Image processing method and device for lumbar spondylolysis detection

Abstract

The present invention discloses an image processing method and device for lumbar spondylolysis detection, the method comprising: obtaining lumbar vertebrae image information to be processed; pre-processing the lumbar vertebrae image information to be processed to obtain initial image processing result information; using a target image recognition model to perform recognition processing on the initial image processing result information to obtain target image recognition result information.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to an image processing method and device for detecting lumbar spondylolysis. Background Art

[0002] Lumbar spondylolysis is a common orthopedic disease. Lumbar spine X-rays, as a conventional detection method, have certain limitations. The anatomical position of the lumbar isthmus and the surrounding tissues is complex, and the changes after the isthmus is broken are often not clearly displayed on the X-ray film, resulting in difficulty in diagnosis. Secondly, the projection angle and imaging quality of the X-ray film may further blur the image, increasing the uncertainty of the diagnosis. Therefore, lumbar spondylolysis is very easy to be missed on lumbar spine X-rays. Although CT scanning can improve the accuracy of diagnosis, it is inefficient, costly, and has a large amount of radiation, and is not suitable for screening. Therefore, an image processing method and device for lumbar spondylolysis detection are provided to improve the image recognition accuracy and efficiency of lumbar spine X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide an image processing method and device for lumbar spondylolysis detection, which is beneficial to improve the image recognition accuracy and recognition efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0004] In order to solve the above technical problems, the first aspect of the embodiment of the present invention discloses an image processing method for lumbar spondylolysis detection, the method comprising:

[0005] Acquire lumbar vertebra image information to be processed;

[0006] Pre-processing the lumbar vertebra image information to be processed to obtain initial image processing result information;

[0007] The target image recognition model is used to perform recognition processing on the initial image processing result information to obtain the target image recognition result information.

[0008] A second aspect of an embodiment of the present invention discloses an image processing device for detecting lumbar spondylolysis, the device comprising:

[0009] An acquisition module, used for acquiring lumbar vertebra image information to be processed;

[0010] A first processing module is used to pre-process the lumbar vertebra image information to be processed to obtain initial image processing result information;

[0011] The second processing module is used to perform recognition processing on the initial image processing result information using a target image recognition model to obtain target image recognition result information.

[0012] A third aspect of the present invention discloses another image processing device for detecting lumbar spondylolysis, the device comprising:

[0013] A memory storing executable program code;

[0014] a processor coupled to the memory;

[0015] The processor calls the executable program code stored in the memory to execute part or all of the steps in the image processing method for lumbar spondylolysis detection disclosed in the first aspect of the embodiment of the present invention.

[0016] The fourth aspect of the present invention discloses a computer-readable storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute some or all of the steps in the image processing method for lumbar spondylolysis detection disclosed in the first aspect of an embodiment of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 It is a flowchart of an image processing method for lumbar spondylolysis detection disclosed in an embodiment of the present invention;

[0019] Figure 2 is a structural schematic diagram of an image processing device for lumbar spondylolysis detection disclosed in an embodiment of the present invention;

[0020] Figure 3 is a structural schematic diagram of another image processing device for lumbar spondylolysis detection disclosed in an embodiment of the present invention;

[0021] Figure 4 is a structural schematic diagram of a target image recognition model disclosed in an embodiment of the present invention;

[0022] Figure 5 is a structural schematic diagram of a first dynamic feature extraction module disclosed in an embodiment of the present invention;

[0023] Figure 6 is a structural schematic diagram of a first convolution feature extraction module disclosed in an embodiment of the present invention;

[0024] Figure 7 It is a structural schematic diagram of a deep feature extraction module disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0025] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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.

[0026] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or equipment that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or equipment.

[0027] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0028] The present invention discloses an image processing method and device for lumbar spondylolysis detection, which is beneficial to improving the image recognition accuracy and recognition efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection. Detailed descriptions are given below.

[0029] Embodiment 1

[0030] See also Figure 1 , Figure 1 1 is a flow chart of an image processing method for detecting lumbar spondylolysis disclosed in an embodiment of the present invention. Figure 1 The image processing method for detecting lumbar spondylolysis described is applied to a management system, such as a local server or a cloud server for management, and the embodiments of the present invention are not limited thereto. Figure 1 As shown, the image processing method for lumbar spondylolysis detection may include the following operations:

[0031] 101. Obtain lumbar vertebra image information to be processed.

[0032] 102. Pre-process the lumbar vertebra image information to be processed to obtain initial image processing result information.

[0033] 103. Use the target image recognition model to perform recognition processing on the initial image processing result information to obtain the target image recognition result information.

[0034] In this optional embodiment, as an optional implementation manner, the above-mentioned pre-processing of the lumbar vertebrae image information to be processed to obtain the initial image processing result information includes:

[0035] Performing correction processing on the lumbar vertebra image information to be processed to obtain corrected lumbar vertebra image information;

[0036] Performing filtering processing on the corrected lumbar vertebrae image information to obtain filtered lumbar vertebrae image information;

[0037] The filtered lumbar vertebrae image information is subjected to specification adjustment processing to obtain initial image processing result information.

[0038] It should be noted that the correction processing of the lumbar vertebrae image information to be processed is implemented based on a quadratic linear interpolation method, which is not limited in the embodiment of the present invention.

[0039] It should be noted that the filtering process performed on the corrected lumbar vertebrae image information is a low-pass filtering process.

[0040] It should be noted that the above-mentioned specification adjustment processing of the filtered lumbar vertebrae image information is to adjust the image size to 256*256, which is not limited in the embodiment of the present invention.

[0041] It should be noted that the above-mentioned target image recognition result information can be used to indicate the analysis of lumbar spondylolysis lesions to reduce dependence on user experience, improve diagnostic accuracy, and thus maximize the quality and efficiency of lumbar spondylolysis detection.

[0042] It can be seen that implementing the image processing method for lumbar spondylolysis detection described in the embodiment of the present invention is beneficial to improving the image recognition accuracy and efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0043] In an optional embodiment, if Figure 4As shown, the target image recognition model includes a first convolution module, a second convolution module, a first normalization module, a second normalization module, a first activation module, a second activation module, a third activation module, a first dynamic feature extraction module, a second dynamic feature extraction module, a third dynamic feature extraction module, a fourth dynamic feature extraction module, a depth feature extraction module, a first convolution feature extraction module, a second convolution feature extraction module, a third convolution feature extraction module, a fourth convolution feature extraction module, a first interactive feature extraction module, a second interactive feature extraction module, a third interactive feature extraction module, a first sampling module, a second sampling module and a first fusion module; wherein,

