Chronic wound assessment system based on artificial intelligence
Through the chronic wound assessment system based on artificial intelligence, combined with the exudate concentration detection and area calculation module, the accuracy and convenience of wound assessment in the existing technology are solved, efficient and accurate assessment of chronic wounds is achieved, and treatment is promoted. Plan development.
Patent Information
- Application Number
- CN202411238576.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-05
- Publication Date
- 2024-12-20
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The prior art lacks accuracy and convenience in evaluating chronic wounds, especially for large and irregular wounds or different colors, conventional measurement methods are difficult to provide reliable evaluation results.
Adopt a chronic wound assessment system based on artificial intelligence, including wound ooze concentration detection module, wound assessment area calculation module and wound assessment comprehensive management server, take photos through cameras, collect odor information and detect viscosity through electronic noses, and use feature extraction networks and volumes The Ji neural network performs automatic evaluation and generates artificial intelligence reports.
It improves the accuracy and convenience of wound assessment, can quickly and accurately understand the patient's wound status, and promotes medical staff to formulate effective treatment plans.
Smart Images
Figure CN119170275A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of wound assessment, and in particular, to an artificial intelligence-based chronic wound assessment system. Background Art
[0002] Wound assessment is a key step in wound management and wound treatment planning. Chronic wounds generally refer to wounds that have not healed or have no tendency to heal for more than one month. The main types are diabetic, venous, arterial, traumatic and pressure. Conventional wound area assessment uses a paper measuring ruler. When the wound area is large and irregular, or when the patient's wound is of different colors, the use of a measuring ruler for assessment will lack accuracy and convenience. Summary of the invention
[0003] In order to make up for the above shortcomings, the present application provides an artificial intelligence-based chronic wound assessment system. By adopting a wound exudate concentration detection module and a wound assessment area measurement module, the viscosity, odor, color, and area size of the wound exudate can be evaluated, and transmitted to the wound assessment intelligent terminal through a wound assessment comprehensive management server, thereby realizing intelligent assessment of the patient's affected area with higher assessment accuracy.
[0004] This application scheme provides an artificial intelligence-based chronic wound assessment system, including a wound exudate concentration detection module, a wound assessment area measurement module, a wound assessment integrated management server, and a wound assessment intelligent terminal;
[0005] The wound exudate concentration detection module is used to detect the viscosity of the wound;
[0006] The wound assessment area calculation module includes a camera for scanning the patient's wound by taking a photo and an electronic nose for collecting wound odor information;
[0007] The wound assessment integrated management server includes a device management program module electrically connected to the camera, the electronic nose and the wound exudate concentration detection module, and also includes a doctor management program module and a patient management program module for recording doctor and patient information respectively;
[0008] The wound assessment intelligent terminal is communicatively connected to the wound assessment integrated management server.
[0009] Preferably, the patient management program module includes patient wound information, and the patient wound information includes artificial intelligence wound assessment report information.
[0010] Preferably, the doctor management program module in the wound assessment comprehensive management server is used to manage hospital information and department information, the hospital information includes a list of hospitals that use the wound assessment area measurement module and the wound exudate concentration detection module; the department information includes a list of departments that use the wound assessment area measurement module and the wound exudate concentration detection module.
[0011] Preferably, the wound assessment area measurement module obtains wound information by measuring the wound assessment area of the wound image taken by the camera based on the trained feature extraction network and convolutional neural network, and the wound assessment area measurement module generates an artificial intelligence wound assessment report based on the wound information and sends it to the wound assessment comprehensive management server.
[0012] Preferably, the feature extraction network is a multi-receptive field pyramid network PSPNet or a feature pyramid network FPN.
[0013] Preferably, the wound assessment intelligent terminal has a report input module for inputting a supplementary report on the patient's wound.
[0014] Preferably, the wound exudate concentration detection module includes
[0015] A control sampling unit, comprising a collection bottle for collecting wound exudate and a sampling cavity connected to the inside thereof, wherein the sampling cavity is connected to a negative pressure wound treatment device, and the height of the wound exudate entering the sampling cavity is controlled by the negative pressure wound treatment device;
[0016] A volume monitoring unit, monitoring the liquid level of the wound exudate in the sampling cavity;
[0017] A pressure detection unit, for detecting the pressure difference of the liquid surface in the sampling cavity;
[0018] The calculation unit calculates the height difference, pressure difference, and time length of the liquid level change of the wound exudate, and calculates the viscosity of the wound exudate.
