Brain blood oxygen saturation monitoring method and system
By obtaining skin pigmentation and ambient light brightness parameters, the neural network model is improved, and the problem of low accuracy of blood oxygen saturation monitoring in the prior art is solved, accurate monitoring in emergency medical care is achieved, and blood oxygen saturation is improved accuracy and intelligence of degree prediction.
Patent Information
- Application Number
- CN202411083714.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-08
- Publication Date
- 2024-11-08
AI Technical Summary
The prior art has low accuracy when monitoring brain blood oxygen saturation in people with different skin colors, and fails to effectively consider the impact of ambient light intensity, resulting in low accuracy in predicting blood oxygen saturation and insufficient samples, affecting emergency medical care.
By obtaining the skin pigmentation and ambient light brightness parameters of the target user's forehead area, the generative adversarial network and particle swarm algorithm are improved, the BP neural network model is optimized, and the environmental image is obtained in combination with the camera is improved to improve the accuracy of blood oxygen saturation monitoring.
It achieves the accuracy of brain blood oxygen saturation monitoring while saving costs, reduces the risk of wrong measurements, enhances the reliability of emergency medical care, and ensures the accuracy and intelligence of blood oxygen saturation monitoring. .
Smart Images

Figure CN118902448A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of blood oxygen saturation monitoring, and in particular to a method and system for monitoring brain blood oxygen saturation. Background Art
[0002] Near-infrared spectroscopy is a technology developed in recent years to monitor brain oxygen supply. It provides a convenient, inexpensive, time-effective and low-cost measurement method for clinical non-invasive testing and is widely used in various occasions of brain oxygen monitoring. Near-infrared spectroscopy is based on the Lambert-Beer law and light scattering theory to measure blood oxygen saturation, using the difference in the light absorption coefficients of reduced and oxygenated hemoglobin.
[0003] However, there is evidence that oximeters are less accurate when monitoring patients with dark skin. Postoperative monitoring of blood oxygen saturation or home monitoring does not require high accuracy, as long as the overall trend of blood oxygen saturation can be determined. However, in emergency medical care, doctors or caregivers need to accurately monitor the patient's blood oxygen saturation. Incorrect measurement results will seriously affect the patient's treatment and recovery. Therefore, how to improve the accuracy of blood oxygen measurement for people with different skin colors is of great significance.
[0004] At present, the detection accuracy is improved by correcting the absorption rate or reflectivity of infrared or near-infrared light by skin color intervals or by correcting the absorption rate or reflectivity of infrared or near-infrared light according to skin color intervals. However, the above methods do not objectively measure the skin pigmentation of the target users. In addition, the existing method of using BP neural network to predict blood oxygen saturation does not consider the impact of ambient light intensity on blood oxygen saturation, and the existing method also has technical problems such as insufficient samples and low accuracy of directly using BP neural network to predict blood oxygen saturation. Finally, integrating nursing cameras into the brain blood oxygen saturation monitoring system to realize intelligent care for patients is also a focus of the current nursing field.
[0005] Based on this, how to achieve the accuracy of blood oxygen saturation monitoring on the basis of realizing intelligent care of the brain blood oxygen saturation monitoring system while considering cost savings is a difficult problem that urgently needs to be solved in this field. Summary of the invention
[0006] The purpose of the present invention is to provide a method and system for monitoring brain blood oxygen saturation to solve the problems raised in the above background technology. The specific scheme is as follows: A brain blood oxygen saturation monitoring method is applied to a brain blood oxygen saturation monitoring system with a nursing function, comprising the following steps: Step S1, obtaining the skin pigmentation degree of the forehead area of the target user; Step S2, obtaining the real-time measured blood oxygen saturation of the target user's forehead area through a brain blood oxygen saturation monitor; Step S3, collecting an environmental image of the area where the target user is located at the same time as the real-time measurement of the blood oxygen saturation, recorded as a first image, the first image is obtained by a camera of the blood oxygen saturation monitoring system, the first image is divided into a plurality of image blocks, each image block is subjected to a DCT transformation, and then the DC components of all the image blocks are averaged to obtain a brightness parameter that can characterize the brightness of the ambient light; Step S4, inputting the skin pigmentation of the target user's forehead area, the brightness parameter that can characterize the ambient light brightness, and the real-time measured blood oxygen saturation of the forehead area into the BP neural network model to obtain the actual blood oxygen saturation of the target user.
