Intelligent surgical anesthesia control method and system
Through intelligent surgical anesthesia control method, combined with medical scene image information and identification information, adaptive fasting, lying posture and emergency prompts are provided, which solves the problem of low intelligence in the existing technology, and improves the nursing efficiency of surgical anesthesia process and Safety enables personalized nursing experience.
Patent Information
- Application Number
- CN202410280686.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-03-12
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-03-12
AI Technical Summary
The existing smart hospital system is relatively low in general anesthesia surgery and cannot effectively meet the personalized needs of different patients during surgical anesthesia. There are problems such as improper preoperative fasting, risk of intraoperative complications, and difficulty in postoperative awakening. .
Using intelligent surgical anesthesia control method based on preset polling cycles, by obtaining medical scene image information, combining identification information for adaptive prompt analysis, identifying target patient objects, determining area identification types, and performing adaptive fasting, lying posture and emergency response Prompt processing, using food, dietary behavior and emergency equipment identification models to provide personalized nursing tips.
It improves the work efficiency of medical staff and the quality of patient care, ensures the safety of the surgical anesthesia process and personalized nursing experience, reduces emergency response time, and enhances the safety of the surgical process.
Smart Images

Figure CN118039185B_ABST
Abstract
Description
Technical Field
[0001] The present application belongs to the field of smart hospital technology and relates to an intelligent surgical anesthesia control method and system. Background Art
[0002] In general anesthesia surgery, on the one hand, there are points that need attention before, during and after the operation. For example: before the operation, the patient did not fast and abstain from drinking as required by the doctor, which led to reflux of gastric contents or vomiting during the operation, increasing the risk of aspiration pneumonia and suffocation; during the operation, the patient may have a variety of complications, the most common of which are respiratory and circulatory system complications, such as upper respiratory tract obstruction, inadequate ventilation, hypotension, hypertension, etc. At this time, it is necessary to ensure that the emergency equipment is complete; and after the operation, the patient may face risks such as tongue prolapse and sputum blocking the airway after waking up, so he needs to take a side-lying position with his head tilted to one side to facilitate medical staff to suction sputum to keep the airway open.
[0003] However, the IoT system used in smart hospitals in the existing technology has a low level of intelligence in the application scenario of general anesthesia surgery, and when faced with different patients and different types of surgeries, it cannot better meet the patients' personalized prompt needs throughout the entire process of surgical anesthesia. Summary of the invention
[0004] In order to overcome the defects and shortcomings of the prior art, the present application provides an intelligent surgical anesthesia control method and system.
[0005] In order to achieve the above objectives, this application adopts the following technical solutions:
[0006] According to one aspect of the present application, an intelligent surgical anesthesia control method is provided, the method comprising:
[0007] Based on a preset polling cycle, identification information is set according to the surgical plan information table; wherein the identification information is specifically any one of a first supervision identification, a second supervision identification, and a non-supervision identification;
[0008] Acquire multiple medical care scene image information, and filter out the medical care scene image information whose area identification type overlaps with the target area type set, wherein each medical care scene image information corresponds to a scene ID;
[0009] All screened medical scene image information is traversed in sequence, and adaptive prompt analysis is performed in combination with the identification information.
[0010] Preferably, in sequentially traversing all the screened medical scene image information and combining the identification information to perform adaptive prompt analysis, specifically including:
[0011] Identify the target patient object based on the currently traversed medical scene image information;
[0012] Determine the area identification type corresponding to the currently traversed medical scene image information;
[0013] If the area identification type is an inpatient department type, and the target patient object has a first supervision identification, an adaptive fasting reminder process is performed;
[0014] If the area identification type is the inpatient department type, and the target patient object has a second supervision identification, an adaptive lying posture prompt process is performed;
[0015] If the area identification type is an operating room type, adaptive emergency prompt processing is performed.
[0016] Preferably, the adaptive fasting reminder processing specifically includes:
[0017] Identify whether there is food in the scene;
[0018] When food is present, if it is identified that the target patient has a diet tendency, a first prompt message is generated and sent to the voice prompter based on the scene ID, wherein the first prompt message is used to prompt the target patient to fast, and the diet tendency of the target patient is specifically indicated as a decrease in the relative distance between the food and the target patient within a preset time period exceeding a preset movement tendency distance;
[0019] When food is present, if it is identified that the target patient has eating behavior, a second prompt message is generated and sent to the medical staff, wherein the second prompt message is used to prompt the medical staff to adjust the surgical plan in time;
[0020] When food is not present, no treatment is done.
