.

[0003] During oral surgery, sedatives are needed to keep the patient at a level of mild to moderate sedation, and occasionally deep sedation can be achieved for a short period of time. However, the use of sedatives has certain risks. For example, an overdose of sedatives can have adverse effects on the patient's health. On the other hand, insufficient sedatives can cause the patient to feel pain and discomfort during treatment, produce a stress response, and affect the treatment process.

[0004] Therefore, how to accurately adjust the dosage of sedative drugs to ensure the health and safety of patients and the smooth progress of treatment has become a problem that needs to be solved urgently. Summary of the invention

[0005] In view of the above-mentioned technical problems and defects, the purpose of the present invention is to provide an AI-based oral sedation treatment auxiliary monitoring method, system and medium. Through the intelligentization of AI technology, not only the safety and accuracy of oral disease patients in sedation treatment are improved, but also the treatment efficiency and patient experience are improved, providing strong guarantees for the health and safety of patients and the smooth progress of treatment.

[0006] In order to achieve the above-mentioned objectives, in a first aspect, the present invention provides an AI-based oral sedation treatment auxiliary monitoring method, which is applied to a host of a sedation monitoring system, and the sedation monitoring system also includes an image recognition module and a speech recognition module. The method includes: when the patient is in a sedated state, receiving a test speech recognition result sent by the speech recognition module, and the test speech recognition result is obtained by the speech recognition module performing speech recognition on a sedation test instruction issued by a doctor; based on the test speech recognition result, issuing a test action recognition instruction to the image recognition module, so that the image recognition module obtains a person image of the patient within a preset time period; receiving a test action recognition result sent by the image recognition module, and the test action recognition result is obtained by the image recognition module performing action recognition on the person image; determining a sedation test reaction result of the patient according to the test speech recognition result and the test action recognition result; and adjusting the dose of sedative drugs currently received by the patient according to the sedation test reaction result.

[0007] The present invention combines speech recognition and image recognition technologies to provide an innovative monitoring and control method for oral disease patients during sedation treatment. First, in terms of safety, by capturing and analyzing the patient's physiological reactions and responses to the doctor's instructions in real time, the system can quickly identify whether the sedation level is appropriate, and promptly issue instructions to medical staff to adjust the drug dosage, thereby effectively avoiding medical risks caused by overdose or underdose of drugs. Secondly, the improvement in accuracy is reflected in the fact that AI technology reduces the reliance on the subjective judgment of medical staff, and analyzes large amounts of data through algorithms to improve the objectivity and accuracy of monitoring. In addition, the improvement in efficiency is reflected in the fact that the automated monitoring process greatly reduces the workload of medical staff, allowing them to focus more on surgical operations and patient communication, while speeding up the response to changes in patient status.

[0008] In some embodiments, after the step of adjusting the sedative drug dosage currently received by the patient according to the sedation test reaction result, it also includes: obtaining the pain expression recognition result sent by the image recognition module, the pain expression recognition result is obtained by the image recognition module by identifying the facial expression image of the patient after the sedative drug dosage is adjusted; according to the pain expression recognition result, adjusting the sedative drug dosage currently received by the patient again.

[0009] By adopting the above embodiment, real-time monitoring and evaluation of the patient's pain response is achieved by obtaining the pain expression recognition results sent by the image recognition module. This process involves a detailed analysis of the patient's facial expressions to identify subtle changes related to pain, thereby providing medical staff with an objective and intuitive pain indicator. Combined with these recognition results, the system can intelligently adjust the patient's current dose of sedatives to ensure that the patient maintains an appropriate level of sedation throughout the treatment process, effectively control pain and improve the patient's comfort. This dynamic adjustment mechanism based on patient feedback not only improves the safety and effectiveness of treatment, but also optimizes the use of medical resources, reduces the risk of excessive sedation, and embodies the advantages of personalized medicine and precision medicine.

[0010] In some embodiments, after the step of adjusting the sedative drug dosage currently received by the patient according to the sedation test reaction result, it also includes: obtaining the pain sound recognition result sent by the voice recognition module, the pain sound recognition result is obtained by the voice recognition module through voice recognition of the sound made by the patient after the sedative drug dosage is adjusted; according to the pain sound recognition result, adjusting the sedative drug dosage currently received by the patient again.

[0011] Using the above embodiment, through the integrated voice recognition module, it is possible to accurately capture and analyze the sounds made by patients after the sedative dosage is adjusted to identify pain-related sound features. The module uses advanced voice processing technology to detect specific parameters in the sound, such as pitch, volume, and rhythm changes, which may be an indication of the patient's pain experience. Based on these pain sound recognition results, the system intelligently adjusts the sedative dosage to achieve precise control of the patient's pain state. This approach not only improves the personalization and accuracy of pain management, but also enhances patient comfort, while providing medical staff with real-time voice feedback to assist them in making quick and accurate clinical decisions. Overall, this solution optimizes the sedation treatment process through voice recognition technology, improving patient safety and treatment quality.

[0012] In some embodiments, the sedation monitoring system also includes a respiratory monitoring module. After the step of adjusting the sedative drug dosage currently received by the patient according to the sedation test reaction result, the system also includes: obtaining a pain breathing recognition result sent by the respiratory monitoring module, the pain breathing recognition result is obtained by the respiratory monitoring module by monitoring the frequency and amplitude of the patient's breathing after the sedative drug dosage is adjusted; and adjusting the sedative drug dosage currently received by the patient again according to the pain breathing recognition result.

[0013] Using the above-mentioned embodiment, an accurate assessment of the patient's pain status is achieved through the respiratory monitoring module. This module carefully monitors and analyzes the patient's respiratory rate and amplitude after the sedative dose is adjusted, capturing changes in breathing patterns that may be related to pain. These painful breathing recognition results provide key physiological feedback to medical staff, guiding them to make targeted adjustments to the sedative dose. This method not only improves the accuracy of pain management, ensures the comfort and safety of patients, but also enhances the personalization and dynamic responsiveness of treatment. Through real-time monitoring and adjustment, this solution optimizes the sedation treatment process, reduces the risk of over-sedation or under-sedation, and significantly improves the clinical treatment effect and the patient's overall medical experience.

[0014] In some embodiments, the sedation monitoring system also includes a pain stimulation module. After the step of adjusting the dosage of sedative drugs currently received by the patient according to the sedation test reaction result, it also includes: sending a pain stimulation instruction to the pain stimulation module so that the pain stimulation module adjusts the intensity of electrical stimulation to the patient; obtaining a stress response recognition result sent by the image recognition module, the stress response recognition result is obtained by the image recognition module by recognizing the character image of the patient when subjected to electrical stimulation of different intensities; determining the sedation level of the patient according to the stress response recognition result; and determining the dosage of sedative drugs received by the patient according to the sedation level.

[0015] By adopting the above embodiment, the intensity of electrical stimulation for the patient is precisely adjusted by sending pain stimulation instructions to the pain stimulation module. Based on the feedback from the electrical stimulation module, the system obtains the real-time recognition results of the patient's facial expressions and body reactions by the image recognition module. These results reflect the patient's stress response to electrical stimulation of different intensities. By analyzing these recognition results, the system can accurately judge the patient's sedation level and then intelligently determine the required sedative drug dosage. This process not only ensures the personalization and accuracy of sedation treatment, but also improves the safety and effectiveness of treatment. Through real-time monitoring and dynamic adjustment, the solution optimizes sedation management, reduces the risk of over-sedation or under-sedation, improves the patient's overall treatment experience, and provides strong decision-making support for medical staff to ensure medical quality and patient safety.

[0016] In some embodiments, before sending a pain stimulation instruction to the pain stimulation module, it also includes: determining the initial electrical stimulation intensity of the pain stimulation module based on the patient's vital signs baseline indicators; generating a pain stimulation instruction based on the initial electrical stimulation intensity, and the pain stimulation instruction is used to enable the pain stimulation module to gradually increase the electrical stimulation intensity to the patient starting from the initial electrical stimulation intensity.

[0017] Using the above embodiment, the initial electrical stimulation intensity of the pain stimulation module is intelligently set by accurately measuring the patient's baseline vital signs, such as heart rate, blood pressure, respiratory rate and blood oxygen saturation. This personalized setting ensures that the electrical stimulation starts at a safe and appropriate level, while providing the patient with a gentle neurological function assessment. As the treatment progresses, the system dynamically generates pain stimulation instructions based on the initial electrical stimulation intensity, so that the pain stimulation module can gradually and orderly increase the electrical stimulation intensity. This method not only improves the safety and effectiveness of the treatment, but also allows medical staff to flexibly adjust the treatment plan based on the patient's real-time response. In this way, the plan achieves precise control of the patient's sedation level, optimizes the management of sedative drug dosage, and provides patients with a more comfortable and humane treatment experience.

[0018] In some embodiments, when the patient is in a sedated state, before the step of receiving the test voice recognition result sent by the voice recognition module, it also includes: obtaining the patient's vital sign baseline indicators; determining whether the vital sign baseline indicators are within the normal range; if so, determining an applicable sedation mode based on the patient's health status; if not, sending an alarm instruction to the alarm module so that the alarm module issues an alarm prompt.