[0044] The input end of the first convolution module is connected to the model input of the target image recognition model, and the output end of the first convolution module is connected to the input end of the first normalization module; the output end of the first normalization module is connected to the input end of the first activation module; the output end of the first activation module is connected to the input end of the first dynamic feature extraction module; the output end of the first dynamic feature extraction module is connected to the input end of the second dynamic feature extraction module; the output end of the second dynamic feature extraction module is respectively connected to the input end of the third dynamic feature extraction module and the input end of the third convolution feature extraction module; the output end of the third dynamic feature extraction module is respectively connected to the input end of the fourth dynamic feature extraction module and the input end of the second convolution feature extraction module; the output end of the fourth dynamic feature extraction module is connected to the input end of the depth feature extraction module; the output end of the depth feature extraction module is respectively connected to the input end of the first convolution feature extraction module and the input end of the first sampling module; the output end of the first convolution feature extraction module is connected to the input end of the third interactive feature extraction module; the third interactive feature extraction The output end of the module is connected to the input end of the first fusion module; the output end of the first sampling module is connected to the input end of the second convolution feature extraction module; the output end of the second convolution feature extraction module is respectively connected to the input end of the second sampling module and the input end of the fourth convolution feature extraction module; the output end of the second sampling module is connected to the input end of the first interactive feature extraction module; the output end of the first interactive feature extraction module is connected to the input end of the first fusion module; the output end of the fourth convolution feature extraction module is respectively connected to the input end of the second convolution module and the input end of the second interactive feature extraction module; the output end of the second convolution module is connected to the input end of the second normalization module; the output end of the second normalization module is connected to the input end of the second activation module; the output end of the second activation module is connected to the input end of the first convolution feature extraction module; the output end of the second interactive feature extraction module is connected to the input end of the first fusion module; the output end of the first fusion module is connected to the input end of the third activation module; the output end of the third activation module is connected to the model output of the target image recognition model.

[0045] It should be noted that the convolution kernel size of the first convolution module and the second convolution module is 1*1, which is not limited in the embodiment of the present invention.

[0046] It should be noted that the above-mentioned first normalization module and second normalization module are constructed based on the BN normalization layer, which is not limited in the embodiment of the present invention.

[0047] It should be noted that the first activation module, the second activation module, and the third activation module are constructed based on the gelu function, which is not limited in the embodiment of the present invention.

[0048] It should be noted that the first sampling module and the second sampling module are up-sampling modules, that is, used for up-sampling operations.

[0049] It should be noted that the above-mentioned first fusion module is constructed based on matrix multiplication, which is not limited in the embodiment of the present invention.

[0050] It can be seen that implementing the image processing method for lumbar spondylolysis detection described in the embodiment of the present invention is beneficial to improving the image recognition accuracy and efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0051] In another optional embodiment, Figure 4 As shown, the first interactive feature extraction module includes a first convolution unit, a second convolution unit, a third convolution unit, a fourth convolution unit, a first normalization unit, a second normalization unit, a first activation unit, a second activation unit and a first fusion unit; wherein,

[0052] The input end of the first convolution unit and the input end of the third convolution unit are both connected to the output end of the third convolution feature extraction module; the output end of the first convolution unit is connected to the input end of the first normalization unit; the output end of the first normalization unit is connected to the input end of the first activation unit; the output end of the first activation unit is connected to the input end of the second convolution unit; the output end of the second convolution unit is connected to the input end of the first fusion unit; the output end of the third convolution unit is connected to the input end of the second normalization unit; the output end of the second normalization unit is connected to the input end of the second activation unit; the output end of the second activation unit is connected to the input end of the fourth convolution unit; the output end of the fourth convolution unit is connected to the input end of the first fusion unit; the output end of the first fusion unit is connected to the input end of the first fusion module.

[0053] It should be noted that the convolution kernel size of the above-mentioned first convolution unit, second convolution unit, third convolution unit, and fourth convolution unit can be one of 1*1, 3*3, 5*5 or 7*7, which is not limited in the embodiment of the present invention.

[0054] It should be noted that the above-mentioned first normalization unit and the second normalization unit are constructed based on the BN normalization layer, which is not limited in the embodiment of the present invention.

[0055] It should be noted that the first activation unit and the second activation unit are constructed based on the sigmoid function, which is not limited in the embodiment of the present invention.

[0056] It should be noted that the above-mentioned first fusion unit is constructed based on matrix summation, which is not limited in the embodiment of the present invention.

[0057] It can be seen that implementing the image processing method for lumbar spondylolysis detection described in the embodiment of the present invention is beneficial to improving the image recognition accuracy and efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0058] In yet another optional embodiment, Figure 5 As shown, the first dynamic feature extraction module includes a fifth convolution unit, a sixth convolution unit, a seventh convolution unit, an eighth convolution unit, a ninth convolution unit, a tenth convolution unit, an eleventh convolution unit, a third normalization unit, a fourth normalization unit, a fifth normalization unit, a sixth normalization unit, a seventh normalization unit, an eighth normalization unit, a ninth normalization unit, a third activation unit, a fourth activation unit, a fifth activation unit, a sixth activation unit, a seventh activation unit, an eighth activation unit, a ninth activation unit, a first slicing unit, a first connection unit, and a second fusion unit; wherein,

[0059] The input end of the fifth convolution unit is connected to the output end of the first activation module, and the output end of the fifth convolution unit is connected to the input end of the third normalization unit; the output end of the third normalization unit is connected to the input end of the third activation unit; the output end of the third activation unit is connected to the input end of the sixth convolution unit; the output end of the sixth convolution unit is connected to the input end of the fourth normalization unit; the output end of the fourth normalization unit is connected to the input end of the fourth activation unit; the output end of the fourth activation unit is connected to the input end of the first slicing unit; the output end of the first slicing unit is respectively connected to the input end of the seventh convolution unit, the input end of the second fusion unit and the input end of the eleventh convolution unit; the output end of the seventh convolution unit is connected to the input end of the fifth normalization unit; the output end of the fifth normalization unit is connected to the input end of the fifth activation unit; the output end of the fifth activation unit is connected to the input end of the eighth convolution unit; the output end of the eighth convolution unit is connected to the input end of the sixth normalization unit; the output end of the sixth normalization unit is connected to the input end of the sixth activation unit The output end of the sixth activation unit is connected to the input end of the second fusion unit; the output end of the second fusion unit is respectively connected to the input end of the ninth convolution unit, the input end of the third fusion unit and the input end of the eleventh convolution unit; the output end of the ninth convolution unit is connected to the input end of the seventh normalization unit; the output end of the seventh normalization unit is connected to the input end of the seventh activation unit; the output end of the seventh activation unit is connected to the input end of the tenth convolution unit; the output end of the tenth convolution unit is connected to the input end of the eighth normalization unit; the output end of the eighth normalization unit is connected to the input end of the third fusion unit; the output end of the third fusion unit is connected to the input end of the first connection unit; the output end of the first connection unit is connected to the input end of the eleventh convolution unit; the eleventh convolution unit is connected to the input end of the ninth normalization unit; the output end of the ninth normalization unit is connected to the input end of the ninth activation unit; the output end of the ninth activation unit is connected to the input end of the second dynamic feature extraction module.

[0060] It should be noted that the convolution kernel size of the fifth convolution unit, the sixth convolution unit, the seventh convolution unit, the eighth convolution unit, the ninth convolution unit, the tenth convolution unit, and the eleventh convolution unit can be one of 1*1, 3*3, 5*5 or 7*7, which is not limited in the embodiment of the present invention.

[0061] It should be noted that the third normalization unit, the fourth normalization unit, the fifth normalization unit, the sixth normalization unit, the seventh normalization unit, the eighth normalization unit, and the ninth normalization unit are constructed based on the BN normalization layer, which is not limited in the embodiment of the present invention.