[0021] Preferably, in identifying whether there is food in the scene, the method specifically includes: using a pre-trained food recognition model to perform recognition to obtain label information of whether there is food, if there is a label containing food, it is determined that there is food in the scene, and if there is a label not containing food, it is determined that there is no food in the scene;
[0022] The food recognition model is obtained through machine learning training using multiple sets of training data, the multiple sets of training data include a first category of data and a second category of data, each set of training data in the first category of data includes: a training image containing food and a label indicating that the training image contains food; each set of training data in the second category of data includes: a training image not containing food and a label indicating that the training image does not contain food.
[0023] Preferably, identifying the target patient subject's dietary behavior specifically includes:
[0024] Based on the currently traversed medical scene image information, the target patient object is captured according to the preset capture area to obtain the image information to be identified;
[0025] Input the image information to be identified into the pre-trained dietary behavior recognition model in sequence for identification to obtain label information of whether it contains dietary behavior;
[0026] The dietary behavior recognition model is obtained by machine learning training using multiple sets of training data, the multiple sets of training data include third category data and fourth category data, each set of training data in the third category data includes: training images containing dietary behavior and labels indicating that the training images contain dietary behavior; each set of training data in the fourth category data includes: training images not containing dietary behavior and labels indicating that the training images do not contain dietary behavior.
[0027] Preferably, the adaptive lying posture prompting process specifically includes:
[0028] Identify the lying posture type of the target patient object in the scene;
[0029] Based on the pre-stored postoperative lying posture rule table, the corresponding preset lying posture type is matched according to the type of surgery;
[0030] If it is determined that the lying posture type of the target patient object in the scene is different from the preset lying posture type, a third prompt message is generated and sent to the medical staff, wherein the third prompt message is used to prompt the medical staff to check the patient's lying posture in time.
[0031] Preferably, in the adaptive emergency prompt processing, it specifically includes:
[0032] Determine the boundaries of the preset emergency placement area according to the scene ID;
[0033] Emergency equipment is identified based on the preset emergency placement area boundary. If emergency equipment exists and its location is within the preset emergency placement area boundary, no processing is performed. Otherwise, a fourth prompt message is generated and sent to medical staff; wherein the fourth prompt message is used to prompt medical staff to arrange emergency equipment in the operating room.
[0034] Preferably, in identifying the emergency equipment based on the preset emergency placement area boundary, it specifically includes: using a pre-trained emergency equipment recognition model to identify whether there is an emergency equipment in the currently traversed medical scene image information;
[0035] Among them, the emergency equipment recognition model is obtained through machine learning training using multiple sets of training data, and the multiple sets of data include fifth category data and sixth category data. Each set of data in the fifth category data includes: training images containing emergency equipment and labels indicating that the training images contain emergency equipment. Each set of data in the sixth category data includes: training images not containing emergency equipment and labels indicating that the training images do not contain emergency equipment.
[0036] Preferably, the identification information is set according to the operation plan information table, specifically including:
[0037] Traversing the surgical plan information table in sequence;
[0038] Filter out surgical items with general anesthesia type as items to be marked;
[0039] Determine the fasting time range and side-lying time range;
[0040] If the current time point is within the fasting time range, a first supervision flag is set for the patient ID corresponding to the patient;
[0041] If the current time point is within the side-lying time range, a second supervision mark is set for the patient ID corresponding to the patient;
[0042] If the current time point is not within the fasting time range or the side-lying time range, the non-supervision flag is set;
[0043] Among them, the surgical plan information table includes multiple surgical items that require anesthesia, each surgical item includes the surgical name, surgical type, anesthesia type, anesthesia time range, and surgical time range, and each surgical item is associated with corresponding patient information, medical information, and family information.
[0044] According to another aspect of the present application, an intelligent surgical anesthesia control system is also provided, the system comprising:
[0045] An identification module is used to set identification information according to the operation plan information table based on a preset polling cycle; wherein the identification information is specifically any one of a first supervision identification, a second supervision identification, and a non-supervision identification;
[0046] A scene screening module is used to obtain multiple medical care scene image information and screen out medical care scene image information whose area identification type overlaps with the target area type set, wherein each medical care scene image information corresponds to a scene ID;
[0047] An adaptive prompt module, used to sequentially traverse all screened medical scene image information and perform adaptive prompt analysis in combination with the identification information;
[0048] In sequentially traversing all the screened medical scene image information and combining the identification information to perform adaptive prompt analysis, specifically including:
[0049] Identify the target patient object based on the currently traversed medical scene image information;
[0050] Determine the area identification type corresponding to the currently traversed medical scene image information;
[0051] If the area identification type is an inpatient department type, and the target patient object has a first supervision identification, an adaptive fasting reminder process is performed;
[0052] If the area identification type is the inpatient department type, and the target patient object has a second supervision identification, an adaptive lying posture prompt process is performed;
[0053] If the area identification type is an operating room type, adaptive emergency prompt processing is performed.