[0019] Preferably, the negative pressure wound therapy device includes a negative pressure zone and a controller, the controller is electrically connected to the volume monitoring unit, the pressure detection unit and the calculation unit, the sampling cavity is a cavity opened on the inner wall of the collecting bottle, is connected to the interior of the collecting bottle, and is connected to the negative pressure zone through a pipeline.
[0020] Preferably, the wound exudate concentration detection module further comprises a constant temperature control unit electrically connected to the controller, and the constant temperature control unit is used to keep the temperature of the wound exudate in the collection bottle unchanged.
[0021] Preferably, the wound exudate concentration detection module further comprises a storage unit electrically connected to the controller for storing data on the viscosity of the wound.
[0022] Beneficial effects: The present application provides an artificial intelligence-based chronic wound assessment system, in which the wound assessment area measurement module uses a camera to take photos and scan the patient's wound, and can automatically measure the area of the wound and the area under different wound colors; the wound assessment area measurement module uses an electronic nose to determine the wound odor information, and then the wound assessment area measurement module sends the wound area, color and odor report to the equipment management program module in the wound assessment comprehensive management server. In addition, the wound exudate concentration detection module also sends the viscosity status of the wound during the detection process to the equipment management program module in the wound assessment comprehensive management server. The doctor management program module in the wound assessment comprehensive management server can send the report information of the corresponding patient to the wound assessment intelligent terminal of the corresponding medical staff. This design makes it easier for medical staff to quickly and accurately understand the patient's wound recovery status, and indirectly promotes the next treatment plan of the medical staff. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] In order to more clearly illustrate the technical solutions of the implementation methods of the present application, the drawings required for use in the implementation methods will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.
[0024] Figure 1 It is a structural block diagram of a chronic wound assessment system based on artificial intelligence provided in an embodiment of the present application;
[0025] Figure 2 This is a structural block diagram of the wound exudate concentration detection module provided in an embodiment of the present application.
[0026] In the figure: 1-wound exudate concentration detection module; 11-control sampling unit; 111-collecting bottle; 112-sampling chamber; 113-negative pressure wound therapy equipment; 12-volume monitoring unit; 13-pressure detection unit; 14-calculation unit; 15-constant temperature control unit; 16-storage unit; 2-wound assessment area measurement module; 21-camera; 22-electronic nose; 3-wound assessment integrated management server; 31-equipment management program module; 32-doctor management program module; 33-patient management program module; 4-wound assessment intelligent terminal. DETAILED DESCRIPTION
[0027] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments.
[0028] See also Figure 1-Figure 2 , the present application provides a chronic wound assessment system based on artificial intelligence, including a wound exudate concentration detection module 1, a wound assessment area measurement module 2, a wound assessment integrated management server 3 and a wound assessment intelligent terminal 4;
[0029] The wound exudate concentration detection module 1 is used to detect the viscosity of the wound;
[0030] The wound assessment area calculation module 2 includes a camera 21 for scanning the patient's wound by taking a photo and an electronic nose 22 for collecting wound odor information;
[0031] The wound assessment integrated management server 3 includes a device management program module 31 electrically connected to the camera 21, the electronic nose 22 and the wound exudate concentration detection module 1, and also includes a doctor management program module 32 and a patient management program module 33 for recording doctor and patient information respectively;
[0032] The wound assessment intelligent terminal 4 is communicatively connected to the wound assessment comprehensive management server 3 .
[0033] In this embodiment, the wound assessment area measurement module 2 uses the camera 21 to take pictures and scan the patient's wound, and can automatically measure the area of the wound and the area under different wound colors; the wound assessment area measurement module 2 uses the electronic nose 22 to determine the wound odor information, and then the wound assessment area measurement module 2 sends the wound area, color and odor report to the equipment management program module 31 in the wound assessment comprehensive management server 3. In addition, the wound exudate concentration detection module 1 also sends the viscosity status of the wound during the detection process to the equipment management program module 31 in the wound assessment comprehensive management server 3. The doctor management program module 32 in the wound assessment comprehensive management server 3 can send the report information of the corresponding patient to the wound assessment intelligent terminal 4 of the corresponding medical staff. This design makes it easy for medical staff to quickly and accurately understand the patient's wound recovery status, and indirectly promote the medical staff's next treatment plan.