[0007] In step S3, the first image is divided into 16*16 image blocks.
[0008] The BP neural network model is obtained in the following way: Step S40, obtaining an original sample set, each sample including skin pigmentation, a brightness parameter that can characterize the brightness of ambient light, and real-time measured blood oxygen saturation of the forehead area, and the sample label is the actual blood oxygen saturation obtained by the blood gas analyzer test; Step S41, improving the original generative adversarial network to obtain an improved generative adversarial network; Step S42, using the improved generative adversarial network to expand the original sample set to obtain an expanded sample set consisting of a mixture of real samples and expanded samples; Step S43, using the improved particle swarm algorithm to optimize the network parameters of the BP neural network to obtain the optimized BP neural network, and using the expanded sample set to train the optimized BP neural network to obtain the trained optimized BP neural network.
[0009] The improved particle swarm algorithm is obtained in the following way: the fixed inertia weight is adaptively adjusted by nonlinear adjustment method, so as to obtain the inertia weight calculation method of the improved particle swarm algorithm: ,in and They represent the maximum and minimum values of the inertia weight, respectively. Their optimal values are 0.19 and -0.29, respectively. The adjustment coefficient , T represents the maximum number of iterations, t represents the current number of iterations; Substituting the above improved inertia weight into the speed iteration formula in the standard particle swarm algorithm, the improved particle swarm algorithm can be obtained.
[0010] The speed iteration formula in the standard particle swarm algorithm is as follows: , where i=1,2,…,N, where N represents the total number of particles, and t represents the current number of iterations. represents the inertia weight, represents the speed of the particle, r1 and r2 represent random numbers, represents the current position of particle i, represents the current individual optimal position of particle i, represents the current optimal position of the particle swarm, and c1 and c2 represent learning factors.
[0011] A brain blood oxygen saturation monitoring system, including a nursing function, includes a brain blood oxygen saturation monitor, a skin pigment detector, a display, a central processing unit and a camera; The central processing unit is communicatively connected with a brain blood oxygen saturation monitor, a skin pigment detector, a display, and a camera; The skin pigment detector is used to obtain the skin pigmentation degree of the target user's forehead area and transmit the value of the skin pigmentation degree of the target user's forehead area to the central processor; The brain blood oxygen saturation monitor is used to obtain the real-time measured blood oxygen saturation of the target user's forehead area, and transmit the real-time measured blood oxygen saturation of the target user's forehead area to the central processor; A camera is used to collect an environmental image of the area where the target user is located at the same time as the real-time measurement of the blood oxygen saturation, which is recorded as a first image; The central processing unit divides the first image into 16*16 image blocks, performs DCT transformation on each image block, and then averages the DC components of all image blocks to obtain a brightness parameter that can characterize the brightness of the ambient light; The skin pigmentation of the target user's forehead area, the brightness parameter that can characterize the brightness of the ambient light, and the real-time measured blood oxygen saturation of the forehead area are input into the BP neural network model to obtain the actual blood oxygen saturation of the target user; The display is used to display the actual blood oxygen saturation of the target user obtained by the central processor in real time.
[0012] The camera is also used to monitor medication feeding for target users.
[0013] A brain blood oxygen saturation monitoring system, the nursing function is mainly achieved through the following steps: (1) Use a camera to capture an image containing the target user and the bed, recorded as the second image; use an edge detection algorithm to detect the left and right edges of the bed in the second image, and fit the straight lines L1 and L2 where the left and right edges are located; set warning lines L3 and L4 parallel to the above two straight lines at a certain distance from the extension of the straight lines L1 and L2; (2) Use a deep learning algorithm to detect and locate the head, arms, and legs of the target user in the second image; (3) Determine whether the center points of the detection frames of the head, unilateral arm, and leg are all above the warning line L3 or L4. If the judgment result is yes, it indicates that the target user is at risk of falling off the bed, and the central processing unit transmits a warning message to the warning unit of the medical station.
[0014] The brain blood oxygen saturation monitoring system also has a switch to turn the nursing function on and off.
[0015] The brain blood oxygen saturation monitor includes a brain blood oxygen saturation probe, which includes a left light receiving unit, a left light emitting unit, a right light emitting unit and a right light receiving unit in left and right directions, the left light receiving unit includes a shallow photoreceptor and a deep photoreceptor, and the right light receiving unit includes a shallow photoreceptor and a deep photoreceptor.
