CN117495818A - Orthopedics auxiliary examination method and system based on image processing - Google Patents

Abstract

The present invention proposes an orthopedic auxiliary examination method and system based on image processing. The orthopedic auxiliary examination method based on image processing includes: monitoring the patient's orthopedic examination data in real time, performing preliminary image processing on the orthopedic examination data, and obtaining orthopedic image data after preliminary image processing; using a boundary line segmentation method to segment the orthopedic examination data. Perform image segmentation on the orthopedic image data to obtain the bone structure area in the orthopedic image data; judge the disease area on the bone structure area to determine whether there is a disease area in the bone structure area; input the disease area into the pre-set Perform image analysis and diagnosis on the trained image recognition model and obtain a diagnosis report. The system includes modules corresponding to the method steps.

Description

An orthopedic auxiliary examination method and system based on image processing

Technical field

The invention proposes an orthopedic auxiliary examination method and system based on image processing, which belongs to the technical field of orthopedic auxiliary examination.

Background technique

The field of orthopedic care requires the use of a variety of auxiliary examination methods to diagnose and evaluate patients' bone health. These methods include X-ray photography, computed tomography (CT scan), magnetic resonance imaging (MRI), ultrasound, etc. Image processing technology plays an important role in orthopedic medicine. This includes enhancement, segmentation, registration, reconstruction and measurement of bone images to extract useful information to assist doctors in diagnosis and surgical planning. The problem of low efficiency of orthopedic image processing in the existing technology.

Contents of the invention

The present invention provides an orthopedic auxiliary examination method and system based on image processing to solve the problem of low orthopedic image processing efficiency in the prior art. The technical solutions adopted are as follows:

An image processing-based orthopedic auxiliary inspection method, the image processing-based orthopedic auxiliary inspection method includes:

Monitor the patient's orthopedic examination data in real time, perform preliminary image processing on the orthopedic examination data, and obtain orthopedic image data after preliminary image processing;

Perform image segmentation on the orthopedic image data using a boundary line segmentation method to obtain bone structure areas in the orthopedic image data;

Perform disease area judgment on the bone structure area to determine whether there is a disease area in the bone structure area;

The lesion area is input into the pre-trained image recognition model for image analysis and diagnosis, and a diagnosis report is obtained.

Further, the patient's orthopedic examination data is monitored in real time, and preliminary image processing is performed on the orthopedic examination data to obtain orthopedic image data after preliminary image processing, including:

Monitor and receive the patient's orthopedic examination data in real time, wherein the orthopedic examination data includes orthopedic image data corresponding to X-rays, CT scans, and MRI;

Perform image noise reduction processing on the orthopedic image data, remove noise and artifacts in the orthopedic image data, and obtain orthopedic image data after image noise reduction processing;

Perform image contrast and clarity enhancement processing on the orthopedic image data after image noise reduction processing to obtain enhanced orthopedic image data;

Image correction is performed on the enhanced orthopedic image data to obtain image-corrected orthopedic image data. The image-corrected orthopedic image data is the orthopedic image data after the preliminary image processing.

Furthermore, when performing image correction, the following algorithm is used:

Step 1: Assume P(x, y) is a point in the deformed image, then its incident angle α with the center of the image is:

Step 2: According to the incident angle calculated in step 1, obtain the isoconcentric projection correction coefficient of the center point P of the deformed image. The calculation formula is:

Among them, r is the equal angle projection correction coefficient, and π is 3.14.

Step 3: Calculate the coordinates of the corrected point P based on the equirectangular projection correction coefficient calculated in step 2. The calculation formula is as follows:

Among them, x′ is the corrected abscissa, y′ is the corrected ordinate, that is, the point P(x, y) in the deformed image is P(x′, y′) after correction.

Further, performing disease area judgment on the bone structure area to determine whether there is a disease area in the bone structure area includes:

Set the first global threshold and the second global threshold;

Compare the grayscale value of each pixel block in the orthopedic image data to the first global threshold and the second global threshold in sequence;

When the grayscale value of the pixel block in the orthopedic image data is less than the first global threshold, then the pixel block corresponding to the grayscale value less than the first global threshold is determined to be normal bone tissue;

When the grayscale value of a pixel block in the orthopedic image data is greater than or equal to the first global threshold, but less than the second global threshold, the pixel block is determined to be a suspected abnormal bone tissue;

When the grayscale value of the pixel block in the orthopedic image data is greater than or equal to the second global threshold, it is determined that the pixel block is abnormal bone tissue;

Divide normal bone tissue, suspected abnormal bone tissue, and areas corresponding to abnormal bone tissue to obtain bone structure areas with possible pathological changes.

Further, setting the first global threshold and the second global threshold includes:

Collect bone image data with diseased areas;

Perform histogram analysis on the bone image data with the diseased area, and obtain a histogram corresponding to the pixel gray value of the bone image with the diseased area;

Extract the gray value corresponding to one or more peak points of the histogram corresponding to the pixel gray value;

When there is only one peak point, the first global threshold and the second global threshold are 0.68H and 0.82H respectively; where H represents the gray value corresponding to the peak point;

When there are multiple peak points, the first global threshold and the second global threshold are 0.63×0.5×(H _min +H _max ) and 0.77×0.5×(H _min +H _max ) respectively; where H _min and H _max respectively represents the lowest gray value and the highest gray value corresponding to multiple peak points.

