Oral anesthesia depth monitoring method based on intelligent perception

Abstract

The present invention relates to the field of oral diagnosis technology, and in particular to an oral anesthesia depth monitoring method based on intelligent perception, comprising: obtaining facial reference points of each frame of face recognition image, recording the facial reference points of the first frame as initial facial reference points; obtaining optical flow facial reference points, and calculating the offset of the facial reference points of each frame of face recognition image according to the Euclidean distance between each optical flow facial reference point and the initial facial reference point; obtaining a number of minimum enclosing circles of each frame of face recognition image; calculating the change directionality of the facial reference points of each frame of face recognition image; obtaining a fitting curve of the offset; calculating the change characteristics of the facial reference points of each frame of face recognition image, and then obtaining the facial expression relief of each frame of face recognition image. According to the facial expression relief, the oral anesthesia depth is finally obtained. The present invention improves the accuracy of intelligent perception oral anesthesia depth monitoring and evaluation by analyzing the facial reference points of face recognition images.

Description

Technical Field

[0001] The present invention relates to the technical field of oral diagnosis, and in particular to an oral anesthesia depth monitoring method based on intelligent perception. Background Art

[0002] The oral anesthesia depth monitoring method based on intelligent perception captures the patient's facial expressions in real time by installing devices such as cameras or sensors, and uses computer vision technology to analyze and identify these expressions to assess the patient's pain and comfort. This can help doctors better control the use of anesthetics, ensure that patients do not feel pain during oral surgery, and avoid side effects caused by excessive anesthesia.

[0003] Analyzing the patient's oral anesthesia depth through facial expression changes is a non-invasive assessment method that can help doctors understand the patient's anesthesia status. During oral surgery, anesthetics relax the patient's facial muscles, thereby reducing pain. However, if the dose of anesthetics is insufficient or too strong, the patient may show a painful expression or a change in consciousness. By observing the changes in the patient's facial expression, it is possible to preliminarily judge whether the patient's anesthesia depth is appropriate. However, continuously opening the mouth may cause facial muscle fatigue and twitching, which will also produce changes in expression, which may interfere with the assessment of intelligent perception of oral anesthesia depth monitoring. Summary of the invention

[0004] The present invention provides an oral anesthesia depth monitoring method based on intelligent perception to solve the existing problems.

[0005] The oral anesthesia depth monitoring method based on intelligent perception of the present invention adopts the following technical solutions:

[0006] An embodiment of the present invention provides a method for monitoring oral anesthesia depth based on intelligent perception, the method comprising the following steps:

[0007] Obtain a face recognition image sequence; in the face recognition image sequence, obtain a number of facial reference points of each frame of the face recognition image;

[0008] Record all facial reference points of the face recognition image of the first frame as initial facial reference points; obtain the optical flow facial reference points in each frame of the face recognition image corresponding to each initial facial reference point, and obtain the offset of the facial reference points in each frame of the face recognition image according to the distance between each initial facial reference point and the optical flow facial reference point corresponding to each frame of the face recognition image;

[0009] Obtaining a number of minimum enclosing circles of each frame of face recognition image; obtaining the change directionality of the facial reference point of each frame of face recognition image according to the difference between the center and area of ​​each minimum enclosing circle of each frame of face recognition image and adjacent face recognition images;

[0010] Performing curve fitting on the offsets of the facial reference points of all frames of face recognition images to obtain an offset fitting curve; obtaining the change characteristics of the facial reference points of each frame of face recognition images according to the change directionality and offset of the facial reference points of each frame of face recognition images and the slope of the data points of the offset of the facial reference points of each frame of face recognition images in the offset fitting curve;

[0011] According to the difference between the change characteristics of the facial reference points of each frame of face recognition image and the adjacent frames of face recognition image, the facial expression relaxation degree of each frame of face recognition image is obtained; according to the facial expression relaxation degree of each frame of face recognition image, the oral anesthesia depth of each frame of face recognition image is obtained.