[0016] Compared with the prior art, the present invention has the following beneficial effects: The present invention realizes the intelligent combination of brain blood oxygen saturation monitoring and user care, and realizes the intelligentization of medical services; Skin pigmentation and ambient light brightness are important factors that affect brain blood oxygen saturation detection. In the prior art, blood oxygen saturation is often corrected by skin grading or according to skin color. Although the influence of skin color on blood oxygen saturation is taken into account, the accuracy is still not high. In addition, in the prior art, brain blood oxygen saturation monitoring rarely considers the influence of ambient brightness. The present invention creatively takes skin pigmentation as a factor to be considered for blood oxygen saturation prediction, and uses the DC component of the image as a brightness parameter to characterize the ambient light brightness for blood oxygen saturation prediction. Although there is a method of using a neural network to predict brain blood oxygen saturation in the prior art, it does not consider the problem of low training accuracy caused by insufficient samples, nor does it consider the problem of poor accuracy when directly using the original BP neural network for prediction. The present invention improves the original generative adversarial network to obtain an improved generative adversarial network, uses the improved generative adversarial network to expand the samples, and further uses an improved particle swarm algorithm to optimize the BP neural network, and uses the optimized BP neural network model to predict the actual brain blood oxygen saturation, thereby improving the accuracy of the actual brain blood oxygen saturation prediction.
[0017] (4) When acquiring the ambient light brightness, the present invention uses an existing camera to acquire the brightness, and does not need to re-set the brightness monitoring device. This saves costs while acquiring the ambient light brightness parameters. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] Figure 1It is a schematic flow chart of a method for monitoring brain blood oxygen saturation of the present invention; Figure 2 This is a plan view of the brain blood oxygen saturation probe; Figure 3 It is a structural schematic diagram of the brain blood oxygen saturation monitoring system of the present invention.
[0019] In the figure: 1, right light emitting unit; 2, left light emitting unit; 3, right light receiving unit; 4, left light receiving unit; 4a, shallow photoreceptor; 4b, deep photoreceptor; 3a, shallow photoreceptor; 3b, deep photoreceptor. DETAILED DESCRIPTION
[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0021] Figure 1 A brain blood oxygen saturation monitoring method of the present invention is applied to a brain blood oxygen saturation monitoring system with a nursing function, comprising the following steps: Step S1, obtaining the skin pigmentation degree of the forehead area of the target user.
[0022] The actual value of brain blood oxygen saturation is affected by many factors. A lot of evidence shows that brain blood oxygen saturation detectors are less accurate in patients with darker skin. Or the skin color is divided into intervals, or the absorption rate or reflectivity of infrared or near-infrared light is corrected according to the skin color interval to improve the accuracy of the detection, but the above method does not objectively measure the skin pigmentation of the target user, and then obtain the actual blood oxygen saturation corresponding to the skin pigmentation one-to-one. Therefore, the above factors need to be considered when determining the actual blood oxygen saturation.
[0023] Among them, in this embodiment, the skin pigmentation degree of the forehead area of the target user is determined by a skin pigmentation detector. Of course, in actual applications, the skin pigmentation degree of the current forehead area of the target user can also be obtained by other methods.
[0024] Among them, the skin pigment detector is based on the principle of spectral absorption. It determines the content of melanin and hemoglobin in the skin by measuring the amount of reflection after a specific wavelength of light is irradiated on the human skin. For example, the transmitter of the skin pigment detector emits three wavelengths of light of 568nm, 660nm, and 880nm to illuminate the surface of the forehead area, and the receiver measures the light reflected by the skin. Since the amount of emitted light is certain, the amount of light absorbed by the skin can be measured, thereby measuring the content of melanin and hemoglobin in the skin.
[0025] Specifically, the skin pigmentation detector emits light to the skin in the forehead area of the target user, and the receiver measures the detected light after being reflected by the skin. The skin pigmentation detector converts the detected light into an electrical signal, determines the skin pigmentation degree in the forehead area based on the electrical signal, and transmits the skin pigmentation value to the central processor.
[0026] Step S2, obtaining the real-time measured blood oxygen saturation of the target user's forehead area through a brain blood oxygen saturation monitor.