An orthopedic auxiliary examination system based on image processing. The orthopedic auxiliary examination system based on image processing includes:

A real-time monitoring module, used to monitor the patient's orthopedic examination data in real time, perform preliminary image processing on the orthopedic examination data, and obtain orthopedic image data after preliminary image processing;

An image segmentation module, used to perform image segmentation on the orthopedic image data using a boundary line segmentation method to obtain the bone structure area in the orthopedic image data;

A diseased area acquisition module, used to determine the diseased area of the bone structure area and determine whether there is a diseased area in the bone structure area;

The recognition and diagnosis module is used to input the lesion area into a pre-trained image recognition model for image analysis and diagnosis, and obtain a diagnosis report.

Further, the real-time monitoring module includes:

The monitoring execution module is used to monitor and receive the patient's orthopedic examination data in real time, where the orthopedic examination data includes orthopedic image data corresponding to X-rays, CT scans and MRI;

An image noise reduction module, used to perform image noise reduction processing on the orthopedic image data, remove noise and artifacts in the orthopedic image data, and obtain orthopedic image data after image noise reduction processing;

An image enhancement module, configured to perform image contrast and clarity enhancement processing on the orthopedic image data after image noise reduction processing, and obtain enhanced orthopedic image data;

The image correction module is used to perform image correction on the enhanced orthopedic image data to obtain image-corrected orthopedic image data. The image-corrected orthopedic image data is the orthopedic image data after the preliminary image processing.

Furthermore, when performing image correction, the following algorithm is used:

Step 1: Assume P(x, y) is a point in the deformed image, then its incident angle α with the center of the image is:

Step 2: According to the incident angle calculated in step 1, obtain the isoconcentric projection correction coefficient of the center point P of the deformed image. The calculation formula is:

Among them, r is the equal angle projection correction coefficient, and π is 3.14.

Step 3: Calculate the coordinates of the corrected point P based on the equirectangular projection correction coefficient calculated in step 2. The calculation formula is as follows:

Among them, x′ is the corrected abscissa, y′ is the corrected ordinate, that is, the point P(x, y) in the deformed image is P(x′, y′) after correction.

Further, the lesion area acquisition module includes:

A threshold setting module, used to set the first global threshold and the second global threshold;

A gray value comparison module, configured to compare the gray value of each pixel block in the orthopedic image data with the first global threshold and the second global threshold in sequence;

A first determination module configured to determine, when the grayscale value of a pixel block in the orthopedic image data is less than a first global threshold, that the pixel block corresponding to a grayscale value less than the first global threshold is normal bone tissue;

A second judgment module configured to determine that the pixel block is a suspected abnormal bone tissue when the grayscale value of the pixel block in the orthopedic image data is greater than or equal to the first global threshold, but less than the second global threshold;

A third determination module, configured to determine that the pixel block is abnormal bone tissue when the grayscale value of the pixel block in the orthopedic image data is greater than or equal to the second global threshold;

The division module is used to divide normal bone tissue and areas corresponding to suspected abnormal bone tissue and abnormal bone tissue to obtain bone structure areas with the possibility of disease.

Further, the threshold setting module includes:

Data collection module, used to collect bone image data with diseased areas;

The histogram analysis module is used to perform histogram analysis on the bone image data with the diseased area, and obtain the histogram corresponding to the pixel gray value of the bone image with the diseased area;

A gray value extraction module, used to extract the gray value corresponding to one or more peak points of the histogram corresponding to the pixel gray value;

The first threshold setting module is used to set the first global threshold and the second global threshold to 0.68H and 0.82H respectively when there is only one peak point; where H represents the gray value corresponding to the peak point;

The second threshold setting module is used to set the first global threshold and the second global threshold to 0.63×0.5×(H _min +H _max ) and 0.77×0.5×(H _min +H respectively) when there are multiple peak points. _max ); where H _min and H _max respectively represent the lowest gray value and the highest gray value corresponding to multiple peak points.

Beneficial effects of the present invention:

The image processing-based orthopedic auxiliary examination method and system proposed by the present invention allow real-time monitoring of the patient's orthopedic condition and can detect potential orthopedic problems earlier. Automated image processing and diagnostic processes can improve diagnostic efficiency and reduce doctors' workload. Using technologies such as deep learning for image analysis and diagnosis can improve the accuracy of diagnosis and reduce the risk of misdiagnosis. The system can generate instant diagnostic reports, providing the medical team with timely information to support treatment decisions.

Description of drawings

Figure 1 is a flow chart of the method of the present invention;

Figure 2 is a system block diagram of the system of the present invention.

Detailed ways

The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described here are only used to illustrate and explain the present invention, and are not intended to limit the present invention.