[0012] Furthermore, the step of obtaining the offset of the facial reference point of each frame of face recognition image according to the distance between each initial facial reference point and the corresponding optical flow facial reference point in each frame of face recognition image includes the following specific steps:

[0013] A rectangular coordinate system is constructed with the vertex at the lower left corner of each frame of face recognition image as the origin, horizontally to the right as the positive direction of the horizontal axis, and vertically upward as the positive direction of the vertical axis, to obtain the rectangular coordinate system of each frame of face recognition image;

[0014] In a rectangular coordinate system, the coordinates of each initial facial reference point and each corresponding optical flow facial reference point in each frame of face recognition image are calculated using the Euclidean distance formula to obtain the Euclidean distance between each initial facial reference point and each corresponding optical flow facial reference point in each frame of face recognition image;

[0015] According to each initial facial reference point in The Euclidean distance between the optical flow facial reference points of the frame face recognition image and the corresponding initial facial reference points is obtained. The offset of the facial reference points of the face recognition image.

[0016] Further, the method according to each initial facial reference point is The Euclidean distance between the optical flow facial reference points of the frame face recognition image and the corresponding initial facial reference points is obtained. The specific calculation formula corresponding to the deviation of the facial reference points of the frame face recognition image is:

[0017] ;

[0018] in, Indicates The deviation of the facial reference points of the frame face recognition image; Indicates the number of initial facial reference points; Indicates The initial facial reference points are The optical flow facial reference points of the frame face recognition image and the The Euclidean distance of the initial facial reference points.

[0019] Furthermore, the step of obtaining the change directionality of the facial reference points of each face recognition image frame according to the difference between the center and area of ​​each minimum enclosing circle of each face recognition image frame and the adjacent face recognition images includes the following specific steps:

[0020] According to Frame face recognition image and adjacent face recognition images, get the first The neighborhood range of the frame face recognition image;

[0021] In the In the neighborhood of the frame face recognition image, the optical flow method is used to match the minimum enclosing circles in all frame face recognition images to obtain several minimum enclosing circle sequences;

[0022] According to The difference between the centers of all the minimum enclosing circles in each minimum enclosing circle sequence of the frame face recognition image is obtained. Several target sequences of frame face recognition images;

[0023] According to The number of target sequences corresponding to the frame face recognition image and the The difference between the areas of adjacent minimum enclosing circles in each minimum enclosing circle sequence corresponding to the frame face recognition image is obtained. The directionality of changes in facial reference points of frame face recognition images.

[0024] Furthermore, according to Frame face recognition image and adjacent face recognition images, get the first The neighborhood range of the frame face recognition image includes the following specific steps:

[0025] The first Frame face recognition image and its temporal sequence before and after The face recognition image of the frame is recorded as The neighborhood range of the frame face recognition image; among them, is the preset range threshold.

[0026] Furthermore, according to The difference between the centers of all the minimum enclosing circles in each minimum enclosing circle sequence of the frame face recognition image is obtained. Several target sequences of frame face recognition images include the following specific steps:

[0027] Jordi If the center coordinates of all the minimum enclosing circles in the minimum enclosing circle sequence are different, then The minimum enclosing circle sequence is recorded as the target sequence.

[0028] Furthermore, according to The number of target sequences corresponding to the frame face recognition image and the The difference between the areas of adjacent minimum enclosing circles in each minimum enclosing circle sequence corresponding to the frame face recognition image is obtained. The specific calculation formula corresponding to the change direction of the facial reference points of the frame face recognition image is:

[0029] ;

[0030] in, Indicates Directionality of changes in facial reference points of frame face recognition images; Indicates The number of target sequences corresponding to the frame face recognition images; Indicates The number of bounding circles of the frame face recognition image; Indicates The minimum number of enclosing circles in each minimum enclosing circle sequence corresponding to the frame face recognition image; Indicates The frame face recognition image corresponds to the The smallest enclosing circle sequence The area of ​​the smallest enclosing circle; Indicates The frame face recognition image corresponds to the The smallest enclosing circle sequence The area of ​​the smallest enclosing circle; for Belong to the interval All internal time The variance of .

[0031] Furthermore, according to the change directionality and offset of the facial reference points of each frame of face recognition image, and the slope of the data point of the offset of the facial reference points of each frame of face recognition image in the offset fitting curve, the specific calculation formula corresponding to the change characteristics of the facial reference points of each frame of face recognition image is obtained as follows:

[0032] ;

[0033] in, Indicates The changing characteristics of facial reference points of frame face recognition images; Indicates The slope of the data point of the deviation degree of the facial reference point of the frame face recognition image in the deviation degree fitting curve; Indicates Directionality of changes in facial reference points of frame face recognition images; Indicates The deviation of the facial reference points of the frame face recognition image; represents the normalization function; Represents an exponential function with a natural constant as its base.

