Operating room anesthesia data monitoring method

Abstract

The invention relates to the technical field of anesthesia data monitoring, and discloses an operating room anesthesia data monitoring method comprising the following steps: 1, installing a high definition camera in an operation area, and collecting an operation video in real time; 2, analyzing the video by using AI; step 3, extracting associated anesthesia data items from a monitor interface, amplifying a display area of the anesthesia data items to a preset proportion, and meanwhile, superposing a current operation stage name, predicted duration and operation risk prompts on a sidebar of the interface; and 4, restoring the display screen of the monitor to a conventional display state. The camera is used for shooting the operation process in real time, the AI server is used for analyzing the operation stage of the operation in real time, then the display interface of the monitor is dynamically adjusted according to data needing to be focused on in the corresponding operation stage, anesthesia data needing to be focused on in the current stage are amplified and displayed, and the operation process is synchronously prompted. Therefore, the cognitive load of the anesthetist is reduced, and the intraoperative safety is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of anesthesia data monitoring, and specifically provides an anesthesia data monitoring method for operating rooms. Background Art

[0002] Anesthesia data monitoring in the operating room plays a crucial role in obstetric and gynecological surgeries. It mainly includes real-time monitoring of the vital signs of parturients (such as blood pressure, heart rate, respiratory rate, and oxygen saturation) to ensure the stability of the basic vital state; monitoring of the depth of anesthesia, using instruments such as bispectral index (BIS) to ensure appropriate anesthesia effect; monitoring of airway patency, especially during general anesthesia, to ensure smooth breathing; and monitoring of the input and output volume, including fluid input, blood loss, and urine output, to reflect the physical condition and adjust the fluid replacement plan. In addition, blood gas analysis can understand the acid-base balance, electrolytes, and oxygen supply status of parturients, further ensuring the safety of the surgery. These anesthesia data monitoring measures jointly ensure the safety and smooth progress of obstetric and gynecological surgeries.

[0003] In the prior art, an anesthesia monitor centrally displays the vital sign data of patients (such as heart rate, blood pressure, oxygen saturation, respiratory rate, etc.) through a display screen. However, due to the large number of data items and dense arrangement, anesthesiologists need to repeatedly switch their visual focus to confirm whether the key indicators are normal. Especially during different stages of the surgery (such as tracheal intubation, skin incision, suture, etc.), some data may fluctuate significantly due to the influence of operations (such as an increase in blood pressure during intubation and changes in heart rate during suture), which requires key monitoring.

[0004] The prior art has the following problems: 1. The fixed display mode requires anesthesiologists to maintain a high degree of concentration for a long time, which is prone to visual fatigue; 2. Anesthesiologists need to rely on experience to manually screen key data and cannot dynamically match the surgical process; 3. The dense data display may delay the rapid identification of abnormal data. Therefore, we propose an anesthesia data monitoring method for operating rooms. Summary of the Invention

[0005] The purpose of the present invention is to provide an anesthesia data monitoring method for operating rooms. By analyzing the surgical operation stage in real time, dynamically adjusting the display interface of the monitor, magnifying and displaying the anesthesia data that needs to be focused on in the current stage, and synchronously prompting the surgical process, the cognitive load of anesthesiologists is reduced, and the intraoperative safety is improved, thus solving the problems raised in the background art.

[0006] To achieve the above purpose, the present invention provides the following technical solution: An anesthesia data monitoring method for operating rooms, including the following steps:

[0007] Step 1: Install a high-definition camera in the surgical area. The camera is aligned with the surgical area, and the data output end of the camera is connected to an AI server. During the surgery, the surgical operation video is collected in real time and transmitted to the AI server;

[0008] Step 2: Denoise and enhance the key areas of the captured video, then use AI to analyze the video to determine the stage of the operation and the surgical operation being performed, and label the ongoing operation with tags (such as tracheal intubation, skin incision, suture).

[0009] Step 3: Based on the surgical operation steps marked by the tags, determine the data that needs to be focused on and monitored, and then extract the associated anesthesia data items from the monitor interface according to the current surgical stage, enlarge the display area of the data items to a preset ratio, reduce or fold the remaining data, and at the same time overlay the name of the current surgical stage, the expected duration, and the operation risk prompt on the sidebar of the interface.