An embodiment of the present invention proposes an auxiliary orthopedic examination method based on image processing, as shown in Figure 1. The auxiliary orthopedic examination method based on image processing includes:

S1. Monitor the patient's orthopedic examination data in real time, perform preliminary image processing on the orthopedic examination data, and obtain orthopedic image data after preliminary image processing;

S2. Use the boundary line segmentation method to perform image segmentation on the orthopedic image data to obtain the bone structure area in the orthopedic image data;

S3. Determine the diseased area of the bone structure area to determine whether there is a diseased area in the bone structure area;

S4. Input the lesion area into the pre-trained image recognition model for image analysis and diagnosis, and obtain a diagnosis report.

The working principle of the above technical solution is: real-time monitoring of the patient's orthopedic examination data: First, the system monitors the patient's orthopedic examination data in real time, which can include imaging data such as X-ray, MRI or CT, which are usually used to examine the patient's bone structure. .

Preliminary image processing: Preliminary image processing of orthopedic examination data acquired in real time, which may include denoising, enhancement, or other image preprocessing methods to obtain clearer and analyzable image data.

Image segmentation using boundary line segmentation: Use boundary line segmentation to segment images to separate bone structure areas in orthopedic images from surrounding tissues. This helps the system focus on the analysis of bone structure.

Lesion area judgment: Within the bone structure area, the system will judge the lesion area to determine whether there is a possible lesion. This may involve detecting abnormal areas, fracture lines, tumors or other abnormalities.

Application of image recognition model: If there are lesion areas, the system inputs these areas into a pre-trained image recognition model. This model could be a deep learning neural network with the ability to identify different orthopedic lesions.

Obtain a diagnostic report: The image recognition model analyzes the image and generates a diagnostic report, which typically includes a detailed description of the patient's orthopedic condition, including lesion type, location, size, and severity.

The effect of the above technical solution is that the image processing-based orthopedic auxiliary examination method proposed in this embodiment can monitor the patient's orthopedic condition in real time and detect potential orthopedic problems earlier. Automated image processing and diagnostic processes can improve diagnostic efficiency and reduce doctors' workload. Using technologies such as deep learning for image analysis and diagnosis can improve the accuracy of diagnosis and reduce the risk of misdiagnosis. The method is able to generate instant diagnostic reports, providing the medical team with timely information to support treatment decisions.

One embodiment of the present invention monitors the patient's orthopedic examination data in real time, performs preliminary image processing on the orthopedic examination data, and obtains orthopedic image data after preliminary image processing, including:

S101. Monitor and receive the patient's orthopedic examination data in real time, where the orthopedic examination data includes orthopedic image data corresponding to X-rays, CT scans and MRI;

S102. Perform image noise reduction processing on the orthopedic image data, remove noise and artifacts in the orthopedic image data, and obtain orthopedic image data after image noise reduction processing;

S103. Perform image contrast and clarity enhancement processing on the orthopedic image data after image noise reduction processing to obtain enhanced orthopedic image data;

S104. Perform image correction on the enhanced orthopedic image data to obtain image-corrected orthopedic image data. The image-corrected orthopedic image data is the orthopedic image data after the preliminary image processing.

Specifically, since the enhanced image data may be deformed and distorted, it will greatly interfere with the subsequent acquisition of the lesion area. In severe cases, it may cause the acquisition of the lesion area to be offset or incomplete, which will affect subsequent diagnosis and surgical planning. There are significant adverse effects, so when performing image correction, the following algorithm is used:

Step 1: Assume P(x, y) is a point in the deformed image, then its incident angle α with the center of the image is:

Step 2: According to the incident angle calculated in step 1, obtain the isoconcentric projection correction coefficient of the center point P of the deformed image. The calculation formula is:

Among them, r is the equal angle projection correction coefficient, and π is 3.14.

Step 3: Calculate the coordinates of the corrected point P based on the equirectangular projection correction coefficient calculated in step 2. The calculation formula is as follows:

Among them, x′ is the corrected abscissa, y′ is the corrected ordinate, that is, the point P(x, y) in the deformed image is P(x′, y′) after correction.

This algorithm uses an isometric angle correction model to correct the image, avoiding the interference of the deformed and distorted image on the subsequent acquisition of the lesion area, facilitating subsequent accurate judgment of the scope of abnormal bone tissue, and providing accurate imaging data and judgment for subsequent diagnosis and surgical planning. in accordance with.

The working principle of the above technical solution is: Real-time monitoring and reception of orthopedic examination data: The system first monitors and receives the patient's orthopedic examination data in real time, which can include various types of orthopedic image data such as X-rays, CT scans and MRIs.

Image noise reduction processing: perform image noise reduction processing on the received orthopedic image data. This step aims to remove noise and artifacts from the image to improve the quality and clarity of the image. Denoising helps the system analyze bone structure more accurately.

Image contrast and sharpness enhancement processing: Perform contrast and sharpness enhancement processing on orthopedic image data that has been processed by noise reduction. This step enhances the contrast in the image, making details in the image more clearly visible, helping doctors better diagnose problems.

Image Correction: Perform image correction on enhanced orthopedic image data. Image correction corrects geometric or other distortions in images to ensure image accuracy and reliability.

