Three-dimensional model construction system and method for wound detection

Abstract

The invention relates to a three-dimensional model construction system and method for wound detection. The system comprises an acquisition module and a processing module, the acquisition module is provided with at least two cameras with different shooting angles so as to acquire images of the wound part of the patient at different shooting angles at the same moment, and the images at different shooting angles reflect the size difference of the same object in the image at different visual angles; the image comprises shooting angle information of a camera acquiring the image; the processing module processes the image by using a modeling unit so as to establish a three-dimensional model of the wound part of the patient; the modeling unit obtains the visual angle difference generated by the image of the wound part of the patient at different shooting angles, and maps the depth information difference of each object in the corresponding image based on the visual angle difference to realize three-dimensional imaging. The three-dimensional model reflects the wound size, blood distribution and the concentration of oxyhemoglobin and deoxyhemoglobin in blood vessels, and evaluation and analysis of characteristic parameters of the wound part of a patient and the healing condition of the wound part are facilitated.

Description

[0001] The original basis of the divisional application is a patent application with the application number 202310360852.0, the application date of April 6, 2023, and the invention name of "a wound detection system, a detection device and a detection method". TECHNICAL FIELD

[0002] The present application relates to the technical field of medical equipment, in particular to a three-dimensional model construction system and method for wound detection. BACKGROUND

[0003] The measurement and evaluation of wounds are the primary link and basis of wound treatment, and are the premise of wound management, playing a crucial role in promoting wound healing. In the process of wound healing, it is necessary to measure and analyze the wound to evaluate the healing of the wound. The existing wound measurement parameters include area, depth, effusion, wound tissue shape, etc. The existing wound detection device usually puts a sterile cotton swab perpendicular to the surface of the wound into the deepest part of the wound, uses a bending forceps or a tweezer to clamp the cotton swab flush with the surface of the wound, takes out the cotton swab, and measures the length of the cotton swab head to the tweezer using a ruler, thereby obtaining the depth of the wound of the patient. This method is simple, but it can easily cause pain to the wounded person, and it can also damage the tissue at the wound site, which is not conducive to wound healing. Accordingly, a non-invasive wound measurement method has emerged, which can avoid damaging the wound tissue while reducing the pain of the patient.

[0004] For example, CN112107331A discloses a medical measurement system, which includes a wound measurement device and a server. The wound measurement device is in communication connection with the server. The wound measurement device includes a wound depth detection subsystem. The wound depth detection subsystem includes an ultrasonic wave generating module for generating ultrasonic waves, an ultrasonic wave receiving module for receiving reflected ultrasonic waves, an ultrasonic wave signal processing module for generating an ultrasonic wave detection image and calculating wound measurement data, and a display module for displaying the ultrasonic wave detection image and the wound measurement data. The server includes a storage module for receiving and storing wound measurement data, a medical record acquisition module for acquiring patient medical records, and a medical record import module for automatically importing detected wound measurement data and time information into a medical record and a preset wound evaluation report.

[0005] CN108814613A discloses an intelligent wound measurement method and a mobile measurement terminal. The method includes: obtaining a patient's identification identifier, obtaining and displaying the corresponding medical history information; acquiring wound image information including a wound image and a reference ruler image; obtaining the wound's length and width based on the relationship between the reference ruler image and the wound image, and obtaining the wound's area based on the identified wound edges; recording the wound's length, width, area, and other input wound and treatment information into the patient's current measurement record; and obtaining a dynamic change map of the patient's wound assessment based on multiple measurement records obtained during multiple measurements. Through this process, effective reference information can be provided for wound treatment and healing monitoring, and the intelligent measurement obtains accurate and comprehensive wound information, enabling dynamic tracking of wound treatment.

[0006] CN113393420A discloses an intelligent wound assessment method based on infrared thermal imaging technology. The method includes: measuring the wound thermal image area based on an irregular area calculation method; establishing an exudate image model based on the relationship between thermal image features and exudate volume; establishing an infrared thermal image model based on the different thermal radiation caused by different wound types; determining the wound's dryness / wetness status based on a mean and variance-based dryness / wetness evaluation method; and outputting the assessment results. This invention has the following advantages and effects: Based on infrared thermal imaging of wounds, this invention can intelligently measure wound area, wound exudate volume, and tissue type, and automatically output the individual scores and total scores for these three items, thus achieving a more objective, accurate, and rapid intelligent assessment.

[0007] Although existing technologies can measure wounds by acquiring ultrasound images, infrared thermal images, etc., the shooting angle of each wound image acquisition is different. This introduces angular errors when comparing the current image of the wound with historical images to assess the wound healing, thus reducing the accuracy of the assessment.

[0008] Furthermore, on the one hand, there are differences in understanding among those skilled in the art; on the other hand, the applicant studied a large number of documents and patents when making this invention, but due to space limitations, not all details and contents were listed in detail. However, this does not mean that the present invention does not possess the features of these prior art. On the contrary, the present invention already possesses all the features of the prior art, and the applicant reserves the right to add relevant prior art to the background art. Summary of the Invention

[0009] In view of the deficiencies of the prior art, the present application provides a wound detection system. The wound detection system comprises at least a collection module and a processing module. The collection module is configured to collect images of a wound site of a patient. The processing module is configured to process the images collected by the collection module. Preferably, the processing of the images by the processing module comprises at least: processing the images by a modeling unit to establish a three-dimensional model of the wound site of the patient; and analyzing the three-dimensional model by an analysis unit to obtain characteristic parameters of the wound site of the patient and / or evaluation information of the healing condition of the wound site of the patient.

