Precise pulmonary blood vessel recognition system and use method thereof

Abstract

The invention is suitable for the technical field of blood vessel recognition, and provides a lung blood vessel precise recognition system and a use method thereof, and the system comprises a control unit, an image acquisition unit, an image processing unit and an image output unit. The control unit is used for realizing interaction between a doctor and the accurate lung blood vessel recognition system, and the doctor can limit a scanning area through the control system, so that the image acquisition unit only obtains a minimum image range containing the lung; the image acquisition unit comprises a CT scanner and computer equipment connected with the CT scanner, and the image acquisition unit is used for acquiring an integral CT image of the lung of a patient and transmitting the acquired integral CT image of the lung to the computer equipment connected with the CT scanner for subsequent processing; according to the device, the problem that the lung blood vessel in the CT image is complex and difficult to segment is solved, and the aim of accurately segmenting the lung blood vessel in the CT image by optimizing the region growing mode is achieved.

Description

Technical Field

[0001] The present invention relates to the field of blood vessel identification technology, and more specifically, to a pulmonary blood vessel accurate identification system and a method for using the same. Background Art

[0002] Accurate pulmonary vascular identification is an important step in obtaining a pulmonary computer-aided detection and diagnosis system. Accurate and clear pulmonary vascular images can clearly display the anatomical structure of the pulmonary artery and its branches, helping to identify vascular-related diseases such as pulmonary embolism and pulmonary hypertension. Through these images, doctors can visually observe whether there are thrombi, stenosis or abnormal dilation in the patient's pulmonary blood vessels, thereby quickly and accurately diagnosing life-threatening diseases such as pulmonary embolism.

[0003] Although obtaining the whole lung CT image can be achieved through ordinary CT scanning, the accurate recognition of lung vascular images is quite challenging. The main reasons for the difficulty in accurately identifying lung vascular images include its complex anatomical structure, numerous blood vessel branches and large diameter variations, resulting in diverse morphologies; at the same time, the contrast between blood vessels and surrounding tissues in CT images is low, and the presence of noise and artifacts makes the boundaries blurred, making it difficult to accurately segment; especially in small blood vessels, capillaries or lesion areas, the segmentation difficulty is further increased.

[0004] In the prior art, some recognition systems use deep learning to distinguish and identify pulmonary blood vessels. This recognition method can segment pulmonary blood vessels from complete lung CT images after simulation through a large number of training sets. However, since pulmonary blood vessels are morphologically similar to structures such as bronchi and tumors, this will cause the model to make misjudgments during segmentation. Although misjudgments can be reduced through multi-task learning, attention mechanisms, etc., it is difficult to completely avoid misjudgments.

[0005] Therefore, the present invention provides a system for accurately identifying pulmonary blood vessels and a method for using the same. Summary of the invention

[0006] In view of the deficiencies in the prior art, an object of the present invention is to provide a pulmonary blood vessel accurate identification system and a method of using the same.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A pulmonary blood vessel accurate identification system comprises a control unit, an image acquisition unit, an image processing unit and an image output unit;

[0009] The control unit is used to realize the interaction between the doctor and the pulmonary blood vessel accurate identification system. The doctor can limit the scanning area through the control system so that the image acquisition unit only obtains the minimum image range including the lungs;

[0010] The image acquisition unit includes a CT scanner and a computer device connected to the CT scanner, and the image acquisition unit is used to acquire a whole lung CT image of the patient, and transmit the acquired whole lung CT image to the computer device connected to the CT scanner for subsequent processing;

[0011] The image processing unit is arranged in a computer device connected to the CT scanner, and is used to perform enhancement processing on the whole lung CT image transmitted by the image acquisition unit, and obtain a lung blood vessel image by processing the whole lung CT image;

[0012] The image processing unit includes an image enhancement unit and an image segmentation unit. The image enhancement unit improves the quality and visualization effect of the whole lung CT image through a series of image processing technologies. The image segmentation unit is used to separately segment and extract the blood vessel image in the whole lung CT image to obtain the lung blood vessel image.

[0013] The image output unit includes a display device, which is connected to the CT scanner and a computer device connected to the CT scanner. The display device is used to represent auxiliary information, an overall lung CT image, and a lung vascular image separately segmented and extracted from the overall lung CT image.

[0014] The present invention is further configured such that the CT scanner has a built-in respiratory gating module, the respiratory gating module includes a strap-type respiratory sensor, the strap-type respiratory sensor is started synchronously with the CT scanner and is wrapped around the patient's abdomen, and the patient's respiratory signal is detected by measuring the volume change caused by respiratory movement. The strap-type respiratory sensor is used to obtain the respiratory signal of the currently scanned patient, and analyze the patient's respiratory signal to obtain the respiratory cycle.

[0015] The present invention is further configured such that the respiratory information transmitted by the strap-type respiratory sensor is analyzed, and the respiratory cycle obtained through the analysis includes: inhalation phase, end of inhalation, exhalation phase, and end of exhalation; the CT scanner uses the end of exhalation as a gating window, and the scanner stops scanning at the inhalation phase, end of inhalation, and exhalation phase, and performs a rapid scan when it is detected that the phase is at the end of exhalation.

[0016] By adopting the above technical solution, accurate scanning is achieved through the built-in respiratory gating module in the CT scanner. Since the pulmonary blood vessels are spread throughout the lungs, the lungs will move when the patient breathes. If the lungs of the patient who is breathing are directly photographed, it may cause excessive artifacts. Therefore, the present invention has a built-in respiratory gating module in the CT scanner. The respiratory gating module can prevent the CT scanner from scanning the patient's lungs while the patient is breathing and obtaining an image with heavy artifacts.