[0019] By adopting the above-mentioned embodiment, accurate monitoring and rapid response to the patient's health status are achieved by acquiring the patient's baseline indicators of vital signs in real time. First, the system measures key physiological parameters such as heart rate, blood pressure, respiratory rate and blood oxygen saturation to determine whether these indicators are within the normal range. If the vital signs are normal, the system will intelligently select the most appropriate sedation mode according to the patient's specific health status, such as nitrous oxide inhalation, intravenous injection or oral medication, to ensure the comfort and safety of the treatment process. If the vital signs are abnormal, the system will immediately send an alarm instruction to the alarm module, triggering an alarm prompt so that medical staff can quickly take intervention measures. This automated monitoring and alarm mechanism significantly improves the efficiency and response speed of patient monitoring, reduces medical risks, ensures patient safety, and also improves the work efficiency of medical staff.

[0020] In some embodiments, before sending an alarm instruction to the alarm module, it also includes: determining the deviation of the vital sign baseline indicator relative to the normal range; determining the alarm level according to the deviation; and generating an alarm instruction based on the alarm level.

[0021] By adopting the above embodiment, real-time monitoring and intelligent early warning of the patient's health status are achieved by accurately measuring the deviation of the patient's vital signs baseline indicators. The system first evaluates the deviation of key physiological parameters such as heart rate, blood pressure, respiratory rate and blood oxygen saturation from the normal range to determine the deviation. According to the size of the deviation, the system automatically divides the alarm level, such as mild, moderate or severe, to ensure that different levels of physiological changes receive corresponding levels of attention. Based on the alarm level, the system quickly generates alarm instructions, which clearly indicate the severity and urgency of the deviation, and provide timely and accurate feedback to medical staff, enabling them to quickly make clinical judgments and take necessary intervention measures. This intelligent alarm system based on data analysis greatly improves the accuracy and response speed of medical monitoring, effectively prevents potential health risks, ensures patient safety, and also optimizes the allocation of medical resources and the workflow of medical staff.

[0022] In a second aspect, an embodiment of the present application provides a sedation monitoring system, comprising a host, an image recognition module and a speech recognition module, the host being connected to the image recognition module and the speech recognition module, respectively; the host comprising: one or more processors and a memory; the memory being coupled to the one or more processors, the memory being used to store computer program code, the computer program code comprising computer instructions, the one or more processors calling the computer instructions so that the host executes the method described in the first aspect and any possible implementation manner of the first aspect.

[0023] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, comprising instructions, which, when executed on an electronic device, enables the electronic device to execute the method described in the first aspect and any possible implementation of the first aspect.

[0024] In a fourth aspect, an embodiment of the present application provides a computer program product comprising instructions, which, when executed on an electronic device, enables the electronic device to execute the method described in the first aspect and any possible implementation of the first aspect.

[0025] It can be understood that the sedation monitoring system provided in the second aspect, the storage medium provided in the third aspect, and the computer program product provided in the fourth aspect are all used to execute the method provided in the embodiment of the present application. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding method, which will not be repeated here.

[0026] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0027] 1. Combining voice recognition and image recognition technology, it provides an innovative monitoring and control method for patients with oral diseases during sedation treatment. First of all, in terms of safety, by capturing and analyzing the patient's physiological reactions and responses to doctors' instructions in real time, the system can quickly identify whether the sedation level is appropriate, and promptly issue instructions to medical staff to adjust the dosage of drugs, thereby effectively avoiding medical risks caused by overdose or underdose of drugs. Secondly, the improvement in accuracy is reflected in the fact that AI technology reduces the reliance on the subjective judgment of medical staff, and analyzes large amounts of data through algorithms to improve the objectivity and accuracy of monitoring. In addition, the improvement in efficiency is reflected in the fact that the automated monitoring process greatly reduces the workload of medical staff, allowing them to focus more on surgical operations and patient communication, while speeding up the response to changes in patient status.

[0028] 2. Real-time monitoring and evaluation of the patient's pain response is achieved by acquiring the pain expression recognition results sent by the image recognition module. The integrated voice recognition module can accurately capture and analyze the sounds made by the patient after the sedative dose is adjusted to identify pain-related sound features. The respiratory monitoring module enables accurate assessment of the patient's pain status. This module carefully monitors and analyzes the patient's respiratory rate and amplitude after the sedative dose is adjusted to capture changes in breathing patterns that may be related to pain.

[0029] 3. By accurately measuring the deviation of the patient's baseline vital signs, real-time monitoring and intelligent early warning of the patient's health status are achieved. The system first evaluates the deviation of key physiological parameters such as heart rate, blood pressure, respiratory rate and blood oxygen saturation from the normal range to determine the deviation. According to the size of the deviation, the system automatically divides the alarm level, such as mild, moderate or severe, to ensure that different levels of physiological changes receive corresponding levels of attention. Based on the alarm level, the system quickly generates alarm instructions, which clearly indicate the severity and urgency of the deviation, providing timely and accurate feedback to medical staff, enabling them to quickly make clinical judgments and take necessary intervention measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The drawings herein are incorporated into and constitute a part of the specification, showing embodiments consistent with the present invention, and together with the specification, are used to explain the principles of the present invention. Obviously, the drawings described below are only some embodiments of the present invention, and for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0031] Figure 1 is an architectural diagram of a sedation monitoring system according to an embodiment of the present application;

[0032] Figure 2 is a step diagram of an AI-based oral sedation treatment auxiliary monitoring method according to an embodiment of the present application;

[0033] Figure 3 This is a schematic diagram of the key feature point recognition of the mouth in an embodiment of the present application;

[0034] Figure 4 This is a schematic diagram of eye key feature point recognition in an embodiment of the present application;

[0035] Figure 5 is a step diagram of another AI-based oral sedation treatment auxiliary monitoring method according to an embodiment of the present application;

[0036] Figure 6 is a flow chart of an AI-based oral sedation treatment auxiliary monitoring method according to an embodiment of the present application;

[0037] Figure 7 This is a schematic diagram of identifying key feature points of eyebrow corners in an embodiment of the present application;

[0038] Figure 8 It is a schematic diagram of the architecture of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION

[0039] The terms used in the following embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to be limiting of the present invention. As used in the specification of the present invention, the singular expressions "a", "a", "above", "the" and "this" are intended to also include plural expressions, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in the present invention refers to any or all possible combinations comprising one or more of the listed items.

[0040] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood as implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present application, unless otherwise specified, the meaning of "plurality" is two or more.

[0041] In the treatment of oral diseases, patients need to be sedated to reduce their fear, anxiety and other emotions, and relieve pain during surgery, which is conducive to the smooth progress of oral surgery.

[0042] Monitoring sedative drug dosage is essential to ensure patient safety, improve the success rate of surgery and treatment, and improve the patient's overall medical experience. The correct level of sedation can reduce patient anxiety and pain, reduce physiological stress responses during surgery, and avoid risks such as respiratory depression, loss of consciousness, or cardiovascular complications caused by excessive sedation. In addition, precise monitoring and adjustment can help shorten patient recovery time, reduce postoperative complications, improve patient satisfaction, and allow medical staff to manage medical resources more effectively. Therefore, the monitoring and control of sedative drugs is an indispensable part of medical practice and is of great significance to maintaining patient health and well-being.

[0043] In the related art, sedation drug dosage monitoring and adjustment usually rely on manual operation and experience judgment of medical staff. Medical staff will assess the sedation effect based on the patient's physiological reactions, vital signs (such as heart rate, blood pressure, respiratory rate and blood oxygen saturation), and state of consciousness, and adjust the drug dosage accordingly. This process may include regular checks of the patient's vital signs and manual adjustments to the drug delivery rate when necessary. In addition, medical staff will also assess the sedation level by observing the patient's facial expressions, body movements, and responses to instructions. However, this method has certain limitations, such as strong subjectivity, possible delayed reaction time, and possible omissions in a busy medical environment, which may lead to inaccurate control of sedation levels.

[0044] Therefore, the embodiment of the present invention provides an AI (artificial intelligence)-based oral sedation treatment auxiliary monitoring solution, which combines speech recognition and image recognition technology to provide an innovative monitoring and control method for oral disease patients during sedation treatment. First, in terms of safety, by capturing and analyzing the patient's physiological reactions and responses to the doctor's instructions in real time, the system can quickly identify whether the sedation level is appropriate, and promptly issue instructions to medical staff to adjust the drug dosage, thereby effectively avoiding medical risks caused by overdose or underdose of drugs. Secondly, the improvement in accuracy is reflected in the fact that AI technology reduces the reliance on the subjective judgment of medical staff, and analyzes a large amount of data through algorithms to improve the objectivity and accuracy of monitoring. In addition, the improvement in efficiency is reflected in the fact that the automated monitoring process greatly reduces the workload of medical staff, allowing them to focus more on surgical operations and patient communication, while speeding up the response to changes in patient status.

[0045] In terms of patient experience, this solution significantly improves patient comfort and satisfaction by ensuring that patients maintain an appropriate level of sedation during treatment, helping to relieve patients' tension and reduce postoperative discomfort. The implementation of personalized treatment allows the system to adjust drug dosages based on the patient's specific responses, making the treatment plan more tailored to the individual differences of the patient and improving the treatment effect. Finally, the monitoring data and adjustment history automatically recorded by the system provide the medical team with valuable information resources, which are helpful for subsequent medical analysis, treatment effect evaluation, and continuous optimization of treatment plans.