[0010] Step 4: After the AI server determines that the surgical operation in this stage is completed based on video recognition, restore the display screen of the monitor to the normal display state, or enlarge the area that needs to be highlighted in the next surgical stage to a preset ratio.

[0011] As a preferred embodiment of the present invention, the camera in Step 1 needs to be disinfected before and after the operation to maintain a sterile environment in the operating room.

[0012] As a preferred embodiment of the present invention, in Step 1, the high-definition camera is installed on the shadowless lamp or the side of the operating table, and the field of view covers the operation area of the surgeon.

[0013] As a preferred embodiment of the present invention, in Step 1, the high-definition camera is connected to the AI processor to ensure that the video stream transmission delay < 200ms.

[0014] As a preferred embodiment of the present invention, the AI server in Step 2 needs to first learn and improve its functions with a large number of surgical videos, and then adapt to the scenario through transfer learning.

[0015] As a preferred embodiment of the present invention, the magnification ratio of the display area in Step 3 is between 150% - 200%.

[0016] As a preferred embodiment of the present invention, in Step 3, while reducing or folding the remaining data, continue to monitor other data, and when other data is abnormal, issue an alarm and also enlarge the display area of the abnormal data.

[0017] As a preferred embodiment of the present invention, the enlarged display area in Step 3 can also be changed manually by the anesthesiologist.

[0018] As a preferred embodiment of the present invention, the signal indicating the end of the surgical stage operation in Step 4 can also be manually switched by the anesthesiologist.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0020] By using a camera to capture the surgical process in real time and an AI server to analyze the operation stage of the surgery in real time, and then dynamically adjusting the display interface of the monitor according to the data that needs to be focused on in the corresponding surgical stage, magnifying and displaying the anesthesia data that needs to be focused on in the current stage, and synchronously prompting the surgical progress, the cognitive load of anesthesiologists is reduced, and the intraoperative safety is improved. Description of the Drawings

[0021] Other features, objects, and advantages of the present invention will become more apparent by reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0022] Figure 1 It is a flowchart of a method for monitoring anesthesia data in an operating room according to the present invention. Detailed Embodiments

[0023] To make the technical means, creative features, achieved purposes, and functions of the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments.

[0024] Embodiment 1

[0025] Step 1: Install a high-definition camera in the surgical area to ensure that the camera can be aimed at the surgical area. The high-definition camera is preferably installed on the side of the shadowless lamp or the operating table, and the field of view covers the operation area of the surgeon. The camera needs to be disinfected before and after the surgery to maintain the sterile environment of the operating room. The data output end of the camera is connected to the AI server, and the high-definition camera is connected to the AI processor of the AI server to ensure that the video stream transmission delay < 200ms. During the surgery, the surgical operation video is collected in real time and transmitted to the AI server;

[0026] Step 2: Build a surgical stage recognition model based on deep learning. The AI server needs to first learn with a large number of surgical videos to improve its functions. Collect a video data set containing different surgical types (such as laparoscopic surgery, thoracotomy), and the surgical operation stage boundaries are marked by surgeons, and then adapt to the scenario through transfer learning. Input the video stream and output the current surgical stage label;

[0027] Model training: Use the labeled surgical video data set (the labeled content includes operating instruments, action types, and stage time series), extract spatial features through a convolutional neural network (CNN), and combine with a time series model (such as LSTM or Transformer) to capture the operation continuity;

[0028] Real-time inference: Ensure low latency through lightweight model deployment (such as MobileNet + time series distillation).

[0029] Denoise and enhance the key areas of the captured video, then use AI to analyze the video to determine the surgical stage and the ongoing surgical operation, and mark tags for the ongoing operation (such as "skin incision", "hemostasis", "suture", etc.);

[0030] Step 3: Establish a "surgical stage - key anesthesia data" mapping table to store the physiological parameters that need to be monitored key points and their normal threshold ranges at different stages (for example, the "blood pressure" and "airway pressure" are associated with the tracheal intubation stage);

[0031] The mapping table can be initialized through historical data analysis or an expert knowledge base and dynamically optimized with actual usage feedback. Furthermore, based on the surgical operation steps marked with tags, it is possible to determine the data that needs to be focused on and monitored. Then, according to the current surgical stage, extract the associated anesthesia data items from the monitor interface, magnify their display area to a preset ratio (150% - 200%), reduce or fold the remaining data, and at the same time overlay the name of the current surgical stage, the expected duration, and the operation risk prompt on the sidebar of the interface;

[0032] Step 4: After the AI server determines that the surgical operation at this stage has ended based on video recognition, restore the display screen of the monitor to the normal display state, or magnify the area that needs to be highlighted in the next surgical stage to a preset ratio.