The effect of the above technical solution is: Improved image quality: Through processing steps such as denoising, enhancement and correction, this technical solution can significantly improve the quality of orthopedic images, allowing doctors to more easily analyze and diagnose the patient's bone structure.

Reduced misdiagnosis: Improvements in image processing help reduce the risk of misdiagnosis and increase diagnostic accuracy.

Improve efficiency: Preliminary image processing can be automatically performed before the doctor's diagnosis, thereby improving the efficiency of diagnosis, shortening waiting time, and making the medical diagnosis process faster.

Overall, this technical solution helps improve the quality of orthopedic images and the accuracy of medical diagnosis, and improves the patient's diagnosis and treatment experience.

In one embodiment of the present invention, performing disease area determination on the skeletal structure area to determine whether there is a diseased area in the skeletal structure area includes:

S301. Set the first global threshold and the second global threshold;

S302. Compare the gray value of each pixel block in the orthopedic image data with the first global threshold and the second global threshold in sequence;

S303. When the gray value of the pixel block in the orthopedic image data is less than the first global threshold, then determine the pixel block corresponding to the gray value less than the first global threshold as normal bone tissue;

S304. When the grayscale value of the pixel block in the orthopedic image data is greater than or equal to the first global threshold, but less than the second global threshold, it is determined that the pixel block is a suspected abnormal bone tissue;

S305. When the gray value of the pixel block in the orthopedic image data is greater than or equal to the second global threshold, it is determined that the pixel block is abnormal bone tissue;

S306. Divide normal bone tissue, suspected abnormal bone tissue, and areas corresponding to abnormal bone tissue to obtain bone structure areas with possible pathological changes.

The working principle of the above technical solution is: setting the first global threshold and the second global threshold: First, before processing the orthopedic image data, the system needs to set two global thresholds, namely the first global threshold and the second global threshold. The selection of these thresholds may be based on previous research or experience.

Comparison of grayscale values of pixel blocks: Each pixel block in the orthopedic image data is processed, and the grayscale value of each pixel block is compared with the first and second global thresholds.

Determine normal bone tissue: When the gray value of a pixel block is less than the first global threshold, the pixel block is determined to be normal bone tissue.

Determining suspected abnormal bone tissue: When the gray value of a pixel block is greater than or equal to the first global threshold but less than the second global threshold, the pixel block is determined to be suspected abnormal bone tissue. This means that the gray value of the pixel block is between normal and abnormal, and further analysis is required.

Determine abnormal bone tissue: When the gray value of a pixel block is greater than or equal to the second global threshold, the pixel block is determined to be abnormal bone tissue. This means that the gray value of the pixel block is significantly higher than the normal range, and there may be obvious bone abnormalities.

Divide bone structure areas with possible pathological changes: Finally, based on the above judgment results, divide the areas corresponding to normal bone tissue, suspected abnormal bone tissue and abnormal bone tissue to obtain bone structure areas with possible pathological changes.

The effects of the above technical solution are: Automated lesion area determination: This technical solution allows automated lesion area determination of orthopedic image data, helping doctors quickly identify possible problem areas and improving the efficiency of orthopedic diagnosis.

Reduce missed diagnosis and misdiagnosis: By classifying bone areas into normal, suspected abnormal, and abnormal based on thresholds, it helps reduce the risk of missed diagnosis and misdiagnosis and improves the accuracy of orthopedic image diagnosis.

Assist doctors in decision-making: This technical solution provides doctors with information to assist in decision-making, making it easier to identify areas of bone structure that may require focus, helping to better formulate treatment plans.

In one embodiment of the present invention, setting the first global threshold and the second global threshold includes:

S3011. Collect bone image data with diseased areas;

S3012. Perform histogram analysis on the bone image data with the diseased area, and obtain a histogram corresponding to the pixel gray value of the bone image with the diseased area;

S3013. Extract the gray value corresponding to one or more peak points of the histogram corresponding to the pixel gray value;

S3014. When there is only one peak point, the first global threshold and the second global threshold are 0.68H and 0.82H respectively; where H represents the gray value corresponding to the peak point;

S3015. When there are multiple peak points, the first global threshold and the second global threshold are 0.63×0.5×(H _min +H _max ) and 0.77×0.5×(H _min +H _max ) respectively; where, H _min and H _max respectively represent the lowest gray value and the highest gray value corresponding to multiple peak points.

The working principle of the above technical solution is as follows: collecting bone image data with diseased areas: first, a set of bone image data with known diseased areas needs to be collected. The data may come from previous cases or studies that include images of known bone lesions.

Histogram analysis: Perform histogram analysis on bone image data with diseased areas. A histogram is a distribution diagram of the number of pixels at different gray levels in an image.

Extract the gray value corresponding to the peak point: In the histogram, one or more peak points will appear, and these peak points correspond to the peak number of pixels with a specific gray value in the image.

Determine the threshold based on the number of peak points: when there is only one peak point (usually corresponding to a single obvious lesion), set the first global threshold to 0.68 times the gray value corresponding to the peak point, and set the second global threshold to 0.82 times the gray value corresponding to the peak point. Grayscale values (these coefficients may be adjusted based on experience). When there are multiple peak points, calculate the lowest and highest grayscale values (Hmin and Hmax) of these peak points, and then set the first global threshold to 0.5 times (Hmin+Hmax) and the second global threshold to 0.77 times (Hmin+ Hmax).