[0034] Furthermore, the specific calculation formula corresponding to the facial expression relaxation degree of each frame of face recognition image is obtained based on the difference between the change characteristics of the facial reference points of each frame of face recognition image and the adjacent frame of face recognition image:

[0035] ;

[0036] in, Indicates The degree of facial expression relaxation of the frame face recognition image; Indicates The number of face recognition images within the neighborhood of the frame face recognition image; represents an exponential function with a natural constant as base; Indicated in The first frame of the face recognition image within the neighborhood Changing features of frame face recognition images.

[0037] Furthermore, the oral anesthesia depth of each face recognition image is obtained according to the facial expression relaxation of each face recognition image, and the specific steps include the following:

[0038] When the facial expression relaxation degree of the last frame of face recognition image is less than the preset empirical threshold, it indicates that the currently monitored oral anesthesia depth is insufficient.

[0039] The beneficial effects of the technical solution of the present invention are:

[0040] Obtain a sequence of face recognition images; obtain several facial reference points of each frame of face recognition images in the sequence of face recognition images; record all facial reference points of the first frame of face recognition images as initial facial reference points; obtain the optical flow facial reference points corresponding to each initial facial reference point in each frame of face recognition images, and obtain the offset of the facial reference points of each frame of face recognition images based on the distance between each initial facial reference point and the optical flow facial reference points corresponding to each frame of face recognition images, which accurately reflects the changes in the patient's facial expressions. Obtain several minimum enclosing circles of each frame of face recognition images; obtain the change direction of the facial reference points of each frame of face recognition images based on the difference between the center and area of ​​each minimum enclosing circle of each frame of face recognition images and adjacent face recognition images, thereby reducing the influence of facial muscle fatigue and twitching caused by continuous opening of the mouth on the results of oral anesthesia depth detection. Curve fitting is performed on the offset of the facial reference points of all frames of face recognition images to obtain an offset fitting curve; based on the change directionality and offset of the facial reference points of each frame of face recognition image, and the slope of the data point of the offset of the facial reference points of each frame of face recognition image in the offset fitting curve, the change characteristics of the facial reference points of each frame of face recognition image are obtained, thereby improving the accuracy of oral anesthesia depth detection. Based on the difference between the change characteristics of the facial reference points of each frame of face recognition image and the adjacent frames of face recognition image, the facial expression relaxation of each frame of face recognition image is obtained, thereby further improving the accuracy of oral anesthesia depth detection. Based on the facial expression relaxation of each frame of face recognition image, the oral anesthesia depth of each frame of face recognition image is obtained. The present invention improves the accuracy of intelligent perception oral anesthesia depth monitoring and evaluation by analyzing the facial reference points of face recognition images. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 The present invention is a flowchart of the steps of the oral anesthesia depth monitoring method based on intelligent perception. DETAILED DESCRIPTION

[0043] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the oral anesthesia depth monitoring method based on intelligent perception proposed by the present invention, its specific implementation method, structure, characteristics and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0044] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0045] The specific scheme of the oral anesthesia depth monitoring method based on intelligent perception provided by the present invention is described in detail below with reference to the accompanying drawings.

[0046] See also Figure 1 , which shows a flowchart of a method for monitoring the depth of oral anesthesia based on intelligent perception according to an embodiment of the present invention, the method comprising the following steps:

[0047] Step S001: Acquire a face recognition image sequence; in the face recognition image sequence, obtain a number of facial reference points of each frame of the face recognition image.

[0048] A camera installed just above the patient's face is used to record the patient's facial expressions during oral treatment in real time. The face recognition image of each frame in the recorded video is intercepted to obtain a face recognition image sequence consisting of continuous frames. Facial recognition technology is used to detect the facial position in each frame of the face recognition image, and align it to eliminate the influence of posture changes, and remove data that does not meet the quality requirements, such as blurred frames, frames with too large a proportion of non-face areas, etc. Among them, facial recognition technology detection is a well-known technology, and the specific method is not introduced here.