[0010] Preferably, the wound detection system can determine the wound changes by establishing the three-dimensional model corresponding to the wound site of the patient and the analysis of the three-dimensional model, and facilitate the monitoring of the healing condition of the wound. The wound detection system can also provide effective reference information for the wound treatment and healing monitoring, and obtain accurate and comprehensive wound information, and dynamically track the wound treatment.

[0011] According to a preferred embodiment, the collection module is provided with at least two cameras having different shooting angles to obtain images of the wound site of the patient at different shooting angles at the same time, and transmit the images at different shooting angles to the processing module.

[0012] Preferably, the processing module can obtain the perspective difference of the images of the wound site of the patient at different shooting angles, and map the depth information difference of each object in the corresponding images based on the perspective difference, to realize three-dimensional imaging.

[0013] According to a preferred embodiment, the modeling unit of the processing module establishes the three-dimensional model of the wound site of the patient based on the images at different shooting angles, and transmits the three-dimensional model to the analysis unit of the processing module to analyze the characteristic parameters of the wound site of the patient and / or the healing condition of the wound site of the patient.

[0014] According to a preferred embodiment, the analysis unit uses a first analysis model trained by deep learning using actual patient wound images as samples to perform image analysis on the three-dimensional model to obtain the characteristic parameters of the wound site of the patient. Preferably, the characteristic parameters include at least one of wound depth, area, and tissue shape.

[0015] Preferably, the analysis unit inputs the three-dimensional model into the first analysis model to obtain the characteristic parameters of the wound site of the patient. Preferably, the first analysis model identifies the visual-based characteristic parameters according to the input three-dimensional model, so that medical personnel can determine the area and depth of the wound, determine whether there is decayed or loose tissue, and determine whether there is a cavity and pus accumulation.

[0016] According to a preferred embodiment, the analysis unit compares the three-dimensional model with historical medical data through a second analysis model to obtain evaluation information of the healing condition of the wound site of the patient. The second analysis model evaluates the healing condition of the wound of the patient by comparing the three-dimensional model with the historical medical data. Preferably, the historical medical data at least includes historical three-dimensional models.

[0017] Preferably, the present application evaluates the healing condition of the wound of the patient through the second analysis model, thereby converting the subjective judgment of medical staff into objective judgment, thereby eliminating the interference of human experience factors on the evaluation result and improving the efficiency and effect of the evaluation of the healing condition of the wound.

[0018] According to a preferred embodiment, the second analysis model compares the three-dimensional model with the historical three-dimensional model in the same viewing direction.

[0019] Preferably, after the second analysis model compares the three-dimensional model with the historical three-dimensional model in the same viewing direction, the healing condition of the wound site of the patient can be judged by comparing the changes of the same parts in the three-dimensional model.

[0020] According to a preferred embodiment, the processing module is further configured with a storage unit for storing historical medical data. Preferably, the analysis unit stores the three-dimensional model as historical medical data in the storage unit after completing the analysis.

[0021] Preferably, the storage unit can optimize the stored historical medical data. The specific optimization method can be to fuse similar cases, thereby reducing the space required for storing the historical medical data while ensuring the diversity of cases in the historical medical data.

[0022] The present application also provides a wound detection method. The wound detection method at least includes:

[0023] Collecting images of the wound of the patient;

[0024] Establishing a three-dimensional model of the wound site of the patient according to the images;

[0025] Analyzing the three-dimensional model through a first analysis model to obtain characteristic parameters of the wound site of the patient.

[0026] According to a preferred embodiment, the wound detection method further includes: analyzing the three-dimensional model through a second analysis model to obtain evaluation information of the healing condition of the wound site of the patient. Preferably, the second analysis model evaluates the healing condition of the wound of the patient by comparing the three-dimensional model with historical medical data. Preferably, the historical medical data at least includes historical three-dimensional models.

[0027] The application also provides a wound detection device. The wound detection device comprises at least a collection module, a processing module and a display module. The collection module is provided with at least two cameras with different shooting angles to obtain images of a wound site of a patient at different shooting angles at the same time, and transmit the images at different shooting angles to the processing module. The processing module processes the images and transmits the processing results to the display module for display. Preferably, the processing of the images by the processing module at least comprises establishing a three-dimensional model of the wound site of the patient according to the images and analyzing the three-dimensional model to obtain characteristic parameters of the wound site of the patient and / or evaluate the healing condition of the wound site of the patient. BRIEF DESCRIPTION OF DRAWINGS

[0028] Fig. 1 is a simplified schematic diagram of a wound detection system according to a preferred embodiment of the application;

[0029] Fig. 2 is a simplified module connection relationship diagram of a collection module according to a preferred embodiment of the application;

[0030] Fig. 3 is a simplified communication connection relationship diagram of a wound detection system according to a preferred embodiment of the application.