[0054] Compared with the prior art, this application has the following advantages and beneficial effects:
[0055] (1) This application analyzes medical care scenarios by applying image recognition and intelligent algorithms. In the adaptive prompt analysis combined with identification information, adaptive fasting prompt processing, adaptive lying posture prompt processing, and adaptive emergency prompt processing are adopted according to the regional identification type and identification information, thereby realizing intelligent nursing prompts during the surgical anesthesia process, improving the nursing efficiency of medical staff when caring for patients. This not only improves the work efficiency of medical staff, but also helps patients receive more accurate and effective care during surgical anesthesia, thereby improving the quality and safety of surgical anesthesia.
[0056] (2) This application uses food recognition models, lying posture recognition models and emergency equipment recognition models to identify in corresponding medical care scenarios, providing smart hospitals with more accurate and personalized nursing experience. This application uses food recognition models and dietary behavior recognition models to implement adaptive fasting reminder processing, thereby achieving the effect of personalized fasting supervision and improving the nursing efficiency of medical staff during ward rounds; the lying posture recognition model is used to implement adaptive lying posture reminder processing to determine whether the lying posture type of the target patient is reasonable, so as to promptly remind medical staff when the lying posture type is unreasonable, so as to help patients correct their lying posture, thereby providing patients with more personalized rehabilitation care effects; the emergency equipment recognition model is used to implement adaptive emergency reminder processing to ensure that the emergency equipment is within the boundaries of the preset emergency placement area, thereby reducing the emergency response time, improving the emergency response efficiency of medical staff during surgery, and increasing the safety of the surgical anesthesia process. BRIEF DESCRIPTION OF THE DRAWINGS
[0057] Figure 1A flowchart of an intelligent surgical anesthesia control method is provided in Example 1 of the present application;
[0058] Figure 2 A schematic block diagram of an intelligent surgical anesthesia control system is provided in Example 2 of the present application;
[0059] Figure 3 A schematic diagram of the structure of a terminal provided in Embodiment 3 of the present application;
[0060] Figure 4 This is a schematic diagram of the structure of a computer device provided in Example 4 of the present application. DETAILED DESCRIPTION
[0061] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.
[0062] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0063] In Example 1, this embodiment provides an intelligent surgical anesthesia control method, which is used to assist medical staff in improving the effect of patrol supervision during surgical anesthesia. The method is applied to a smart hospital system, which includes a server, multiple cameras, and multiple voice prompters. The server is wirelessly connected to the multiple cameras, the multiple voice prompters, the multiple first user terminals, and the multiple second user terminals, respectively. Each camera corresponds to a voice prompter, each camera corresponds to a scene ID, and each voice prompter also has a corresponding scene ID.
[0064] In this embodiment, the first user terminal is a terminal device used by medical staff, such as a mobile phone, a computer, a smart bracelet, etc., and the second user terminal is a terminal device used by the patient's family, such as a mobile phone, a smart bracelet.
[0065] In this embodiment, each camera is set in a space within the hospital to capture and obtain medical care scene image information. The medical care scene image information is specifically an image frame sequence, and is distinguished according to the spatial distribution within the hospital. Each medical care scene information corresponds to an area identification type, and the area identification type can be an outpatient department type, an inpatient department type, an operating room type, an examination area type, an emergency department type, etc.
[0066] It should be noted that this embodiment mainly identifies the scenarios of inpatient department types and operating room types to provide more personalized nursing tips for patients who need general anesthesia surgery.
[0067] In this embodiment, if Figure 1 As shown, the method includes:
[0068] Step S100: Based on a preset polling cycle, identification information is set according to the surgical plan information table; wherein the identification information is specifically any one of a first supervision identification, a second supervision identification, and a non-supervision identification.
[0069] In this embodiment, the preset polling period is set to 10 minutes, which can be adjusted to 30 minutes or the like according to actual conditions.