[0049] In the face recognition image sequence, the AAM algorithm is used to operate each frame of the face recognition image to obtain a number of facial reference points of each frame of the face recognition image. The AAM algorithm is a well-known technology, and the specific method is not introduced here.

[0050] Step S002: record all facial reference points of the face recognition image of the first frame as initial facial reference points; obtain the optical flow facial reference point corresponding to each initial facial reference point in each frame of the face recognition image, and obtain the offset of the facial reference point of each frame of the face recognition image according to the distance between each initial facial reference point and the optical flow facial reference point corresponding to each frame of the face recognition image.

[0051] It should be noted that the effective time of oral anesthetic drugs varies depending on the type of drug and dosage, but usually starts to work within a few minutes after injection. During oral surgery, the doctor will select the appropriate anesthetic drug and dosage based on the patient's specific situation and surgical needs to ensure the safety and comfort of the surgical process. At the beginning of the operation, the doctor has judged that the anesthesia effect is appropriate, so the position of the facial reference point in the face recognition image at the beginning of the operation is used as a reference position to evaluate the depth of oral anesthesia during subsequent operations.

[0052] It is further necessary to explain that as the operation progresses, the anesthetic effect may gradually weaken and the patient's expression may also change. For example, when the anesthetic effect weakens, the patient may frown, tighten his expression, twist his face, and other expression changes. These changes are usually caused by the patient trying to relieve pain or discomfort. Therefore, during oral surgery, it is necessary to closely observe the changes in the patient's facial expression. The position changes of the patient's facial reference points in each frame of the image are used as the basis for judging the changes in facial expression. It is necessary to first obtain the offset of the patient's facial reference points.

[0053] In the face recognition image sequence, all facial reference points of the first frame of the face recognition image are recorded as initial facial reference points.

[0054] A rectangular coordinate system is constructed with the vertex at the lower left corner of each face recognition image frame as the origin, horizontally to the right as the positive direction of the horizontal axis, and vertically upward as the positive direction of the vertical axis, to obtain the rectangular coordinate system of each face recognition image frame.

[0055] The optical flow method is used to calculate each initial facial reference point in the face recognition image sequence of continuous frames to obtain the optical flow facial reference point corresponding to each initial facial reference point in each frame of face recognition image. The optical flow method is a well-known technology and the specific method is not introduced here.

[0056] In a rectangular coordinate system, the coordinates of each initial facial reference point and each corresponding optical flow facial reference point in each frame of face recognition image are calculated using the Euclidean distance formula to obtain the Euclidean distance between each initial facial reference point and each corresponding optical flow facial reference point in each frame of face recognition image. The Euclidean distance formula is a well-known technology, and the specific method is not introduced here.

[0057] First Take the face recognition image as an example, then The specific calculation formula corresponding to the deviation of the facial reference points of the frame face recognition image is:

[0058] ;

[0059] in, Indicates The deviation of the facial reference points of the frame face recognition image; Indicates the number of initial facial reference points; Indicates The initial facial reference points are The optical flow facial reference points of the frame face recognition image and the The Euclidean distance of the initial facial reference points.

[0060] What needs to be explained is: Indicates The degree of change in the patient's expression in the frame face recognition image, The larger the value of , the greater the deviation of the facial reference points, which means that the patient's expression changes more.

[0061] Through the above process, we get The offset of the facial reference points of the face recognition image.

[0062] Each frame of face recognition image is operated according to the above process to obtain the offset of the facial reference point of each frame of face recognition image.

[0063] It should be noted that the above calculation is not performed on the face recognition image of the first frame.

[0064] Step S003: obtaining several minimum enclosing circles of each face recognition image frame; obtaining the change directionality of the facial reference points of each face recognition image frame according to the difference between the center and area of ​​each minimum enclosing circle of each face recognition image frame and adjacent face recognition images.

[0065] It should be noted that when the anesthesia effect weakens as the operation progresses, the change in the position of the facial reference points is manifested as concentrated tightening and distortion of the facial muscles. The general public's reaction to pain is frowning, while the muscle spasm caused by the continuous opening of the mouth is a local twitch. The change of the facial reference points in the local area is roughly parallel movement and reset. Therefore, the changes in facial expressions due to different reasons are also different, and it is necessary to analyze the directionality of the movement of the facial reference points to eliminate interference.