[0046] To achieve the above effects, an embodiment of the present invention provides an AI-based oral sedation treatment auxiliary monitoring method, which is applied to a host of a sedation monitoring system, such as Figure 1 As shown, the sedation monitoring system also includes an image recognition module and a speech recognition module.

[0047] Among them, the host is the executor of the method of this embodiment, responsible for coordinating and managing the workflow of the voice recognition module, image recognition module and other functional modules. It has sufficient computing power to process data from the voice recognition module, image recognition module and other functional modules in real time, and use relevant algorithms for analysis and decision-making. The host is not only responsible for the collection, storage and management of data, but also has a high degree of user interaction interface, so that medical staff can easily monitor the patient's sedation state and make timely adjustments based on the information provided by other modules. In addition, the host is also designed with advanced security protocols and data encryption technology to ensure the privacy of patient information and the stable operation of the system. By integrating the latest communication interface, the host can seamlessly connect with the hospital information system to achieve data sharing and remote monitoring, greatly improving the quality and efficiency of medical services.

[0048] The image recognition module uses computer vision and machine learning technology to capture and analyze the patient's facial expressions or body movements in real time when in a sedated state. The module can accurately identify and evaluate the patient's response to the doctor's instructions, such as opening the mouth, opening the eyes, clenching the fist, bending the legs, etc. By analyzing the key feature points and movement patterns in the character image, the system can determine the patient's sedation level. In addition, the image recognition module can also monitor the patient's physiological state, such as the tension of the facial muscles and eye movements, to provide medical staff with intuitive and quantitative information about the patient's sedation state, thereby helping them make more accurate decisions on drug dosage adjustments.

[0049] Among them, the scope of recognition of the image recognition module includes but is not limited to opening eyes, eyes wide open, eyes closed, opening mouth, opening mouth wide, closing mouth, frowning, normal eyebrows, clenching fists, opening hands, raising hands, bending legs and other actions, and cooperating with doctor's instructions, vital monitoring data, pain stimulation, respiratory rate and other indicators, with the help of professional models to assist doctors in diagnosis in real time, and display early warning information in real time.

[0050] Due to the local strong light in the actual scene, the face may not be recognized. In order to solve this problem, the image recognition module of this embodiment uses a dynamic environmental light sensitivity adjustment model and is trained based on TensorFlow (a symbolic mathematical system based on data flow programming, which is widely used in the programming implementation of various machine learning algorithms). It supports face recognition, key facial feature points, target detection and tracking, and image classification under strong light. According to the intensity of local light, it dynamically adjusts the light sensitivity parameters, contrast, brightness, etc. to achieve the recognition of facial features; the occluded parts are simulated and created so that the model can still complete recognition under complex conditions.

[0051] The training process of the dynamic ambient light sensitivity adjustment model is as follows:

[0052] (1) Environment setup and deployment of TensorFlow.

[0053] (2) Collect a large number of face images under different lighting conditions, including indoor, outdoor, strong light, shadow, occlusion, etc. Ensure that the dataset is diverse to cover all possible lighting and environmental conditions.

[0054] (3) Use image processing techniques such as rotation, scaling, cropping, brightness adjustment, etc. to increase the diversity of the data set. Strong light and occlusion effects can be simulated to enhance the model's adaptability to these situations.

[0055] (4) Standardize the image, such as normalizing pixel values. Face detection and alignment are required to ensure that the data input to the model is consistent.

[0056] (5) Choose a suitable deep learning architecture, such as a convolutional neural network (CNN). You can consider using an existing pre-trained model as a starting point, such as VGG, ResNet, etc., and perform transfer learning.

[0057] (6) Train the model to automatically learn features in the image. Use convolutional layers to extract key features of the image, such as edges, textures, etc.

[0058] (7) Train the model to recognize the lighting conditions in the image and dynamically adjust the parameters. This may involve learning to adjust the contrast, brightness, etc. of the image.

[0059] (8) The model is trained to recognize key feature points of the face, such as eyes, nose, mouth, etc. Accurate detection of key points should be maintained even under occlusion or strong light conditions.

[0060] (9) Define a loss function to measure the difference between the model prediction and the true label. Customize the loss function to specifically address recognition problems under lighting and occlusion conditions.

[0061] (10) Use optimization algorithms such as gradient descent to train the model. Adjust hyperparameters such as learning rate and batch size to achieve optimal performance.

[0062] (11) Evaluate the performance of the model on the validation set. Use indicators such as accuracy, recall, and F1 score to measure the model effect.

[0063] (12) Adjust the model structure or parameters based on the evaluation results. Multiple iterations may be required to continuously improve the model performance.

[0064] (13) Test the model in actual application scenarios to ensure its effectiveness under real-world conditions.

[0065] (14) Collect new data and continuously learn and update the model to adapt to new lighting and environmental conditions.

[0066] The image recognition module is equipped with a facial recognition model, which is mainly implemented through MediaPipe technology. MediaPipe is an open source computer vision processing framework that trains models based on TensorFlow and supports face recognition, face key points, human posture recognition, target detection and tracking, image classification, portrait segmentation, gesture recognition, text classification, speech classification, etc. The facial recognition model of this embodiment adopts an existing open source model algorithm without special restrictions.

[0067] Taking facial recognition as an example, the input facial image will first be processed through a neural network model to extract key feature points. These feature points may include the positions of facial parts such as eyes, nose, and mouth. Each feature point will be described as having a specific coordinate position and feature information. This information helps to identify and track different parts of the face. By comparing the extracted feature points with the feature points in the pre-trained model, specific features of the face, such as eyes, nose, mouth, etc., can be identified, thereby determining the outline and key features of the portrait.

[0068] Based on the detected feature points, the pose and position of the portrait can be inferred, including information such as rotation, tilt and scale, to further improve the recognition and tracking of the portrait. By continuously detecting and updating the position of the feature points, the system can track the movement and changes of the portrait in real time and maintain accurate recognition of the face.

[0069] The voice recognition module is responsible for receiving and processing the test instructions issued by the doctor, such as the voice command "open your mouth" or "open your eyes". The voice recognition module can communicate with the host through an infrared transceiver.

[0070] The voice recognition module exchanges information with the host through an infrared transceiver, and the infrared transceiver is connected to the host's system service through a driver-free TTL (transistor-transistor logic) to USB chip. The infrared transceiver uses a driver-free TTL to USB chip, so it can communicate with the host directly through the USB interface without the need for an additional driver.

[0071] The speech recognition module uses the latest natural language processing (NLP) technology to accurately recognize and transcribe human voice signals, maintaining high accuracy even in noisy medical environments. Through speech recognition models and algorithms, the module is able to understand the doctor's command intent and convert it into commands that the system can understand. In addition, the speech recognition module has the ability to self-learn and adapt, and can be optimized according to the doctor's language habits and command style to improve the efficiency and accuracy of recognition. By working closely with the image recognition module, the speech recognition module ensures the responsiveness and interactivity of the entire monitoring system, providing strong support for the realization of intelligent sedation monitoring and management.

[0072] Specifically, the voice recognition module mainly implements voice recognition technology through a TTL communication module (i.e., a driver-free TTL to USB chip). The voice recognition technology involves transmitting the voice signal to the voice recognition module for processing, and then transmitting the recognition result to the program through the communication module. First, a far-field audio acquisition device is needed to collect the doctor's voice signal, which is converted into a digital signal through an analog-to-digital converter (ADC). Next, the digitized voice signal needs to be pre-processed, such as noise reduction, voice signal enhancement, etc., to improve the accuracy of voice recognition. The processed voice signal is transmitted to the voice recognition module, which will recognize the voice signal and convert the voice into text. Using the TTL communication module, the recognition result can be transmitted back to the actual business system through serial communication. The supported signal types include but are not limited to mouth opening commands, eye opening commands, pain stimulation commands, etc. After issuing the corresponding command, the business system will determine whether the patient in the picture captured by the camera has completed the action under the command.

[0073] The voice recognition module can use the uniOneUpdateTool tool to flash voice wake-up words and alarm words, and access the host system service through a driver-free TTL-to-USB chip.

[0074] Host system service, use koa2 to build system service, receive TTL hardware incoming data through serialport, push the received data to redis, use the interface to enable image recognition, and play the corresponding wake-up / alarm prompt tone. The host uses the interface to enable image recognition and play the corresponding wake-up / alarm prompt tone.

[0075] In some embodiments, the sedation monitoring system may also include a respiratory monitoring module, a pain stimulation module, and a vital signs module.

[0076] The respiratory monitoring module can use abdominal respiratory monitoring technology and be tied to the patient's abdomen through a belt. It can capture the patient's abdominal movement, quantify abdominal tension and monitor respiratory intensity, and then accurately assess the patient's respiratory rate and amplitude, providing key physiological data for medical staff.

[0077] In the respiratory monitoring module, a USB Bluetooth transceiver driver is configured, which is connected to the host through a Bluetooth transmitter. The patient wears a belt-type acquisition module, and the data is synchronized to the USB Bluetooth transceiver and connected to the host system service.

[0078] Host system service, use koa2 to build system service, use interface function to trigger the start and end of collection, receive USB Bluetooth transceiver data through serialport, push the received data to redis, and display it in a visual way in the host interface system.