[0027] Among them, the brain blood oxygen saturation detector of this embodiment fixes the brain blood oxygen saturation probe of the brain blood oxygen saturation detector on the forehead area of the target user to continuously measure the blood oxygen saturation of the target user, and then obtains the real-time measured blood oxygen saturation of the forehead area of the target user.
[0028] The brain blood oxygen saturation monitor is based on near-infrared spectroscopy. Near-infrared spectroscopy measures blood oxygen saturation based on the Lambert-Beer law and light scattering theory, using the difference in the light absorption coefficients of reduced and oxygenated hemoglobin.
[0029] Preferably, see Figure 2, is a brain blood oxygen saturation monitor, including a brain blood oxygen saturation probe, the brain blood oxygen saturation probe includes a right light-emitting unit 1 and a left light-emitting unit 2, and a right light-receiving unit 3 and a left light-receiving unit 4 as light-receiving units, the left light-receiving unit 4, the left light-emitting unit 2, the right light-emitting unit 1, and the right light-receiving unit 3 are arranged in sequence along the left-right direction, and the right light-emitting unit 2 and the left light-emitting unit 1 are symmetrically arranged along the straight line L. The left light-receiving unit 4 includes a shallow photoreceptor 4a and a deep photoreceptor 4b, and the right light-receiving unit 3 includes a shallow photoreceptor 3a and a deep photoreceptor 3b. By including shallow photoreceptors 3a and 4a and deep photoreceptors 3b and 4b, more accurate measurements can be performed. Preferably, the distance d2 of the deep photoreceptors 3b and 4b from the light-emitting units 1 and 2 is farther than the distance d1 of the shallow photoreceptors 3a and 4a from the light-emitting units 1 and 2. The distance d1 from the light emitting units 1 and 2 to the superficial photoreceptors 3a and 4a is the distance that the light emitted by the light emitting units 1 and 2 passes through the superficial part of the brain to reach the forehead surface, for example, preferably greater than 0.1 mm and less than 35 mm, more preferably greater than 0.5 mm and less than 30 mm. The distance d2 from the light emitting units 1 and 2 to the deep photoreceptors 3b and 4b is the distance that the light emitted from the light emitting units 1 and 2 passes through the deep part of the brain to reach the forehead surface. It is preferably greater than 35 mm and not greater than 60 mm, more preferably not less than 40 mm and not greater than 50 mm. The right light emitting unit 1 and the left light emitting unit 2 are respectively a light emitting diode emitting red light with a wavelength of 660 nm and a light emitting diode emitting infrared light with a wavelength of 810 nm. The light processing unit processes the two diffuse light signals to obtain the cerebral blood oxygen saturation value.
[0030] Preferably, the calculation formula for brain blood oxygen saturation is: .
[0031] Among them, R1 is the red light signal received by the shallow photoreceptor 3a, R2 is the red light signal received by the deep photoreceptor 3b, N1 is the infrared light signal received by the shallow photoreceptor 4a, and N2 is the infrared light signal received by the deep photoreceptor 4b. A, B, C, and D are coefficients respectively. , E N It is the absorption coefficient of oxygenated hemoglobin HbO2 and reduced hemoglobin Hb under 810nm wavelength infrared light. and They are the absorption coefficients under 660nm wavelength red light. The absorption coefficients can be read from the infrared absorption spectra of oxygenated hemoglobin and reduced hemoglobin. B and C are obtained through experimental summary.
[0032] Step S3, collecting an environmental image of the area where the target user is located at the same time as the real-time measurement of the blood oxygen saturation, recorded as the first image, the first image is obtained through the camera of the blood oxygen saturation monitoring system, the first image is divided into 16*16 image blocks, each image block is subjected to DCT transformation, and then the DC components of all the image blocks are averaged to obtain a brightness parameter that can characterize the brightness of the ambient light.