[0034] The patient management program module 33 includes patient wound information, and the patient wound information includes artificial intelligence wound assessment report information. Specifically, the wound assessment integrated management server 3 records the wound condition information obtained from the wound exudate concentration detection module 1 and the wound assessment area measurement module 2 to the patient management program module 33 to form the patient wound information. Such a design also facilitates the use of the patient management program module 33 to quickly compare the wound detection data before and after.
[0035] The doctor management program module 32 in the wound assessment integrated management server 3 is used to manage hospital information and department information. The hospital information includes a list of hospitals that use the wound assessment area measurement module 2 and the wound exudate concentration detection module 1; the department information includes a list of departments that use the wound assessment area measurement module 2 and the wound exudate concentration detection module 1. Specifically, the patient information and doctor information are recorded in the wound assessment integrated management server 3, so that the wound report of the corresponding patient can be quickly and accurately sent to the corresponding medical staff.
[0036] The wound assessment area calculation module 2 obtains wound information by calculating the wound assessment area of the wound image taken by the camera 21 and based on the trained feature extraction network and convolutional neural network. The wound assessment area calculation module 2 generates an artificial intelligence wound assessment report based on the wound information and sends it to the wound assessment comprehensive management server 3.
[0037] The feature extraction network is a multi-receptive field pyramid network PSPNet or a feature pyramid network FPN.
[0038] The specific steps of wound assessment in the wound assessment area calculation module 2 are as follows:
[0039] 1. Construct a feature extraction network and convolutional neural network for wound assessment area measurement;
[0040] 2. Take a picture of the wound and use deep learning annotation tools to annotate the wound picture;
[0041] 3. Generate wound image training set based on labeled wound images;
[0042] Fourth, input the wound image training set into the feature extraction network and convolutional neural network for image data training to generate a training model;
[0043] 5. Based on the trained feature extraction network and convolutional neural network, the wound assessment area is measured on the captured wound images to obtain trauma information.
[0044] It should also be noted that PSPNet uses a pyramid pooling module to aggregate contextual information. The pyramid pooling module has 4 layers; pyramid pooling module level 1, kernel size is 60*60, step size is 60; pyramid pooling module level 2, kernel size is 30*30, step size is 30; pyramid pooling module level 3, kernel size is 20*20, step size is 20; pyramid pooling module level 4, kernel size is 10*10, step size is 10; the convolutional neural network has 101 layers;
[0045] Step 1: Use the tensorflow-gpu framework to build PSPNet and a 101-layer convolutional neural network. Step 2: Use the deep learning annotation tool Labelme to annotate wound images. Step 3: Use the deep learning annotation tool Labelme to batch generate josn files as a wound image training set.
[0046] The deep learning parameters of PSPNet and convolutional neural network were set as 30000 training iterations, 0.9 stochastic gradient descent method, 0.1 learning rate dynamic decay, and 0.0001 weight decay.
[0047] Step 4: Input the wound image training set into PSPNet and convolutional neural network. After image data enhancement and image data training, continuously adjust the deep learning parameters of PSPNet and convolutional neural network to improve the accuracy and generate a training model.
[0048] Step 5: Extracting feature maps using the feature extraction network PSPNet includes the following steps:
[0049] 1) Input wound image data into PSPNet;
[0050] 2) By applying the pre-trained dilated convolution, the output feature map is 1 / 8 size;
[0051] 3) The output size of the pyramid pooling module level 1 is (1, 1, 2048) feature map, the output size of level 2 is (2, 2, 2048) feature map, the output size of level 3 is (3, 3, 2048) feature map, and the output size of level 4 is (4, 4, 2048) feature map;
[0052] 4) Through the PSPNet convolution layer, the predicted feature map of the wound image is obtained.
[0053] The wound assessment intelligent terminal 4 has a report input module for inputting a patient's wound supplementary report.