[0033] Embodiment 2

[0034] Step 1: Install a high - definition camera in the surgical area to ensure that the camera can be aligned with the surgical area. The high - definition camera is preferably installed on the side of the shadowless lamp or the operating table, and the field of view covers the operation area of the surgeon. The camera needs to be disinfected before and after the operation to maintain the sterile environment of the operating room. The data output end of the camera is connected to the AI server, and the high - definition camera is connected to the AI processor of the AI server to ensure that the video stream transmission delay < 200ms. During the operation, the surgical operation video is collected in real - time and transmitted to the AI server;

[0035] Step 2: Build a surgical stage recognition model based on deep learning. The AI server needs to first use a large number of surgical videos to learn and improve its functions. Collect a video dataset containing different surgical types (such as laparoscopic surgery, thoracotomy), and have surgeons mark the boundaries of the operation stages. Then, adapt to the scenario through transfer learning, input the video stream, and output the current surgical stage label;

[0036] Model training: Use the labeled surgical video dataset (the labeled content includes operating instruments, action types, stage time series), extract spatial features through a convolutional neural network (CNN), and combine with a time series model (such as LSTM or Transformer) to capture the continuity of operations;

[0037] Real-time inference: Ensure low latency through lightweight model deployment (such as MobileNet+temporal distillation).

[0038] De-noise and enhance key areas of the captured video, then use AI to analyze the video to determine the stage of the surgery and the ongoing surgical operation, and label the ongoing operation (such as "skin incision", "hemostasis", "suturing", etc.);

[0039] Step 3: Establish a "surgical stage-key anesthesia data" mapping table to store the physiological parameters that need to be monitored at different stages and their normal threshold ranges (for example, "blood pressure" and "airway pressure" are associated with the tracheal intubation stage);

[0040] The mapping table can be initialized through historical data analysis or expert knowledge base, and dynamically optimized with actual usage feedback. It can then determine the data that needs to be monitored based on the labeled surgical operation steps.

[0041] The difference from the first embodiment is that an additional display screen is provided as a secondary display screen, and according to the current surgical stage, the associated anesthesia data items are extracted from the monitor interface as the main display screen, and its display area is enlarged to a preset ratio, preferably a 200% enlargement ratio, and the remaining data are uniformly transferred to the secondary display screen for display, and at the same time, the current surgical stage name, estimated duration and operation risk prompt are superimposed on the interface sidebar of the main display screen;

[0042] Step 4: After the AI server determines that the surgical operation at this stage is completed based on video recognition, the display of the monitor serving as the main display screen is restored to the normal display state, or the key display area required for the next surgical stage is enlarged to a preset ratio. When the normal display state is restored, the secondary display screen is divided into the screen-off state.

[0043] Embodiment 3

[0044] Step 1: Install a high-definition camera in the operating area and ensure that the camera can be aimed at the operating area. The high-definition camera is preferably installed on the side of the shadowless lamp or the operating table, and the field of view covers the operating area of the surgeon. The camera needs to be disinfected before and after the operation to maintain the sterile environment of the operating room. The data output end of the camera is connected to the AI server. The high-definition camera is connected to the AI processor of the AI server to ensure that the video stream transmission delay is less than 200ms. During the operation, the surgical operation video is collected in real time and transmitted to the AI server;

[0045] Step 2: Build a surgical stage recognition model based on deep learning. The AI server needs to first learn with a large number of surgical videos to improve its functions. Collect a video dataset containing different surgical types (such as laparoscopic surgery, thoracotomy), and have surgeons mark the boundaries of operation stages. Then, adapt to the scenario through transfer learning, input the video stream, and output the current surgical stage label;

[0046] Model training: Use the labeled surgical video dataset (the labeled content includes operating instruments, action types, and stage time series), extract spatial features through a convolutional neural network (CNN), and combine with a time series model (such as LSTM or Transformer) to capture the continuity of operations;

[0047] Real-time inference: Ensure low latency through lightweight model deployment (such as MobileNet + time series distillation).