[0062] It should be noted that the third activation unit, the fourth activation unit, the fifth activation unit, the sixth activation unit, the seventh activation unit, the eighth activation unit, and the ninth activation unit are constructed based on the sigmoid function, which is not limited in the embodiment of the present invention.

[0063] It should be noted that the above-mentioned first slicing unit is constructed based on the slice program, which is not limited in the embodiment of the present invention.

[0064] It should be noted that the above-mentioned first connection unit is constructed based on the fully connected layer, which is not limited in the embodiment of the present invention.

[0065] It should be noted that the above-mentioned second fusion unit is constructed based on matrix summation, which is not limited in the embodiment of the present invention.

[0066] It can be seen that implementing the image processing method for lumbar spondylolysis detection described in the embodiment of the present invention is beneficial to improving the image recognition accuracy and efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0067] In yet another optional embodiment, Figure 6 As shown, the first convolution feature extraction module includes a second connection unit, a third connection unit, a second slicing unit, a twelfth convolution unit, a thirteenth convolution unit, a fourteenth convolution unit, a fifteenth convolution unit, a sixteenth convolution unit, a tenth normalization unit, an eleventh normalization unit, a twelfth normalization unit, a thirteenth normalization unit, a fourteenth normalization unit, a tenth activation unit, an eleventh activation unit, a twelfth activation unit, a thirteenth activation unit, a fourteenth activation unit, a fourth fusion unit and a fifth fusion unit; wherein,

[0068] The input end of the second connection unit is connected to the output end of the deep feature extraction module, and the output end of the second connection unit is connected to the input end of the second slicing unit; the output end of the second slicing unit is respectively connected to the input end of the twelfth convolution unit, the input end of the fourth fusion unit, and the input end of the sixteenth convolution unit; the output end of the twelfth convolution unit is connected to the input end of the tenth normalization unit; the output end of the tenth normalization unit is connected to the input end of the tenth activation unit; the output end of the tenth activation unit is connected to the input end of the thirteenth convolution unit; the output end of the thirteenth convolution unit is connected to the input end of the eleventh normalization unit; the output end of the eleventh normalization unit is connected to the input end of the eleventh activation unit; the output end of the eleventh activation unit is connected to the input end of the fourth fusion unit; the output end of the fourth fusion unit is respectively connected to the input end of the sixteenth convolution unit, the input end of the fourteenth convolution unit, and the input end of the fifth The input end of the fusion unit; the output end of the fourteenth convolution unit is connected to the input end of the twelfth normalization unit; the output end of the twelfth normalization unit is connected to the input end of the twelfth activation unit; the output end of the twelfth activation unit is connected to the input end of the fifteenth convolution unit; the output end of the fifteenth convolution unit is connected to the input end of the thirteenth normalization unit; the output end of the thirteenth normalization unit is connected to the input end of the thirteenth activation unit; the output end of the thirteenth activation unit is connected to the input end of the fifth fusion unit; the output end of the fifth fusion unit is connected to the input end of the third connection unit; the output end of the third connection unit is connected to the input end of the sixteenth convolution unit; the output end of the sixteenth convolution unit is connected to the input end of the fourteenth normalization unit; the output end of the tenth normalization unit is connected to the input end of the fourteenth activation unit; the output end of the fourteenth activation unit is connected to the input end of the third interactive feature extraction module.

[0069] It should be noted that the second connection unit and the third connection unit are constructed based on the fully connected layer, which is not limited in the embodiment of the present invention.

[0070] It should be noted that the second slicing unit is constructed based on the slice program, which is not limited in the embodiment of the present invention.

[0071] It should be noted that the convolution kernel size of the twelfth convolution unit, the thirteenth convolution unit, the fourteenth convolution unit, the fifteenth convolution unit and the sixteenth convolution unit can be one of 1*1, 3*3, 5*5 or 7*7, which is not limited in the embodiment of the present invention.

[0072] It should be noted that the tenth normalization unit, the eleventh normalization unit, the twelfth normalization unit, the thirteenth normalization unit, and the fourteenth normalization unit are constructed based on the BN normalization layer, which is not limited in the embodiment of the present invention.

[0073] It should be noted that the tenth activation unit, the eleventh activation unit, the twelfth activation unit, the thirteenth activation unit, and the fourteenth activation unit are constructed based on the gelu function, which is not limited in the embodiment of the present invention.

[0074] It should be noted that the fourth fusion unit and the fifth fusion unit are constructed based on matrix summation, which is not limited in the embodiment of the present invention.

[0075] It can be seen that implementing the image processing method for lumbar spondylolysis detection described in the embodiment of the present invention is beneficial to improving the image recognition accuracy and efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0076] In an optional embodiment, if Figure 7 As shown, the above-mentioned deep feature extraction module includes a seventeenth convolution unit, an eighteenth convolution unit, a nineteenth convolution unit, a twentieth convolution unit, a twenty-first convolution unit, a fourth connection unit, a fifth connection unit, a fifteenth normalization unit, a sixteenth normalization unit, a fifteenth activation unit, a sixteenth activation unit, a first pooling unit, a second pooling unit, and a third pooling unit; wherein,

[0077] The output end of the fourth dynamic feature extraction module is respectively connected to the input end of the seventeenth convolution unit, the input end of the eighteenth convolution unit and the input end of the fourth connection unit; the output end of the seventeenth convolution unit is respectively connected to the input end of the eighteenth convolution unit and the input end of the fourth connection unit; the output end of the eighteenth convolution unit is respectively connected to the input end of the nineteenth convolution unit and the input end of the fourth connection unit; the output end of the nineteenth convolution unit is connected to the input end of the fourth connection unit; the output end of the fourth connection unit is connected to the input end of the twentieth convolution unit; the output end of the twentieth convolution unit is connected to the input end of the fifteenth normalization unit; the output end of the fifteenth normalization unit is connected to the input end of the fifteenth activation unit; The output end of the activation unit is respectively connected to the input end of the first pooling unit, the input end of the fifth connection unit and the input end of the second pooling unit; the output end of the first pooling unit is connected to the input end of the second pooling unit; the output end of the second pooling unit is respectively connected to the input end of the third pooling unit and the input end of the fifth connection unit; the output end of the third pooling unit is connected to the input end of the fifth connection unit; the output end of the fifth connection unit is connected to the input end of the twenty-first convolution unit; the output end of the twenty-first convolution unit is connected to the input end of the sixteenth normalization unit; the output end of the sixteenth normalization unit is connected to the input end of the sixteenth activation unit; the output end of the sixteenth activation unit is connected to the input end of the first convolution feature extraction module.

[0078] It should be noted that the convolution kernel size of the seventeenth convolution unit, the eighteenth convolution unit, the nineteenth convolution unit, the twentieth convolution unit, and the twenty-first convolution unit can be one of 3*3, 5*5 or 7*7, which is not limited in the embodiment of the present invention. Furthermore, the above-mentioned convolution unit has the following characteristics, namely, explicitly generating multiple features with various large receptive fields, which makes the subsequent kernel selection process more flexible. Secondly, compared with directly applying a single larger kernel, the effect of sequential decomposition is more significant. While maintaining the same theoretical receptive field, decomposition can greatly improve the accuracy with only a small increase in parameters.

