Digital image intelligent system for urinary surgery

Abstract

The invention discloses a digital image intelligent system for urinary surgery, which relates to the technical field of digital image processing, and comprises a first image acquisition module for acquiring an initial image of a urinary disease part of a patient; the image analysis module is used for screening out an initial region with stones; the second image acquisition module is used for acquiring a high-resolution image of the initial region; and the recognition module is used for recognizing the stone in the high-resolution image through a deep learning technology and extracting position information and form information of the stone. According to the urinary calculus recognition system and method, efficient and accurate recognition of urinary calculus is achieved by integrating a preliminary image acquisition module, a high-resolution image acquisition module, a deep learning analysis module, a morphological analysis module, a type judgment module, a position positioning module, a comprehensive data integration module and the like.

Description

Technical Field

[0001] The invention relates to the technical field of digital image processing, in particular to an intelligent digital image system for urology surgery. Background Art

[0002] The digital imaging system of urology plays an important role in the detection and diagnosis of urinary stones, but it currently has the following shortcomings, which affect the recognition efficiency and diagnostic effect of urinary stones.

[0003] Urinary stones are often obscured by surrounding tissues or other high-density structures (such as bone or other tissue calcification) in images, making it difficult to identify stones. For some types of stones, the image contrast is not enough to distinguish the stones from the surrounding soft tissues. In particular, some low-density stones may be difficult to identify in conventional X-ray images, and some digital imaging systems have low imaging resolution and cannot clearly display the boundaries and morphological characteristics of stones. Summary of the invention

[0004] 1) Technical issues solved

[0005] The present invention provides a urology digital imaging intelligent system, which can distinguish the boundaries and morphological features of stones in images.

[0006] 2) Technical solution

[0007] To achieve the above object, the present invention provides the following technical solution: a urology digital imaging intelligent system, comprising:

[0008] A first image acquisition module is used to obtain an initial image of the patient's urinary disease site;

[0009] An image analysis module, used for receiving the initial image transmitted from the first image acquisition module for processing and initial analysis, wherein the image analysis module performs noise removal, contrast enhancement and boundary detection preprocessing on the initial image based on an image processing algorithm to screen out an initial area where stones exist;

[0010] A second image acquisition module is configured to screen out the initial area of ​​the stone according to the image analysis module and obtain a high-resolution image of the initial area;

[0011] The recognition module is used to receive the high-resolution image of the initial area provided by the second image acquisition module. The recognition module recognizes the stones in the high-resolution image through deep learning technology and extracts the location information and morphological information of the stones.

[0012] Furthermore, the image analysis module pre-processes the initial image based on an image processing algorithm to screen out the initial area where stones exist, specifically including the following steps:

[0013] Using a filtering algorithm to remove noise in the initial image and enhance the contrast between the stone and surrounding tissue;

[0014] The Canny edge detection is used to detect the boundary of the image after noise removal and contrast enhancement, and the boundary contour of the potential stone is extracted;

[0015] The initial area where stones exist is initially screened out based on grayscale value, shape and edge features.

[0016] Furthermore, the second image acquisition module automatically adjusts the imaging device to focus on the initial area, and automatically adjusts the focal length and scanning angle to obtain a high-resolution image of the initial area.

[0017] Furthermore, the imaging device identifies the depth of the initial area, and dynamically adjusts the focal length based on an autofocus algorithm, while rotating the scanning angle to avoid bone occlusion, ultimately generating a high-resolution image containing the complete structure of the initial area.

[0018] Furthermore, the high-resolution image of the initial area transmitted by the second image acquisition module is used as input to the recognition module, and the pre-trained deep learning model is run in the recognition module to identify the stone area on the high-resolution image and generate a binary image with the stone boundary marked.

[0019] Furthermore, the recognition module uses image segmentation technology to extract the area of ​​the stone from the binary image, performs morphological analysis on the binary image based on mathematical morphology image operations, and extracts the size and shape characteristics of the stone from the binary image.

[0020] Furthermore, the recognition module determines the type of the stone based on the shape characteristics and imaging characteristics of the stone.

[0021] Furthermore, the recognition module calculates the position of the stone in the binary image plane based on the segmentation result of the binary image.