[0079] The pain stimulation module can use a medium frequency therapeutic device to evaluate the nerve response and consciousness level of oral disease patients in a sedated state through current stimulation. The module is able to emit currents of specific frequencies and intensities to stimulate specific parts of the patient, such as the skin or muscles, to simulate the stimulation that may be encountered during surgery. By monitoring the patient's response to electrical stimulation, such as muscle contractions or changes in consciousness, a more accurate judgment of the patient's sedation depth can be made. The pain stimulation module is designed with patient safety and comfort in mind, ensuring that the current intensity is within a safe range and allowing medical staff to adjust the stimulation parameters according to the patient's specific situation.

[0080] Specifically, the pain stimulation module sends the host-specified signal through the infrared transceiver, and the infrared transceiver accesses the system service through a driver-free TTL-to-USB chip. The infrared transceiver is responsible for sending specific infrared signals to achieve functions such as switching control of the pain stimulation module and adjusting the stimulation intensity. The infrared transceiver uses a driver-free TTL-to-USB chip, which can communicate with the host directly through the USB interface without the need for an additional driver.

[0081] The host system service builds a lightweight server application based on the koa2 framework, which is responsible for processing requests from the client and triggering corresponding operation instructions. In order to test and verify the accuracy of the signal, the sscom tool is used to test and confirm the specific hex encoding of the signal. Once confirmed, the system service will convert the hex encoding into the corresponding control instructions through the defined interface function. These control instructions are then sent through the serialport library, and the USB-TTL conversion module is used to convert the digital signal into a format suitable for reception by the infrared transceiver. Finally, these signals are sent to the medium frequency therapeutic instrument through the infrared transceiver to perform the switch control of electrical stimulation and adjust the stimulation gear.

[0082] The automation and integrated design of the entire process not only improves the convenience and accuracy of operation, but also enhances the stability and maintainability of the system through centralized management of system services. In addition, the use of modern development tools such as the koa2 framework and the serialport library ensures the efficient operation and good scalability of system services, providing convenience for possible functional upgrades and maintenance in the future.

[0083] The vital signs module synchronizes the vital signs data collected by the monitor in real time to the central processing monitoring host through the data collector. These data include vital sign parameters such as blood pressure, blood oxygen saturation, heart rate, etc., and support patients to perform baseline measurements. The host monitors the vital signs data in real time and compares it with the baseline information. When the data exceeds the normal range, the system will issue an abnormal reminder. Doctors can view these reminder information in real time on the host screen.

[0084] The data collector can improve the efficiency and stability of data transmission. As an intermediate link, the data collector ensures that the data collected by the vital signs module can be accurately transmitted to the host. This design not only enhances the reliability of data transmission, but also makes it possible to perform preliminary data preprocessing through the data collector, reduce the burden on the host, and improve the response speed of the entire system. In addition, the use of the data collector also facilitates the scalability and upgrade of the equipment, making it easier to add new monitoring modules or improve existing modules in the future to adapt to changing medical needs and technological advances.

[0085] The following specifically describes the AI-based oral sedation treatment auxiliary monitoring method of this embodiment to achieve the effect of the above technical solution. Figure 2 As shown, the following steps are included:

[0086] Step 101, when the patient is in a sedated state, receiving a test speech recognition result sent by the speech recognition module, the test speech recognition result is obtained by the speech recognition module performing speech recognition on a sedation test instruction issued by a doctor.

[0087] Specifically, the host is connected to the speech recognition module through a communication interface, and can instantly receive the test speech recognition results sent by the speech recognition module. These results are obtained by the speech recognition module actively capturing, analyzing and recognizing the sedation test instructions issued by the doctor. The speech recognition module uses the built-in speech recognition algorithm and model to convert the doctor's voice instructions into digital signals, which are then encoded into a specific data format and sent to the host. After receiving these data, the host will further process and analyze them to determine the patient's sedation state and adjust the treatment strategy based on the analysis results. The sedation test instructions issued by the doctor can be "open mouth", "open eyes", "clench fist", etc.

[0088] For example, when the doctor issues the sedation test instruction "open your mouth", the speech recognition module will immediately capture this voice signal and convert it into the text message "open your mouth" using its speech recognition function. Subsequently, this text message is sent to the host as the test speech recognition result. The host receives the test speech recognition result for further processing.

[0089] Step 102: sending a test action recognition instruction to the image recognition module based on the test voice recognition result, so that the image recognition module obtains the person image of the patient within a preset time period.

[0090] Specifically, when the host receives the test speech recognition results sent by the speech recognition module, it will send the corresponding test action recognition instructions to the image recognition module based on these results. This instruction will trigger the image recognition module to start the camera within the system preset time period, capture and analyze the patient's image in real time, and identify whether the patient has completed specific actions according to the doctor's voice instructions.

[0091] For example, the doctor issues a voice command of "open your mouth", which is captured and recognized by the voice recognition module, and then the recognition result of the action command of "open your mouth" is sent to the host. After receiving this command, the host immediately sends a test action recognition command to the image recognition module, instructing it to start capturing and analyzing the patient's image in the next few seconds (for example, 5 seconds). The image recognition module is then started, and the camera is used to capture the patient's facial movements, especially the opening and closing of the mouth.

[0092] Step 103, receiving the test action recognition result sent by the image recognition module, where the test action recognition result is obtained by the image recognition module performing action recognition on the person image.

[0093] Specifically, when the image recognition module completes the motion analysis of the patient's image, it will send the motion recognition results to the host. These results are based on the image recognition module's capture and analysis of the patient's actions in executing the doctor's instructions. After receiving this data, the host will further analyze and process it.

[0094] For example, the doctor issues an "open mouth" command, and the voice recognition module has recognized this command and sent it to the host. Subsequently, the image recognition module captures the patient's facial movements, especially the opening and closing of the mouth, within a preset time period. If the patient opens his mouth as instructed, the image recognition module will recognize this action and send the test action recognition result of "open mouth action confirmation" to the host.

[0095] The image recognition module can call the mouth opening model, which belongs to the facial recognition model and can further obtain the key feature points of the mouth. The program evaluates the degree of mouth opening by calculating the vertical distance or horizontal distance between the key feature points of the mouth. According to specific needs and actual conditions, a dynamic threshold is used to determine whether the degree of mouth opening meets the conditions. The user's actions are identified based on the logic judgment algorithm based on threshold comparison. When the calculated mouth opening degree exceeds the threshold, the recognition of the mouth opening action is triggered. In real-time applications, the video frames captured by the camera can be continuously obtained, and face detection, key feature point detection and mouth opening degree calculation are performed on each frame to achieve real-time mouth opening action recognition.

[0096] Exemplarily, the mouth opening recognition algorithm in the image recognition module is as follows:

[0097] First, multiple key feature points of the mouth in the human image are identified through the open - mouth model, the coordinate values of each key feature point are obtained, and each key feature point is numbered. The recognition result of the key feature points is as follows Figure 3 shown.

[0098] Then, calculate the aspect ratio of the open mouth:

[0099] MOUTH_RATIO = (|| M51 - M59 || + || M53 - M57 ||) / 2 * || M55 - M49 || ; where M49, M51, M53, M55, M57 and M59 are Figure 3 the coordinate values of key feature points 49, 51, 53, 55 and 57. ||M51 - M59 || represents the vertical distance of the patient's lips at the right position of the philtrum, || M53 - M57 || represents the vertical distance of the patient's lips at the left position of the philtrum, and || M55 - M49 || represents the horizontal distance between the left and right sides of the patient's mouth, that is, the width of the mouth.

[0100] Threshold for wide open mouth: MOUTH_THRESHOLD = 0.55;

[0101] Threshold for open mouth: MOUTH_THRESHOLD_HALF = 0.4;

[0102] When MOUTH_RATIO >= MOUTH_THRESHOLD, it is judged as a wide open mouth;

[0103] When MOUTH_THRESHOLD > MOUTH_RATIO >= MOUTH_THRESHOL, it is judged as an open mouth;

[0104] When MOUTH_RATIO < MOUTH_THRESHOL, it is judged as not having an open mouth.

[0105] The image recognition module can call the open - eye model. The open - eye model belongs to the face recognition model and can further obtain the key feature points of the mouth. The program evaluates the degree of eye opening by calculating the vertical or horizontal distance between the key feature points of the eyes, calculates the eye aspect ratio EAR. When the human eye is open, EAR fluctuates around a specified threshold. When the human eye is closed, EAR drops rapidly and theoretically approaches zero. Therefore, we consider that when EAR is lower than a certain threshold, the eyes are in a closed state.

[0106] Exemplarily, the open - eye recognition algorithm in the image recognition module is as follows:

[0107] First, identify multiple key feature points of the eyes in the human image through the eye-opening model, obtain the coordinate values of each key feature point, and label each key feature point with a serial number. The key feature point recognition result is as Figure 4 shown.

[0108] Aspect ratio of the eye: EYE_RATIO = (||P2 - P6|| + ||P3 - P5||) / 2 * ||P4 - P1||; where P1, P2, P3, P4, P5, and P6 are Figure 4 the coordinate values of the key feature points P1, P2, P3, P4, P5, and P6. ||P2 - P6|| represents the vertical distance at the right position of the pupil of the patient's eye, ||P3 - P5|| represents the vertical distance at the left position of the pupil of the patient's eye, and ||P4 - P1|| represents the horizontal distance between the left and right sides of the patient's eye, that is, the width of the eye.