[0079] It should be noted that the fourth connection unit and the fifth connection unit are constructed based on a fully connected layer, which is not limited in the embodiment of the present invention.

[0080] It should be noted that the fifteenth normalization unit and the sixteenth normalization unit are constructed based on the BN normalization layer, which is not limited in the embodiment of the present invention.

[0081] It should be noted that the fifteenth activation unit and the sixteenth activation unit are constructed based on the sigmoid function, which is not limited in the embodiment of the present invention.

[0082] It should be noted that the first pooling unit, the second pooling unit and the third pooling unit are constructed based on the maximum pooling layer, which is not limited in the embodiment of the present invention.

[0083] It can be seen that implementing the image processing method for lumbar spondylolysis detection described in the embodiment of the present invention is beneficial to improving the image recognition accuracy and efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0084] In another optional embodiment, the target image recognition model is obtained based on the following steps:

[0085] Obtaining X-ray image information of lesions;

[0086] The lesion X-ray image information is screened and processed to obtain first X-ray image information; the first X-ray image information includes M first X-ray images; M is a positive integer;

[0087] Performing data enhancement processing on the first X-ray image information to obtain second X-ray image information; the second X-ray image information includes N second X-ray images; N is a positive integer not less than M;

[0088] performing image size optimization processing on the second X-ray image information to obtain third X-ray image information;

[0089] Dividing the third X-ray image information into a data set to obtain an X-ray image set; the X-ray image set includes a plurality of X-ray images to be used;

[0090] Acquire training condition information; the training condition information includes the number of sample images and the iteration threshold;

[0091] Initializing the initial image recognition model using the training condition information to obtain a first training image recognition model;

[0092] The number of sample images is used to screen and process the unused X-ray images in the X-ray image set to obtain target X-ray image information; the target X-ray image information includes L target X-ray images; L is consistent with the number of sample images;

[0093] Using the target X-ray image information to train the first training image recognition model, and obtaining a second training image recognition model and training result parameter information;

[0094] Determine whether the training result parameter information meets the model training termination condition and obtain the conditional judgment result; the model training termination condition represents the comparison between the number of training times and the iteration threshold;

[0095] When the result of the conditional judgment is no, the first training image recognition model is updated by using the second training image recognition model, and the screening process of the unused X-ray images in the X-ray image set by using the number of sample images is triggered to obtain target X-ray image information;

[0096] When the conditional judgment result is yes, the second training image recognition model is determined to be the target image recognition model.

[0097] It should be noted that the number of sample images may be set by a user or may be a preset default value, such as 64, which is not limited in the embodiment of the present invention.

[0098] It should be noted that the termination condition of model training is that the number of training times is equal to the iteration threshold. The iteration threshold may be input by a user and is not limited in the embodiment of the present invention. Further, the above number of training times represents the number of training times to complete the first training image recognition model.

[0099] It should be noted that the model architecture of the above initial image recognition model is consistent with that of the target image recognition model.

[0100] It should be noted that the above-mentioned lesion X-ray image information represents an X-ray image of a lumbar spondylolysis lesion.

[0101] It should be noted that the above-mentioned image size optimization processing for the second X-ray image information is to adjust the image size to 256*256, which is not limited in the embodiment of the present invention.

[0102] It should be noted that the above-mentioned screening process of the unused X-ray images in the X-ray image set by using the number of sample images is to randomly select the unused X-ray images of a number corresponding to the number of samples.

[0103] It should be noted that the above-mentioned initialization processing of the initial image recognition model using the training condition information is to set the initialization parameters of the code used to train the initial image recognition model, that is, to replace the system default number of sample images and iteration threshold with the number of sample images and iteration threshold in the obtained training condition information.

[0104] It should be noted that the above screening process for the lesion X-ray image information is to remove the lesion X-ray images that are obviously unclear in the lesion X-ray image information.

[0105] Preferably, the above M is a positive integer not greater than 500. Further, N is a positive integer not less than 2000.

[0106] It should be noted that the above-mentioned data enhancement processing of the first X-ray image information includes image segmentation, shearing, angle rotation, etc., which is not limited in the embodiment of the present invention.

[0107] It can be seen that implementing the image processing method for lumbar spondylolysis detection described in the embodiment of the present invention is beneficial to improving the image recognition accuracy and efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0108] Embodiment 2

[0109] See also Figure 2 , Figure 2 : is a schematic diagram of the structure of an image processing device for detecting lumbar spondylolysis disclosed in an embodiment of the present invention. Figure 2 The described device can be applied to a warehouse management system, such as a local server or cloud server for image processing management of warehouse logistics for lumbar spondylolysis detection, etc., and the embodiments of the present invention are not limited thereto. Figure 2 As shown, the device may include:

[0110] An acquisition module 201 is used to acquire lumbar vertebra image information to be processed;

[0111] The first processing module 202 is used to pre-process the lumbar vertebra image information to be processed to obtain initial image processing result information;

[0112] The second processing module 203 is used to perform recognition processing on the initial image processing result information using the target image recognition model to obtain the target image recognition result information.

[0113] It can be seen that the implementation Figure 2The described image processing device for lumbar spondylolysis detection is beneficial to improving the image recognition accuracy and recognition efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0114] In another optional embodiment, Figure 2 As shown, the target image recognition model includes a first convolution module, a second convolution module, a first normalization module, a second normalization module, a first activation module, a second activation module, a third activation module, a first dynamic feature extraction module, a second dynamic feature extraction module, a third dynamic feature extraction module, a fourth dynamic feature extraction module, a depth feature extraction module, a first convolution feature extraction module, a second convolution feature extraction module, a third convolution feature extraction module, a fourth convolution feature extraction module, a first interactive feature extraction module, a second interactive feature extraction module, a third interactive feature extraction module, a first sampling module, a second sampling module and a first fusion module; wherein,

[0115] The input end of the first convolution module is connected to the model input of the target image recognition model, and the output end of the first convolution module is connected to the input end of the first normalization module; the output end of the first normalization module is connected to the input end of the first activation module; the output end of the first activation module is connected to the input end of the first dynamic feature extraction module; the output end of the first dynamic feature extraction module is connected to the input end of the second dynamic feature extraction module; the output end of the second dynamic feature extraction module is respectively connected to the input end of the third dynamic feature extraction module and the input end of the third convolution feature extraction module; the output end of the third dynamic feature extraction module is respectively connected to the input end of the fourth dynamic feature extraction module and the input end of the second convolution feature extraction module; the output end of the fourth dynamic feature extraction module is connected to the input end of the depth feature extraction module; the output end of the depth feature extraction module is respectively connected to the input end of the first convolution feature extraction module and the input end of the first sampling module; the output end of the first convolution feature extraction module is connected to the input end of the third interactive feature extraction module; the third interactive feature extraction The output end of the module is connected to the input end of the first fusion module; the output end of the first sampling module is connected to the input end of the second convolution feature extraction module; the output end of the second convolution feature extraction module is respectively connected to the input end of the second sampling module and the input end of the fourth convolution feature extraction module; the output end of the second sampling module is connected to the input end of the first interactive feature extraction module; the output end of the first interactive feature extraction module is connected to the input end of the first fusion module; the output end of the fourth convolution feature extraction module is respectively connected to the input end of the second convolution module and the input end of the second interactive feature extraction module; the output end of the second convolution module is connected to the input end of the second normalization module; the output end of the second normalization module is connected to the input end of the second activation module; the output end of the second activation module is connected to the input end of the first convolution feature extraction module; the output end of the second interactive feature extraction module is connected to the input end of the first fusion module; the output end of the first fusion module is connected to the input end of the third activation module; the output end of the third activation module is connected to the model output of the target image recognition model.