[0070] In this embodiment, the surgical plan information table is pre-entered by medical staff. The surgical plan information table includes multiple surgical items that require anesthesia. Each surgical item includes the name of the operation, the type of operation, the type of anesthesia, the anesthesia time range, and the surgical time range. Each surgical item is associated with the corresponding patient information, medical information, and family information. Exemplarily, surgical item a1 is a surgical plan for patient x1. Surgical item a1 is associated with patient x1, doctor x2, doctor x3, nurse x4, nurse x5, patient family member x6, and patient family member x7. Here, a1, x1...x7 are only used for examples and are not specifically limited. Among them, patient information, medical information, and family information are all the names and patient IDs of the corresponding personnel. The patient name can be extracted from the patient information;
[0071] The identification information is set according to the operation plan information table, specifically including:
[0072] Step S110: traverse the surgical plan information table in sequence;
[0073] Step S120: Filter out surgical items whose anesthesia type is general anesthesia as items to be marked;
[0074] Step S130: Determine the fasting time range and the side-lying time range;
[0075] In this embodiment, the process of determining the fasting time range specifically includes: based on the pre-stored anesthesia fasting and eating rules table, determining the corresponding preset fasting time threshold according to the type of surgery, subtracting the preset fasting time threshold from the starting time point of the anesthesia time range as the timestamp obtained as the starting critical point of the fasting time range, and using the timestamp corresponding to the starting time point of the anesthesia time range as the end critical point of the fasting time range.
[0076] In this embodiment, the process of determining the lateral lying time range specifically includes: based on the pre-stored anesthesia rule table, determining the corresponding preset awakening time threshold according to the anesthesia type, adding the end time point of the anesthesia time range plus the preset awakening time threshold to obtain the timestamp as the end critical point of the lateral lying time range, and using the timestamp corresponding to the end time point of the operation time range as the starting critical point of the lateral lying time range.
[0077] In actual application, anesthesia treatment needs to be completed before surgery, so the end time point of the anesthesia time range is smaller than the start time point of the surgery time range.
[0078] Step S140: If the current time point is within the fasting time range, a first supervision flag is set for the patient ID corresponding to the patient;
[0079] Step S150: if the current time point is within the side-lying time range, a second monitoring flag is set for the patient ID corresponding to the patient;
[0080] Step S160: If the current time point is not within the fasting time range or the side-lying time range, a non-supervision flag is set.
[0081] Step S200: Acquire multiple medical care scene image information, and filter out the medical care scene image information with overlapping area identification types and target area type sets; each medical care scene image information corresponds to a scene ID. Exemplarily, the target area type set in this embodiment includes an inpatient department type and an operating room type to ensure that the processed scenes are pre-operative, intraoperative, and post-operative situations of general anesthesia surgery.
[0082] Step S300: traverse all screened medical scene image information in sequence, and perform adaptive prompt analysis in combination with identification information;
[0083] In the process of sequentially traversing all the screened medical scene image information and combining the identification information for adaptive prompt analysis, the following steps are specifically included:
[0084] Step S310: Identify the target patient object based on the currently traversed medical scene image information;
[0085] In identifying the target patient object based on the currently traversed medical scene image information, specifically including:
[0086] Step S311: Based on the pre-stored patient face database, facial feature comparison is performed on the personnel in the medical scene image information in turn, and the similarity between the facial features is calculated. If the similarity is greater than or equal to the preset similarity threshold, it is judged as a successful recognition, otherwise it is judged as a failed recognition;.
[0087] Step S312: For the successfully identified person, determine whether he is the target patient object according to the surgery plan information table. If the patient ID of the person is in the surgery plan information table, he is determined to be the target patient object, otherwise he is determined to be a non-target patient object.
[0088] The patient face database includes multiple patient face feature information, each of which corresponds to a patient ID, so that all target patient objects are matched according to whether the patient ID is in the surgery plan information table.
[0089] In this embodiment, the facial feature information of the patient is obtained by the same feature extraction algorithm, such as the LBP feature extraction algorithm or the Haar feature extraction algorithm. It should be noted that the LBP feature extraction algorithm or the Haar feature extraction algorithm are both prior arts. For example, the LBP feature extraction algorithm is an algorithm for describing the local texture features of an image. It generates a binary number by comparing the grayscale values of the central pixel and its neighboring pixels, thereby obtaining a histogram representing the local texture features. In face recognition, LBP can be used to extract local texture features of the face. The Haar feature extraction algorithm is an image feature descriptor that obtains features by calculating the pixel value difference between adjacent areas in an image. In face recognition, the Haar feature can be used to describe the shape and texture information of the face.
[0090] When performing facial feature comparison, feature vectors extracted from facial images are used, and then the similarity between two feature vectors is calculated to measure the degree of match between them, for example, cosine similarity is used to measure the degree of match between them.