The effects of the above technical solution are: Threshold selection based on image data: This technical solution adopts a threshold selection strategy based on specific image data instead of a static fixed threshold. This means that the threshold can be adjusted according to the characteristics of the actual image, helping to better adapt to the image processing needs of different situations.

Improved threshold accuracy: By analyzing histograms of bone image data with diseased areas, the system can more accurately determine thresholds, thereby helping to better determine normal and diseased areas and improving the accuracy of image segmentation.

An embodiment of the present invention proposes an image processing-based orthopedic auxiliary examination system, as shown in Figure 2. The image processing-based orthopedic auxiliary examination system includes:

A real-time monitoring module, used to monitor the patient's orthopedic examination data in real time, perform preliminary image processing on the orthopedic examination data, and obtain orthopedic image data after preliminary image processing;

An image segmentation module, used to perform image segmentation on the orthopedic image data using a boundary line segmentation method to obtain the bone structure area in the orthopedic image data;

A diseased area acquisition module, used to determine the diseased area of the bone structure area and determine whether there is a diseased area in the bone structure area;

The recognition and diagnosis module is used to input the lesion area into a pre-trained image recognition model for image analysis and diagnosis, and obtain a diagnosis report.

The working principle of the above technical solution is: real-time monitoring of the patient's orthopedic examination data: First, the system monitors the patient's orthopedic examination data in real time, which can include imaging data such as X-ray, MRI or CT, which are usually used to examine the patient's bone structure. .

Preliminary image processing: Preliminary image processing of orthopedic examination data acquired in real time, which may include denoising, enhancement, or other image preprocessing methods to obtain clearer and analyzable image data.

Image segmentation using boundary line segmentation: Use boundary line segmentation to segment images to separate bone structure areas in orthopedic images from surrounding tissues. This helps the system focus on the analysis of bone structure.

Lesion area judgment: Within the bone structure area, the system will judge the lesion area to determine whether there is a possible lesion. This may involve detecting abnormal areas, fracture lines, tumors or other abnormalities.

Application of image recognition model: If there are lesion areas, the system inputs these areas into a pre-trained image recognition model. This model could be a deep learning neural network with the ability to identify different orthopedic lesions.

Obtain a diagnostic report: The image recognition model analyzes the image and generates a diagnostic report, which typically includes a detailed description of the patient's orthopedic condition, including lesion type, location, size, and severity.

The effect of the above technical solution is that the image processing-based orthopedic auxiliary examination method proposed in this embodiment can monitor the patient's orthopedic condition in real time and detect potential orthopedic problems earlier. Automated image processing and diagnostic processes can improve diagnostic efficiency and reduce doctors' workload. Using technologies such as deep learning for image analysis and diagnosis can improve the accuracy of diagnosis and reduce the risk of misdiagnosis. The method is able to generate instant diagnostic reports, providing the medical team with timely information to support treatment decisions.

In one embodiment of the present invention, the real-time monitoring module includes:

The monitoring execution module is used to monitor and receive the patient's orthopedic examination data in real time, where the orthopedic examination data includes orthopedic image data corresponding to X-rays, CT scans and MRI;

An image noise reduction module, used to perform image noise reduction processing on the orthopedic image data, remove noise and artifacts in the orthopedic image data, and obtain orthopedic image data after image noise reduction processing;

An image enhancement module, configured to perform image contrast and clarity enhancement processing on the orthopedic image data after image noise reduction processing, and obtain enhanced orthopedic image data;

The image correction module is used to perform image correction on the enhanced orthopedic image data to obtain image-corrected orthopedic image data. The image-corrected orthopedic image data is the orthopedic image data after the preliminary image processing.

Specifically, since the enhanced image data may be deformed and distorted, it will greatly interfere with the subsequent acquisition of the lesion area. In severe cases, it may cause the acquisition of the lesion area to be offset or incomplete, which will affect subsequent diagnosis and surgical planning. There are significant adverse effects, so when performing image correction, the following algorithm is used:

Step 1: Assume P(x, y) is a point in the deformed image, then its incident angle α with the center of the image is:

Step 2: According to the incident angle calculated in step 1, obtain the isoconcentric projection correction coefficient of the center point P of the deformed image. The calculation formula is:

Among them, r is the equal angle projection correction coefficient, and π is 3.14.

Step 3: Calculate the coordinates of the corrected point P based on the equirectangular projection correction coefficient calculated in step 2. The calculation formula is as follows:

Among them, x′ is the corrected abscissa, y′ is the corrected ordinate, that is, the point P(x, y) in the deformed image is P(x′, y′) after correction.

This algorithm uses an isometric angle correction model to correct the image, avoiding the interference of the deformed and distorted image on the subsequent acquisition of the lesion area, facilitating subsequent accurate judgment of the scope of abnormal bone tissue, and providing accurate imaging data and judgment for subsequent diagnosis and surgical planning. in accordance with.