[0116] It can be seen that implementation Figure 2 The described image processing device for lumbar spondylolysis detection is beneficial to improving the image recognition accuracy and recognition efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0117] In yet another optional embodiment, Figure 2 As shown, the first interactive feature extraction module includes a first convolution unit, a second convolution unit, a third convolution unit, a fourth convolution unit, a first normalization unit, a second normalization unit, a first activation unit, a second activation unit and a first fusion unit; wherein,

[0118] The input end of the first convolution unit and the input end of the third convolution unit are both connected to the output end of the third convolution feature extraction module; the output end of the first convolution unit is connected to the input end of the first normalization unit; the output end of the first normalization unit is connected to the input end of the first activation unit; the output end of the first activation unit is connected to the input end of the second convolution unit; the output end of the second convolution unit is connected to the input end of the first fusion unit; the output end of the third convolution unit is connected to the input end of the second normalization unit; the output end of the second normalization unit is connected to the input end of the second activation unit; the output end of the second activation unit is connected to the input end of the fourth convolution unit; the output end of the fourth convolution unit is connected to the input end of the first fusion unit; the output end of the first fusion unit is connected to the input end of the first fusion module.

[0119] It can be seen that the implementation Figure 2 The described image processing device for lumbar spondylolysis detection is beneficial to improving the image recognition accuracy and recognition efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0120] In yet another optional embodiment, Figure 2 As shown, the first dynamic feature extraction module includes a fifth convolution unit, a sixth convolution unit, a seventh convolution unit, an eighth convolution unit, a ninth convolution unit, a tenth convolution unit, an eleventh convolution unit, a third normalization unit, a fourth normalization unit, a fifth normalization unit, a sixth normalization unit, a seventh normalization unit, an eighth normalization unit, a ninth normalization unit, a third activation unit, a fourth activation unit, a fifth activation unit, a sixth activation unit, a seventh activation unit, an eighth activation unit, a ninth activation unit, a first slicing unit, a first connection unit, and a second fusion unit; wherein,

[0121] The input end of the fifth convolution unit is connected to the output end of the first activation module, and the output end of the fifth convolution unit is connected to the input end of the third normalization unit; the output end of the third normalization unit is connected to the input end of the third activation unit; the output end of the third activation unit is connected to the input end of the sixth convolution unit; the output end of the sixth convolution unit is connected to the input end of the fourth normalization unit; the output end of the fourth normalization unit is connected to the input end of the fourth activation unit; the output end of the fourth activation unit is connected to the input end of the first slicing unit; the output end of the first slicing unit is respectively connected to the input end of the seventh convolution unit, the input end of the second fusion unit and the input end of the eleventh convolution unit; the output end of the seventh convolution unit is connected to the input end of the fifth normalization unit; the output end of the fifth normalization unit is connected to the input end of the fifth activation unit; the output end of the fifth activation unit is connected to the input end of the eighth convolution unit; the output end of the eighth convolution unit is connected to the input end of the sixth normalization unit; the output end of the sixth normalization unit is connected to the input end of the sixth activation unit The output end of the sixth activation unit is connected to the input end of the second fusion unit; the output end of the second fusion unit is respectively connected to the input end of the ninth convolution unit, the input end of the third fusion unit and the input end of the eleventh convolution unit; the output end of the ninth convolution unit is connected to the input end of the seventh normalization unit; the output end of the seventh normalization unit is connected to the input end of the seventh activation unit; the output end of the seventh activation unit is connected to the input end of the tenth convolution unit; the output end of the tenth convolution unit is connected to the input end of the eighth normalization unit; the output end of the eighth normalization unit is connected to the input end of the third fusion unit; the output end of the third fusion unit is connected to the input end of the first connection unit; the output end of the first connection unit is connected to the input end of the eleventh convolution unit; the eleventh convolution unit is connected to the input end of the ninth normalization unit; the output end of the ninth normalization unit is connected to the input end of the ninth activation unit; the output end of the ninth activation unit is connected to the input end of the second dynamic feature extraction module.

[0122] It can be seen that the implementation Figure 2 The described image processing device for lumbar spondylolysis detection is beneficial to improving the image recognition accuracy and recognition efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0123] In yet another optional embodiment, Figure 2 As shown, the first convolution feature extraction module includes a second connection unit, a third connection unit, a second slicing unit, a twelfth convolution unit, a thirteenth convolution unit, a fourteenth convolution unit, a fifteenth convolution unit, a sixteenth convolution unit, a tenth normalization unit, an eleventh normalization unit, a twelfth normalization unit, a thirteenth normalization unit, a fourteenth normalization unit, a tenth activation unit, an eleventh activation unit, a twelfth activation unit, a thirteenth activation unit, a fourteenth activation unit, a fourth fusion unit and a fifth fusion unit; wherein,

[0124] The input end of the second connection unit is connected to the output end of the deep feature extraction module, and the output end of the second connection unit is connected to the input end of the second slicing unit; the output end of the second slicing unit is respectively connected to the input end of the twelfth convolution unit, the input end of the fourth fusion unit, and the input end of the sixteenth convolution unit; the output end of the twelfth convolution unit is connected to the input end of the tenth normalization unit; the output end of the tenth normalization unit is connected to the input end of the tenth activation unit; the output end of the tenth activation unit is connected to the input end of the thirteenth convolution unit; the output end of the thirteenth convolution unit is connected to the input end of the eleventh normalization unit; the output end of the eleventh normalization unit is connected to the input end of the eleventh activation unit; the output end of the eleventh activation unit is connected to the input end of the fourth fusion unit; the output end of the fourth fusion unit is respectively connected to the input end of the sixteenth convolution unit, the input end of the fourteenth convolution unit, and the input end of the fifth The input end of the fusion unit; the output end of the fourteenth convolution unit is connected to the input end of the twelfth normalization unit; the output end of the twelfth normalization unit is connected to the input end of the twelfth activation unit; the output end of the twelfth activation unit is connected to the input end of the fifteenth convolution unit; the output end of the fifteenth convolution unit is connected to the input end of the thirteenth normalization unit; the output end of the thirteenth normalization unit is connected to the input end of the thirteenth activation unit; the output end of the thirteenth activation unit is connected to the input end of the fifth fusion unit; the output end of the fifth fusion unit is connected to the input end of the third connection unit; the output end of the third connection unit is connected to the input end of the sixteenth convolution unit; the output end of the sixteenth convolution unit is connected to the input end of the fourteenth normalization unit; the output end of the tenth normalization unit is connected to the input end of the fourteenth activation unit; the output end of the fourteenth activation unit is connected to the input end of the third interactive feature extraction module.