[0031] LIST OF REFERENCE NUMBERS

[0032] 100: wound detection system; 110: collection module; 111: camera; 112: mounting shell; 120: processing module; 121: modeling unit; 122: analysis unit; 123: storage unit; 130: display module. DETAILED DESCRIPTION

[0033] The application will be described in detail below with reference to the accompanying drawings. Figs. 1 to 3

[0034] Example 1

[0035] This embodiment provides a wound detection system 100. Referring to Fig. 1 , preferably, the wound detection system 100 can comprise a collection module 110, a processing module 120 and a display module 130. Preferably, the collection module 110 is used to collect images of a wound site of a patient. Preferably, the processing module 120 is used to process the images collected by the collection module 110. Preferably, the display module 130 is used to display the processing results of the processing module 120.

[0036] Preferably, the processing module 120 can be connected with the collection module 110 and the display module 130 in a wired or wireless manner.

[0037] ​Preferably, the acquisition module 110 can acquire images of the wound site of the patient from two or more shooting angles. Preferably, the processing module 120 processes the images acquired by the acquisition module 110 to establish a three-dimensional model corresponding to the wound site of the patient, and the processing module 120 can also analyze the three-dimensional model to obtain characteristic parameters of the wound site of the patient and / or evaluation information of the healing of the wound site of the patient.

[0038] Preferably, the acquisition module 110 can include two or more cameras 111 with different shooting angles. Preferably, the acquisition module 110 can acquire images of the wound site of the patient at different shooting angles at the same time through the cameras 111 with different shooting angles.

[0039] Referring to Fig. 2 Preferably, the acquisition module 110 of the embodiment can be provided with three cameras 111. Preferably, the three cameras 111 are provided on the arc-shaped curved mounting shell 112 in a spaced manner. Preferably, the three cameras 111 include a first camera provided at the center of the curved top and a second camera and a third camera provided around the first camera. Preferably, the second camera has a first distance from the first camera, and the third camera has a second distance from the first camera, wherein the second distance is greater than the first distance. Referring to Fig. 2 Preferably, the first camera is provided at the center of the arc-shaped curved surface, the second camera has an angle of 15° with the first camera about the center of the curved surface on the arc-shaped curved surface, and the third camera has an angle of 30° with the first camera about the center of the curved surface on the arc-shaped curved surface.

[0040] Preferably, the acquisition module 110 can be a handheld acquisition probe provided with three cameras 111 with non-equidistant intervals. Preferably, the three cameras 111 on the acquisition module 110 can acquire three images with different and non-mirror shooting angles.

[0041] Preferably, the acquisition module 110 of the embodiment can acquire images of the wound site of the patient at three shooting angles at the same time when acquiring images of the wound site of the patient at any shooting angle, so that the processing module 120 can establish a three-dimensional model corresponding to the wound site of the patient according to the images of the wound site of the patient at three shooting angles at the same time.

[0042] Preferably, the images acquired by the acquisition module 110 at three shooting angles can reflect the size differences of the same object in the image from different perspectives. An image acquired from a single shooting angle can only capture two-dimensional information of the patient's wound. By acquiring images from three shooting angles, the acquisition module 110 makes the patient's wound appear at different sizes in the three images. Specifically, the size of the patient's wound in the image changes with the shooting angle. Since the three shooting angles are known, the processing module 120 can deduce the three-dimensional size of the patient's wound from the size in the different images, thereby obtaining the three-dimensional information of the patient's wound. This allows the processing module 120 to build a three-dimensional model corresponding to the patient's wound and ensures the accuracy of the three-dimensional model.

[0043] See Fig. 3 Preferably, the processing module 120 may include a modeling unit 121, an analysis unit 122, and a storage unit 123. Preferably, the analysis unit 122 is data-connected to both the modeling unit 121 and the storage unit 123. Preferably, the processing module 120 may be, for example, a logic gate array, a controller and arithmetic logic unit, a digital signal processor, a microcomputer, a programmable logic controller, a field-programmable gate array, a programmable logic array, a microprocessor, or any other means or combination of means configured to respond to and execute instructions in a defined manner to achieve a desired result.

[0044] Preferably, the processing module 120's processing of the images acquired by the acquisition module 110 may include: the processing module 120 using the modeling unit 121 to process the images to establish a three-dimensional model of the patient's wound site; and the processing module 120 using the analysis unit 122 to analyze the three-dimensional model to obtain characteristic parameters of the patient's wound site and / or assessment information on the healing status of the patient's wound site. Preferably, the storage unit 123 is used to store historical medical data.

[0045] Preferably, the modeling unit 121 is data-connected to the acquisition module 110 to receive images of the patient's wound site acquired by the acquisition module 110. Preferably, the modeling unit 121 receives the images of the patient's wound site acquired by the acquisition module 110 and establishes a three-dimensional model of the patient's wound site based on images acquired by the acquisition module 110 from different shooting angles. Preferably, the modeling unit 121 transmits the established three-dimensional model to the analysis unit 122 configured in the processing module 120 for analysis.