[0109] Threshold for wide-open eyes: EYE_THRESHOLD = 0.2;

[0110] Threshold for eye-opening: EYE_THRESHOLD_HALF = 0.12;

[0111] When EYE_RATIO >= EYE_THRESHOLD, it is judged as wide-open eyes;

[0112] When EYE_THRESHOLD > EYE_RATIO >= EYE_THRESHOL, it is judged as eye-opening;

[0113] When EYE_RATIO < EYE_THRESHOL, it is judged as not eye-opening.

[0114] In some embodiments, the image recognition module can also identify whether there are occlusions in the human image, which may affect the recognition of the patient's actions. The image recognition module detects abnormal elements in the image, such as surgical instruments, clothing, or other objects. These occlusions may interfere with the clear observation of the patient's expression and actions. Once an occlusion is detected, the analysis parameters will be automatically adjusted or the recognition will be retried after the occlusion is removed to ensure the accuracy and reliability of the action recognition. In addition, the host can also record the occlusion event, providing the medical staff with complete monitoring process information to help them better understand the context of the patient's reaction, thus making more accurate clinical judgments and decisions.

[0115] Step 104, determine the sedation test response result of the patient according to the test speech recognition result and the test action recognition result.

[0116] Specifically, the host compares the test speech recognition results with the test action recognition results to determine whether the patient's actual response is consistent with the doctor's instructions, thereby evaluating the patient's sedation state. This step is a key step in determining whether the patient is at an appropriate sedation level.

[0117] Continuing with the example of the above steps, if the doctor's instruction is "open your mouth", and the image recognition module confirms that the patient has indeed opened his mouth, the host will match this action recognition result with the voice recognition result of the voice recognition module. If the two are consistent, that is, the patient has indeed responded to the voice instruction, the host will determine that the patient's sedation test reaction result is positive, indicating that the patient's emotions and responsiveness are normal; conversely, if the test voice recognition result does not match the test action recognition result, the patient did not complete the command action within the specified time according to the doctor's instructions, or the action was performed incorrectly, then the sedation test reaction result is that the patient is in a state of excessive sedation.

[0118] Step 105: adjusting the sedative drug dosage currently received by the patient according to the sedation test response result.

[0119] Specifically, the host conducts a comprehensive assessment of the patient's sedation state based on the received sedation test response results, through built-in algorithms and preset medical protocols. If the test response results show that the patient's response to instructions is slow or not as expected, the host may determine that the dose of sedatives (such as midazolam, remimazolam, nitrous oxide) needs to be reduced to improve the patient's wakefulness and responsiveness. On the contrary, if the patient's response is too active or the sedation level is insufficient, the host may instruct to increase the drug dose to ensure that the patient reaches an appropriate sedation state.

[0120] When adjusting the dose, the host also considers the patient's physiological parameters, historical drug response data, and other relevant medical information to ensure that the adjusted dose is both safe and effective.

[0121] Adjustment instructions will be automatically sent to the drug delivery system to achieve accurate drug management, and all adjustment actions and parameter changes will be recorded and archived for medical staff to review and future treatment reference. Among them, the host will automatically send instructions to the drug delivery system to adjust the dosage of sedative drugs to change the settings of the infusion pump in the drug delivery system to accurately control the flow rate of the drug.

[0122] In order to explain the technical effect of this embodiment in more detail, Figure 5 and Figure 6 To illustrate the method of this embodiment, the method specifically includes the following steps:

[0123] Step 201, preparation work before starting the sedation auxiliary monitoring task.

[0124] After the user clicks the "Start" button on the operation interface, enters the patient number, and clicks the "OK" button, the host determines whether the patient is registered. If not, it guides the patient to register; if already registered, the host detects whether the camera is available. After confirmation, it starts the sedation assisted monitoring task.

[0125] Step 202, obtain the camera image, then the user adjusts the shooting angle, performs the picture test, and completes the test.

[0126] Step 203, obtaining the patient's vital sign baseline indicators.

[0127] Among them, the baseline indicators of vital signs include but are not limited to heart rate, blood oxygen, blood pressure, and respiratory rate. Heart rate reflects the frequency of heart activity, blood oxygen index shows the oxygen content in the blood, blood pressure indicates the pressure of blood on the blood vessel wall when blood flows in the blood vessel, and respiratory rate measures the speed of breathing.

[0128] The host collects the patient's physiological data through the connected data collector to obtain the baseline indicators of vital signs. The data collector is connected to the vital signs collection module, which can collect data such as heart rate, blood oxygen, blood pressure, and respiratory rate.

[0129] The host uses advanced sensors and data processing technology to accurately record these vital signs in real time, providing medical staff with a reference to the patient's physical condition before sedation, and serving as an important basis for monitoring and adjusting sedation plans in subsequent treatment. By continuously tracking and analyzing these key physiological parameters, the host can assist medical staff in promptly detecting any abnormal conditions of the patient and taking appropriate medical measures.

[0130] Step 204, determine whether the baseline index of the vital sign is within the normal range. If yes, proceed to step 205; if not, proceed to step 212.

[0131] Step 205: Determine an applicable sedation mode according to the health condition of the patient.

[0132] Among them, sedation modes include nitrous oxide mode, intravenous injection mode, oral sedation mode and no sedation mode.

[0133] Specifically, the host will consider factors such as the patient's health status, surgical needs, patient preferences, and possible drug interactions when determining the applicable sedation mode. First, the host will analyze the patient's baseline vital signs, such as heart rate, blood oxygen, blood pressure, and respiratory rate, as well as any medical history and drug allergies that may exist. Then, based on this information, the host will evaluate which sedation mode is most suitable for the patient. If the patient requires mild sedation, the nitrous oxide mode may be selected; for situations requiring moderate sedation, the intravenous mode may be used; if the patient prefers non-invasive sedation, the oral sedation mode may be considered; in certain low-risk or non-sedation-free surgeries, the no sedation mode may also be selected. The host will also consider the patient's age, weight, and overall health to ensure that the selected sedation mode can meet the needs of the surgery while minimizing potential risks and side effects. Through this comprehensive evaluation, the host is able to recommend the most appropriate sedation treatment plan for each patient in a personalized manner.

[0134] In the Nitrous Oxide mode, the Nitrous Oxide concentration is set. In the Intravenous Infusion mode, the medication dose and delivery rate are set. In the Oral Sedation mode, the medication dose is recorded.

[0135] Afterwards, the doctor performs sedation operations on the patient according to the set sedation mode to put the patient into a sedated state.

[0136] Step 206, receiving the test speech recognition result sent by the speech recognition module, wherein the test speech recognition result is obtained by the speech recognition module performing speech recognition on the sedation test instruction issued by the doctor.

[0137] This step refers to step 101 and will not be repeated here.

[0138] Step 207: Send a test action recognition instruction to the image recognition module based on the test speech recognition result. The test action recognition instruction can enable the image recognition module to obtain the person image of the patient within a preset time period.

[0139] This step refers to step 102 and will not be described in detail here.

[0140] Step 208, receiving the test action recognition result sent by the image recognition module, wherein the test action recognition result is obtained by the image recognition module performing action recognition on the person image.

[0141] This step refers to step 103 and will not be described again here.

[0142] Step 209, determining the sedation test reaction result of the patient according to the test speech recognition result and the test action recognition result.

[0143] This step refers to step 104 and will not be described in detail here.

[0144] Step 210, adjusting the sedative drug dosage currently received by the patient according to the sedation test response result.

[0145] This step refers to step 105 and will not be described again here.

[0146] Step 211, the sedation monitoring task ends.

[0147] After the task is completed, you can view the report, which summarizes all the key data and events of the patient during the treatment process, including voice recognition results, image recognition results, respiratory monitoring data, electrical stimulation intensity adjustment records, and sedative drug dosage adjustment history. The report is presented in a clear and structured manner, which is convenient for medical staff to quickly grasp the patient's sedation state and treatment response. In addition, the system also has a one-click export function, allowing medical staff to export reports in the form of electronic documents, such as PDF or Excel format, through simple operations. This not only facilitates medical record archiving, but also facilitates communication with patients or their families, ensuring the traceability and sharing of information. This function improves work efficiency and reduces the time spent by medical staff on report generation and compilation, allowing them to focus more on patient treatment and care.

[0148] Step 212: Send an alarm instruction to the alarm module so that the alarm module issues an alarm prompt.

[0149] First, the host determines the deviation of the baseline index of the vital sign from the normal range; then determines the alarm level according to the deviation; and then generates an alarm instruction based on the alarm level. Finally, the host sends an alarm instruction to the alarm module, so that the alarm module issues an alarm prompt.

[0150] Specifically, when the host monitors the patient's baseline vital signs, it will first compare them with the preset normal physiological range to determine the degree of deviation of each indicator. For example, if the heart rate is too fast or the blood oxygen saturation is too low, the host will calculate the deviation of these indicators from the normal range. According to the degree of these deviations, the host will determine the corresponding alarm level, such as mild, moderate or severe. Once the alarm level is determined, the host will generate alarm instructions based on this level, which clearly indicate the type and urgency of the deviation. The host will then send these alarm instructions to the alarm module, triggering the corresponding alarm prompts, such as flashing visual indicator lights, sound alarms, or displaying warning messages of different colors on the monitoring screen, to immediately notify medical staff to pay attention to the patient's abnormal state and take necessary countermeasures.