[0125] It can be seen that the implementation Figure 2 The described image processing device for lumbar spondylolysis detection is beneficial to improving the image recognition accuracy and recognition efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0126] In yet another optional embodiment, Figure 2 As shown, the deep feature extraction module includes a seventeenth convolution unit, an eighteenth convolution unit, a nineteenth convolution unit, a twentieth convolution unit, a twenty-first convolution unit, a fourth connection unit, a fifth connection unit, a fifteenth normalization unit, a sixteenth normalization unit, a fifteenth activation unit, a sixteenth activation unit, a first pooling unit, a second pooling unit, and a third pooling unit; wherein,

[0127] The output end of the fourth dynamic feature extraction module is respectively connected to the input end of the seventeenth convolution unit, the input end of the eighteenth convolution unit and the input end of the fourth connection unit; the output end of the seventeenth convolution unit is respectively connected to the input end of the eighteenth convolution unit and the input end of the fourth connection unit; the output end of the eighteenth convolution unit is respectively connected to the input end of the nineteenth convolution unit and the input end of the fourth connection unit; the output end of the nineteenth convolution unit is connected to the input end of the fourth connection unit; the output end of the fourth connection unit is connected to the input end of the twentieth convolution unit; the output end of the twentieth convolution unit is connected to the input end of the fifteenth normalization unit; the output end of the fifteenth normalization unit is connected to the input end of the fifteenth activation unit; The output end of the activation unit is respectively connected to the input end of the first pooling unit, the input end of the fifth connection unit and the input end of the second pooling unit; the output end of the first pooling unit is connected to the input end of the second pooling unit; the output end of the second pooling unit is respectively connected to the input end of the third pooling unit and the input end of the fifth connection unit; the output end of the third pooling unit is connected to the input end of the fifth connection unit; the output end of the fifth connection unit is connected to the input end of the twenty-first convolution unit; the output end of the twenty-first convolution unit is connected to the input end of the sixteenth normalization unit; the output end of the sixteenth normalization unit is connected to the input end of the sixteenth activation unit; the output end of the sixteenth activation unit is connected to the input end of the first convolution feature extraction module.

[0128] It can be seen that implementation Figure 2 The described image processing device for lumbar spondylolysis detection is beneficial to improving the image recognition accuracy and recognition efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0129] In yet another optional embodiment, Figure 2 As shown, the target image recognition model is obtained based on the second processing module 203 executing the following steps:

[0130] Obtaining X-ray image information of lesions;

[0131] The lesion X-ray image information is screened and processed to obtain first X-ray image information; the first X-ray image information includes M first X-ray images; M is a positive integer;

[0132] Performing data enhancement processing on the first X-ray image information to obtain second X-ray image information; the second X-ray image information includes N second X-ray images; N is a positive integer not less than M;

[0133] performing image size optimization processing on the second X-ray image information to obtain third X-ray image information;

[0134] Dividing the third X-ray image information into a data set to obtain an X-ray image set; the X-ray image set includes a plurality of X-ray images to be used;

[0135] Acquire training condition information; the training condition information includes the number of sample images and the iteration threshold;

[0136] Initializing the initial image recognition model using the training condition information to obtain a first training image recognition model;

[0137] The number of sample images is used to screen and process the unused X-ray images in the X-ray image set to obtain target X-ray image information; the target X-ray image information includes L target X-ray images; L is consistent with the number of sample images;

[0138] Using the target X-ray image information to train the first training image recognition model, and obtaining a second training image recognition model and training result parameter information;

[0139] Determine whether the training result parameter information meets the model training termination condition and obtain the conditional judgment result; the model training termination condition represents the comparison between the number of training times and the iteration threshold;

[0140] When the result of the conditional judgment is no, the first training image recognition model is updated by using the second training image recognition model, and the screening process of the unused X-ray images in the X-ray image set by using the number of sample images is triggered to obtain target X-ray image information;

[0141] When the conditional judgment result is yes, the second training image recognition model is determined to be the target image recognition model.

[0142] It can be seen that the implementation Figure 2 The described image processing device for lumbar spondylolysis detection is beneficial to improving the image recognition accuracy and recognition efficiency of lumbar X-ray films, thereby solving the limitations of traditional methods in lumbar spondylolysis detection.

[0143] Embodiment 3

[0144] See also Figure 3 , Figure 3 is a schematic diagram of the structure of another image processing device for detecting spondylolysis in the embodiment of the present invention. Figure 3 The described device can be applied to a warehouse management system, such as a local server or cloud server for image processing management of warehouse logistics for lumbar spondylolysis detection, etc., and the embodiments of the present invention are not limited thereto. Figure 3 As shown, the device may include:

[0145] A memory 301 storing executable program codes;

[0146] a processor 302 coupled to the memory 301;

[0147] The processor 302 calls the executable program code stored in the memory 301 to execute the steps of the image processing method for lumbar spondylolysis detection described in the first embodiment.

[0148] Embodiment 4

[0149] An embodiment of the present invention discloses a computer-readable storage medium storing a computer program for electronic data exchange, wherein the computer program enables a computer to execute the steps of the image processing method for lumbar spondylolysis detection described in the first embodiment.

[0150] Embodiment 5

[0151] An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to enable a computer to execute the steps of the image processing method for lumbar spondylolysis detection described in the first embodiment.

[0152] The device embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0153] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or partly contributed to the prior art in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0154] Finally, it should be noted that the image processing method and device for lumbar spondylolysis detection disclosed in the embodiment of the present invention only disclose the preferred embodiments of the present invention, which are only used to illustrate the technical scheme of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical schemes described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical schemes from the spirit and scope of the technical schemes of the embodiments of the present invention.