[0046] Preferably, the modeling unit 121 can obtain the depth information of each object in the image by scanning the images of the patient's wound acquired by the acquisition module 110 from different shooting angles. Preferably, the modeling unit 121 can obtain the depth information of each object in the image based on the viewing angle differences caused by the different shooting angles of each camera 111, thereby realizing three-dimensional imaging. Preferably, the multiple images transmitted from the acquisition module 110 to the modeling unit 121 include the shooting angle information of the camera 111 that acquired the image. Since the shooting angles of each camera 111 are different, the images acquired by different cameras 111 contain the depth information differences of each object in the image caused by the viewing angle differences. Therefore, the modeling unit 121 can obtain the viewing angle differences generated by the patient's wound image at different shooting angles, and map the depth information differences of each object in the image based on the viewing angle differences to achieve three-dimensional imaging.

[0047] Preferably, the modeling unit 121 can identify the size of the patient's wound from an image taken from a single shooting angle, thereby obtaining the size of the patient's wound mapped at three shooting angles. Since the three shooting angles are known, the modeling unit 121 can inversely deduce the actual size of the patient's wound by obtaining the size mapped at the three shooting angles, thereby obtaining the three-dimensional information of the patient's wound.

[0048] Preferably, the analysis unit 122 uses a deep learning-based analysis model to analyze the three-dimensional model. Preferably, the analysis unit 122 may be configured with a first analysis model for extracting feature parameters of the patient's wound site and a second analysis model for assessing the patient's wound healing status.

[0049] Preferably, the first analysis model and the second analysis model can be neural network models trained by deep learning using a large number of actual patient wound images as samples.

[0050] Preferably, the analysis unit 122 performs image-based segmentation of the three-dimensional model using the first analysis model to obtain characteristic parameters of the patient's wound site. Preferably, the characteristic parameters include at least one or a combination of wound depth, area, and tissue shape.

[0051] Preferably, the analysis unit 122 inputs the three-dimensional model into the first analysis model to obtain characteristic parameters of the patient's wound site. Preferably, the first analysis model identifies vision-based characteristic parameters based on the input three-dimensional model, enabling medical personnel to determine the area and depth of the wound, whether there is necrotic or loose tissue, and whether there are cavities and pus accumulation.

[0052] Because the human tissue at the patient's wound site has distinct image characteristics compared to the patient's normal human tissue, and because necrotic or loose tissue, cavities, and pus accumulation in the wound exhibit image characteristics different from healthy tissue, the first analysis model can preferably be a neural network model trained using actual patient wound images as samples through deep learning. Preferably, the sample images used during training of the first analysis model can include images of sinuses, necrotic tissue, pus accumulation, wound edges, and normal human tissue annotated by medical personnel. Preferably, when analyzing the three-dimensional model, the first analysis model can slice the three-dimensional model along any direction, thereby converting the three-dimensional model into a stack of images. Preferably, the first analysis model can identify the sliced ​​images, distinguishing between injured and normal tissue, thereby determining the area and depth of the wound. Further, the first analysis model identifies the images of the injured tissue to determine whether it contains pathological features such as necrotic or loose tissue, cavities, and pus accumulation.

[0053] Preferably, the analysis unit 122 analyzes the three-dimensional model using a second analysis model to obtain assessment information on the healing status of the patient's wound. The second analysis model assesses the wound healing status by comparing the three-dimensional model with historical medical data. Preferably, the historical medical data includes at least historical three-dimensional models. Preferably, the historical medical data may include the three-dimensional model of the patient's previous wound examination and three-dimensional models of wounds from other patients with similar wounds.

[0054] Preferably, the analysis unit 122 is data-connected to the display module 130, so that the analysis unit 122 can transmit the analysis results of the first analysis model and the second analysis model and the three-dimensional model to the display module 130 for display.

[0055] Preferably, after completing the analysis, the analysis unit 122 stores the three-dimensional model as historical medical data in the storage unit 123.

[0056] Preferably, the storage unit 123 can optimize the stored historical medical data. Specifically, the optimization method can be to merge similar cases, thereby reducing the space required to store the historical medical data while ensuring the diversity of cases in the historical medical data.

[0057] Preferably, the wound detection system 100 provided in this embodiment can be used to measure the characteristic parameters of a patient's wound and to assess the healing status of the patient's wound.

[0058] Preferably, when the trauma detection system 100 provided in this embodiment is used to measure the characteristic parameters of a patient's wound, medical personnel can acquire images of the patient's wound site from any shooting angle using the acquisition module 110. Preferably, the three cameras 111 on the acquisition module 110 can acquire images of the patient's wound site at three shooting angles at the same time. Preferably, the cameras 111 transmit the acquired images to the modeling unit 121 of the processing module 120 to establish a three-dimensional model. Preferably, the trauma detection system 100 can acquire near-infrared images of the patient's wound site using the near-infrared imaging principle. Preferably, the cameras 111 can acquire optical image information of the patient's wound site under near-infrared light illumination. Preferably, the cameras 111 can acquire the distribution and direction of blood vessels in the patient's wound by acquiring near-infrared images of the patient's wound site. Because hemoglobin in the blood can absorb near-infrared light, the near-infrared light reflected by hemoglobin to the camera 111 is attenuated.

