Recognition and detection method and system for orthopedic instrument

Abstract

The invention discloses an orthopedic instrument identification and detection method and system, and the method comprises the steps: obtaining image information which comprises an initial image and a final image; the category of the image information is recognized, a corresponding recognition and detection model is distributed to the image information based on the category, and the recognition and detection model can be used for recognizing and obtaining target features in the image information; and establishing a data set A including the target features of the initial image and a data set B including the target features of the final image, comparing the data set A with the data set B and generating a difference set C, and generating use data of the orthopedic instrument based on the difference set C. The problems that in the recognition process of an existing orthopaedic implantation instrument, the instrument recognition effect is poor, and used specific orthopaedic instrument data are difficult to accurately obtain can be effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of identification and detection of orthopedic machinery, and in particular to an identification and detection method and system for orthopedic equipment. Background Art

[0002] Orthopedic implants require high temperature and high pressure sterilization before surgery. Since orthopedic instruments and equipment are in an unpacked and uncoded (naked) state, it is impossible to accurately identify, accept, manage and trace the instruments, which brings greater management and medical risks.

[0003] At present, the identification and management of existing orthopedic instruments mainly rely on electronic magnifiers and manual reading, which are difficult to identify. At the same time, the identification of instruments through attached labels or coding has problems such as low efficiency, inaccurate identification, and misidentification. In addition, it is difficult to accurately control the actual usage of existing orthopedic instruments during actual use. Summary of the invention

[0004] In view of the above shortcomings of the prior art, the purpose of the present invention is to disclose an orthopedic instrument identification and detection method and system to improve the problem that the existing orthopedic implant instruments have poor instrument identification effect and it is difficult to accurately obtain the specific orthopedic instrument data used during the identification process.

[0005] To achieve the above-mentioned purpose and other related purposes, the present invention discloses an identification and detection method of orthopedic instruments, which comprises:

[0006] Acquiring image information, wherein the image information includes an initial image and a final image;

[0007] Identify the category of the image information, and assign a corresponding recognition detection model to the image information based on the category, wherein the recognition detection model can be used to identify and obtain target features in the image information;

[0008] Establishing a data set A including target features of the initial image and a data set B including target features of the final image,

[0009] The data set A and the data set B are compared to generate a difference set C, and based on the difference set C, usage data of the orthopedic instrument is generated.

[0010] In one embodiment of the present invention, the recognition and detection model includes a bone nail recognition model, a bone plate recognition model and a surgical tool recognition model, and the recognition and detection model adopts a convolutional neural network model architecture;

[0011] The categories of the image information include bone nail image categories, bone plate image categories and surgical tool categories;

[0012] Among them, the bone screw image category corresponds to the bone screw recognition model, the bone plate image category corresponds to the bone plate recognition model, and the surgical tool category corresponds to the surgical tool recognition model.

[0013] In one embodiment of the present invention, the step of identifying the category of the image information and assigning a corresponding recognition detection model to the image information based on the category, wherein the recognition detection model can be used to identify and obtain target features in the image information, includes:

[0014] Extracting identification element information from the image information, and identifying and reading category information covered by the identification element information;

[0015] If the identification element information cannot be identified, an image enhancement identification step is performed, which includes:

[0016] Acquire high-definition image information, perform identification element information processing on the high-definition image information, and acquire category information;

[0017] Based on the category information, a corresponding recognition detection model is assigned to the image information.

[0018] In one embodiment of the present invention, the identification and detection model labels the bone screws, bone plates and surgical instruments, and defines a specific identification code for each bone screw, each bone plate and each surgical instrument.

[0019] In one embodiment of the present invention, it comprises:

[0020] The obtained initial image and the final image are respectively adjusted to the same size, and divided into M*N grids;

[0021] Obtaining the probability of the target feature existing in the grid, and if the target feature exists in the grid, predicting the bounding box of the target feature in the area, wherein the target feature includes bone screws, bone plates, and surgical instruments;

[0022] The redundant part of the bounding box is removed by using the predicted value of the bounding box, and the estimated bounding box of the orthopedic surgical instrument is obtained by performing a non-maximum suppression operation on the bounding box whose category confidence exceeds the threshold point;

[0023] Each bounding box is assigned a specific position code.