[0033] Through experiments, it was found that since the brain blood oxygen saturation detector uses near-infrared spectroscopy to measure blood oxygen saturation, the ambient light brightness will have a certain impact on the measurement of brain blood oxygen saturation. For example, when the patient is given light therapy, the blood oxygen saturation value displayed by the brain blood oxygen saturation detector will drop sharply to 36%, and the alarm will begin. When the light therapy is stopped, the blood oxygen saturation will rise to 96%. In addition, even if the patient is not given light therapy, the blood oxygen saturation values during the day and at night will be different due to the ambient light brightness. In order to improve the accuracy of brain blood oxygen saturation measurement, the influencing factor of ambient light brightness needs to be taken into account when predicting the actual blood oxygen saturation of the target user. The existing ambient light brightness is generally measured by a photometer or estimated based on an illumination model. The former requires a separate photometer, while the latter is complex to calculate. In order to solve the above problems and effectively utilize the existing camera in the blood oxygen saturation monitoring system, the present invention collects the environmental image of the area where the target user is located, divides the image into 16*16 image blocks, performs DCT transformation on each image block, and then averages the DC components of all image blocks to obtain a brightness parameter that can characterize the ambient light brightness. It should be noted that the DC component of the DCT transformation is not directly equivalent to the actual brightness value of the environment, but it can reflect the brightness distribution of the ambient light in the image space domain to a certain extent. Therefore, the average value of the DC components of all image blocks is used here to characterize the ambient light brightness, thereby obtaining a brightness parameter that can characterize the ambient light brightness.
[0034] The blood oxygen saturation monitoring system includes a brain blood oxygen saturation detector, a camera, a central processing unit, etc. The camera can monitor the status of the target user and obtain the environmental image of the area where the target user is located, namely the first image. The camera can transmit the captured first image to the central processing unit. The central processing unit performs the above operations on the first image and finally obtains the brightness parameters that can characterize the brightness of the ambient light.
[0035] The brain blood oxygen saturation detector can transmit the real-time measured blood oxygen saturation of the target user's forehead area to the central processor.
[0036] Step S4, inputting the skin pigmentation of the target user's forehead area, the brightness parameter that can characterize the ambient light brightness, and the real-time measured blood oxygen saturation of the forehead area into the trained and optimized BP neural network model to obtain the actual blood oxygen saturation of the target user.
[0037] The brain blood oxygen saturation monitoring method of this embodiment, when monitoring the actual blood oxygen saturation of the target user, combines the real-time measured blood oxygen saturation of the forehead area obtained by the brain blood oxygen saturation detector, and then combines the actual skin pigmentation of the forehead area of the target user when the target user is in the monitoring state, and the brightness parameter of the ambient light brightness of the target user's environment area in real time when the target user is in the monitoring state, and obtains the real-time, real and accurate actual blood oxygen saturation of the target user through the trained optimized BP neural network model. The present invention can measure the blood oxygen saturation in real time according to the brain blood oxygen saturation detector and the brightness parameter of the ambient light brightness of the target user's environment area obtained at the same time as the brain blood oxygen saturation detector, and then obtain the real-time, real and accurate actual blood oxygen saturation, and then realize that no matter how the indoor environment of the target user changes and how to treat, such as the adjustment of the brightness of the ambient light in the target user's environment, the brightness change caused by day or night or light therapy, etc., the real and actual brain blood oxygen saturation of the target user can be displayed in real time, avoiding false alarms and improving the treatment effect. The present invention can also provide the target user with the actual brain blood oxygen saturation according to the difference of melanin in the target user's skin. In this embodiment, through the camera integrated in the brain blood oxygen saturation monitoring system, while monitoring the target user, the ambient brightness of the area where the target user is located can also be obtained, thereby achieving cost saving while being able to obtain the real, accurate and actual blood oxygen saturation.
[0038] According to one embodiment of the present application, the trained optimized BP neural network model is obtained by: Step S40, obtaining an original sample set, each sample including skin pigmentation, a brightness parameter that can characterize the brightness of ambient light, and real-time measured blood oxygen saturation of the forehead area, and the sample label is the actual blood oxygen saturation obtained by the blood gas analyzer test; Step S41, improving the original generative adversarial network to obtain an improved generative adversarial network: The loss function of the original generative adversarial network is: , where G represents the generator, D represents the discriminator, V(D, G) represents the objective function, D(x) represents the discriminator's judgment result, and D(x) z represents random noise, represents the noise distribution, x represents the real data, In order to further improve the quality of the fake samples generated by the generator and enable the generator to continuously learn new knowledge, the loss function is improved to obtain the loss function of the improved generative adversarial network: ,in Represents the penalty factor. The generator generates some forged samples G(z) by adding noise z, and a random number that conforms to the uniform distribution Construct interpolation samples ,Right now , use the gradient of the interpolated sample for gradient penalty, so that the discriminator satisfies the 1-lipchiz condition. It should be noted that the generative adversarial network can also be improved in other ways, which are not limited to this.