[0059] Preferably, when the modeling unit 121 in the processing module 120 establishes a three-dimensional model, it can model the blood distribution in the three-dimensional model based on the near-infrared image collected by the camera 111.

[0060] Preferably, the first analysis model of the analysis unit 122 can determine whether there is internal bleeding in the wound by detecting the distribution of blood in the patient's wound. Preferably, when there is internal bleeding in the patient's wound, the first analysis model of the analysis unit 122 can detect the presence of blood distribution outside the patient's blood vessels. Preferably, the trauma detection system 100 can detect bleeding points inside the wound before bleeding occurs, facilitating timely treatment of the patient's wound bleeding by medical personnel.

[0061] As the wound heals, oxygen metabolism in the injured area increases, oxygen in the blood vessels is consumed, and hemoglobin changes. Specifically, oxygen metabolism increases in the wound area, and blood oxygen is consumed; the concentration of oxyhemoglobin in hemoglobin decreases, while the concentration of deoxyhemoglobin increases.

[0062] Camera 111 can acquire near-infrared images that characterize the levels of oxyhemoglobin and deoxyhemoglobin in the patient's wound site.

[0063] Preferably, when the modeling unit 121 in the processing module 120 builds a 3D model, it can render the blood vessels in the 3D model based on the near-infrared image acquired by the camera 111. The depth of the blood vessel color reflects the concentration of oxyhemoglobin and deoxyhemoglobin. Preferably, when the second analysis model of the analysis unit 122 compares the 3D model with the 3D model in historical medical data, it can determine the concentration changes of oxyhemoglobin and deoxyhemoglobin by comparing the changes in blood vessel color, thereby determining the healing status of the wound site.

[0064] Preferably, when the trauma detection system 100 provided in this embodiment is used to measure the characteristic parameters of a patient's wound, medical personnel can acquire images of the patient's wound site from any shooting angle using the acquisition module 110. Preferably, the three cameras 111 on the acquisition module 110 can acquire images of the patient's wound site at three shooting angles at the same time. Preferably, the cameras 111 transmit the acquired images to the modeling unit 121 of the processing module 120 to establish a three-dimensional model. Preferably, the three-dimensional model includes the distribution of blood vessels. Preferably, the first analysis model of the analysis unit 122 identifies the three-dimensional model, making it easier for medical personnel to clearly see the internal condition of the sinus tract wound, and also making it easier for medical personnel to more accurately measure the depth, width, and size of erosion grooves or holes on the sidewall of the sinus tract wound.

[0065] The present invention provides a wound detection system 100 that, in addition to measuring the characteristic parameters of the wound, can also assess the wound healing status of the patient.

[0066] In current technologies, the assessment of a patient's wound healing is mostly based on judgments made by medical personnel according to measurement parameters. The accuracy of the judgment results depends primarily on the accuracy of the measurement parameters and the experience of the medical personnel.

[0067] Preferably, the present invention acquires images of the patient's wound from three shooting angles through the acquisition module 110 and models them, thereby ensuring the accuracy of the three-dimensional model.

[0068] Preferably, the analysis unit 122 obtains the measurement parameters through the first analysis model. The first analysis model, trained by deep learning, identifies the three-dimensional model modeled from the fused multi-angle images, thereby ensuring the accuracy of the measurement parameters.

[0069] Preferably, the present invention also uses a second analytical model to assess the healing status of the patient's wound, thereby converting the subjective judgment of medical personnel into an objective judgment, thus eliminating the interference of human experience factors on the assessment results and improving the efficiency and effectiveness of wound healing assessment.

[0070] For patients with chronic wounds, after applying medication to the wound, medical staff need to conduct irregular examinations of the wound to monitor its healing process and prevent adverse conditions such as infection.

[0071] Preferably, when medical personnel are caring for a patient's wound, they can use the wound detection system 100 provided by the present invention to detect the patient's wound and assess the wound healing status.

[0072] Preferably, before treating the patient's wound, medical personnel can use the wound detection system 100 to obtain measurement parameters of the patient's wound site, thereby providing data support for medical personnel to perform wound cleaning, suturing, medication application, and other treatments.

[0073] Preferably, before treating the patient's wound, medical personnel can acquire images of the wound site from three different shooting angles at the same time using the acquisition module 110. The modeling unit 121 builds a three-dimensional model based on the images acquired by the acquisition module 110. The first analysis model in the analysis unit 122 analyzes the three-dimensional model to obtain the feature parameters of the patient's wound site.

[0074] Preferably, for patients requiring full-process care, medical personnel can obtain a three-dimensional model of the wound site after debridement, suturing, and medication application using the trauma detection system 100, and store this three-dimensional model in the storage unit 123. Preferably, the analysis unit 122 of the trauma detection system 100 can compare this three-dimensional model with historical medical data using a second analysis model, and select historical cases that are closest to the three-dimensional model from the historical medical data. Preferably, after selecting historical cases, the analysis unit 122 can set the historical cases as reference cases to set detection strategies. Preferably, the examination plan includes at least the time interval between the current examination and the next examination. Preferably, in the detection strategy set by the analysis unit 122, the time interval between each examination is the same as the time interval between each examination in the reference case. For example, the analysis unit 122 can set the time interval between each examination in the detection strategy to 2 days based on the reference case.