The working principle of the above technical solution is: Real-time monitoring and reception of orthopedic examination data: The system first monitors and receives the patient's orthopedic examination data in real time, which can include various types of orthopedic image data such as X-rays, CT scans and MRIs.

Image noise reduction processing: perform image noise reduction processing on the received orthopedic image data. This step aims to remove noise and artifacts from the image to improve the quality and clarity of the image. Denoising helps the system analyze bone structure more accurately.

Image contrast and clarity enhancement processing: Contrast and clarity enhancement processing is performed on the orthopedic image data that has been processed by noise reduction. This step enhances the contrast in the image, making details in the image more clearly visible, helping doctors better diagnose problems.

Image Correction: Perform image correction on enhanced orthopedic image data. Image correction corrects geometric or other distortions in images to ensure image accuracy and reliability.

The effect of the above technical solution is: Improved image quality: Through processing steps such as denoising, enhancement and correction, this technical solution can significantly improve the quality of orthopedic images, allowing doctors to more easily analyze and diagnose the patient's bone structure.

Reduced misdiagnosis: Improvements in image processing help reduce the risk of misdiagnosis and increase diagnostic accuracy.

Improve efficiency: Preliminary image processing can be automatically performed before the doctor's diagnosis, thereby improving the efficiency of diagnosis, shortening waiting time, and making the medical diagnosis process faster.

Overall, this technical solution helps improve the quality of orthopedic images and the accuracy of medical diagnosis, and improves the patient's diagnosis and treatment experience.

In one embodiment of the present invention, the lesion area acquisition module includes:

A threshold setting module, used to set the first global threshold and the second global threshold;

A gray value comparison module, configured to compare the gray value of each pixel block in the orthopedic image data with the first global threshold and the second global threshold in sequence;

A first determination module configured to determine, when the grayscale value of a pixel block in the orthopedic image data is less than a first global threshold, that the pixel block corresponding to a grayscale value less than the first global threshold is normal bone tissue;

A second judgment module configured to determine that the pixel block is a suspected abnormal bone tissue when the grayscale value of the pixel block in the orthopedic image data is greater than or equal to the first global threshold, but less than the second global threshold;

A third determination module, configured to determine that the pixel block is abnormal bone tissue when the grayscale value of the pixel block in the orthopedic image data is greater than or equal to the second global threshold;

The division module is used to divide normal bone tissue and areas corresponding to suspected abnormal bone tissue and abnormal bone tissue to obtain bone structure areas with the possibility of disease.

The working principle of the above technical solution is: setting the first global threshold and the second global threshold: First, before processing the orthopedic image data, the system needs to set two global thresholds, namely the first global threshold and the second global threshold. The selection of these thresholds may be based on previous research or experience.

Comparison of grayscale values of pixel blocks: Each pixel block in the orthopedic image data is processed, and the grayscale value of each pixel block is compared with the first and second global thresholds.

Determine normal bone tissue: When the gray value of a pixel block is less than the first global threshold, the pixel block is determined to be normal bone tissue.

Determining suspected abnormal bone tissue: When the gray value of a pixel block is greater than or equal to the first global threshold but less than the second global threshold, the pixel block is determined to be suspected abnormal bone tissue. This means that the gray value of the pixel block is between normal and abnormal, and further analysis is required.

Determine abnormal bone tissue: When the gray value of a pixel block is greater than or equal to the second global threshold, the pixel block is determined to be abnormal bone tissue. This means that the gray value of the pixel block is significantly higher than the normal range, and there may be obvious bone abnormalities.

Divide bone structure areas with possible pathological changes: Finally, based on the above judgment results, divide the areas corresponding to normal bone tissue, suspected abnormal bone tissue and abnormal bone tissue to obtain bone structure areas with possible pathological changes.

The effects of the above technical solution are: Automated lesion area determination: This technical solution allows automated lesion area determination of orthopedic image data, helping doctors quickly identify possible problem areas and improving the efficiency of orthopedic diagnosis.

Reduce missed diagnosis and misdiagnosis: By classifying bone areas into normal, suspected abnormal, and abnormal based on thresholds, it helps reduce the risk of missed diagnosis and misdiagnosis and improves the accuracy of orthopedic image diagnosis.

Assist doctors in decision-making: This technical solution provides doctors with information to assist in decision-making, making it easier to identify areas of bone structure that may require focus, helping to better formulate treatment plans.

In one embodiment of the present invention, the threshold setting module includes:

Data collection module, used to collect bone image data with diseased areas;

The histogram analysis module is used to perform histogram analysis on the bone image data with the diseased area, and obtain the histogram corresponding to the pixel gray value of the bone image with the diseased area;

A gray value extraction module, used to extract the gray value corresponding to one or more peak points of the histogram corresponding to the pixel gray value;

The first threshold setting module is used to set the first global threshold and the second global threshold to 0.68H and 0.82H respectively when there is only one peak point; where H represents the gray value corresponding to the peak point;

The second threshold setting module is used to set the first global threshold and the second global threshold to 0.63×0.5×(H _min +H _max ) and 0.77×0.5×(H _min +H respectively) when there are multiple peak points. _max ); where H _min and H _max respectively represent the lowest gray value and the highest gray value corresponding to multiple peak points.