[0039] Step S42, using the improved generative adversarial network to expand the original sample set to obtain an expanded sample set consisting of a mixture of real samples and expanded samples; Step S43, using the improved particle swarm algorithm to optimize the network parameters of the BP neural network to obtain the optimized BP neural network, and using the expanded sample set to train the optimized BP neural network to obtain the trained optimized BP neural network.
[0040] Among them, the network parameters of the BP neural network include weights, hidden layer thresholds and output layer thresholds.
[0041] According to one embodiment of the present application, the improved particle swarm algorithm is obtained by: The speed iteration formula in the standard particle swarm algorithm is as follows:
[0042] Where i=1,2,…,N, N represents the total number of particles, t represents the current number of iterations, represents the inertia weight, represents the velocity of the particle, Represents a random number, represents the current position of particle i, represents the current individual optimal position of particle i, represents the current optimal position of the particle swarm, Represents the learning factor.
[0043] In order to accelerate the convergence speed of the particle swarm, a nonlinear adjustment method is used to adaptively adjust the fixed inertia weight, thereby obtaining the inertia weight calculation method of the improved particle swarm algorithm. ;in and The optimal values of are 0.19 and -0.29 respectively, and the adjustment coefficient ,T represents the maximum number of iterations, and t represents the current number of iterations. Substituting the improved inertia weight into the speed iteration formula of the standard particle swarm algorithm, the improved particle swarm algorithm can be obtained.
[0044] According to one embodiment of the present application, a blood oxygen saturation monitoring system is integrated with a camera, and the camera is used to capture an environmental image of the area where the target user is located at the same time as the real-time measurement of the blood oxygen saturation.
[0045] The blood oxygen saturation monitoring system includes a brain blood oxygen saturation detector, a camera, a central processing unit, etc. The camera can monitor the state of the target user and obtain an environmental image of the area where the target user is located, namely, a first image.
[0046] Figure 3 A schematic diagram of a brain blood oxygen saturation monitoring system provided in an embodiment of the present application.
[0047] like Figure 3 As shown, the brain blood oxygen saturation monitoring system has a nursing function, which includes a brain blood oxygen saturation monitor, a skin pigment detector, a display, a central processing unit, and a camera. The central processing unit is connected to the brain blood oxygen saturation monitor, the skin pigment detector, the display, and the camera in communication; The skin pigmentation detector is used to obtain the skin pigmentation degree of the forehead area of the target user and transmit the skin pigmentation value to the central processor.
[0048] The brain blood oxygen saturation monitor is used to obtain the real-time measured blood oxygen saturation of the target user's forehead area, and transmit the real-time measured blood oxygen saturation of the target user's forehead area to the central processor.
[0049] A camera is used to collect an environmental image of the area where the target user is located at the same time as the real-time measurement of the blood oxygen saturation, which is recorded as a first image; The central processing unit divides the first image into 16*16 image blocks, performs DCT transformation on each image block, and then averages the DC components of all image blocks to obtain a brightness parameter that can characterize the brightness of the ambient light, and inputs the skin pigmentation of the forehead area of the target user, the brightness parameter that can characterize the brightness of the ambient light, and the real-time measured blood oxygen saturation of the forehead area into the trained and optimized BP neural network model to obtain the actual blood oxygen saturation of the target user; The display is used to display the actual blood oxygen saturation of the target user obtained by the central processor in real time.
[0050] According to one embodiment of the present application, the nursing function is mainly implemented through the following steps: Use a camera to capture an image containing the target user and the bed, which is recorded as the second image; use an edge detection algorithm to detect the left and right edges of the bed in the second image, and fit the straight lines L1 and L2 where the left and right edges are located; set warning lines L3 and L4 parallel to the above two straight lines at a certain distance from the extension of the straight lines L1 and L2; Using a deep learning algorithm to detect and locate the head, arms, and legs of the target user in the second image; It is determined whether the center points of the detection frames of the head, unilateral arm and legs are all beyond the warning line L3 or L4. If the judgment result is yes, it indicates that the target user is at risk of falling from the bed, and the central processor transmits a warning message to the warning unit of the medical station.