[0075] Preferably, the analysis unit 122 transmits the detection strategy to the display module 130 for display to medical personnel. Preferably, medical personnel can detect the patient's wound site according to the detection strategy set by the trauma detection system 100.

[0076] Preferably, when medical personnel examine the patient's wound site according to the time interval of the detection strategy, the analysis unit 122 can compare the three-dimensional model obtained in this detection with the three-dimensional model of the patient in the previous detection through the second analysis model, thereby assessing the patient's wound healing status.

[0077] Preferably, during the detection process, medical personnel can acquire images of the patient's wound site from any shooting angle using the acquisition module 110. The modeling unit 121 of the processing module 120 can build a three-dimensional model based on the images acquired by the acquisition module 110. Preferably, when comparing the three-dimensional model acquired in this detection with the three-dimensional model from the patient's previous detection, the second analysis model of the analysis unit 122 can adjust the three-dimensional model and the historical three-dimensional model to the same viewing angle for comparison, thereby determining the healing status of the patient's wound site.

[0078] Because the image capturing angle of the acquisition module 110 may change during each test due to variations in the patient's posture or the medical staff's shooting angle, the viewing angle of the 3D model created by the modeling unit 121 will differ. Preferably, the 3D model established by the modeling unit 121 includes vascular information. Since the distribution of blood vessels in human tissue is relatively fixed, the second analysis model can use the vascular positions in the 3D model as a reference frame. By aligning the blood vessels in the 3D model created in the current test with those in the 3D model created in the previous test, the viewing direction of the 3D model is adjusted for comparison.

[0079] Preferably, the second analytical model compares the three-dimensional model with historical three-dimensional models to assess the healing status of the patient's wound. Preferably, the assessment of the patient's wound healing status by the second analytical model includes determining whether new adverse healing conditions have occurred, such as subcutaneous abscesses, cavities, tissue necrosis, or bleeding; and determining changes in characteristic parameters of the wound, such as a decrease or increase in wound depth or area.

[0080] Preferably, the second analysis model compares the 3D model with the historical 3D model to assess whether the patient's wound is deteriorating or healing well. Preferably, when the second analysis model's assessment result indicates wound deterioration, such as the formation of a subcutaneous abscess or cavity, or partial tissue necrosis, the analysis unit 122 sends the assessment result to the display module 130, alerting medical personnel to treat the patient's wound. Preferably, after treating the patient's wound, medical personnel can use the trauma detection system 100 to inspect the treated wound and formulate a new detection strategy.

[0081] Preferably, when the evaluation result of the second analysis model indicates good healing, i.e., the wound area, wound depth, and other parameters have decreased, and no adverse healing conditions such as abscess, cavity, or sinus tract have occurred, the analysis unit 122 adjusts the detection strategy and sends the adjusted detection strategy and evaluation result to the display module 130 for display to medical personnel. Preferably, when the evaluation result of the second analysis model indicates good healing, the analysis unit 122 can compare the evaluation result of the second analysis model with the healing condition of a reference case to adjust the detection strategy. Preferably, when the evaluation result of the second analysis model indicates good healing, but the patient's wound healing condition is worse than that of the reference case, i.e., the reduction in parameters such as the wound area and wound depth is smaller than that of the reference case, indicating that the patient's wound healing speed is slower, the analysis unit 122 can extend the time interval between each detection in the detection strategy. Preferably, the analysis unit 122 can extend the time interval between each detection in the detection strategy to 3 days. Preferably, when the evaluation result of the second analysis model indicates good healing and the reduction in parameters such as wound area and wound depth is greater than that of the reference case, it indicates that the patient's wound is healing faster, and the analysis unit 122 can shorten the time interval between each test in the detection strategy. Preferably, the analysis unit 122 can shorten the time interval between each test in the detection strategy to 1 day.

[0082] Preferably, the trauma detection system 100 can compare the three-dimensional model with the historical three-dimensional model through the second analysis model to assess the healing status of the patient's wound. Furthermore, the analysis unit 122 can adjust the time interval of each detection in the detection strategy according to the assessment results, thereby rationally allocating medical resources based on the patient's wound healing status.

[0083] Example 2

[0084] This embodiment is a further improvement on embodiment 1, and repeated content will not be described again.

[0085] The present invention also provides a wound detection device. The wound detection device includes at least: an acquisition module 110, a processing module 120, and a display module 130. The acquisition module 110 is equipped with at least two cameras 111 with different shooting angles to acquire images of the patient's wound site at different shooting angles at the same time, and transmits the images from different shooting angles to the processing module 120. The processing module 120 processes the images and transmits the processing results to the display module 130 for display. Preferably, the image processing by the processing module 120 includes at least: establishing a three-dimensional model of the patient's wound site based on the images and analyzing the three-dimensional model to obtain characteristic parameters of the patient's wound site and / or assessing the healing status of the patient's wound site.