[0048] Denoise and enhance the key areas of the captured video, and then use AI to analyze the video to determine the surgical stage and the ongoing surgical operation, and mark tags for the ongoing operations (such as "skin incision", "hemostasis", "suture", etc.);

[0049] Step 3: Establish a mapping table of "surgical stage - key anesthesia data", store the physiological parameters that need to be monitored key points and their normal threshold ranges at different stages (for example: the tracheal intubation stage is associated with "blood pressure", "airway pressure");

[0050] The mapping table can be initialized through historical data analysis or an expert knowledge base, and dynamically optimized with actual usage feedback. Furthermore, based on the surgical operation steps marked with tags, it is possible to determine the data that needs to be focused on and monitored.

[0051] The difference from Embodiment 1 is that an additional display screen is set as a secondary display screen. According to the current surgical stage, extract the associated anesthesia data items from the monitor interface serving as the main display screen, and magnify their display areas to a preset ratio, preferably a magnification ratio of 200%. The remaining data is uniformly transferred to the secondary display screen for display. At the same time, the name of the current surgical stage, the expected duration, and operation risk prompts are superimposed on the sidebar of the main display screen interface. The difference from Embodiment 2 is that while switching the remaining data to the secondary display screen for display, continue to monitor other data, and when these data are abnormal, issue an alarm and also switch the corresponding data to the monitor interface serving as the main display screen for key display;

[0052] Step 4: After the AI server determines that the surgical operation in this stage has ended based on video recognition, restore the display of the monitor serving as the main display screen to the normal display state, or magnify the key display area required for the next surgical stage to a preset ratio. When restoring to the normal display state, the secondary display screen goes into the screen-off state.

[0053] Example 4

[0054] Step 1: Install a high-definition camera in the surgical area to ensure that the camera can be aligned with the surgical area. The high-definition camera is preferably installed on the shadowless lamp or the side of the operating table, and the field of view covers the operation area of the surgeon. The camera needs to be disinfected before and after the operation to maintain the aseptic environment of the operating room. The data output end of the camera is connected to the AI server, and the high-definition camera is connected to the AI processor of the AI server to ensure that the video stream transmission delay < 200ms. During the operation, the surgical operation video is collected in real time and transmitted to the AI server;

[0055] Step 2: Build a surgical stage recognition model based on deep learning. The AI server needs to first use a large number of surgical videos to learn and improve its functions. Collect a video dataset containing different surgical types (such as laparoscopic surgery, thoracotomy), and have surgeons mark the boundaries of the operation stages. Then, adapt to the scenario through transfer learning, input the video stream and output the current surgical stage label;

[0056] Model training: Use the labeled surgical video dataset (the labeled content includes operating instruments, action types, stage time series), extract spatial features through a convolutional neural network (CNN), and combine with a time series model (such as LSTM or Transformer) to capture the continuity of operations;

[0057] Real-time inference: Ensure low latency through lightweight model deployment (such as MobileNet + time series distillation).

[0058] Denoise and enhance the key areas of the captured video, and then use AI to analyze the video to determine the surgical stage and the ongoing surgical operation, and mark the ongoing operation with labels (such as "skin incision", "hemostasis", "suture", etc.);

[0059] Step 3: Establish a "surgical stage - key anesthesia data" mapping table, and store the physiological parameters that need to be monitored key points and their normal threshold ranges at different stages (for example: the tracheal intubation stage is associated with "blood pressure", "airway pressure");

[0060] The mapping table can be initialized through historical data analysis or expert knowledge base, and dynamically optimized with the feedback of actual use. Furthermore, it can judge the data that needs to be focused on according to the surgical operation steps marked by the labels.