[0022] 3) Beneficial effects:

[0023] Compared with the prior art, the invention has the following beneficial effects:

[0024] The present invention realizes efficient and accurate identification of urinary stones by integrating modules such as preliminary image acquisition, high-resolution image acquisition, deep learning analysis, morphological analysis, type judgment, position positioning and comprehensive data integration. The system can effectively solve the problem that stones are difficult to identify due to the occlusion of surrounding tissues or high-density structures, and low-resolution images cannot clearly display the boundaries and morphology of stones. The image quality is optimized by adaptive imaging technology, and the location and boundaries of stones are accurately identified by combining deep learning models. At the same time, morphological analysis and type judgment are used to provide detailed stone feature information, and a panoramic diagnostic report is generated through comprehensive analysis. The system not only improves the accuracy and efficiency of stone identification, but also provides a comprehensive reference for clinical treatment, helps personalized diagnosis and treatment decisions, and significantly improves the intelligent level of urological diagnosis and treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A urology digital imaging intelligent system provided by an embodiment of the present invention;

[0026] Figure 2 A urology digital imaging intelligent system provided by an embodiment of the present invention;

[0027] In the figure:

[0028] 10. First image acquisition module; 20. Image analysis module; 30. Second image acquisition module; 40. Recognition module. DETAILED DESCRIPTION

[0029] The technical solutions 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, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0030] In the description of the present invention, it should be understood that the terms "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside" and "outside" etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0031] In addition, the terms “first”, “second”, etc., if used, are merely used to distinguish between the descriptions and should not be understood as indicating or implying relative importance.

[0032] It should be noted that, in the absence of conflict, the features in the embodiments of the present invention may be combined with each other.

[0033] Combination Figure 1 to Figure 2 A urology digital imaging intelligent system is shown, specifically, the first image acquisition module 10 is used to obtain preliminary image data of the patient's urinary system, and the first image acquisition module 10 may include imaging equipment such as CT, X-ray, and ultrasound, and first provides preliminary images of the patient's urethra, bladder, kidneys, etc. This module is connected to the image analysis module 20, and transmits the preliminary acquired images to the image analysis module 20 for preliminary processing and boundary extraction.

[0034] In some embodiments of the present invention, conventional digital imaging equipment (such as an X-ray machine or a low-resolution CT) is used to obtain an initial image of the patient's urinary system, and the image acquisition device generates a two-dimensional image of the patient's urinary area according to preset parameters (such as exposure time and radiation dose) as basic data for subsequent analysis.

[0035] The image analysis module 20 is responsible for preprocessing and preliminary analysis of the image from the first image acquisition module 10. This module includes image denoising, contrast enhancement, boundary detection and preliminary stone area screening. The specific processing includes: image noise removal and local contrast enhancement; multi-scale edge detection algorithm to extract possible stone areas; and preliminary stone screening based on grayscale value and morphological characteristics. The image analysis module 20 receives the preliminary image transmitted by the first image acquisition module 10, and after processing, transmits the result to the second image acquisition module 30 for further high-resolution image acquisition, and completes the final identification of the stone through the identification module 40.

[0036] In summary, the image analysis module 20 uses the initial image data to perform preliminary analysis through the image processing algorithm, quickly detects areas suspected of stones, and marks key areas for further verification. More specifically, the image analysis module 20 performs operations such as denoising, contrast enhancement, and edge smoothing on the collected initial image to improve the image quality. Then, an edge detection algorithm (such as the Canny algorithm or the Sobel operator) is used to identify the boundaries of high-density areas in the image. Through conditions such as brightness and morphological features, key areas suspected of stones are quickly screened, and annotated images are generated to indicate areas where stones may exist and transmitted to the second image acquisition module 30.

[0037] For example, the system processes X-ray images and detects a high-density structure in the kidney area with blurred boundaries but a shape close to an ellipse, and preliminarily marks the area as a suspected stone area.

[0038] The second image acquisition module 30 is used to obtain high-resolution images of the candidate areas of the stones preliminarily screened by the image analysis module 20. According to the specific location and morphological characteristics of the stone area, a higher-resolution scanning device (such as a high-resolution CT or ultrasound device) is used for detailed imaging to accurately locate the stone and its boundaries. It can be understood that this module receives the candidate area information transmitted by the image analysis module 20, obtains high-resolution images, and transmits these images to the recognition module 40 for further analysis and recognition.