[0024] In one solution of the present invention, it includes: acquiring data information of a preset orthopedic instrument, and establishing a mapping relationship between the position code and the corresponding orthopedic instrument.

[0025] In one solution of the present invention, based on the initial image, a mapping relationship between the identification code and the position code of the target feature is established, and the existence of the position code and its corresponding bounding box in the final image is retained.

[0026] In one embodiment of the present invention, in the step of establishing a data set A including target features of the initial image and a data set B including target features of the final image,

[0027] When there is a target feature in the bounding box, the label can be defined as 1; when there is no target feature in the bounding box, the label can be defined as 0.

[0028] In one embodiment of the present invention, in the step of comparing the data set A and the data set B to generate a difference set C, and generating the usage data of the orthopedic instrument based on the difference set C,

[0029] The difference set C is the position encoding information of all bounding boxes whose labels are 1 in the initial image and 0 in the final image;

[0030] By identifying the position code, the usage data of the orthopedic instrument is obtained.

[0031] The present invention also discloses an orthopedic instrument identification and detection system applied to any of the above-mentioned orthopedic instrument identification and detection methods, comprising:

[0032] A data acquisition module, which is used to acquire image information, wherein the image information includes an initial image and a final image;

[0033] A data processing module, which is used to identify the category of the image information and assign a corresponding recognition detection model to the image information based on the category, wherein the recognition detection model can be used to identify and obtain target features in the image information;

[0034] a data set generation module, which is used to establish a data set A including the target features of the initial image and a data set B including the target features of the final image, and

[0035] A data comparison module is used to compare the data set A and the data set B and generate a difference set C, and based on the difference set C, generate usage data of the orthopedic instrument.

[0036] In summary, the present invention discloses an orthopedic instrument recognition and detection method and system, which recognizes and processes image categories, so as to facilitate different processing methods for different categories of image information and assign corresponding recognition and detection models. When defining the recognition code for each instrument, the recognition and detection model establishes a mapping relationship between the recognition code and the preset basic information. Then, the data information of the preset orthopedic instrument is obtained, and a mapping relationship between the position code and the corresponding orthopedic instrument is established. A position code-recognition code-data information comparison table is established, and in the subsequent final image, the specific usage of the medical device can be obtained by identifying the position code. This can effectively improve the problem that the existing orthopedic implant instruments have poor instrument recognition effect and are difficult to accurately obtain the specific orthopedic instrument data used during the recognition process. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.

[0038] Figure 1 A schematic diagram of a flow chart of an orthopedic instrument identification and detection method in one embodiment of the present invention;

[0039] Figure 2 The figure is a schematic diagram of modules of an orthopedic instrument identification and detection system in one embodiment of the present invention.

[0040] Component number description

[0041] 100. Identification and detection of orthopedic instruments; 10. Data acquisition module; 20. Data processing module; 30. Data set generation module; 40. Data comparison module.

[0042] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0043] 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.

[0044] In addition, if "and / or" or "and / or" appears in the full text, its meaning includes three parallel solutions. Taking "A and / or B" as an example, it includes solution A, solution B, or solutions that satisfy both A and B. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the protection scope required by the present invention.

[0045] See also Figure 1 The present invention discloses an orthopedic instrument recognition and detection method, which can be used to improve the problem that the instrument recognition effect is poor and it is difficult to accurately obtain the specific orthopedic instrument data used during the recognition process of existing orthopedic implant instruments.

[0046] The present invention provides an orthopedic instrument identification and detection method, which is applied to an orthopedic instrument identification device, and the orthopedic instrument identification device at least includes an interactive interface and an image acquisition unit. The interactive interface can be used to display data information and realize interaction with people. At the same time, the image acquisition unit is used to realize image acquisition of the orthopedic instrument.

[0047] Among them, the interactive end may be allowed to be located in one or more electronic devices, and the electronic device is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable gate arrays (Field-Programmable Gate Array, FPGA), digital processors (Digital Signal Processor, DSP), embedded devices, etc.

[0048] The electronic device may be any electronic product that can perform human-computer interaction with a user, such as a personal computer, a tablet computer, a smart phone, a personal digital assistant (PDA), an interactive network television (IPTV), etc.