[0066] K-means clustering is used to calculate the positions of all facial reference points in each frame of face recognition image, and facial reference points with close distances in each frame of face recognition image are clustered into one category, thereby obtaining several clusters of each frame of face recognition image. K-means clustering is a well-known technology, and the specific method is not introduced here.

[0067] It should be noted that: in this embodiment, the number of clusters in K-means clustering is set to 10, and this is taken as an example for description.

[0068] In each frame of face recognition image, the minimum enclosing circle of each cluster is drawn to obtain several minimum enclosing circles of each frame of face recognition image.

[0069] It should be noted that the minimum enclosing circle is the smallest circle that contains all facial reference points in the cluster. The minimum enclosing circle is a well-known technology, and the specific method is not introduced here.

[0070] Still Frame face recognition image as an example, Frame face recognition image and its temporal sequence before and after The face recognition image of the frame is recorded as The neighborhood range of the frame face recognition image. Among them, is a preset range threshold. The preset range threshold in this embodiment is The value is 2, and this example is used for description.

[0071] Through the above process, the neighborhood range of each frame of face recognition image is obtained.

[0072] In the In the neighborhood of the frame face recognition image, the optical flow method is used to match the minimum enclosing circles in all the frame face recognition images to obtain a number of minimum enclosing circle sequences. The optical flow method is a well-known technology and the specific method is not introduced here.

[0073] It should be noted that all the minimum enclosing circles in each minimum enclosing circle sequence correspond to different frames of face recognition images.

[0074] Jordi If the center coordinates of all the minimum enclosing circles in the minimum enclosing circle sequence are different, then The minimum enclosing circle sequence is recorded as the target sequence.

[0075] The first The specific calculation formula corresponding to the change direction of the facial reference points of the frame face recognition image is:

[0076] ;

[0077] in, Indicates Directionality of changes in facial reference points of frame face recognition images; Indicates The number of target sequences corresponding to the frame face recognition images; Indicates The number of bounding circles of the frame face recognition image; Indicates The minimum number of enclosing circles in each minimum enclosing circle sequence corresponding to the frame face recognition image; Indicates The frame face recognition image corresponds to the The smallest enclosing circle sequence The area of ​​the smallest enclosing circle; Indicates The frame face recognition image corresponds to the The smallest enclosing circle sequence The area of ​​the smallest enclosing circle; for Belong to the interval All internal time The variance of .

[0078] What needs to be explained is: The larger the value of , the more the entire expression is changing. The facial reference points are moving globally, and the greater the direction of change of the facial reference points. The larger the value is, the more changes are seen in all parts of the face. The changes in facial expression caused by the pain effect caused by the weakening of the anesthesia effect act on the entire face. The greater the change in the concentration of the facial reference points, the greater the direction of the change in the facial reference points.

[0079] Through the above process, we get The directionality of changes in facial reference points of frame face recognition images.

[0080] Each frame of face recognition image is operated according to the above process to obtain the change direction of the facial reference points of each frame of face recognition image.

[0081] Step S004: curve fitting is performed on the offsets of the facial reference points of all frames of face recognition images to obtain an offset fitting curve; based on the change directionality and offset of the facial reference points of each frame of face recognition images and the slope of the data points of the offsets of the facial reference points of each frame of face recognition images in the offset fitting curve, the change characteristics of the facial reference points of each frame of face recognition images are obtained.

[0082] It should be noted that the change of facial expression is a continuous process. As the effect of anesthesia weakens, slight pain will cause changes in facial expression. However, the mouth remains open during the operation, which will cause spasms in the surrounding muscles and cause twitching changes. The position of the facial reference points will also change. This change has nothing to do with the effect of anesthesia and needs to be excluded. Analyze the difference between the changes in facial expression caused by interference factors and anesthesia. The pain caused by the weakening of the anesthesia effect will cause overall changes in facial expression such as frowning, tightening, and facial distortion, while twitching is caused by muscle spasms in local areas. The range and degree of change of the two are different. The time of twitching is fast and short, while the expression changes caused by pain are relatively long. At the same time, by analyzing the offset of the facial reference points, the type and intensity of facial expression can be determined. The greater the offset of the facial reference points, the greater the change in expression. Therefore, by analyzing the changes in facial expression under different influences, the change characteristics of facial reference points are obtained.