[0091] Step S320: determining the area identification type corresponding to the currently traversed medical scene image information;
[0092] Step S330: If the area identification type is an inpatient department type, and the target patient object has a first supervision identification, then an adaptive fasting reminder process is performed;
[0093] In actual application, a camera is used to shoot the target patient object, and an adaptive fasting reminder is performed when the target patient object has a first supervision mark, so as to timely remind the target patient object of the fasting requirement before anesthesia to avoid affecting the anesthetic effect. The adaptive fasting reminder can be used to personalize the fasting time according to the patient's surgery type, thereby improving the supervision efficiency for the patient and reducing the surgical risk. The adaptive fasting reminder can be used to arrange a reasonable fasting time according to the surgery time and surgery type, thereby avoiding surgery postponement or emergency surgery requiring additional preparation work, thereby improving the efficiency of the patient in completing the surgical process.
[0094] In the adaptive fasting cue processing, specifically including:
[0095] Step S331: Identify whether there is food in the scene;
[0096] In this embodiment, a pre-trained food recognition model is used for recognition to obtain label information of whether food is contained. The food recognition model is obtained by machine learning training using multiple sets of training data. The multiple sets of training data include a first category of data and a second category of data. Each set of training data in the first category of data includes: a training image containing food and a label indicating that the training image contains food; each set of training data in the second category of data includes: a training image not containing food and a label indicating that the training image does not contain food; and then, it is judged whether there is food in the scene by using the label containing food and the label not containing food.
[0097] Step S332: When food is present, if it is identified that the target patient object has a dietary tendency, a first prompt message is generated and sent to the voice prompter based on the scene ID, wherein the first prompt message is used to prompt the target patient object to fast; wherein the dietary tendency of the target patient object is specifically indicated as a decrease in the relative distance between the food and the target patient object within a preset time exceeding a preset exercise trend distance, and the first prompt message is obtained by text splicing of the patient name and the fasting reminder text.
[0098] In actual application, the preset time and the preset trend change distance can be set according to actual needs. For example, the preset time is set to 2 seconds, and the preset trend change distance is set to 20 centimeters. The specific values are not limited here.
[0099] Step S333: when food is present, if it is identified that the target patient has eating behavior, a second prompt message is generated and sent to the medical staff, wherein the second prompt message is used to prompt the medical staff to adjust the surgical plan in time;
[0100] Specifically, by identifying the target patient's eating behavior, the medical staff can be prompted to adjust the surgical plan in time when the target patient's eating behavior is identified, so as to avoid affecting the operation due to the patient's unexpected fasting. Since the medical staff cannot take care of the patient specifically, the risk of surgical anesthesia is reduced by identifying the target patient's eating behavior.
[0101] In this embodiment, the identification of the target patient's dietary behavior specifically includes:
[0102] Step S333a: Based on the currently traversed medical scene image information, a screen capture of the target patient object is performed according to a preset capture area to obtain image information to be identified;
[0103] Step S333b: sequentially inputting the image information to be identified into a pre-trained dietary behavior recognition model for identification to obtain label information of whether it contains dietary behavior, wherein the dietary behavior recognition model is obtained by machine learning training using multiple sets of training data, wherein the multiple sets of training data include third-category data and fourth-category data, wherein each set of training data in the third-category data includes: training images containing dietary behavior and labels indicating that the training images contain dietary behavior; and each set of training data in the fourth-category data includes: training images not containing dietary behavior and labels indicating that the training images do not contain dietary behavior;
[0104] The second prompt information includes the current time point, the patient's name and a screenshot image of the eating behavior, so as to provide timely feedback to medical staff on the patient's eating behavior during the fasting period; wherein, the screenshot image of the eating behavior is obtained by taking a screenshot of the currently traversed medical scene image information when the eating behavior recognition model recognizes that the target patient object has eating behavior.
[0105] Step S334: When no food exists, no processing is performed.
[0106] Step S340: If the area identification type is the inpatient department type, and the target patient object has a second supervision identification, then an adaptive lying posture prompt process is performed;
[0107] In actual application, in the adaptive lying posture prompt processing, the corresponding preset lying posture type is matched according to the type of surgery, and it is judged whether the lying posture type of the target patient is reasonable, so that medical staff can be prompted in time when the lying posture type is unreasonable to help patients correct their lying posture, thereby providing patients with more personalized rehabilitation care effects and improving patients' personalized rehabilitation effects.