[0086] Preferably, the acquisition module 110 of this embodiment can be equipped with three cameras 111. Preferably, the three cameras 111 are arranged at intervals on the arc-shaped mounting housing 112. Preferably, the three cameras 111 include a first camera disposed at the center of the top of the arc surface and a second camera and a third camera disposed around the first camera. Preferably, the distance between the second camera and the first camera is a first distance, and the distance between the third camera and the first camera is a second distance, wherein the second distance is greater than the first distance.

[0087] Preferably, the processing module 120 may include a modeling unit 121, an analysis unit 122, and a storage unit 123. Preferably, the analysis unit 122 is data-connected to both the modeling unit 121 and the storage unit 123.

[0088] Preferably, the modeling unit 121 can obtain the depth information of each object in the image by scanning the images of the patient's wound acquired by the acquisition module 110 from different shooting angles. Preferably, the modeling unit 121 can obtain the depth information of each object in the image based on the viewing angle differences caused by the different shooting angles of each camera 111, thereby realizing three-dimensional imaging. Preferably, the multiple images transmitted from the acquisition module 110 to the modeling unit 121 include the shooting angle information of the camera 111 that acquired the image. Since the shooting angles of each camera 111 are different, the images acquired by different cameras 111 contain the depth information differences of each object in the image caused by the viewing angle differences. Therefore, the modeling unit 121 can obtain the viewing angle differences generated by the patient's wound image at different shooting angles, and map the depth information differences of each object in the image based on the viewing angle differences to achieve three-dimensional imaging.

[0089] Preferably, the analysis unit 122 uses a deep learning-based analysis model to analyze the three-dimensional model. Preferably, the analysis unit 122 may be configured with a first analysis model for extracting feature parameters of the patient's wound site and a second analysis model for assessing the patient's wound healing status.

[0090] Preferably, the analysis unit 122 inputs the three-dimensional model into the first analysis model to obtain characteristic parameters of the patient's wound site. Preferably, the first analysis model identifies vision-based characteristic parameters based on the input three-dimensional model, enabling medical personnel to determine the area and depth of the wound, whether there is necrotic or loose tissue, and whether there are cavities and pus accumulation.

[0091] Preferably, when the second analysis model of the analysis unit 122 compares the three-dimensional model with the three-dimensional model in historical medical data, it can determine the concentration changes of oxyhemoglobin and deoxyhemoglobin by comparing the changes in blood vessel color, and thus determine the healing status of the wound site.

[0092] Example 3

[0093] This embodiment is a further improvement on Embodiments 1 and 2, and the repeated content will not be described again.

[0094] The present invention also provides a trauma detection method. The trauma detection method includes at least:

[0095] Acquire images of the patient's wound site;

[0096] A three-dimensional model of the patient's wound site is created based on the images;

[0097] The first analysis model is used to analyze the three-dimensional model to obtain the characteristic parameters of the patient's wound site.

[0098] According to a preferred embodiment, the trauma detection method further includes: analyzing the three-dimensional model using a second analysis model to obtain assessment information on the healing status of the patient's wound. Preferably, the second analysis model assesses the healing status of the patient's wound by comparing the three-dimensional model with historical medical data. Preferably, the historical medical data includes at least historical three-dimensional models.

[0099] Preferably, in this embodiment, images of the patient's wound site can be acquired from two or more shooting angles.

[0100] Preferably, when building a 3D model of the patient's wound site based on images of the wound site, this embodiment can obtain the depth information of each object in the image based on the perspective differences caused by different shooting angles, thereby achieving 3D imaging. Since the shooting angles are different, the images used for modeling contain the depth information differences of each object in the image caused by the perspective differences. Therefore, during modeling, the perspective differences generated by the patient's wound site image at different shooting angles can be obtained, and the depth information differences of each object in the image can be mapped based on the perspective differences to achieve 3D imaging.

[0101] Preferably, the present invention utilizes a deep learning-based analysis model to analyze the three-dimensional model. Preferably, the analysis model used in the present invention for analyzing the three-dimensional model includes a first analysis model for extracting feature parameters of the patient's wound site and a second analysis model for assessing the patient's wound healing status.

[0102] Preferably, the first analysis model identifies visual feature parameters based on the input 3D model, enabling medical personnel to determine the area and depth of the wound, whether there is necrotic or loose tissue, and whether there are cavities or pus accumulation. Preferably, the second analysis model can adjust the 3D model to the same viewpoint for comparison, thereby assessing the healing status of the patient's wound.

[0103] It should be noted that the specific embodiments described above are exemplary. Those skilled in the art can devise various solutions inspired by the disclosure of this invention, and these solutions all fall within the scope of this invention and its protection. Those skilled in the art should understand that this specification and its accompanying drawings are illustrative and do not constitute a limitation on the claims. The scope of protection of this invention is defined by the claims and their equivalents. Throughout the text, features introduced by "preferred" are merely optional and should not be construed as mandatory. Therefore, the applicant reserves the right to abandon or delete relevant preferred features at any time. This specification contains multiple inventive concepts. Phrases such as "preferred," "according to a preferred embodiment," or "optionally" indicate that the corresponding paragraph discloses an independent concept. The applicant reserves the right to file divisional applications based on each inventive concept.