[0151] In some embodiments, while the patient is in a sedated state, it is necessary to continuously monitor his baseline vital signs.

[0152] When the patient's vital signs baseline indicators are normal, that is, all calculation results are within the normal range, there is no abnormal information or the abnormal information is within the acceptable range, the patient's physiological indicators meet the health standards. On the host operation interface, the display color of the vital signs baseline indicators is green.

[0153] Some calculated results of the patient's vital signs baseline indicators are beyond the normal range, but have not reached a dangerous level; slight abnormalities or fluctuations may require further observation and investigation; some physiological indicators slightly deviate from the normal range, but are still within the controllable range. When the above situations occur, it means that the user has an abnormal situation, but has not yet reached a dangerous state and needs attention. At this time, the display color of the vital signs baseline indicators on the host's operation interface is yellow, indicating a yellow warning state, and an alarm sounds.

[0154] An example of the criteria for a yellow alert status is as follows:

[0155] (1) Abnormal yellow blood oxygen:

[0156] Normal state (without oxygen): blood oxygen level is less than 95%.

[0157] Oxygen status: blood oxygen level is less than 97%.

[0158] (2) Abnormal heart rate yellow:

[0159] 20% above baseline;

[0160] 30% below baseline;

[0161] Less than 70 times / min.

[0162] (3) Abnormal yellow blood pressure:

[0163] Diastolic and systolic blood pressure are 20% higher than baseline;

[0164] Diastolic and systolic blood pressures were 20% lower than baseline.

[0165] (4) Mouth opening command yellow abnormality:

[0166] After issuing a mouth-opening command, no face or mouth feature points were recognized;

[0167] After issuing the mouth-opening command, the mouth opens only normally.

[0168] (5) Abnormal yellow color of the eye-opening command:

[0169] After issuing an eye-opening command, no facial or mouth feature points were recognized;

[0170] After issuing the eye-opening command, the eyes open only to a normal degree.

[0171] (6) Frowning yellow abnormality:

[0172] Through camera monitoring, when a frowning event is detected in real time, the system records the frown as a yellow abnormality. The probability of false triggering of frowning is relatively high, so it is only set as a yellow abnormality here.

[0173] (7) Abnormal yellow color in sedated state

[0174] Midazolam rate is greater than 0.5 mg / min and is administered continuously;

[0175] Remimazolam is administered at a rate greater than 0.5 mg / min and continuously;

[0176] The nitrous oxide dose is greater than 30% for 30 minutes, greater than 40% for 20 minutes, or greater than 50% for 10 minutes;

[0177] No "eyes wide open" detected for 5 minutes;

[0178] Continue to check for "eyes closed" for 1 minute.

[0179] (8) Abnormal yellow breath

[0180] Compared with the baseline, the fluctuation amplitude decreased by more than 50% or increased by more than 150%.

[0181] (9) Abnormal yellow color due to pain stimulation

[0182] The patient still has no response to a pain stimulus baseline value that is 2 times greater than the expected value of the pain stimulus.

[0183] It is understandable that the numerical values ​​of the above evaluation criteria are only an optional example and can be specifically adjusted according to actual conditions.

[0184] Some calculated results of the patient's vital signs baseline indicators are obviously beyond the normal range, indicating a serious problem; a critical condition or a sharp deterioration; a serious abnormality of physiological indicators, which may threaten the user's health or life safety. When the above situation occurs, it means that the patient has a serious abnormality and needs to take immediate action, and may need to be treated by a doctor in a timely manner. At this time, the display color of the vital signs baseline indicators on the host's operation interface is red, indicating a red warning state, and a rapid alarm sounds.

[0185] An example of a red alert status criterion is as follows:

[0186] (1) Abnormal blood oxygen red color:

[0187] Not receiving oxygen: blood oxygen level is less than 93%.

[0188] Oxygen status: blood oxygen level is less than 95%.

[0189] (2) Abnormal heart rate red:

[0190] 30% above baseline;

[0191] 50% below baseline;

[0192] Less than 50 times / min.

[0193] (3) Abnormal blood pressure red:

[0194] Diastolic and systolic blood pressure are 30% higher than baseline;

[0195] Diastolic and systolic blood pressures were 30% lower than baseline.

[0196] (4) The red color of the mouth opening command is abnormal:

[0197] When a face is recognized, the user does not open his mouth for more than 10 seconds.

[0198] (5) The red color of the eye-opening command is abnormal:

[0199] When a face is recognized, the user does not open his eyes for more than 10 seconds.

[0200] (6) Frowning with abnormal redness:

[0201] Through camera monitoring, when a frowning event is detected in real time, the system records the frown as a yellow abnormality. The probability of false triggering of frowning is relatively high, so it is only set as a yellow abnormality here.

[0202] (7) Abnormal red breathing

[0203] Compared with the baseline, the fluctuation range is reduced by more than 75% or increased by more than 200%.

[0204] (9) Abnormal redness due to painful stimulation

[0205] The patient still has no response to a baseline pain stimulus value that is 4 times greater than the expected value of the pain stimulus.

[0206] It is understandable that the numerical values ​​of the above evaluation criteria are only an optional example and can be specifically adjusted according to actual conditions.

[0207] In some embodiments, after step 210, the following steps are also included:

[0208] (11) Obtaining a pain expression recognition result sent by the image recognition module, wherein the pain expression recognition result is obtained by the image recognition module through recognizing the facial expression image of the patient after the sedative drug dosage is adjusted.

[0209] Specifically, when the host receives the pain expression recognition results sent by the image recognition module, this process involves a detailed analysis of the patient's facial expression images. The image recognition module uses advanced image processing and machine learning algorithms to capture and recognize the patient's facial expressions after the sedative dose is adjusted to detect whether there are signs of pain. These pain expression recognition results reflect the patient's possible discomfort or pain level, providing important non-verbal feedback information to medical staff. Based on these results, the host can further evaluate the effect of the sedative dose and decide whether further adjustments are needed.

[0210] Suppose that during oral surgery, a patient receives an adjustment in the dosage of midazolam. The image recognition module starts working after the medication is adjusted, capturing the patient's facial expressions through the camera and analyzing whether there are any expressions that may indicate pain, such as frowning, closing the eyes, or drooping the corners of the mouth. If the image recognition module detects these expressions and identifies them as expressions of pain, it sends this result to the host.

[0211] The image recognition module can call the frown model, which belongs to the facial recognition model. It can further obtain the key feature points of the eyebrows. By analyzing the position and shape changes of the key feature points of the eyebrows, the frown state can be evaluated. Currently, the slope and convexity of the eyebrow area are calculated to determine the degree of frowning. According to the actual situation and needs, a threshold is set to distinguish between frowning and normal state.

[0212] Exemplarily, the frown recognition algorithm in the image recognition module is as follows:

[0213] First, the frown model is used to identify multiple key feature points of the eyebrows and eye corners in the character image, the coordinate values ​​of each key feature point are obtained, and each key feature point is marked with a serial number. The key feature point recognition results are shown in the figure. Figure 7 shown.

[0214] Then calculate the aspect ratio of the eyebrow to the corner of the eye:

[0215] BROW_POS_LOCATION_RATIO = || B285 - B55 || / || B417 - B285 ||; B285, B55, B417 are Figure 7 The coordinate values ​​of the key feature points 285, 55 and 417. || B285 - B55 || represents the distance between the left and right brow vertices, and || B417 - B285 || represents the distance between the eyebrow head and the brow vertex.

[0216] Frowning threshold: BROW_THRESHOLD = 0.45;

[0217] When BROW_POS_LOCATION_RATIO >= BROW_THRESHOLD, it is determined as frowning.

[0218] When BROW_POS_LOCATION_RATIO < BROW_THRESHOLD, it is determined as not frowning.

[0219] (12)Based on the pain expression recognition result, readjust the dose of the sedative drug currently received by the patient.

[0220] Specifically, after receiving the pain expression recognition result, the host will readjust the dose of the sedative drug for the patient based on these results. This adjustment takes into account the patient's response to the current drug dose and the discomfort or pain state that the pain expression may indicate. The intelligent algorithm built into the host will evaluate the severity of the pain expression and combine it with the patient's overall physiological data to determine the most appropriate dose adjustment strategy. This process aims to ensure the patient's comfort and safety while maintaining the sedation level required for the operation or treatment.

[0221] Continuing with the above example, if the image recognition module detects pain expressions such as frowning in the patient after the midazolam dose adjustment, the host will analyze the recognition results of these expressions. If it is determined that the patient's pain expression indicates the need for a deeper sedation level, the host may decide to increase the dose of midazolam to relieve the patient's pain and provide sufficient sedation. After adjusting the dose, the host will continue to monitor the patient's vital signs and facial expressions to ensure that the new dose can produce the expected effect and respond promptly to any further changes.

[0222] Such an adjustment method can obtain the lowest-level sedation and analgesia measures required for the patient to achieve the ideal sedation effect.

[0223] In some embodiments, after step 210, the following steps are further included:

[0224] (21)Obtain the pain voice recognition result sent by the voice recognition module, where the pain voice recognition result is obtained by the voice recognition module performing voice recognition on the voice emitted by the patient after adjusting the sedative drug dose.