[0051] Specifically, the camera in the present invention can monitor the target user and the bed position area in real time, and transmit the real-time video and image of the camera to the central processor. The central processor can obtain a first image and a second image based on the video and image obtained by the camera. The first image is used to obtain a brightness parameter that can characterize the brightness of the ambient light, and the second image is used to determine the risk of the target user falling from the bed. The present invention can not only realize the intelligent care of the brain blood oxygen saturation monitoring system, but also provide parameters for accurately obtaining brain blood oxygen saturation, thereby reducing costs.
[0052] According to one embodiment of the present application, the brain blood oxygen saturation monitoring system also has a switch for turning the nursing function on and off.
[0053] For critically ill patients or children, there is a risk of unconsciously falling from the bed, while for patients with normal consciousness, after staying in bed for a long time, due to the need to get out of bed or go to the bathroom, the center point of the detection frame where the head, unilateral arm and leg are located may exceed the warning line L3 or L4, thus causing a false alarm. Based on this, medical staff or administrators can determine whether to turn on or off the nursing function of the brain blood oxygen saturation monitoring system according to the actual situation of the target user.
[0054] According to one embodiment of the present application, the warning unit further includes a monitor capable of displaying the bed position of the target user.
[0055] When monitoring multiple target users, the locations of the target users, such as room numbers and bed numbers, can also be displayed.
[0056] According to one embodiment of the present application, medical personnel or administrators can also remotely reset the warning line while watching the monitor.
[0057] According to one embodiment of the present application, there is also a medicine feeding monitoring function, which is mainly achieved through the following steps: (1) Use a camera to shoot a video of the target user's medication process. By detecting the video frames, determine the first frame image of the target user's palm with drugs, recorded as image F1. Detect image F1 to obtain the size and number of each drug in the palm; It should be noted that when detecting video frames, there may be multiple consecutive frames of images showing drugs on the palm of the target user. Here, only the first frame of the image showing drugs on the palm is taken.
[0058] The video frames after image F1 are detected. If there is a first frame image in which no medicine is on the palm of the target user in the video frame, it is determined that the target user has taken the medicine; the size and quantity of each medicine determined from image F1 are compared with the size and quantity of the medicine that should be taken that were previously stored to determine whether the medicine has been taken correctly. If the judgment result is that the medicine has not been taken correctly, a warning message of "wrong medication" is output to the monitor of the warning unit.
[0059] This embodiment also provides a computer-readable storage medium on which a computer program is stored, characterized in that when the program is executed by a processor, the above-mentioned brain blood oxygen saturation monitoring method is implemented.
[0060] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.
[0061] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.
[0062] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.
[0063] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or otherwise processing in a suitable manner if necessary, and then stored in a computer memory.
[0064] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0065] A person skilled in the art may understand that all or part of the steps in the above-mentioned embodiment method may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.
[0066] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.
[0067] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.
Claims
1. A brain blood oxygen saturation monitoring method, applied to a brain blood oxygen saturation monitoring system with nursing function, characterized in that: The steps include: Step S1, obtaining the skin pigmentation degree of the forehead area of the target user; Step S2, obtaining the real-time measured blood oxygen saturation of the target user's forehead area through a brain blood oxygen saturation monitor; Step S3, collecting an environmental image of the area where the target user is located at the same time as the real-time measurement of the blood oxygen saturation, recorded as a first image, the first image is obtained by a camera of the blood oxygen saturation monitoring system, the first image is divided into a plurality of image blocks, each image block is subjected to a DCT transformation, and then the DC components of all the image blocks are averaged to obtain a brightness parameter that can characterize the brightness of the ambient light; Step S4, inputting the skin pigmentation of the target user's forehead area, the brightness parameter that can characterize the ambient light brightness, and the real-time measured blood oxygen saturation of the forehead area into the BP neural network model to obtain the actual blood oxygen saturation of the target user.
2. A method for monitoring brain blood oxygen saturation as claimed in claim 1, characterized in that: In step S3, the first image is divided into 16*16 image blocks.
3. A method for monitoring brain blood oxygen saturation as claimed in any one of claims 1 or 2, characterized in that: The BP neural network model is obtained in the following way: Step S40, obtaining an original sample set, each sample including skin pigmentation, a brightness parameter that can characterize the brightness of ambient light, and real-time measured blood oxygen saturation of the forehead area, and the sample label is the actual blood oxygen saturation obtained by the blood gas analyzer test; Step S41, improving the original generative adversarial network to obtain an improved generative adversarial network; Step S42, using the improved generative adversarial network to expand the original sample set to obtain an expanded sample set consisting of a mixture of real samples and expanded samples; Step S43, using the improved particle swarm algorithm to optimize the network parameters of the BP neural network to obtain the optimized BP neural network, and using the expanded sample set to train the optimized BP neural network to obtain the trained optimized BP neural network.