[0054] The wound exudate concentration detection module 1 includes a control sampling unit 11, a volume monitoring unit 12, a pressure detection unit 13 and a calculation unit 14;
[0055] The control sampling unit 11 includes a collection bottle 111 for collecting wound exudate and a sampling cavity 112 connected thereto. The sampling cavity 112 is connected to a negative pressure wound treatment device 113, and the height of the wound exudate entering the sampling cavity 112 is controlled by the negative pressure wound treatment device 113.
[0056] The volume monitoring unit 12 monitors the liquid level of the wound exudate in the sampling cavity 112;
[0057] The pressure detection unit 13 detects the pressure difference of the liquid surface in the sampling chamber 112;
[0058] The calculation unit 14 calculates the height difference, pressure difference, and time length of the liquid level change of the wound exudate, and calculates the viscosity of the wound exudate.
[0059] The negative pressure wound therapy device 113 includes a negative pressure zone and a controller. The controller is electrically connected to the volume monitoring unit 12, the pressure detection unit 13 and the calculation unit 14. The sampling chamber 112 is a cavity opened on the inner wall of the collecting bottle 111, which is connected to the interior of the collecting bottle 111 and connected to the negative pressure zone through a pipeline.
[0060] Specifically, a height value h is set during the sampling process, and the volume monitoring unit 12 monitors in real time. When the height difference between the liquid level of the wound exudate extracted from the sampling chamber 112 and the liquid level of the wound exudate in the collecting bottle 111 is h, the valve body provided between the sampling chamber 112 and the negative pressure zone is adjusted to maintain the liquid level in the sampling chamber 112 unchanged, thereby maintaining the height difference h unchanged. At the same time, the pressure detection unit 13 detects the pressure difference ΔP of the liquid level of the wound exudate in the sampling chamber 112, and then controls the valve body of the collecting bottle 111 and the sampling end in the sampling unit 11, and the valve body of the sampling chamber 112 and the negative pressure zone to be adjusted to be consistent. At the same time, the volume monitoring unit 12 monitors the sampling chamber 112. The liquid level of the wound exudate in the cavity 112 changes. When the liquid level of the wound exudate in the sampling cavity 112 is at the same horizontal plane as the liquid level of the wound exudate in the collecting bottle 111, the time length Δt for the liquid level to drop is recorded. Then, the calculation unit 14 calculates the flow rate of the wound exudate flowing through the sampling cavity 112 through a formula based on the radius R of the sampling cavity 112, the height difference h and the time length data. The calculation unit 14 uses the least squares method to perform curve fitting based on the pressure difference and the flow data to obtain the a and n values, and obtains a curve in which the static viscosity of the wound exudate changes with the shear rate. The slope of this curve is calculated, which is the viscosity of the wound exudate.
[0061] The wound exudate concentration detection module 1 further includes a constant temperature control unit 15 electrically connected to the controller, and the constant temperature control unit 15 is used to keep the temperature of the wound exudate in the collection bottle 111 constant. Specifically, the constant temperature control unit 15 is a constant temperature water bath, and the collection bottle 111 is placed in the constant temperature control unit 15, so that the temperature of the wound exudate in the collection bottle 111 remains constant, and the temperature difference is negligible.
[0062] The wound exudate concentration detection module 1 further includes a storage unit 16 electrically connected to the controller for storing and storing data on the viscosity of the wound. This design facilitates the data in the wound exudate concentration detection module 1 to be directly retrieved and viewed through the storage unit 16 .
[0063] When used in this application:
[0064] The wound assessment area measurement module 2 uses the camera 21 to take pictures and scan the patient's wound, and can automatically measure the area of the wound and the area under different wound colors; the wound assessment area measurement module 2 uses the electronic nose 22 to determine the wound odor information, and then the wound assessment area measurement module 2 sends the wound area, color and odor report to the equipment management program module 31 in the wound assessment comprehensive management server 3. In addition, the wound exudate concentration detection module 1 also sends the viscosity status of the wound during the detection process to the equipment management program module 31 in the wound assessment comprehensive management server 3. The doctor management program module 32 in the wound assessment comprehensive management server 3 can send the report information of the corresponding patient to the wound assessment intelligent terminal 4 of the corresponding medical staff. This design makes it easy for medical staff to quickly and accurately understand the patient's wound recovery status, and indirectly promote the medical staff's next treatment plan.