[0225] Specifically, the host can obtain the pain sound recognition results of the sound made by the patient after the sedative drug dosage is adjusted through integration with the voice recognition module. This process involves the real-time collection and analysis of the patient's voice by the voice recognition module, using sound processing technology and pattern recognition algorithms to detect possible signs of pain expressed in the sound, such as changes in pitch, increased volume, or specific groans. These pain sound recognition results are then transmitted to the host, providing medical staff with another important dimension for assessing the patient's pain status, helping to more fully understand the patient's sedation effect and pain perception.

[0226] For example, during a dental treatment, the patient received a dose of midazolam for sedation. During the treatment, the voice recognition module was activated to monitor any sounds that the patient might make. If the patient groaned or had unusual pitch changes in the voice during the treatment, the voice recognition module would identify these sound features and send the pain sound recognition results to the host.

[0227] (22) Based on the pain sound recognition result, the dosage of the sedative drug currently being received by the patient is adjusted again.

[0228] Specifically, based on the received pain sound recognition results, the host will activate its built-in intelligent analysis system to adjust the patient's current sedative dosage. This adjustment process is based on the analysis of the pain sound recognition results, combined with the patient's vital signs and other clinical data, to calculate a new sedative dosage. The host's goal is to optimize the sedation effect, reduce the patient's pain, and maintain a suitable treatment environment while ensuring the patient's safety.

[0229] Continuing with the dental treatment example, if the voice recognition module detects that the patient groans during treatment and identifies this as an indication of pain, the host will receive this pain sound recognition result. The host analyzes these results and, taking into account the patient's vital signs data such as heart rate and blood pressure, may decide to appropriately increase the dose of sedative drugs to better control the patient's pain response. Medical staff adjust drug delivery based on the host's recommendations to ensure that the patient remains in an appropriately sedated state throughout the treatment while reducing pain and discomfort.

[0230] This adjustment method can provide the lowest level of sedation and analgesia required for the patient to achieve the ideal sedation effect.

[0231] In some embodiments, after step 210, the following steps are further included:

[0232] (31) Obtaining the pain breathing recognition result sent by the breathing monitoring module, wherein the pain breathing recognition result is obtained by the breathing monitoring module by monitoring the frequency and amplitude of the patient's breathing after adjusting the sedative drug dosage.

[0233] Specifically, the host uses the respiratory monitoring module to obtain the patient's painful breathing recognition results after the sedative drug dosage is adjusted. This process involves the respiratory monitoring module carefully monitoring the patient's breathing rate and amplitude, using sensors to capture breathing activities, and applying specific algorithms to analyze changes in breathing patterns that may indicate pain. For example, faster or irregular breathing may be a manifestation of pain. The respiratory monitoring module converts these analysis results into painful breathing recognition results and sends them to the host so that medical staff can assess the patient's pain level and sedation effect based on this.

[0234] Assume that during a medical procedure, a patient is given a dose of sedative medication. A respiratory monitoring module is equipped on the patient and tracks the patient's breathing pattern in real time. If the patient's breathing rate increases significantly or the breathing amplitude becomes erratic after the sedative medication dose is adjusted, this may indicate that the patient is in pain or discomfort. The respiratory monitoring module identifies these changes as painful breathing signals and sends the results to the host computer for further analysis by medical staff.

[0235] (32) Based on the painful breathing identification result, the dosage of the sedative drug currently received by the patient is adjusted again.

[0236] Specifically, based on the painful breathing recognition results sent by the respiratory monitoring module, the host will re-evaluate and adjust the patient's current sedative dosage. The host's built-in intelligent decision support system will analyze the painful breathing recognition results and combine them with the patient's other vital signs and clinical data to determine whether the sedation regimen needs to be adjusted. If the analysis results show that the patient's breathing pattern changes are related to pain, the host may recommend increasing the dose of sedatives to achieve better analgesia and maintain an appropriate level of sedation.

[0237] Taking the above medical procedure as an example, when the respiratory monitoring module detects that the patient's breathing rate has increased and the amplitude has changed after the sedative dose is adjusted, this information is transmitted to the host. The host analyzes these painful breathing recognition results and refers to the patient's heart rate, blood pressure and other vital signs data to intelligently determine that the current sedative dose may not be enough to control the patient's pain. Therefore, the host may instruct medical staff to increase the sedative dose appropriately to better control the patient's pain response and ensure that the patient remains comfortable and stable during the procedure.

[0238] This adjustment method can provide the lowest level of sedation and analgesia required for the patient to achieve the ideal sedation effect.

[0239] In some embodiments, after step 210, the following steps are further included:

[0240] (41) Sending a pain stimulation instruction to the pain stimulation module so that the pain stimulation module adjusts the intensity of electrical stimulation to the patient.

[0241] Specifically, during the process of monitoring and evaluating the patient's sedation state, the host will send a pain stimulation instruction to the pain stimulation module. This instruction contains specific parameters for adjusting the intensity of electrical stimulation. After receiving the instruction, the pain stimulation module will change the intensity of electrical stimulation according to the instructions of the host to adapt to the patient's current sedation needs and treatment goals. This adjustment helps to simulate the stimulation that may be encountered during surgery and evaluate the patient's neural response and sedation level.

[0242] For example, during an oral surgery, the patient needs to receive different degrees of electrical stimulation to test his sedation level. The host sends instructions to the pain stimulation module to adjust the intensity of the electrical stimulation based on the patient's current state and treatment needs. If the patient does not respond significantly to the current intensity of electrical stimulation, the host may decide to increase the intensity of electrical stimulation to further test the patient's response.

[0243] (42) Obtaining a stress response recognition result sent by the image recognition module, wherein the stress response recognition result is obtained by the image recognition module by recognizing the patient's image when the patient is subjected to electrical stimulation of different intensities.

[0244] Specifically, after the pain stimulation module adjusts the intensity of electrical stimulation, the host will obtain the stress response recognition results sent by the image recognition module. These results are obtained by recognizing the patient's image when subjected to electrical stimulation of different intensities. The image recognition module uses computer vision technology to capture the patient's facial expressions and body movements, analyzes whether there is a stress response, such as frowning, body tension, etc., and sends these analysis results to the host.

[0245] Assume that during the electrical stimulation, the image recognition module captures subtle changes in the patient's facial expressions, such as slightly wrinkled eyebrows or pursed lips, which may be stress responses to the electrical stimulation. The image recognition module sends these recognition results to the host for further analysis and decision-making.

[0246] (43) Determine the patient’s sedation level based on the stress response identification result.

[0247] Specifically, the host analyzes the patient's stress response based on the stress response recognition results sent by the image recognition module to determine the patient's sedation level. This process involves a detailed assessment of the patient's facial expressions and body language, as well as an analysis of the relationship between these responses and the intensity of electrical stimulation. The host will take these factors into consideration to accurately determine the patient's sedation state.

[0248] For example, if the image recognition module detects that the patient has a frown and slight body movement after electrical stimulation, the host uses these stress response recognition results as the basis for assessing the sedation level. The host analyzes the intensity and duration of these reactions to determine whether the patient is at an appropriate level of sedation.

[0249] The pain stimulation module gives the patient pain stimulation through electric current. The minimum pain stimulation that can be perceived in the awake state is the pain stimulation baseline value, and the minimum pain stimulation that can be perceived in different sedation states is the pain measurement value corresponding to different sedation states. The ideal pain measurement value in the sedation state is a multiple of the pain baseline value, the pain stimulation expected value, and the specific value is determined based on a large number of experimental tests and clinical experience.

[0250] (44) Determine the dose of sedative medication that the patient receives based on the level of sedation.

[0251] Specifically, once the host determines the patient's sedation level, it will determine the dose of sedative medication the patient receives based on this level. This decision-making process is based on the relationship between the patient's sedation level and medication dose, as well as the need to maintain the patient's safety and comfort. The host will calculate the necessary medication dose adjustments to ensure that the patient remains appropriately sedated throughout the treatment.

[0252] Assume that the host determines through the stress response recognition results that the patient's current sedation level is low and may not be sufficient to cope with the surgical process. Based on this judgment, the host may decide to increase the dose of sedatives, such as midazolam, to improve the patient's sedation level. The host converts this decision into a specific dose adjustment instruction and sends it to the drug delivery system to achieve precise drug management.

[0253] In some embodiments, before sending the pain stimulation instruction to the pain stimulation module, the method further includes:

[0254] (S1) Determine the initial electrical stimulation intensity of the pain stimulation module according to the patient's baseline vital signs.

[0255] At the beginning of treatment, the host will determine the initial electrical stimulation intensity of the pain stimulation module based on the patient's baseline vital signs. This process involves analyzing key physiological parameters such as the patient's heart rate, blood pressure, respiratory rate, and blood oxygen saturation to assess the patient's initial health status and possible response to electrical stimulation. The host's built-in algorithm will combine these data to calculate an initial electrical stimulation intensity that is both safe and effective, aiming to conduct effective neurological function assessment without causing discomfort to the patient.

[0256] For example, in preparation for a medical procedure that requires an electrical stimulation assessment, the patient's baseline heart rate is 70 beats / minute, blood pressure is 120 / 80 mmHg, respiratory rate is 15 breaths / minute, and oxygen saturation is 98%. The host analyzes this data and determines that the initial electrical stimulation intensity is low to moderate to ensure that starting at this intensity will not cause excessive stress reactions in the patient.