[0049] The electronic device may also include a network device and / or a user device, wherein the network device includes, but is not limited to, a single network server, a server group consisting of multiple network servers, or a cloud consisting of a large number of hosts or network servers based on cloud computing.

[0050] The network where the electronic device is located includes but is not limited to the Internet, a wide area network, a metropolitan area network, a local area network, a virtual private network (VPN), etc.

[0051] The following will be combined Figure 1 The identification and detection method of the orthopedic instrument of the present invention is described in detail.

[0052] Specifically, the identification and detection method of orthopedic instruments disclosed in the present invention may include the following steps.

[0053] First, step S10 is executed to obtain image information, where the image information includes an initial image and a final image.

[0054] It is understandable that the initial image and the final image are acquired through the image acquisition unit. The image acquisition unit allows the use of a self-aligning AI focus module to support curved surface features and 360-degree comprehensive image acquisition and synthesis. The quality of image information acquisition can be effectively improved through the self-aligning AI focus module. At the same time, the initial image is taken before the use of the medical device, and the final image is taken after some of the medical devices are used. There is a certain time difference between the initial image and the final image during the shooting process.

[0055] Next, step S20 is executed to identify the category of the image information, and to assign a corresponding recognition detection model to the image information based on the category. The recognition detection model can be used to identify and obtain target features in the image information.

[0056] In this technical solution, medical devices in actual use mainly include bone screw image categories, bone plate image categories and surgical tool categories. Therefore, the image categories can be identified and processed to facilitate different processing methods for different categories of image information and to assign corresponding recognition detection models.

[0057] Specifically, the recognition detection model includes a bone screw recognition model, a bone plate recognition model and a surgical tool recognition model, and the recognition detection model adopts a convolutional neural network model architecture. At the same time, the bone screw image category corresponds to the bone screw recognition model, the bone plate image category corresponds to the bone plate recognition model, and the surgical tool category corresponds to the surgical tool recognition model.

[0058] Since the recognition detection model uses a convolutional neural network architecture, it can automatically extract and learn features. For example, the convolutional neural network can automatically extract and learn important features in the input data, such as edges and textures in images. At the same time, the convolutional neural network model can visualize the input data to help understand how the model works and the optimization process. It helps to debug and optimize the model and improve the performance of the model.

[0059] Specifically, in step S20, the following steps may be included.

[0060] S210, extracting identification element information in the image information, and identifying and reading category information covered by the identification element information.

[0061] In the present technical solution, the medical devices mainly include bone nails, bone plates and surgical instruments. Among them, the bone nails, bone plates and surgical instruments are usually placed in a sterilized placement tray. Therefore, it is allowed to set an identification code that can be used for identification on the placement tray. The identification code can be a bar code or a QR code, but it is not limited to this. The identification code can be determined according to actual needs. Therefore, after the image acquisition unit acquires the initial image, it is allowed to obtain the specific category of the initial image by identifying the identification code.

[0062] S220: If the identification element information cannot be identified, then perform an image enhancement identification step, which includes:

[0063] Obtain high-definition image information, perform identification element information processing on the high-definition image information, and obtain category information. Among them, in this technical solution, it is allowed to adjust the shooting environment to have sufficient and uniform light to avoid shadows and reflections. Secondly, adjust the camera's ISO, shutter speed, aperture and other parameters according to the shooting object and environment to obtain the best exposure and clarity.

[0064] S230: Based on the category information, assign a corresponding recognition detection model to the image information.

[0065] Furthermore, the bone screw recognition model, the bone plate recognition model and the surgical tool recognition model are used to recognize the bone screws, the bone plates and the surgical tools respectively.

[0066] Specifically, the bone screw recognition model can be used to identify the size, dimensions and specific location of the bone screw on the placement plate. The bone plate recognition model can be used to identify the size, location and specific category of the bone plate. In addition, the surgical tool recognition model can be used to identify the specific category of surgical tools, and give each surgical tool a name and effect, and the specific information can also be displayed through the interactive terminal.

[0067] Furthermore, during the actual detection process, the identification detection model can also allow bone screws, bone plates and surgical instruments to be labeled, and a specific identification code can be defined for each bone screw, each bone plate and each surgical instrument to facilitate the positioning of an instrument through the identification code.