[0039] In some feasible embodiments of the present invention, the functions of the second image acquisition module 30 include automatically adjusting the imaging device to focus on the stone area according to the candidate area provided by the first image acquisition and image analysis module 20. Automatically adjust the focal length and scanning angle to obtain a high-resolution image of the area. If there is a situation where the image resolution is insufficient, super-resolution reconstruction technology is used to enhance the image details to ensure that the boundaries and morphology of the stone can be clearly presented. And based on the feedback of the real-time imaging results, the system can continue to adjust the imaging parameters to ensure that the acquired image meets the recognition requirements. The second image acquisition module 30 transmits the acquired high-resolution image to the recognition module 40 for in-depth analysis to ensure accurate recognition and morphological extraction of the stone.

[0040] Specifically, the second image acquisition module 30 uses a high-resolution imaging device (such as a high-definition CT or a micro-focus X-ray machine) to perform a detailed scan of the initially marked key areas to obtain high-resolution images for subsequent precise analysis. According to the annotation information provided by the image analysis module 20, the imaging device is focused on the suspected stone area, and adaptive imaging technology is introduced to adjust imaging parameters such as resolution and scanning speed according to the characteristics of the target area (such as density and volume). Obtain a high-definition image of the target area to ensure that the boundaries and morphological features are clear, and pass the high-resolution image to the recognition module 40 for further processing.

[0041] For example, based on the preliminary annotation, the high-definition CT scanner of the second image acquisition module 30 scans the suspected stone area in the kidney, generating a higher resolution image that clearly shows the outline of the stone.

[0042] Finally, the recognition module 40 is the core of the system and is mainly responsible for automatic recognition and analysis of stones based on the high-resolution images provided by the second image acquisition module 30. This module recognizes stones through deep learning technology and extracts information such as the accurate location, size, shape, type, etc. of the stones. Its specific functions include the following.

[0043] Deep learning algorithms such as convolutional neural networks (CNN) are used to automatically analyze the boundaries, size, shape and other features of stones in the image.

[0044] More specifically, deep learning technology is used to analyze the stone features in high-resolution images and automatically identify the location and boundaries of the stones. Here, a convolutional neural network (CNN), including multiple layers of convolution, pooling, and fully connected layers, can be used to extract the texture, edge, and shape features of the image. Prior to this, a large number of annotated urinary stone image datasets were used for training, so that the model can learn and accurately distinguish the characteristics of stones and surrounding tissues. The brightness, contrast, and shape features in the image are identified by the deep learning model, and the boundary detection method is combined to accurately locate the stone.

[0045] In summary, the high-resolution image transmitted by the second image acquisition module 30 is used as input, and then input into the recognition module 40 to run the pre-trained deep learning model (such as UNet, ResNet, etc.) to perform stone area recognition, and generate a binary image marked with stone boundaries (stone is 1, non-stone area is 0).

[0046] The morphology of the stones was analyzed, including the edge smoothness, contrast, volume, etc.

[0047] More specifically, a morphological analysis is performed on the identified stone area to extract the geometric features of the stone, including size, boundary smoothness, volume, and morphological structure. This includes calculating the area of ​​the stone to provide an accurate size, analyzing whether the edge of the stone is regular to assist in determining the type of stone, and determining the shape of the stone (such as elliptical, irregular, etc.) and whether there are multiple stones. Here, the stone area can be extracted using image segmentation technology, and image operations based on mathematical morphology (such as dilation and erosion) can be used to further optimize the stone boundary.

[0048] In summary, the stone area images output by deep learning analysis are subjected to morphological analysis, and the size and shape features are extracted through image processing algorithms (such as regional projection, boundary fitting, etc.), and finally the detailed geometric parameters of the stone are provided.

[0049] Based on the morphology and texture characteristics of the stones, determine the type of stones (such as calcium oxalate stones, uric acid stones, etc.).

[0050] More specifically, the recognition module 40 determines the type of stone based on the morphological characteristics, texture characteristics and imaging characteristics (such as density) of the stone, providing a reference for subsequent treatment. Regarding the types of stones, they include calcium oxalate stones (high density, clear boundaries), uric acid stones (low density, irregular shape) and phosphate stones (usually multiple, more regular shape). Here, an image classification model can be used, combined with texture analysis and density measurement, and image characteristic analysis (such as CT value range) can be introduced to accurately classify stones.

[0051] In summary, texture analysis and statistical methods are combined to extract features such as stone density, brightness, and contrast, and classification algorithms or pre-trained deep learning models are used to determine the stone type. Finally, the stone type and corresponding probability can be output into the diagnosis report.