[0257] (S2) generating a pain stimulation instruction according to the initial electrical stimulation intensity, wherein the pain stimulation instruction is used to enable the pain stimulation module to gradually increase the electrical stimulation intensity for the patient starting from the initial electrical stimulation intensity.

[0258] Specifically, based on the determined initial electrical stimulation intensity, the host generates pain stimulation instructions that detail how the pain stimulation module should start from the initial intensity and gradually increase the electrical stimulation intensity for the patient according to a preset program. This instruction generation process takes into account the safety and comfort of the patient, ensuring that the increase in electrical stimulation is gradual, allowing medical staff and patients enough time to adapt to the change in intensity and monitor any possible adverse reactions.

[0259] Assume that in the above medical procedure, the host determines that the initial electrical stimulation intensity is 2 mA based on the patient's baseline indicators. The host then generates a series of pain stimulation instructions, instructing the pain stimulation module to start from 2 mA and increase the intensity by 0.5 mA every 30 seconds until the predetermined maximum intensity is reached or the patient responds. Such instructions allow medical staff to monitor the patient's response to electrical stimulation and adjust the intensity of electrical stimulation or stop electrical stimulation if necessary.

[0260] This embodiment achieves comprehensive monitoring and intelligent management of the sedation treatment process for patients with oral diseases by integrating advanced speech recognition, image recognition, respiratory monitoring, electrical stimulation control and intelligent data processing technologies. In the scheme, the speech recognition module can accurately capture and identify the doctor's instructions and the patient's response sound, the image recognition module analyzes the patient's facial expressions and movements in real time, the respiratory monitoring module accurately tracks the patient's breathing pattern, and the pain stimulation module provides controllable neurological function assessment when needed. The host, as the core of the system, integrates the data of these modules, intelligently judges the patient's sedation level, and adjusts the dosage of sedative drugs accordingly to ensure that the patient is safe and comfortable throughout the treatment process. Overall, the solution significantly improves the level of personalization of treatment, the convenience of operation, the accuracy of monitoring and the safety of medical treatment, optimizes the utilization of medical resources, improves the patient's treatment experience, and provides strong decision-making support for medical staff.

[0261] The method provided in the above embodiment can be executed by a host of a sedation monitoring system, which is an electronic device. The electronic device in the embodiment of the present invention is described below from the perspective of hardware processing. Figure 8 , is a schematic diagram of a physical device structure of an electronic device in an embodiment of the present application.

[0262] It should be noted that Figure 8 The structure of the electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0263] like Figure 8 As shown, the electronic device includes a central processing unit (CPU) 401, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 402 or the program loaded from the storage part 408 to the random access memory (RAM) 403, such as executing the method described in the above embodiment. In RAM 403, various programs and data required for system operation are also stored. CPU 401, ROM 402 and RAM 403 are connected to each other through bus 404. Input / output (I / O) interface 405 is also connected to bus 404.

[0264] The following components are connected to the I / O interface 405: an input section 406 including an audio input device, a button switch, etc.; an output section 407 including a liquid crystal display (LCD) and an audio output device, an indicator light, etc.; a storage section 408 including a hard disk, etc.; and a communication section 409 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the I / O interface 405 as needed. A removable medium 411, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 410 as needed so that a computer program read therefrom is installed into the storage section 408 as needed.

[0265] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication part 409, and / or installed from a removable medium 411. When the computer program is executed by the central processing unit (CPU) 401, various functions defined in the present invention are performed.

[0266] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, apparatus, or device.

[0267] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram may represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in an order different from that marked in the accompanying drawings.

[0268] Specifically, the electronic device of this embodiment includes a processor and a memory, the memory is coupled to one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and one or more processors call the computer instructions to enable the electronic device to execute the method provided by the above embodiment.

[0269] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the electronic device described in the above embodiment; or may exist independently without being assembled into the electronic device. The above storage medium carries one or more computer programs, and when the above one or more computer programs are executed by a processor of the electronic device, the electronic device implements the method provided in the above embodiment.

[0270] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0271] As used in the above embodiments, the term "when..." may be interpreted to mean "if..." or "after..." or "in response to determining..." or "in response to detecting...", depending on the context. Similarly, the phrases "upon determining..." or "if (the stated condition or event) is detected" may be interpreted to mean "if determining..." or "in response to determining..." or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)", depending on the context.

[0272] Those skilled in the art can understand that to implement all or part of the processes in the above-mentioned embodiments, the processes can be completed by computer programs to instruct related hardware, and the programs can be stored in computer-readable storage media. When the programs are executed, they can include the processes of the above-mentioned method embodiments. The aforementioned storage media include: ROM or random access memory RAM, magnetic disk or optical disk and other media that can store program codes.

Claims

1. An AI-based oral sedation treatment auxiliary monitoring method, characterized in that: A host applied to a sedation monitoring system, wherein the sedation monitoring system further comprises an image recognition module and a speech recognition module, and the method comprises: When the patient is in a sedated state, receiving a test voice recognition result sent by the voice recognition module, wherein the test voice recognition result is obtained by the voice recognition module performing voice recognition on a sedation test instruction issued by a doctor; Based on the test speech recognition result, a test action recognition instruction is issued to the image recognition module, so that the image recognition module acquires the person image of the patient within a preset time period; receiving a test action recognition result sent by the image recognition module, wherein the test action recognition result is obtained by the image recognition module performing action recognition on the person image; Determining a sedation test reaction result of the patient according to the test speech recognition result and the test action recognition result; The dosage of the sedative drug currently received by the patient is adjusted according to the sedation test response result.

2. The method according to claim 1, characterized in that After the step of adjusting the dosage of the sedative drug currently received by the patient according to the sedation test reaction result, the method further includes: Acquire the pain expression recognition result sent by the image recognition module, wherein the pain expression recognition result is obtained by the image recognition module through recognizing the facial expression image of the patient after the sedative drug dosage is adjusted; According to the pain expression recognition result, the sedative drug dosage currently received by the patient is adjusted again.

3. The method according to claim 1, characterized in that: After the step of adjusting the dosage of the sedative drug currently received by the patient according to the sedation test reaction result, the method further includes: Acquiring a pain sound recognition result sent by the speech recognition module, wherein the pain sound recognition result is obtained by the speech recognition module performing speech recognition on a sound emitted by the patient after the dosage of the sedative drug is adjusted; According to the pain sound recognition result, the dosage of the sedative drug currently received by the patient is adjusted again.

4. The method according to claim 1, characterized in that The sedation monitoring system further includes a respiratory monitoring module, and after the step of adjusting the dosage of the sedative drug currently received by the patient according to the sedation test response result, further includes: Obtaining the painful breathing recognition result sent by the breathing monitoring module, wherein the painful breathing recognition result is obtained by the breathing monitoring module by monitoring the frequency and amplitude of the patient's breathing after adjusting the sedative drug dosage; According to the painful breathing recognition result, the dosage of the sedative drug currently received by the patient is adjusted again.

5. The method according to claim 1, characterized in that The sedation monitoring system further includes a pain stimulation module, and after the step of adjusting the dosage of the sedative drug currently received by the patient according to the sedation test response result, further includes: sending a pain stimulation instruction to the pain stimulation module so that the pain stimulation module adjusts the intensity of electrical stimulation to the patient; Acquiring a stress response recognition result sent by the image recognition module, wherein the stress response recognition result is obtained by the image recognition module through recognizing a person image of the patient when the patient is subjected to electrical stimulation of different intensities; determining the patient's sedation level according to the stress response identification result; The dosage of sedative medication received by the patient is determined based on the sedation level.

6. The method according to claim 5, characterized in that Before sending the pain stimulation instruction to the pain stimulation module, the method further includes: Determining the initial electrical stimulation intensity of the pain stimulation module according to the patient's baseline vital signs; A pain stimulation instruction is generated according to the initial electrical stimulation intensity, and the pain stimulation instruction is used to enable the pain stimulation module to gradually increase the electrical stimulation intensity for the patient starting from the initial electrical stimulation intensity.

7. The method according to any one of claims 1 to 6, characterized in that: When the patient is in a sedated state, before the step of receiving the test speech recognition result sent by the speech recognition module, the step further includes: Obtaining baseline vital signs of the patient; Determining whether the baseline indicators of the vital signs are within a normal range; If so, the applicable sedation mode is determined according to the health condition of the patient; if not, an alarm instruction is sent to the alarm module so that the alarm module issues an alarm prompt.

8. The method according to claim 7, characterized in that Before sending the alarm instruction to the alarm module, the method further includes: Determining the deviation of the baseline vital sign index from the normal range; determining an alarm level according to the degree of deviation; An alarm instruction is generated based on the alarm level.

9. A sedation monitoring system, characterized in that: The method comprises a host, an image recognition module and a speech recognition module, wherein the host is connected to the image recognition module and the speech recognition module respectively; the host comprises: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code comprises computer instructions, and the one or more processors call the computer instructions so that the host executes the method as described in any one of claims 1 to 8.

10. A computer-readable storage medium storing computer instructions, characterized in that: When the computer instructions are executed on an electronic device, the electronic device is caused to execute the method as claimed in any one of claims 1 to 8.

Images