[0108] In the adaptive lying posture prompt processing, it specifically includes:
[0109] Step S341: Identify the lying posture type of the target patient object in the scene;
[0110] In this embodiment, a pre-trained lying posture recognition model is used to perform recognition to obtain label information of the lying posture type of the target object. The lying posture recognition model is obtained by machine learning training using multiple sets of training data, and each set of training data in the multiple sets of training data includes a training image containing a lying posture of a person and a label identifying the lying posture type of the person;
[0111] Step S342: Based on the pre-stored postoperative supine posture rule table, the corresponding preset supine posture type is matched according to the operation type; in this embodiment, the postoperative supine posture rule table is specifically a mapping relationship table between the operation type and the preset supine posture type; for example, when the operation type is general anesthesia, the corresponding preset supine posture type is supine position without pillow, by tilting the head to one side to ensure the airway is unobstructed and prevent aspiration of vomit; when the operation type is subarachnoid anesthesia, the corresponding preset supine posture type is supine position or head down position to reduce the risk of headache and hypotension; when the operation type is cranial surgery, The corresponding preset lying position type is a 15° to 30° head-high, feet-low slope position to promote intracranial venous return and reduce cerebral edema; when the surgery type is cervical and thoracic surgery, the corresponding preset lying position type is a high slope position to promote breathing and circulation; when the surgery type is abdominal surgery, the corresponding preset lying position type is a low semi-sitting position to reduce abdominal tension and relieve pain; when the surgery type is spinal or hip surgery, the corresponding preset lying position type is a prone or supine position, etc. The specific mapping relationship can be determined by medical staff according to expert experience and is not specifically limited here.
[0112] Step S343: If it is determined that the lying posture type of the target patient object in the scene is different from the preset lying posture type, a third prompt message is generated and sent to the medical staff, wherein the third prompt message is used to prompt the medical staff to check the patient's lying posture in time, and the third prompt message includes the current time point, the patient's name and the lying posture screenshot image.
[0113] In this embodiment, the lying posture screenshot image is obtained by taking a screenshot of the currently traversed medical care scene image information when the lying posture type of the target patient object in the scene is different from the preset lying posture type.
[0114] Step S350: If the area identification type is an operating room type, an adaptive emergency prompt process is performed;
[0115] In actual application, through adaptive emergency prompt processing, when the location of the emergency equipment is outside the boundary of the preset emergency placement area, the medical staff will be reminded in time to arrange the location of the emergency equipment, thereby ensuring that it is within the boundary of the preset emergency placement area, reducing the emergency response time, which helps to improve the emergency response efficiency of medical staff during surgery and increase the safety of the surgical process.
[0116] In the adaptive emergency prompt processing, it specifically includes:
[0117] Step S351: Determine the boundary of the preset emergency placement area according to the scene ID;
[0118] In this embodiment, the preset emergency placement area boundary is obtained by medical staff marking each medical care scene in advance, and the preset emergency placement area boundary is stored in the emergency placement area boundary image set. Since the camera setting method is fixed, the preset emergency placement area boundary does not need to be changed before each recognition, and each medical care scene image information corresponds to a scene ID, and then the preset emergency placement area boundary corresponding to the current medical care scene is found according to the scene ID.
[0119] Step S352: Identify emergency equipment based on the preset emergency placement area boundary. If emergency equipment exists and the location of the emergency equipment is within the preset emergency placement area boundary, no processing is performed. Otherwise, a fourth prompt message is generated and sent to medical staff; the fourth prompt message is used to prompt medical staff to arrange emergency equipment in the operating room, and the fourth prompt message includes the current time point, scene ID and operating room screenshot image.
[0120] In identifying emergency equipment based on the boundary of a preset emergency placement area, specifically including: using a pre-trained emergency equipment recognition model to identify whether there is emergency equipment in the currently traversed medical scene image information; wherein the emergency equipment recognition model is obtained by machine learning training using multiple sets of training data, the multiple sets of data include fifth category data and sixth category data, each set of data in the fifth category data includes: a training image containing emergency equipment and a label indicating that the training image contains emergency equipment, and each set of data in the sixth category data includes: a training image not containing emergency equipment and a label indicating that the training image does not contain emergency equipment;
[0121] In this embodiment, the judgment condition for whether the position of the emergency device is within the boundary of the preset emergency placement area is specifically: mark the bottom edge corresponding to the emergency device, and if the bottom edge is within the boundary of the preset emergency placement area, it is judged that the position of the emergency device is within the boundary of the preset emergency placement area; otherwise, the position of the emergency device is outside the boundary of the preset emergency placement area.
[0122] Exemplarily, in marking the bottom edge corresponding to the emergency equipment, it specifically includes: first finding the bottom edge of the emergency equipment through an edge detection algorithm, then using the average grayscale value in the grayscale histogram of the currently traversed medical scene image information as the image threshold, performing binarization processing according to the image threshold, and marking the bottom edge of the emergency equipment after the binarization processing based on the region growing method.