[0083] In the face recognition image sequence, the least square method is used to perform curve fitting on the offset of the facial reference points of all frames of face recognition images to obtain the offset fitting curve. The least square method is a well-known technology and the specific method is not introduced here.

[0084] Still Take the face recognition image as an example, then The specific calculation formula corresponding to the change characteristics of the facial reference points of the frame face recognition image is:

[0085] ;

[0086] in, Indicates The changing characteristics of facial reference points of frame face recognition images; Indicates The slope of the data point of the deviation degree of the facial reference point of the frame face recognition image in the deviation degree fitting curve; Indicates Directionality of changes in facial reference points of frame face recognition images; Indicates The deviation of the facial reference points of the frame face recognition image; represents the normalization function; Represents an exponential function with a natural constant as its base.

[0087] What needs to be explained is: The smaller the value of The slower the change speed of the facial reference point offset of the frame face recognition image, the greater the expression change, which means that the change is more likely to be caused by the weakening of anesthesia, and the greater the change characteristics; The larger the value of The greater the variation characteristics of the facial reference points of the frame face recognition image; The larger the value of is, the greater the change in the expression in the frame image is. The greater the variation characteristics of the facial reference points of the frame face recognition image.

[0088] Through the above process, we get The changing characteristics of facial reference points in frame face recognition images.

[0089] Each frame of face recognition image is operated according to the above process to obtain the change characteristics of the facial reference points of each frame of face recognition image.

[0090] Step S005: Obtain the degree of relief of facial expression of each face recognition image frame according to the difference between the change characteristics of facial reference points of each face recognition image frame and adjacent face recognition image frames; obtain the depth of oral anesthesia of each face recognition image frame according to the degree of relief of facial expression of each face recognition image frame.

[0091] It should be noted that oral anesthetics can reduce the patient's pain and discomfort, making the facial expression more relaxed and comfortable. However, as the anesthetic effect weakens, the patient may begin to feel some tingling or discomfort, which may cause changes in facial expression, such as frowning, facial distortion, etc. These changes are usually caused by the patient's attempt to relieve pain or discomfort. Therefore, when monitoring the depth of oral anesthesia, it is necessary to obtain the patient's facial expression relief based on the patient's facial expression change characteristics. The smaller the change characteristics of the facial reference points, the greater the patient's facial expression relief. At the same time, the expression change is a continuous process, so it is also necessary to consider the change characteristics of the facial reference points of the face recognition image within the neighborhood range, so as to obtain the facial expression relief of the face recognition image.

[0092] Still Take the face recognition image as an example, then The specific calculation formula corresponding to the facial expression relaxation degree of the frame face recognition image is:

[0093] ;

[0094] in, Indicates The degree of facial expression relaxation of the frame face recognition image; Indicates The number of face recognition images within the neighborhood of the frame face recognition image; represents an exponential function with a natural constant as base; Indicated in The first frame of the face recognition image within the neighborhood Changing features of frame face recognition images.

[0095] What needs to be explained is: The larger the value, the greater the expression change. The smaller the facial expression relaxation of the frame face recognition image; Indicates The average value of the change characteristics of the overall surface reference points of the frame face recognition image within the neighborhood range. The expression is changing continuously. When the change characteristics of the facial reference points within the neighborhood range are large, then The smaller the facial expression relaxation of the frame face recognition image.

[0096] Through the above process, we get Facial expression relaxation of frame face recognition images.

[0097] Each frame of face recognition image is operated according to the above process to obtain the facial expression relaxation degree of each frame of face recognition image.

[0098] When the facial expression relaxation degree of the last frame of face recognition image is less than the preset empirical threshold, it indicates that the currently monitored oral anesthesia depth is insufficient; the preset empirical threshold in this embodiment is 0.8, which is used as an example for description.

[0099] Through the above process, the depth of oral anesthesia of each frame of face recognition image is determined.

[0100] It should be noted that when the currently monitored oral anesthesia depth is insufficient, the system will immediately notify the doctor, who can then adjust the anesthesia dose or take other relief measures based on this feedback, such as adding local anesthetics or using sedatives. This method can improve patient comfort and avoid accidental awakening or confusion caused by improper anesthesia, thereby improving surgical safety and treatment effectiveness. In addition, this monitoring method reduces the error of subjective judgment, allowing doctors to control the depth of anesthesia more accurately, ensuring that patients remain painless and maximally comfortable during oral treatment.