[0052] Based on the stone boundaries in the image, the specific location of the stone can be accurately located, including the left or right kidney, ureter or bladder.

[0053] More specifically, the recognition module 40 accurately locates the position of the stone according to the image data, marks its specific location in the urinary system (such as the kidney, ureter, bladder, etc.), and can calculate the coordinates of the stone in the image plane based on the image segmentation results.

[0054] In summary, the recognition module 40 collects all the results of the aforementioned functional modules, obtains comprehensive diagnostic information through the rule base and data mining technology, and finally presents the comprehensive analysis results in the form of charts and text, and provides them to doctors and patients. In order to achieve efficient and accurate recognition of urinary stones by integrating modules such as preliminary image acquisition, high-resolution image acquisition, deep learning analysis, morphological analysis, type judgment, position positioning and comprehensive data integration. The system can effectively solve the problem that stones are difficult to identify due to the occlusion of surrounding tissues or high-density structures, and low-resolution images cannot clearly display the boundaries and morphology of stones. The image quality is optimized through adaptive imaging technology, and the location and boundaries of stones are accurately identified by combining deep learning models. At the same time, morphological analysis and type judgment are used to provide detailed stone feature information, and a panoramic diagnosis report is generated through comprehensive analysis. The system not only improves the accuracy and efficiency of stone identification, but also provides a comprehensive reference for clinical treatment, helps personalized diagnosis and treatment decisions, and significantly improves the intelligent level of urology diagnosis and treatment.

[0055] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. The patent protection scope of the present invention shall be based on the claims. All equivalent structural changes made using the contents of the description and drawings of the present invention should also be included in the protection scope of the present invention.

Claims

1. A urology digital imaging intelligent system, characterized in that: include: A first image acquisition module is used to obtain an initial image of the patient's urinary disease site; An image analysis module, used for receiving the initial image transmitted from the first image acquisition module for processing and initial analysis, wherein the image analysis module performs noise removal, contrast enhancement and boundary detection preprocessing on the initial image based on an image processing algorithm to screen out an initial area where stones exist; A second image acquisition module is configured to screen out the initial area of ​​the stone according to the image analysis module and obtain a high-resolution image of the initial area; The recognition module is used to receive the high-resolution image of the initial area provided by the second image acquisition module. The recognition module recognizes the stones in the high-resolution image through deep learning technology and extracts the location information and morphological information of the stones.

2. The urology digital imaging intelligent system according to claim 1, characterized in that: The image analysis module pre-processes the initial image based on the image processing algorithm to screen out the initial area where the stones exist, specifically including the following steps: Using a filtering algorithm to remove noise in the initial image and enhance the contrast between the stone and surrounding tissue; The Canny edge detection is used to detect the boundary of the image after noise removal and contrast enhancement, and the boundary contour of the potential stone is extracted; The initial area where stones exist is initially screened out based on grayscale value, shape and edge features.

3. The urology digital imaging intelligent system according to claim 1, characterized in that: The second image acquisition module automatically adjusts the imaging device to focus on the initial area, and automatically adjusts the focal length and scanning angle to obtain a high-resolution image of the initial area.

4. The urology digital imaging intelligent system according to claim 3, characterized in that: The imaging device identifies the depth of the initial area, dynamically adjusts the focal length based on an autofocus algorithm, and rotates the scanning angle to avoid bone occlusion, ultimately generating a high-resolution image containing the complete structure of the initial area.

5. The urology digital imaging intelligent system according to claim 1, characterized in that: The high-resolution image of the initial area transmitted by the second image acquisition module is used as input to the recognition module, and the pre-trained deep learning model is run in the recognition module to identify the stone area on the high-resolution image and generate a binary image with the stone boundary marked.

6. The urology digital imaging intelligent system according to claim 5, characterized in that: The recognition module uses image segmentation technology to extract the area of ​​the stone from the binary image, performs morphological analysis on the binary image based on mathematical morphology image operations, and extracts the size and shape characteristics of the stone from the binary image.

7. The urology digital imaging intelligent system according to claim 6, characterized in that: The recognition module determines the type of the stone based on the shape characteristics and imaging characteristics of the stone.

8. The urology digital imaging intelligent system according to claim 6, characterized in that: The recognition module calculates the position of the stone in the binary image plane based on the segmentation result of the binary image.

Images