[0061] The difference from the first embodiment is that an additional display screen is set as a secondary display screen. According to the current surgical stage, relevant anesthetic data items are extracted from the monitor interface serving as the main display screen, and the display area thereof is enlarged to a preset ratio, preferably a magnification ratio of 200%. The remaining data is uniformly transferred to the secondary display screen for display. At the same time, the name of the current surgical stage, the estimated duration, and the operation risk prompt are superimposed on the sidebar of the interface of the main display screen. The difference from the second embodiment is that while switching the remaining data to the secondary display screen for display, the monitoring of other data continues. When these data are abnormal, an alarm is issued and the corresponding data is also switched to the monitor interface serving as the main display screen for key display;

[0062] Step 4: After the AI server determines that the surgical operation of this stage is completed based on video recognition, the display of the monitor serving as the main display screen is restored to the normal display state, or the key display area required for the next surgical stage is enlarged to a preset ratio. When restoring to the normal display state, the secondary display screen enters the screen-off state.

[0063] The difference from the third embodiment is that: the enlarged display area in the third embodiment can also be changed manually by the anesthesiologist to ensure the controllability when the anesthesiologist needs to pay attention to the corresponding data according to experience. Additionally, the signal indicating the end of the surgical stage operation in Step 4 can also be switched manually by the anesthesiologist to achieve the purpose of human-machine collaborative work with higher reliability.

[0064] Furthermore, all data monitored during the surgical process needs to be recorded in the database. These data are of great significance for evaluating the anesthetic effect, the patient's condition, and the formulation of subsequent treatment plans. The preservation of anesthetic data helps to trace the medical quality and can be used as important materials for scientific research and teaching.

[0065] The above shows and describes the basic principles, main features, and advantages of the present invention. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic features of the present invention, the present invention can be implemented in other specific forms. Therefore, in any regard, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be encompassed within the present invention. Any reference signs in the claims should not be regarded as limiting the claimed rights.

[0066] In addition, it should be understood that although this specification is described in terms of embodiments, not every embodiment contains only one independent technical solution. This narrative style of the specification is only for clarity. Those skilled in the art should regard the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. An operating room anesthesia data monitoring method, characterized in that It includes the following steps: Step 1: Install a high-definition camera in the surgical area. The camera is aimed at the surgical area, and the data output end of the camera is connected to the AI server. During the operation, the surgical operation video is collected in real time and transmitted to the AI server; Step 2: Denoise and enhance the key areas of the captured video, and then use AI to analyze the video to determine the stage of the operation and the ongoing surgical operation, and mark the ongoing operation with labels (such as tracheal intubation, skin incision, suture); Step 3: According to the surgical operation steps marked by the labels, determine the data that needs to be focused on and monitored. Then, according to the current surgical stage, extract the associated anesthesia data items from the monitor interface, enlarge their display area to a preset ratio, reduce or fold the remaining data, and at the same time superimpose the name of the current surgical stage, the expected duration, and the operation risk prompt on the sidebar of the interface; Step 4: After the AI server determines that the surgical operation at this stage is over based on the video recognition, restore the display screen of the monitor to the normal display state, or enlarge the area that needs to be focused on in the next surgical stage to a preset ratio.

2. The operating room anesthesia data monitoring method according to claim 1, wherein: The camera in Step 1 needs to be disinfected before and after the operation to maintain a sterile environment in the operating room.

3. The method for monitoring anesthesia data in an operating room according to claim 1, wherein: In Step 1, the high-definition camera is installed on the side of the shadowless lamp or the operating table, and the field of view covers the operation area of the surgeon.

4. A method for monitoring anesthesia data in an operating room according to claim 1, characterized in that: In Step 1, the high-definition camera is connected to the AI processor to ensure that the video stream transmission delay < 200ms.

5. The method for monitoring anesthesia data in an operating room according to claim 1, characterized in that: The AI server in Step 2 needs to first learn and improve its functions with a large number of surgical videos, and then adapt to the scenario through transfer learning.

6. The method for monitoring anesthesia data in an operating room according to claim 1, characterized in that: The magnification ratio of the display area in Step 3 is between 150% - 200%.

7. The surgical anesthesia data monitoring method according to claim 1, characterized in that: In Step 3, while reducing or folding the remaining data, continue to monitor other data, and when other data is abnormal, send an alarm and also enlarge the display area of the abnormal data.

8. A method for monitoring anesthesia data in an operating room according to claim 1, characterized in that: The enlarged display area in Step 3 can also be changed manually by the anesthesiologist.

9. The method for monitoring anesthesia data in an operating room according to claim 1, characterized in that: The signal indicating the end of the surgical stage operation in Step 4 can also be switched manually by the anesthesiologist.

Images