[0065] It is obvious to those skilled in the art that the present application is not limited to the details of the exemplary embodiments described above, and that the present application can be implemented in other specific forms without departing from the spirit or essential features of the present application. Therefore, the embodiments should be regarded as exemplary and non-limiting in all respects, and the scope of the present application is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims be included in the present application. Any reference numeral in a claim should not be regarded as limiting the claim to which it relates.
Claims
1. An artificial intelligence-based chronic wound assessment system, characterized in that: include A wound exudate concentration detection module (1) for detecting the viscosity of the wound; A wound assessment area calculation module (2), comprising a camera (21) for scanning a patient's wound by taking a photo and an electronic nose (22) for collecting wound odor information; A wound assessment integrated management server (3), comprising a device management program module (31) electrically connected to the camera (21), the electronic nose (22) and the wound exudate concentration detection module (1), and also comprising a doctor management program module (32) and a patient management program module (33) for recording doctor and patient information respectively; The wound assessment intelligent terminal (4) is communicatively connected to the wound assessment integrated management server (3).
2. The artificial intelligence-based chronic wound assessment system according to claim 1, characterized in that: The patient management program module (33) includes patient wound information, and the patient wound information includes artificial intelligence wound assessment report information.
3. The artificial intelligence-based chronic wound assessment system according to claim 1, characterized in that: The doctor management program module (32) in the wound assessment integrated management server (3) is used to manage hospital information and department information. The hospital information includes a list of hospitals that use the wound assessment area measurement module (2) and the wound exudate concentration detection module (1); and the department information includes a list of departments that use the wound assessment area measurement module (2) and the wound exudate concentration detection module (1).
4. The artificial intelligence-based chronic wound assessment system according to claim 1, characterized in that: The wound assessment area calculation module (2) obtains wound information by performing wound assessment area calculation on the wound image taken by the camera (21) based on the trained feature extraction network and convolutional neural network, and the wound assessment area calculation module (2) generates an artificial intelligence wound assessment report based on the wound information and sends it to the wound assessment integrated management server (3).
5. The artificial intelligence-based chronic wound assessment system according to claim 4, characterized in that: The feature extraction network is a multi-receptive field pyramid network PSPNet or a feature pyramid network FPN.
6. The artificial intelligence-based chronic wound assessment system according to claim 1, characterized in that: The wound assessment intelligent terminal (4) has a report input module for inputting a patient's wound supplementary report.
7. The artificial intelligence-based chronic wound assessment system according to claim 1, characterized in that: The wound exudate concentration detection module (1) comprises A control sampling unit (11) comprises a collection bottle (111) for collecting wound exudate and a sampling cavity (112) connected to the inside of the collection bottle (111), wherein the sampling cavity (112) is connected to a negative pressure wound treatment device (113), and the height of the wound exudate entering the sampling cavity (112) is controlled by the negative pressure wound treatment device (113); A volume monitoring unit (12) for monitoring the liquid level of the wound exudate in the sampling cavity (112); A pressure detection unit (13) for detecting the pressure difference of the liquid surface in the sampling cavity (112); The calculation unit (14) calculates the height difference, pressure difference, and time length of the liquid level change of the wound exudate, and calculates the viscosity of the wound exudate.
8. The artificial intelligence-based chronic wound assessment system according to claim 7, characterized in that: The negative pressure wound therapy device (113) comprises a negative pressure area and a controller, wherein the controller is electrically connected to the volume monitoring unit (12), the pressure detection unit (13) and the calculation unit (14); the sampling cavity (112) is a cavity opened on the inner wall of the collecting bottle (111), connected to the interior of the collecting bottle (111), and connected to the negative pressure area through a pipeline.
9. The artificial intelligence-based chronic wound assessment system according to claim 8, characterized in that: The wound exudate concentration detection module (1) further comprises a constant temperature control unit (15) electrically connected to the controller, wherein the constant temperature control unit (15) is used to keep the temperature of the wound exudate in the collection bottle (111) unchanged.
10. The artificial intelligence-based chronic wound assessment system according to claim 9, characterized in that: The wound exudate concentration detection module (1) also includes a storage unit (16) electrically connected to the controller, for storing data on the viscosity of the wound.