The working principle of the above technical solution is as follows: collecting bone image data with diseased areas: first, a set of bone image data with known diseased areas needs to be collected. The data may come from previous cases or studies that include images of known bone lesions.

Histogram analysis: Perform histogram analysis on bone image data with diseased areas. A histogram is a distribution diagram of the number of pixels at different gray levels in an image.

Extract the gray value corresponding to the peak point: In the histogram, one or more peak points will appear, and these peak points correspond to the peak number of pixels with a specific gray value in the image.

Determine the threshold based on the number of peak points: when there is only one peak point (usually corresponding to a single obvious lesion), set the first global threshold to 0.68 times the gray value corresponding to the peak point, and set the second global threshold to 0.82 times the gray value corresponding to the peak point. Grayscale values (these coefficients may be adjusted based on experience). When there are multiple peak points, calculate the lowest and highest grayscale values (Hmin and Hmax) of these peak points, and then set the first global threshold to 0.5 times (Hmin+Hmax) and the second global threshold to 0.77 times (Hmin+ Hmax).

The effects of the above technical solution are: Threshold selection based on image data: This technical solution adopts a threshold selection strategy based on specific image data instead of a static fixed threshold. This means that the threshold can be adjusted according to the characteristics of the actual image, helping to better adapt to the image processing needs of different situations.

Improved threshold accuracy: By analyzing histograms of bone image data with diseased areas, the system can more accurately determine thresholds, thereby helping to better determine normal and diseased areas and improving the accuracy of image segmentation.

Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the invention. In this way, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims (10)

1. An orthopedics auxiliary inspection method based on image processing is characterized by comprising the following steps:

monitoring orthopedics examination data of a patient in real time, and performing preliminary image processing on the orthopedics examination data to obtain orthopedics image data after the preliminary image processing;

image segmentation is carried out on the orthopedics image data by utilizing a boundary line segmentation mode, and a bone structure area in the orthopedics image data is obtained;

judging the pathological change area of the skeletal structure area, and determining whether the pathological change area exists in the skeletal structure area;

and inputting the lesion area into a pre-trained image recognition model for image analysis and diagnosis, and obtaining a diagnosis report.

2. The image processing-based orthopedics auxiliary examination method as defined in claim 1, wherein monitoring orthopedics examination data of a patient in real time and performing preliminary image processing on the orthopedics examination data to obtain orthopedics image data after the preliminary image processing, includes:

monitoring and receiving orthopedics examination data of a patient in real time, wherein the orthopedics examination data comprises orthopedics image data corresponding to X-rays, CT scanning and MRI;

Performing image noise reduction processing on the orthopedics image data, removing noise and artifacts in the orthopedics image data, and obtaining orthopedics image data after the image noise reduction processing;

image contrast and definition enhancement processing is carried out on the orthopedics image data subjected to the image noise reduction processing, so that enhanced orthopedics image data is obtained;

and carrying out image correction on the enhanced orthopaedics image data to obtain image corrected orthopaedics image data, wherein the image corrected orthopaedics image data is the orthopaedics image data after preliminary image processing.

3. The image processing-based orthopedics auxiliary examination method as defined in claim 1, wherein the following algorithm is adopted in the image correction:

step one: let P (x, y) be a point in the deformed image, the angle of incidence α with the center of the image is:

step two: according to the incidence angle calculated in the step one, the equivalent angle projection correction coefficient of the midpoint P of the deformed image is calculated, and the calculation formula is as follows:

wherein r is an isosceles angle projection correction coefficient, and pi is 3.14;

step three: and (3) calculating coordinates of the corrected point P according to the calculated equivalent angle projection correction coefficient in the step (II), wherein the calculation formula is as follows:

Where x 'is the corrected abscissa and y' is the corrected ordinate, i.e. the point P (x, y) in the deformed image is corrected to P (x ', y').

4. The image processing-based orthopedics auxiliary examination method as defined in claim 1, wherein the step of performing lesion area judgment on the bone structure area to determine whether a lesion area exists in the bone structure area comprises the steps of:

setting a first global threshold and a second global threshold;

comparing the gray value of each pixel block in the orthopaedics image data with a first global threshold value and a second global threshold value in sequence;

when the gray value of the pixel block in the orthopaedics image data is smaller than a first global threshold value, judging the pixel block corresponding to the gray value smaller than the first global threshold value as normal bone tissue;

when the gray value of the pixel block in the orthopaedics image data is larger than or equal to the first global threshold but smaller than the second global threshold, judging that the pixel block is suspected abnormal bone tissue;

when the gray value of the pixel block in the orthopaedics image data is larger than or equal to a second global threshold, judging that the pixel block is abnormal bone tissue;

and dividing the normal bone tissue into regions corresponding to the suspected abnormal bone tissue and the abnormal bone tissue to obtain a bone structure region with the possibility of pathological changes.