Claims

1. An image processing method for detecting lumbar spondylolysis, characterized in that: The method comprises: Acquire lumbar vertebra image information to be processed; Pre-processing the lumbar vertebra image information to be processed to obtain initial image processing result information; Using a target image recognition model to perform recognition processing on the initial image processing result information to obtain target image recognition result information; Wherein, the target image recognition model includes a first convolution module, a second convolution module, a first normalization module, a second normalization module, a first activation module, a second activation module, a third activation module, a first dynamic feature extraction module, a second dynamic feature extraction module, a third dynamic feature extraction module, a fourth dynamic feature extraction module, a depth feature extraction module, a first convolution feature extraction module, a second convolution feature extraction module, a third convolution feature extraction module, a fourth convolution feature extraction module, a first interactive feature extraction module, a second interactive feature extraction module, a third interactive feature extraction module, a first sampling module, a second sampling module and a first fusion module; wherein, The input end of the first convolution module is connected to the model input of the target image recognition model, and the output end of the first convolution module is connected to the input end of the first normalization module; the output end of the first normalization module is connected to the input end of the first activation module; the output end of the first activation module is connected to the input end of the first dynamic feature extraction module; the output end of the first dynamic feature extraction module is connected to the input end of the second dynamic feature extraction module; the output end of the second dynamic feature extraction module is respectively connected to the input end of the third dynamic feature extraction module and the input end of the third convolution feature extraction module; the output end of the third dynamic feature extraction module is respectively connected to the input end of the fourth dynamic feature extraction module and the input end of the second convolution feature extraction module; the output end of the fourth dynamic feature extraction module is connected to the input end of the depth feature extraction module; the output end of the depth feature extraction module is respectively connected to the input end of the first convolution feature extraction module and the input end of the first sampling module; the output end of the first convolution feature extraction module is connected to the input end of the third interactive feature extraction module; the third interactive feature extraction module The output end of the block is connected to the input end of the first fusion module; the output end of the first sampling module is connected to the input end of the second convolution feature extraction module; the output end of the second convolution feature extraction module is respectively connected to the input end of the second sampling module and the input end of the fourth convolution feature extraction module; the output end of the second sampling module is connected to the input end of the first interactive feature extraction module; the output end of the first interactive feature extraction module is connected to the input end of the first fusion module; the output end of the fourth convolution feature extraction module is respectively connected to the input end of the second convolution module and the input end of the second interactive feature extraction module; the output end of the second convolution module is connected to the input end of the second normalization module; the output end of the second normalization module is connected to the input end of the second activation module; the output end of the second activation module is connected to the input end of the first convolution feature extraction module; the output end of the second interactive feature extraction module is connected to the input end of the first fusion module; the output end of the first fusion module is connected to the input end of the third activation module; the output end of the third activation module is connected to the model output of the target image recognition model.

2. The image processing method for lumbar spondylolysis detection according to claim 1, characterized in that: The first interactive feature extraction module includes a first convolution unit, a second convolution unit, a third convolution unit, a fourth convolution unit, a first normalization unit, a second normalization unit, a first activation unit, a second activation unit and a first fusion unit; wherein, The input end of the first convolution unit and the input end of the third convolution unit are both connected to the output end of the third convolution feature extraction module; the output end of the first convolution unit is connected to the input end of the first normalization unit; the output end of the first normalization unit is connected to the input end of the first activation unit; the output end of the first activation unit is connected to the input end of the second convolution unit; the output end of the second convolution unit is connected to the input end of the first fusion unit; the output end of the third convolution unit is connected to the input end of the second normalization unit; the output end of the second normalization unit is connected to the input end of the second activation unit; the output end of the second activation unit is connected to the input end of the fourth convolution unit; the output end of the fourth convolution unit is connected to the input end of the first fusion unit; the output end of the first fusion unit is connected to the input end of the first fusion module.

3. The image processing method for lumbar spondylolysis detection according to claim 1, characterized in that: The first dynamic feature extraction module includes a fifth convolution unit, a sixth convolution unit, a seventh convolution unit, an eighth convolution unit, a ninth convolution unit, a tenth convolution unit, an eleventh convolution unit, a third normalization unit, a fourth normalization unit, a fifth normalization unit, a sixth normalization unit, a seventh normalization unit, an eighth normalization unit, a ninth normalization unit, a third activation unit, a fourth activation unit, a fifth activation unit, a sixth activation unit, a seventh activation unit, an eighth activation unit, a ninth activation unit, a first slicing unit, a first connection unit, and a second fusion unit; wherein, The input end of the fifth convolution unit is connected to the output end of the first activation module, and the output end of the fifth convolution unit is connected to the input end of the third normalization unit; the output end of the third normalization unit is connected to the input end of the third activation unit; the output end of the third activation unit is connected to the input end of the sixth convolution unit; the output end of the sixth convolution unit is connected to the input end of the fourth normalization unit; the output end of the fourth normalization unit is connected to the input end of the fourth activation unit; the output end of the fourth activation unit is connected to the input end of the first slicing unit; the output end of the first slicing unit is respectively connected to the input end of the seventh convolution unit, the input end of the second fusion unit and the input end of the eleventh convolution unit; the output end of the seventh convolution unit is connected to the input end of the fifth normalization unit; the output end of the fifth normalization unit is connected to the input end of the fifth activation unit; the output end of the fifth activation unit is connected to the input end of the eighth convolution unit; the output end of the eighth convolution unit is connected to the input end of the sixth normalization unit; the output end of the sixth normalization unit is connected to the input end of the The input end of the sixth activation unit; the output end of the sixth activation unit is connected to the input end of the second fusion unit; the output end of the second fusion unit is respectively connected to the input end of the ninth convolution unit, the input end of the third fusion unit and the input end of the eleventh convolution unit; the output end of the ninth convolution unit is connected to the input end of the seventh normalization unit; the output end of the seventh normalization unit is connected to the input end of the seventh activation unit; the output end of the seventh activation unit is connected to the input end of the tenth convolution unit; the output end of the tenth convolution unit is connected to the input end of the eighth normalization unit; the output end of the eighth normalization unit is connected to the input end of the third fusion unit; the output end of the third fusion unit is connected to the input end of the first connection unit; the output end of the first connection unit is connected to the input end of the eleventh convolution unit; the output end of the eleventh convolution unit is connected to the input end of the ninth normalization unit; the output end of the ninth normalization unit is connected to the input end of the ninth activation unit; the output end of the ninth activation unit is connected to the input end of the second dynamic feature extraction module.

4. The image processing method for lumbar spondylolysis detection according to claim 1, characterized in that: The first convolution feature extraction module includes a second connection unit, a third connection unit, a second slicing unit, a twelfth convolution unit, a thirteenth convolution unit, a fourteenth convolution unit, a fifteenth convolution unit, a sixteenth convolution unit, a tenth normalization unit, an eleventh normalization unit, a twelfth normalization unit, a thirteenth normalization unit, a fourteenth normalization unit, a tenth activation unit, an eleventh activation unit, a twelfth activation unit, a thirteenth activation unit, a fourteenth activation unit, a fourth fusion unit and a fifth fusion unit; wherein, The input end of the second connection unit is connected to the output end of the deep feature extraction module, and the output end of the second connection unit is connected to the input end of the second slicing unit; the output end of the second slicing unit is respectively connected to the input end of the twelfth convolution unit, the input end of the fourth fusion unit, and the input end of the sixteenth convolution unit; the output end of the twelfth convolution unit is connected to the input end of the tenth normalization unit; the output end of the tenth normalization unit is connected to the input end of the tenth activation unit; the output end of the tenth activation unit is connected to the input end of the thirteenth convolution unit; the output end of the thirteenth convolution unit is connected to the input end of the eleventh normalization unit; the output end of the eleventh normalization unit is connected to the input end of the eleventh activation unit; the output end of the eleventh activation unit is connected to the input end of the fourth fusion unit; the output end of the fourth fusion unit is respectively connected to the input end of the sixteenth convolution unit, the input end of the fourteenth convolution unit and the input end of the The input end of the fifth fusion unit; the output end of the fourteenth convolution unit is connected to the input end of the twelfth normalization unit; the output end of the twelfth normalization unit is connected to the input end of the twelfth activation unit; the output end of the twelfth activation unit is connected to the input end of the fifteenth convolution unit; the output end of the fifteenth convolution unit is connected to the input end of the thirteenth normalization unit; the output end of the thirteenth normalization unit is connected to the input end of the thirteenth activation unit; the output end of the thirteenth activation unit is connected to the input end of the fifth fusion unit; the output end of the fifth fusion unit is connected to the input end of the third connection unit; the output end of the third connection unit is connected to the input end of the sixteenth convolution unit; the output end of the sixteenth convolution unit is connected to the input end of the fourteenth normalization unit; the output end of the tenth normalization unit is connected to the input end of the fourteenth activation unit; the output end of the fourteenth activation unit is connected to the input end of the third interactive feature extraction module.