[0101] So far, the present invention is completed.

[0102] In summary, in an embodiment of the present invention, facial reference points of each frame of face recognition image are obtained, and the facial reference points of the first frame are recorded as initial facial reference points; optical flow facial reference points are obtained, and the offset of the facial reference points of each frame of face recognition image is calculated according to the Euclidean distance between each optical flow facial reference point and the initial facial reference point; a number of minimum enclosing circles of each frame of face recognition image are obtained; the directionality of change of the facial reference points of each frame of face recognition image is calculated; a fitting curve of the offset degree is obtained; the change characteristics of the facial reference points of each frame of face recognition image are calculated, and then the facial expression relaxation of each frame of face recognition image is obtained. According to the facial expression relaxation, the depth of oral anesthesia is finally obtained. The present invention improves the accuracy of intelligent perception of oral anesthesia depth monitoring and evaluation by analyzing the facial reference points of face recognition images.

[0103] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for monitoring the depth of oral anesthesia based on intelligent perception, characterized in that: The method comprises the following steps: Obtain a face recognition image sequence; in the face recognition image sequence, obtain a number of facial reference points of each frame of the face recognition image; Record all facial reference points of the face recognition image of the first frame as initial facial reference points; obtain the optical flow facial reference points in each frame of the face recognition image corresponding to each initial facial reference point, and obtain the offset of the facial reference points in each frame of the face recognition image according to the distance between each initial facial reference point and the optical flow facial reference point corresponding to each frame of the face recognition image; Obtaining a number of minimum enclosing circles of each frame of face recognition image; obtaining the change directionality of the facial reference point of each frame of face recognition image according to the difference between the center and area of ​​each minimum enclosing circle of each frame of face recognition image and adjacent face recognition images; Performing curve fitting on the offsets of the facial reference points of all frames of face recognition images to obtain an offset fitting curve; obtaining the change characteristics of the facial reference points of each frame of face recognition images according to the change directionality and offset of the facial reference points of each frame of face recognition images and the slope of the data points of the offset of the facial reference points of each frame of face recognition images in the offset fitting curve; According to the difference between the change characteristics of the facial reference points of each frame of face recognition image and the adjacent frames of face recognition image, the facial expression relaxation degree of each frame of face recognition image is obtained; according to the facial expression relaxation degree of each frame of face recognition image, the oral anesthesia depth of each frame of face recognition image is obtained; According to the change directionality and offset of the facial reference points of each frame of face recognition image, and the slope of the data point of the offset of the facial reference points of each frame of face recognition image in the offset fitting curve, the specific calculation formula corresponding to the change characteristics of the facial reference points of each frame of face recognition image is obtained as follows: ； in, Indicates The changing characteristics of facial reference points of frame face recognition images; Indicates The slope of the data point of the deviation degree of the facial reference point of the frame face recognition image in the deviation degree fitting curve; Indicates Directionality of changes in facial reference points of frame face recognition images; Indicates The deviation of the facial reference points of the frame face recognition image; represents the normalization function; represents an exponential function with a natural constant as base; The specific calculation formula corresponding to the facial expression relaxation degree of each face recognition image frame is obtained based on the difference between the change characteristics of the facial reference points of each face recognition image frame and the adjacent face recognition image frames: ； in, Indicates The degree of facial expression relaxation of the frame face recognition image; Indicates The number of face recognition images within the neighborhood of the frame face recognition image; Indicated in The first frame of the face recognition image within the neighborhood Changing features of frame face recognition images.

2. The method for monitoring the depth of oral anesthesia based on intelligent perception according to claim 1, characterized in that: The step of obtaining the offset of the facial reference point of each frame of face recognition image according to the distance between each initial facial reference point and the corresponding optical flow facial reference point in each frame of face recognition image comprises the following specific steps: A rectangular coordinate system is constructed with the vertex at the lower left corner of each frame of face recognition image as the origin, horizontally to the right as the positive direction of the horizontal axis, and vertically upward as the positive direction of the vertical axis, to obtain the rectangular coordinate system of each frame of face recognition image; In a rectangular coordinate system, the coordinates of each initial facial reference point and each corresponding optical flow facial reference point in each frame of face recognition image are calculated using the Euclidean distance formula to obtain the Euclidean distance between each initial facial reference point and each corresponding optical flow facial reference point in each frame of face recognition image; According to each initial facial reference point in The Euclidean distance between the optical flow facial reference points of the frame face recognition image and the corresponding initial facial reference points is obtained. The offset of the facial reference points of the face recognition image.