[0123] In this embodiment, the edge detection algorithm can adopt Canny edge detection, Sobel operator, etc., by detecting the edges in the image, that is, the places where the pixel values change dramatically, so as to find the bottom edge of the emergency equipment; this embodiment performs binarization processing according to the image threshold to make the bottom edge of the emergency equipment in the image more obvious, thereby improving the accuracy of edge detection; it should also be noted that the edge detection algorithm and the region growing method are both existing technologies in image processing.
[0124] In Example 2, this embodiment provides an intelligent surgical anesthesia control system, which corresponds to the above-mentioned intelligent surgical anesthesia control.
[0125] like Figure 2 As shown, the system includes: an identification module, which is used to set identification information according to the operation plan information table based on a preset polling cycle; wherein the identification information is specifically any one of the first supervision identification, the second supervision identification, and the non-supervision identification;
[0126] A scene screening module is used to obtain multiple medical care scene image information and screen out medical care scene image information whose area identification type overlaps with the target area type set, wherein each medical care scene image information corresponds to a scene ID;
[0127] The adaptive prompt module is used to traverse all the screened medical scene image information in sequence and perform adaptive prompt analysis in combination with the identification information.
[0128] In Example 3, Figure 3 As shown, this embodiment provides a terminal, including: at least one memory and at least one processor; wherein, at least one memory is used to store program codes, and at least one processor is used to call the program codes stored in at least one memory to execute any one of the intelligent surgical anesthesia control methods in the above embodiments.
[0129] In embodiment 4, this embodiment provides a computer device, which may be a server, and its internal structure diagram may be as follows: Figure 4 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is a physical layer for storing various databases. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, an intelligent surgical anesthesia control method is implemented.
[0130] Those skilled in the art will understand that Figure 4 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.
[0131] In Example 5, this embodiment provides a storage medium, which is used to store program codes, and the program codes are used to execute the above-mentioned intelligent surgical anesthesia control method.
[0132] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other to form a technical solution.
[0133] In addition, although each operation is described in a specific order, this should not be understood as requiring these operations to be performed in the specific order shown or in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be interpreted as limiting the scope of the present disclosure. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.
[0134] The above embodiments are preferred implementation modes of the present application, but the implementation modes of the present application are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present application should be equivalent replacement methods and are included in the protection scope of the present application.
Claims
1. An intelligent surgical anesthesia control method, It is characterized in that include: Based on a preset polling cycle, setting identification information according to the surgical plan information table; The identification information is specifically any one of the first supervision identification, the second supervision identification, and the non-supervision identification; Acquire multiple medical care scene image information, and filter out the medical care scene image information whose area identification type overlaps with the target area type set, wherein each medical care scene image information corresponds to a scene ID; All the screened medical scene image information is traversed in sequence, and adaptive prompt analysis is performed in combination with the identification information; In sequentially traversing all the screened medical scene image information and combining the identification information to perform adaptive prompt analysis, specifically including: Identify the target patient object based on the currently traversed medical scene image information; Determine the area identification type corresponding to the currently traversed medical scene image information; If the area identification type is an inpatient department type, and the target patient object has a first supervision identification, an adaptive fasting reminder process is performed; If the area identification type is the inpatient department type, and the target patient object has a second supervision identification, an adaptive lying posture prompt process is performed; If the area identification type is an operating room type, an adaptive emergency prompt process is performed; The identification information is set according to the operation plan information table, specifically including: Traversing the surgical plan information table in sequence; Filter out surgical items with general anesthesia type as items to be marked; Determine the fasting time range and side-lying time range; If the current time point is within the fasting time range, a first supervision flag is set for the patient ID corresponding to the patient; If the current time point is within the side-lying time range, a second supervision mark is set for the patient ID corresponding to the patient; If the current time point is not within the fasting time range or the side-lying time range, the non-supervision flag is set; The surgical plan information table includes multiple surgical items that require anesthesia, and each surgical item includes the name of the surgery, the type of surgery, the type of anesthesia, the anesthesia time range, and the surgical time range; in the adaptive fasting prompt processing, it specifically includes: Identify whether there is food in the scene; When food is present, if it is identified that the target patient has a diet tendency, a first prompt message is generated and sent to the voice prompter based on the scene ID, wherein the first prompt message is used to prompt the target patient to fast, and the diet tendency of the target patient is specifically indicated as a decrease in the relative distance between the food and the target patient within a preset time period exceeding a preset movement tendency distance; When food is present, if it is identified that the target patient has eating behavior, a second prompt message is generated and sent to the medical staff, wherein the second prompt message is used to prompt the medical staff to adjust the surgical plan in time; When food is not present, no treatment is done; In the adaptive lying posture prompt processing, it specifically includes: Identify the lying posture type of the target patient object in the scene; Based on the pre-stored postoperative lying posture rule table, the corresponding preset lying posture type is matched according to the type of surgery; If it is determined that the lying posture type of the target patient object in the scene is different from the preset lying posture type, a third prompt message is generated and sent to the medical staff, wherein the third prompt message is used to prompt the medical staff to check the patient's lying posture in time; In the adaptive emergency prompt processing, it specifically includes: Determine the boundaries of the preset emergency placement area according to the scene ID; Emergency equipment is identified based on the preset emergency placement area boundary. If emergency equipment exists and its location is within the preset emergency placement area boundary, no processing is performed. Otherwise, a fourth prompt message is generated and sent to medical staff; wherein the fourth prompt message is used to prompt medical staff to arrange emergency equipment in the operating room.