5. The image processing method for lumbar spondylolysis detection according to claim 1, characterized in that: The deep feature extraction module includes a seventeenth convolution unit, an eighteenth convolution unit, a nineteenth convolution unit, a twentieth convolution unit, a twenty-first convolution unit, a fourth connection unit, a fifth connection unit, a fifteenth normalization unit, a sixteenth normalization unit, a fifteenth activation unit, a sixteenth activation unit, a first pooling unit, a second pooling unit, and a third pooling unit; wherein, The output end of the fourth dynamic feature extraction module is respectively connected to the input end of the seventeenth convolution unit, the input end of the eighteenth convolution unit and the input end of the fourth connection unit; the output end of the seventeenth convolution unit is respectively connected to the input end of the eighteenth convolution unit and the input end of the fourth connection unit; the output end of the eighteenth convolution unit is respectively connected to the input end of the nineteenth convolution unit and the input end of the fourth connection unit; the output end of the nineteenth convolution unit is connected to the input end of the fourth connection unit; the output end of the fourth connection unit is connected to the input end of the twentieth convolution unit; the output end of the twentieth convolution unit is connected to the input end of the fifteenth normalization unit; the output end of the fifteenth normalization unit is connected to the input end of the fifteenth activation unit; the fifteenth activation unit is connected to the input end of the fifteenth activation unit. The output end of the activation unit is respectively connected to the input end of the first pooling unit, the input end of the fifth connection unit and the input end of the second pooling unit; the output end of the first pooling unit is connected to the input end of the second pooling unit; the output end of the second pooling unit is respectively connected to the input end of the third pooling unit and the input end of the fifth connection unit; the output end of the third pooling unit is connected to the input end of the fifth connection unit; the output end of the fifth connection unit is connected to the input end of the twenty-first convolution unit; the output end of the twenty-first convolution unit is connected to the input end of the sixteenth normalization unit; the output end of the sixteenth normalization unit is connected to the input end of the sixteenth activation unit; the output end of the sixteenth activation unit is connected to the input end of the first convolution feature extraction module.

6. The image processing method for lumbar spondylolysis detection according to claim 1, characterized in that: The target image recognition model is obtained based on the following steps: Obtaining X-ray image information of lesions; The lesion X-ray image information is screened to obtain first X-ray image information; the first X-ray image information includes M first X-ray images; M is a positive integer; performing data enhancement processing on the first X-ray image information to obtain second X-ray image information; The second X-ray image information includes N second X-ray images; N is a positive integer not less than M; performing image size optimization processing on the second X-ray image information to obtain third X-ray image information; Dividing the third X-ray image information into a data set to obtain an X-ray image set; the X-ray image set includes a plurality of X-ray images to be used; Acquiring training condition information; the training condition information includes the number of sample images and an iteration threshold; Initializing the initial image recognition model using the training condition information to obtain a first training image recognition model; Using the number of sample images, screening the unused X-ray images in the X-ray image set to obtain target X-ray image information; The target X-ray image information includes L target X-ray images; L is consistent with the number of sample images; Using the target X-ray image information to train the first training image recognition model to obtain a second training image recognition model and training result parameter information; Determine whether the training result parameter information meets the model training termination condition, and obtain a conditional judgment result; The model training termination condition represents the comparison between the number of training times and the iteration threshold; When the result of the conditional judgment is no, the first training image recognition model is updated by using the second training image recognition model, and the method of screening the unused X-ray images in the X-ray image set by using the number of sample images is triggered to obtain target X-ray image information; When the conditional judgment result is yes, the second training image recognition model is determined to be the target image recognition model.

7. An image processing device for detecting lumbar spondylolysis, characterized in that: The device comprises: An acquisition module, used for acquiring lumbar vertebra image information to be processed; A first processing module is used to pre-process the lumbar vertebra image information to be processed to obtain initial image processing result information; A second processing module is used to perform recognition processing on the initial image processing result information using a target image recognition model to obtain target image recognition result information; Wherein, the target image recognition model includes a first convolution module, a second convolution module, a first normalization module, a second normalization module, a first activation module, a second activation module, a third activation module, a first dynamic feature extraction module, a second dynamic feature extraction module, a third dynamic feature extraction module, a fourth dynamic feature extraction module, a depth feature extraction module, a first convolution feature extraction module, a second convolution feature extraction module, a third convolution feature extraction module, a fourth convolution feature extraction module, a first interactive feature extraction module, a second interactive feature extraction module, a third interactive feature extraction module, a first sampling module, a second sampling module and a first fusion module; wherein, The input end of the first convolution module is connected to the model input of the target image recognition model, and the output end of the first convolution module is connected to the input end of the first normalization module; the output end of the first normalization module is connected to the input end of the first activation module; the output end of the first activation module is connected to the input end of the first dynamic feature extraction module; the output end of the first dynamic feature extraction module is connected to the input end of the second dynamic feature extraction module; the output end of the second dynamic feature extraction module is respectively connected to the input end of the third dynamic feature extraction module and the input end of the third convolution feature extraction module; the output end of the third dynamic feature extraction module is respectively connected to the input end of the fourth dynamic feature extraction module and the input end of the second convolution feature extraction module; the output end of the fourth dynamic feature extraction module is connected to the input end of the depth feature extraction module; the output end of the depth feature extraction module is respectively connected to the input end of the first convolution feature extraction module and the input end of the first sampling module; the output end of the first convolution feature extraction module is connected to the input end of the third interactive feature extraction module; the third interactive feature extraction module The output end of the block is connected to the input end of the first fusion module; the output end of the first sampling module is connected to the input end of the second convolution feature extraction module; the output end of the second convolution feature extraction module is respectively connected to the input end of the second sampling module and the input end of the fourth convolution feature extraction module; the output end of the second sampling module is connected to the input end of the first interactive feature extraction module; the output end of the first interactive feature extraction module is connected to the input end of the first fusion module; the output end of the fourth convolution feature extraction module is respectively connected to the input end of the second convolution module and the input end of the second interactive feature extraction module; the output end of the second convolution module is connected to the input end of the second normalization module; the output end of the second normalization module is connected to the input end of the second activation module; the output end of the second activation module is connected to the input end of the first convolution feature extraction module; the output end of the second interactive feature extraction module is connected to the input end of the first fusion module; the output end of the first fusion module is connected to the input end of the third activation module; the output end of the third activation module is connected to the model output of the target image recognition model.

8. An image processing device for detecting lumbar spondylolysis, characterized in that: The device comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the image processing method for lumbar spondylolysis detection as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, which, when called, are used to execute the image processing method for lumbar spondylolysis detection as described in any one of claims 1-6.

Images