Claims

1. A three-dimensional model construction system for trauma detection, characterized in that, The system includes: The acquisition module (110) is equipped with at least two cameras (111) with different shooting angles to acquire images of the patient's wound at the same time from different shooting angles. The images from different shooting angles reflect the size difference of the same object in the image from different perspectives. The image contains the shooting angle information of the camera (111) that acquires the image. The processing module (120) uses the modeling unit (121) to process the image to establish a three-dimensional model of the patient's wound site; The modeling unit (121) acquires the perspective differences generated by the patient's wound site image under different shooting angles, and maps the depth information differences of each object in the corresponding image based on the perspective differences to achieve three-dimensional imaging.

2. The system according to claim 1, characterized in that, The three cameras (111) on the acquisition module (110) acquire images of the patient's wound at three shooting angles at the same time. The modeling unit (121) identifies the size of the patient's wound from an image taken from a single shooting angle, thereby obtaining the size of the patient's wound mapped from three shooting angles; The modeling unit (121) obtains the actual size of the patient's wound by back-calculating the size of the wound at three shooting angles, thereby obtaining the three-dimensional information of the patient's wound.

3. The system according to claim 1 or 2, characterized in that, The camera (111) acquires the distribution and direction of blood vessels in the patient's wound by collecting near-infrared images of the wound site; Since hemoglobin in the blood can absorb near-infrared light, the near-infrared light reflected by the hemoglobin to the camera (111) is attenuated. The modeling unit (121) models the blood distribution in a three-dimensional model based on the near-infrared image collected by the camera (111).

4. The system according to any one of claims 1 to 3, characterized in that, The camera (111) acquires near-infrared images characterizing the content of oxyhemoglobin and deoxyhemoglobin in the patient's wound site; When building a three-dimensional model, the modeling unit (121) renders the blood vessels in the three-dimensional model based on the near-infrared image acquired by the camera (111); The color of blood vessels reflects the concentration of oxyhemoglobin and deoxyhemoglobin.

5. The system according to any one of claims 1 to 4, characterized in that, The three cameras (111) are able to capture three non-mirror images from different shooting angles; Three cameras (111) are arranged at intervals on an arc-shaped mounting housing (112); The three cameras (111) include a first camera positioned at the center of the top of the curved surface and a second and a third camera positioned around the first camera; The distance between the second camera and the first camera is called the first distance, and the distance between the third camera and the first camera is called the second distance, wherein the second distance is greater than the first distance.

6. A method for constructing a three-dimensional model for trauma detection, characterized in that, The method includes: The acquisition module (110) is equipped with at least two cameras (111) with different shooting angles to acquire images of the patient's wound at different shooting angles at the same time. The images with different shooting angles reflect the size difference of the same object in the image from different perspectives. The image contains the shooting angle information of the camera (111) that acquires the image. The acquisition module (110) sends the image to the processing module (120). The modeling unit (121) in the processing module (120) receives images of the patient's wound site acquired by the acquisition module (110) and establishes a three-dimensional model of the patient's wound site based on images acquired by the acquisition module (110) from different shooting angles. The modeling unit (121) acquires the perspective differences generated by the patient's wound site image under different shooting angles, and maps the depth information differences of each object in the corresponding image based on the perspective differences to achieve three-dimensional imaging.

7. The method according to claim 6, characterized in that, The method further includes: The three cameras (111) on the acquisition module (110) acquire images of the patient's wound at three shooting angles at the same time. The modeling unit (121) identifies the size of the patient's wound from an image taken from a single shooting angle, thereby obtaining the size of the patient's wound mapped at three shooting angles. By obtaining the size of the patient's wound mapped at three shooting angles, the actual size of the patient's wound is derived, thereby obtaining the three-dimensional information of the patient's wound.

8. The method according to claim 6 or 7, characterized in that, The method further includes: The camera (111) acquires the distribution and direction of blood vessels in the patient's wound by collecting near-infrared images of the wound site; Since hemoglobin in the blood can absorb near-infrared light, the near-infrared light reflected by the hemoglobin to the camera (111) is attenuated. The modeling unit (121) models the blood distribution in a three-dimensional model based on the near-infrared image collected by the camera (111).

9. The method according to any one of claims 6 to 8, characterized in that, The method further includes: the camera (111) acquiring near-infrared images characterizing the content of oxyhemoglobin and deoxyhemoglobin in the patient's wound site; When building a three-dimensional model, the modeling unit (121) renders the blood vessels in the three-dimensional model based on the near-infrared image acquired by the camera (111); The color of blood vessels reflects the concentration of oxyhemoglobin and deoxyhemoglobin.

10. The method according to any one of claims 6 to 9, characterized in that, The method further includes: setting three cameras (111) at intervals on an arc-shaped mounting housing (112) so that the three cameras (111) can acquire three non-mirror images with different shooting angles; The three cameras (111) include a first camera positioned at the center of the top of the curved surface and a second and a third camera positioned around the first camera; The distance between the second camera and the first camera is called the first distance, and the distance between the third camera and the first camera is called the second distance, wherein the second distance is greater than the first distance.

Images