[0068] It should be noted that before using a medical device, basic data information is usually preset, including the type, batch, function and size of the medical device. When defining the identification code for each device, the recognition detection model also establishes a mapping relationship between the identification code and the preset basic information.

[0069] Further, the acquired initial image and the acquired final image are respectively adjusted to the same size and divided into M*N grids. The probability of the target feature existing in the grid is obtained, and if the target feature exists in the grid, the bounding box of the target feature in the area is predicted, and the target feature includes bone screws, bone plates and surgical instruments.

[0070] By dividing the initial image and the final image into multiple grids, determining the specific position of each device on the initial image, and marking the positions, the specific usage of the medical device can be obtained.

[0071] It is understandable that the predicted value of the bounding box can be used to remove the redundant part of the bounding box, and the predicted bounding box of the orthopedic surgical instrument is obtained by applying a non-maximum suppression operation to the bounding box whose category confidence exceeds the threshold point. At the same time, a specific position code is assigned to each bounding box.

[0072] Therefore, for the initial image, each instrument in the image is labeled with a corresponding position code. Then, the data information of the preset orthopedic instruments is obtained, and a mapping relationship between the position code and the corresponding orthopedic instrument is established.

[0073] It should be noted that based on the initial image, a mapping relationship between the identification code and the position code of the target feature is established, and the existence of the position code and its corresponding bounding box in the final image is retained. A position code-identification code-data information comparison table can be established. And in the subsequent final image, by identifying the position code, the specific usage of the medical device can be obtained.

[0074] Furthermore, step S30 is performed to establish a data set A including target features of the initial image and a data set B including target features of the final image.

[0075] When there is a target feature in the bounding box, the label can be defined as 1; when there is no target feature in the bounding box, the label can be defined as 0.

[0076] Finally, step S40 is executed to compare the data set A and the data set B to generate a difference set C, and based on the difference set C, the usage data of the orthopedic instrument is generated.

[0077] Therefore, in the actual comparison process, the difference set C is the position codes of all bounding boxes whose labels of the bounding boxes of the initial image are 1 and the bounding boxes of the final image are 0. And by identifying the position codes, the usage data of the orthopedic instrument is obtained.

[0078] It should be noted that the step division of the above methods is only for clear description. When implemented, they can be combined into one step or some steps can be split and decomposed into multiple steps as long as they contain the same logical relationship.

[0079] See also Figure 2 The present invention also discloses an orthopedic instrument recognition and detection system 100 applied to any of the above-mentioned orthopedic instrument recognition and detection methods, which includes a data acquisition module 10, a data processing module 20, a data set generation module 30 and a data comparison module 40. The module referred to in the present invention refers to a series of computer program segments that can be executed by a processor and can complete fixed functions, which are stored in a memory.

[0080] Specifically, the data acquisition module 10 is used to acquire image information, and the image information includes an initial image and a final image. The data processing module 20 is used to identify the category of the image information, and assign a corresponding recognition detection model to the image information based on the category, and the recognition detection model can be used to identify and acquire target features in the image information. The data set generation module 30 is used to establish a data set A including the target features of the initial image and a data set B including the target features of the final image. And the data comparison module 40 is used to compare the data set A and the data set B and generate a difference set C, and generate the use data of the orthopedic instrument based on the difference set C.

[0081] In summary, the present invention discloses an orthopedic instrument recognition and detection method and system, which recognizes and processes image categories, so as to facilitate different processing methods for different categories of image information and assign corresponding recognition and detection models. When defining the recognition code for each instrument, the recognition and detection model establishes a mapping relationship between the recognition code and the preset basic information. Then, the data information of the preset orthopedic instrument is obtained, and a mapping relationship between the position code and the corresponding orthopedic instrument is established. A position code-recognition code-data information comparison table is established, and in the subsequent final image, the specific usage of the medical device can be obtained by identifying the position code.

[0082] Therefore, the problem that the existing orthopedic implant device identification process has poor device identification effect and is difficult to accurately obtain the specific orthopedic device data used can be effectively improved. Therefore, the present invention effectively overcomes some practical problems in the prior art and has high utilization value and use significance.

[0083] The above description is only an exemplary embodiment of the present invention, and does not limit the patent scope of the present invention. All equivalent structural changes made by using the contents of the present invention specification and drawings under the technical concept of the present invention, or directly / indirectly applied in other related technical fields are included in the patent protection scope of the present invention.