2. The method according to claim 1, It is characterized in that In identifying whether there is food in the scene, specifically comprising: using a pre-trained food recognition model to identify whether there is food label information, if there is a label containing food, it is determined that there is food in the scene, and if there is a label not containing food, it is determined that there is no food in the scene; The food recognition model is obtained through machine learning training using multiple sets of training data, the multiple sets of training data include a first category of data and a second category of data, each set of training data in the first category of data includes: a training image containing food and a label indicating that the training image contains food; each set of training data in the second category of data includes: a training image not containing food and a label indicating that the training image does not contain food.
3. The method according to claim 1, It is characterized in that In identifying the dietary behaviors of target patients, specifically: Based on the currently traversed medical scene image information, the target patient object is captured according to the preset capture area to obtain the image information to be identified; Input the image information to be identified into the pre-trained dietary behavior recognition model in sequence for identification to obtain label information of whether it contains dietary behavior; The dietary behavior recognition model is obtained by machine learning training using multiple sets of training data, the multiple sets of training data include third category data and fourth category data, each set of training data in the third category data includes: training images containing dietary behavior and labels indicating that the training images contain dietary behavior; each set of training data in the fourth category data includes: training images not containing dietary behavior and labels indicating that the training images do not contain dietary behavior.
4. The method according to claim 1, It is characterized in that In identifying the emergency equipment based on the preset emergency placement area boundary, specifically including: using a pre-trained emergency equipment recognition model to identify whether there is an emergency equipment in the currently traversed medical scene image information; Among them, the emergency equipment recognition model is obtained through machine learning training using multiple sets of training data, and the multiple sets of data include fifth category data and sixth category data. Each set of data in the fifth category data includes: training images containing emergency equipment and labels indicating that the training images contain emergency equipment. Each set of data in the sixth category data includes: training images not containing emergency equipment and labels indicating that the training images do not contain emergency equipment.
5. An intelligent surgical anesthesia control system, It is characterized in that For executing the method according to claim 1, the system comprises: An identification module is used to set identification information according to the operation plan information table based on a preset polling cycle; wherein the identification information is specifically any one of the first supervision identification, the second supervision identification, and the non-supervision identification; A scene screening module is used to obtain multiple medical care scene image information and screen out the medical care scene image information whose area identification type overlaps with the target area type set, wherein each medical care scene image information corresponds to a scene ID; An adaptive prompt module, used to sequentially traverse all screened medical scene image information and perform adaptive prompt analysis in combination with the identification information; In sequentially traversing all the screened medical scene image information and combining the identification information to perform adaptive prompt analysis, specifically including: Identify the target patient object based on the currently traversed medical scene image information; Determine the area identification type corresponding to the currently traversed medical scene image information; If the area identification type is an inpatient department type, and the target patient object has a first supervision identification, an adaptive fasting reminder process is performed; If the area identification type is the inpatient department type, and the target patient object has a second supervision identification, an adaptive lying posture prompt process is performed; If the area identification type is an operating room type, adaptive emergency prompt processing is performed.
Citation Information
Patent Citations
Automatic anesthesia control method based on Bayesian network
CN103432650A
Speech recognition intelligent anesthesia system based on big data
CN110232961A
Intelligent operating room operation management method based on digitization
CN117079787A
Scene-adaptive diet monitoring method and device, equipment and storage medium
CN117393109A
Method for controlling depth and appropriate level of anesthesia in surgical assistance
RU2131213C1