5. The image processing-based orthopaedic auxiliary examination method of claim 4, wherein setting a first global threshold and a second global threshold comprises:

collecting bone image data having a lesion area;

performing histogram analysis on bone image data with a lesion area to obtain a histogram corresponding to a pixel gray value of a bone image with the lesion area;

extracting gray values corresponding to one or more peak points of the histogram corresponding to the pixel gray values;

when only one peak point exists, the first global threshold value and the second global threshold value are respectively 0.68H and 0.82H; wherein H represents the gray scale value corresponding to the peak point;

when there are a plurality of peak points, the first and second global thresholds are 0.63×0.5× (H _min +H _max ) And 0.77 x 0.5 x (H _min +H _max ) The method comprises the steps of carrying out a first treatment on the surface of the Wherein H is _min And H _max The lowest gray scale value and the highest gray scale value corresponding to the plurality of peak points are respectively represented.

6. An image processing-based orthopaedic auxiliary inspection system, the image processing-based orthopaedic auxiliary inspection system comprising:

the real-time monitoring module is used for monitoring the orthopedics examination data of the patient in real time, and performing preliminary image processing on the orthopedics examination data to obtain orthopedics image data after the preliminary image processing;

The image segmentation module is used for carrying out image segmentation on the orthopedics image data in a boundary line segmentation mode to obtain a skeleton structure area in the orthopedics image data;

the lesion region acquisition module is used for judging the lesion region of the bone structure region and determining whether the bone structure region has the lesion region or not;

and the identification and diagnosis module is used for inputting the lesion area into a pre-trained image identification model to perform image analysis and diagnosis, and obtaining a diagnosis report.

7. The image processing-based orthopaedic auxiliary inspection system of claim 6, wherein the real-time monitoring module comprises:

the monitoring execution module is used for monitoring and receiving orthopedics examination data of a patient in real time, wherein the orthopedics examination data comprise orthopedics image data corresponding to X-rays, CT scanning and MRI;

the image noise reduction module is used for carrying out image noise reduction on the orthopedics image data, removing noise and artifacts in the orthopedics image data and obtaining orthopedics image data after the image noise reduction;

the image enhancement module is used for carrying out image contrast and definition enhancement processing on the orthopedics image data subjected to the image noise reduction processing to obtain enhanced orthopedics image data;

The image correction module is used for carrying out image correction on the enhanced orthopaedics image data to obtain image corrected orthopaedics image data, wherein the image corrected orthopaedics image data is the orthopaedics image data after preliminary image processing.

8. The image processing-based orthopedics auxiliary examination method as defined in claim 1, wherein the following algorithm is adopted in the image correction:

step one: let P (x, y) be a point in the deformed image, the angle of incidence α with the center of the image is:

step two: according to the incidence angle calculated in the step one, the equivalent angle projection correction coefficient of the midpoint P of the deformed image is calculated, and the calculation formula is as follows:

wherein r is an isosceles angle projection correction coefficient, and pi is 3.14;

step three: and (3) calculating coordinates of the corrected point P according to the calculated equivalent angle projection correction coefficient in the step (II), wherein the calculation formula is as follows:

where x 'is the corrected abscissa and y' is the corrected ordinate, i.e. the point P (x, y) in the deformed image is corrected to P (x ', y').

9. The image processing-based orthopaedic auxiliary inspection system of claim 6, wherein the lesion region acquisition module comprises:

The threshold setting module is used for setting a first global threshold and a second global threshold;

the gray value comparison module is used for comparing the gray value of each pixel block in the orthopaedics image data with a first global threshold value and a second global threshold value in sequence;

the first judging module is used for judging the pixel blocks corresponding to the gray values smaller than the first global threshold value as normal bone tissues when the gray values of the pixel blocks in the orthopaedics image data are smaller than the first global threshold value;

the second judging module is used for judging that the pixel block is suspected abnormal bone tissue when the gray value of the pixel block in the orthopaedics image data is larger than or equal to the first global threshold but smaller than the second global threshold;

the third judging module is used for judging that the pixel block is abnormal bone tissue when the gray value of the pixel block in the orthopedic image data is larger than or equal to the second global threshold;

the division module is used for dividing the normal bone tissue into regions corresponding to the suspected abnormal bone tissue and the abnormal bone tissue, and obtaining a bone structure region with the possibility of pathological changes.

10. The image processing-based orthopaedic auxiliary inspection system of claim 8, wherein the threshold setting module comprises:

A data collection module for collecting bone image data having a lesion area;

the histogram analysis module is used for carrying out histogram analysis on the bone image data with the lesion area to obtain a histogram corresponding to the pixel gray value of the bone image with the lesion area;

the gray value extraction module is used for extracting gray values corresponding to one or more peak points of the histogram corresponding to the pixel gray values;

the first threshold setting module is used for setting the first global threshold and the second global threshold to be 0.68H and 0.82H respectively when only one peak point exists; wherein H represents the gray scale value corresponding to the peak point;

a second threshold setting module for, when there are multiple peak points, setting the first and second global thresholds to 0.63×0.5× (H _min +H _max ) And 0.77 x 0.5 x (H _min +H _max ) The method comprises the steps of carrying out a first treatment on the surface of the Wherein H is _min And H _max The lowest gray scale value and the highest gray scale value corresponding to the plurality of peak points are respectively represented.