4. A method for monitoring brain blood oxygen saturation as claimed in claim 3, characterized in that: The improved particle swarm algorithm is obtained in the following way: The fixed inertia weight is adaptively adjusted by nonlinear adjustment method, so as to obtain the inertia weight calculation method of the improved particle swarm algorithm: ,in and They represent the maximum and minimum values of the inertia weight, respectively. Their optimal values are 0.19 and -0.29, respectively. The adjustment coefficient ,T represents the maximum number of iterations, t represents the current number of iterations; Substituting the above improved inertia weight into the speed iteration formula in the standard particle swarm algorithm, the improved particle swarm algorithm can be obtained.
5. A method for monitoring brain blood oxygen saturation as claimed in claim 4, characterized in that: The speed iteration formula in the standard particle swarm algorithm is as follows: , where i=1,2,…,N, where N represents the total number of particles, and t represents the current number of iterations. represents the inertia weight, represents the speed of the particle, r1 and r2 represent random numbers, represents the current position of particle i, represents the current individual optimal position of particle i, represents the current optimal position of the particle swarm, and c1 and c2 represent learning factors.
6. A brain blood oxygen saturation monitoring system, characterized in that: Includes care functions, including a brain oxygen saturation monitor, skin pigment detector, display, central processor and camera; The central processing unit is communicatively connected with a brain blood oxygen saturation monitor, a skin pigment detector, a display, and a camera; The skin pigment detector is used to obtain the skin pigmentation degree of the target user's forehead area and transmit the value of the skin pigmentation degree of the target user's forehead area to the central processor; The brain blood oxygen saturation monitor is used to obtain the real-time measured blood oxygen saturation of the target user's forehead area, and transmit the real-time measured blood oxygen saturation of the target user's forehead area to the central processor; A camera is used to collect an environmental image of the area where the target user is located at the same time as the real-time measurement of the blood oxygen saturation, which is recorded as a first image; The central processing unit divides the first image into 16*16 image blocks, performs DCT transformation on each image block, and then averages the DC components of all image blocks to obtain a brightness parameter that can characterize the brightness of the ambient light; The skin pigmentation of the target user's forehead area, the brightness parameter that can characterize the brightness of the ambient light, and the real-time measured blood oxygen saturation of the forehead area are input into the BP neural network model to obtain the actual blood oxygen saturation of the target user; The display is used to display the actual blood oxygen saturation of the target user obtained by the central processor in real time.
7. A brain blood oxygen saturation monitoring system as claimed in claim 6, characterized in that: The nursing function is mainly achieved through the following steps: (1) Use a camera to capture an image containing the target user and the bed, recorded as the second image; use an edge detection algorithm to detect the left and right edges of the bed in the second image, and fit the straight lines L1 and L2 where the left and right edges are located; set warning lines L3 and L4 parallel to the above two straight lines at a certain distance from the extension of the straight lines L1 and L2; (2) Use a deep learning algorithm to detect and locate the head, arms, and legs of the target user in the second image; (3) Determine whether the center points of the detection frames of the head, unilateral arm, and leg are all above the warning line L3 or L4. If the judgment result is yes, it indicates that the target user is at risk of falling off the bed, and the central processing unit transmits a warning message to the warning unit of the medical station.
8. A brain blood oxygen saturation monitoring system as claimed in claim 6, characterized in that: It also has a switch to turn the nursing function on and off.
9. A brain blood oxygen saturation monitoring system as claimed in claim 6, characterized in that: The brain blood oxygen saturation monitor includes a brain blood oxygen saturation probe, which includes a left light receiving unit, a left light emitting unit, a right light emitting unit and a right light receiving unit in left and right directions, the left light receiving unit includes a shallow photoreceptor and a deep photoreceptor, and the right light receiving unit includes a shallow photoreceptor and a deep photoreceptor.
Citation Information
Cited By
Near-infrared detection method and device for blood oxygen saturation of local brain tissue
CN122642901A
A method and device for detecting local cerebral tissue oxygen saturation by near-infrared
CN122642901B