Claims

1. A method for identifying and detecting orthopedic instruments, characterized in that: include: Acquiring image information, wherein the image information includes an initial image and a final image; Identify the category of the image information, and assign a corresponding recognition detection model to the image information based on the category, wherein the recognition detection model can be used to identify and obtain target features in the image information; Establishing a data set A including target features of the initial image and a data set B including target features of the final image, The data set A and the data set B are compared to generate a difference set C, and based on the difference set C, usage data of the orthopedic instrument is generated.

2. The orthopedic instrument identification and detection method according to claim 1, characterized in that: The recognition and detection model includes a bone nail recognition model, a bone plate recognition model and a surgical tool recognition model, and the recognition and detection model adopts a convolutional neural network model architecture; The categories of the image information include bone nail image categories, bone plate image categories and surgical tool categories; Among them, the bone screw image category corresponds to the bone screw recognition model, the bone plate image category corresponds to the bone plate recognition model, and the surgical tool category corresponds to the surgical tool recognition model.

3. The orthopedic instrument identification and detection method according to claim 1, characterized in that: In the step of identifying the category of the image information and assigning a corresponding recognition detection model to the image information based on the category, wherein the recognition detection model can be used to identify and obtain target features in the image information, the step includes: Extracting identification element information from the image information, and identifying and reading category information covered by the identification element information; If the identification element information cannot be identified, an image enhancement identification step is performed, which includes: Acquire high-definition image information, perform identification element information processing on the high-definition image information, and acquire category information; Based on the category information, a corresponding recognition detection model is assigned to the image information.

4. The orthopedic instrument identification and detection method according to claim 3, characterized in that: The identification detection model labels the bone screws, bone plates and surgical instruments, and defines a specific identification code for each bone screw, each bone plate and each surgical instrument.

5. The orthopedic instrument identification and detection method according to claim 4, characterized in that: Also includes: The obtained initial image and the final image are respectively adjusted to the same size, and divided into M*N grids; Obtaining the probability of the target feature existing in the grid, and if the target feature exists in the grid, predicting the bounding box of the target feature in the area, wherein the target feature includes bone screws, bone plates, and surgical instruments; The redundant part of the bounding box is removed by using the predicted value of the bounding box, and the estimated bounding box of the orthopedic surgical instrument is obtained by performing a non-maximum suppression operation on the bounding box whose category confidence exceeds the threshold point; Each bounding box is assigned a specific position code.

6. The orthopedic instrument identification and detection method according to claim 4, characterized in that: Also includes: The data information of the preset orthopedic instrument is obtained, and a mapping relationship between the position code and the corresponding orthopedic instrument is established.

7. The orthopedic instrument identification and detection method according to claim 4, characterized in that: Based on the initial image, a mapping relationship between the identification code and the position code of the target feature is established, and the existence of the position code and its corresponding bounding box in the final image is retained.

8. The orthopedic instrument identification and detection method according to claim 4, characterized in that: In the step of establishing a data set A including target features of the initial image and a data set B including target features of the final image, the steps include: When there is a target feature in the bounding box, the label can be defined as 1; when there is no target feature in the bounding box, the label can be defined as 0.

9. The orthopedic instrument identification and detection method according to claim 4, characterized in that: The step of comparing the data set A and the data set B to generate a difference set C, and generating the usage data of the orthopedic instrument based on the difference set C includes: The difference set C is the position encoding information of all bounding boxes whose labels are 1 in the initial image and 0 in the final image; By identifying the position code, the usage data of the orthopedic instrument is obtained.

10. An orthopedic instrument recognition and detection system applied to the orthopedic instrument recognition and detection method according to any one of claims 1 to 9, characterized in that: include: A data acquisition module, which is used to acquire image information, wherein the image information includes an initial image and a final image; A data processing module, which is used to identify the category of the image information and assign a corresponding recognition detection model to the image information based on the category, wherein the recognition detection model can be used to identify and obtain target features in the image information; a data set generation module, which is used to establish a data set A including the target features of the initial image and a data set B including the target features of the final image, and A data comparison module is used to compare the data set A and the data set B and generate a difference set C, and based on the difference set C, generate usage data of the orthopedic instrument.

Images