3. The method for monitoring the depth of oral anesthesia based on intelligent perception according to claim 2, characterized in that: According to each initial facial reference point The Euclidean distance between the optical flow facial reference points of the frame face recognition image and the corresponding initial facial reference points is obtained. The specific calculation formula corresponding to the deviation of the facial reference points of the frame face recognition image is: ； in, Indicates The deviation of the facial reference points of the frame face recognition image; Indicates the number of initial facial reference points; Indicates The initial facial reference points are The optical flow facial reference points of the frame face recognition image and the The Euclidean distance of the initial facial reference points.

4. The method for monitoring the depth of oral anesthesia based on intelligent perception according to claim 1, characterized in that: The step of obtaining the change directionality of the facial reference points of each frame of face recognition image according to the difference between the center and area of ​​each minimum enclosing circle of each frame of face recognition image and adjacent face recognition images includes the following specific steps: According to Frame face recognition image and adjacent face recognition images, get the first The neighborhood range of the frame face recognition image; In the In the neighborhood of the frame face recognition image, the optical flow method is used to match the minimum enclosing circles in all frame face recognition images to obtain several minimum enclosing circle sequences; According to The difference between the centers of all the minimum enclosing circles in each minimum enclosing circle sequence of the frame face recognition image is obtained. Several target sequences of frame face recognition images; According to The number of target sequences corresponding to the frame face recognition image and the The difference between the areas of adjacent minimum enclosing circles in each minimum enclosing circle sequence corresponding to the frame face recognition image is obtained. The directionality of changes in facial reference points of frame face recognition images.

5. The method for monitoring the depth of oral anesthesia based on intelligent perception according to claim 4, characterized in that: According to the Frame face recognition image and adjacent face recognition images, get the first The neighborhood range of the frame face recognition image includes the following specific steps: The first Frame face recognition image and its temporal sequence before and after The face recognition image of the frame is recorded as The neighborhood range of the frame face recognition image; among them, is the preset range threshold.

6. The method for monitoring the depth of oral anesthesia based on intelligent perception according to claim 4, characterized in that: According to the The difference between the centers of all the minimum enclosing circles in each minimum enclosing circle sequence of the frame face recognition image is obtained. Several target sequences of frame face recognition images include the following specific steps: Jordi If the center coordinates of all the minimum enclosing circles in the minimum enclosing circle sequence are different, then The minimum enclosing circle sequence is recorded as the target sequence.

7. The method for monitoring oral anesthesia depth based on intelligent perception according to claim 4, characterized in that: According to the The number of target sequences corresponding to the frame face recognition image and the The difference between the areas of adjacent minimum enclosing circles in each minimum enclosing circle sequence corresponding to the frame face recognition image is obtained. The specific calculation formula corresponding to the change direction of the facial reference points of the frame face recognition image is: ； in, Indicates Directionality of changes in facial reference points of frame face recognition images; Indicates The number of target sequences corresponding to the frame face recognition images; Indicates The number of bounding circles of the frame face recognition image; Indicates The minimum number of enclosing circles in each minimum enclosing circle sequence corresponding to the frame face recognition image; Indicates The frame face recognition image corresponds to the The smallest enclosing circle sequence The area of ​​the smallest enclosing circle; Indicates The frame face recognition image corresponds to the The smallest enclosing circle sequence The area of ​​the smallest enclosing circle; for Belong to the interval All internal time The variance of .

8. The method for monitoring oral anesthesia depth based on intelligent perception according to claim 1, characterized in that: The method of obtaining the oral anesthesia depth of each face recognition image according to the facial expression relaxation degree of each face recognition image frame includes the following specific steps: When the facial expression relaxation degree of the last frame of face recognition image is less than the preset empirical threshold, it indicates that the currently monitored oral anesthesia depth is insufficient.

Images