[0017] The present invention is further configured such that the image enhancement unit processes the whole lung CT image by contrast stretching and gamma correction;

[0018] Among them, contrast stretching maps the grayscale value of the whole lung CT image to a wider range through linear transformation; gamma correction adjusts the brightness of the whole lung CT image through nonlinear changes.

[0019] The present invention is further configured that the image segmentation unit segments the whole lung CT image by optimizing the region growing method, wherein the optimized region growing method selects seed points and gradually merges neighborhoods similar to the seed points in the whole lung CT image;

[0020] The steps of optimizing the region growing method are:

[0021] A1, construct a seed point set and select the seed point, set the initial gray value threshold Tthreshold, and initialize the region R;

[0022] A2, check the grayscale of all points in the neighborhood of the seed point; if the grayscale value of the neighborhood meets the merging condition, merge it into the region R, and add the pixel to the seed point set; if the grayscale value of the neighborhood does not meet the merging condition, ignore the point; if the termination condition is met, do not continue to execute downward;

[0023] A3, select a random point from the seed point set and execute S2;

[0024] Termination condition: The grayscale values ​​of all neighboring pixels do not meet the merging condition;

[0025] The seed point is selected by the operator or doctor through the control unit. Due to the special structure of the pulmonary blood vessels, a point in the pulmonary aorta is selected as the seed point.

[0026] The present invention is further configured that the merging condition for the stepwise merging in the optimized region growing method is grayscale similarity, and the grayscale value similarity is determined by calculating the grayscale difference between the neighborhood pixel and the seed point to determine whether it meets the merging standard, and the merging standard is whether the difference between the current grayscale value and the grayscale value of the seed point is less than the grayscale similarity threshold, and the merging standard calculation formula is:

[0027] |I neighborhood - I seed |≤T threshold

[0028] Among them, Ineighborhood is the gray value of the neighborhood pixels, Iseed is the gray value of the seed point, and Tthreshold is the gray similarity threshold.

[0029] The present invention is further configured such that the grayscale similarity threshold is set to a non-fixed value, and the grayscale similarity threshold is dynamically optimized and adjusted. By real-time monitoring of the grayscale value of the merged area, the grayscale similarity threshold is dynamically adjusted to obtain a more accurate segmentation solution.

[0030] The present invention is further configured such that the grayscale similarity threshold is dynamically adjusted according to the grayscale value mean of all pixels currently merged, and the grayscale value mean is calculated by:

[0031]

[0032] Among them, μ is the mean gray value of the pixel, R is the current area, N is the number of pixels in the area, P is the current point, T is the gray value, and PT is the T value at the P point position.

[0033] The present invention is further configured such that the auxiliary information represented by the image output unit through the display device includes the patient's name and age.

[0034] A method for using a pulmonary blood vessel accurate identification system, according to the above-mentioned pulmonary blood vessel accurate identification system, comprises the following steps:

[0035] S1. The doctor sends a control command through the control system to set the scanning position according to the patient's body shape;

[0036] S2, the image acquisition unit scans the patient's lungs through a CT scanner to obtain an overall CT image of the patient's lungs, and transmits the obtained overall CT image of the lungs to a computer device connected to the CT scanner for subsequent processing;

[0037] S3, the image processing unit receives the whole CT image of the patient's lungs;

[0038] S31, the image enhancement unit performs contrast stretching and gamma correction on the whole lung CT image for processing, and inputs the enhanced whole lung CT image into the image segmentation unit for pulmonary blood vessel segmentation;

[0039] S32, the image segmentation unit performs pulmonary vascular segmentation on the whole lung CT image by optimizing the region growing method to obtain a pulmonary vascular image;

[0040] S4. The image output unit represents the auxiliary information, the whole lung CT image, and the lung blood vessel image.

[0041] In summary, the present application includes at least one of the following beneficial technical effects:

[0042] 1. The present invention realizes accurate scanning through a respiratory gating module built into a CT scanner. Since the pulmonary blood vessels are distributed throughout the lungs, the lungs will move when the patient breathes. If the lungs of the patient who is breathing are directly photographed, it may cause excessive artifacts. Therefore, the present invention has a built-in respiratory gating module in the CT scanner. The respiratory gating module can prevent the CT scanner from scanning the patient's lungs while the patient is breathing and obtaining an image with heavy artifacts.

[0043] 2. The present invention obtains a pulmonary vascular image by setting an image segmentation unit. The image segmentation unit adopts an optimized regional growing method. The merging condition for gradual merging is grayscale similarity. The grayscale value similarity is determined by calculating the grayscale difference between the neighborhood pixels and the seed point to determine whether it meets the merging standard. The merging standard is whether the difference between the current grayscale value and the grayscale value of the seed point is less than the grayscale similarity threshold, and the similarity threshold is adjusted in real time feedback to obtain an accurate pulmonary vascular image. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a schematic diagram of a module of a system for accurately identifying pulmonary blood vessels according to the present invention.

[0045] Figure 2 The figure is a flow chart of a method for using a system for accurately identifying pulmonary blood vessels according to the present invention. DETAILED DESCRIPTION

[0046] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application may be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0047] It should be noted that, unless otherwise specified, all technical and scientific terms used in this application have the same meanings as commonly understood by ordinary technicians in the technical field to which this application belongs.

[0048] See also Figure 1 , Figure 2 , the present invention provides the following technical solutions:

[0049] Embodiment 1, a pulmonary blood vessel accurate identification system includes a control unit, an image acquisition unit, an image processing unit and an image output unit.

[0050] The control unit is used to realize the interaction between the doctor and the lung blood vessel precision identification system. Through the control system, the doctor can limit the scanning area so that the image acquisition unit only obtains the minimum image range that includes the lungs; the doctor can limit the range according to the patient's body shape. For example, when the patient is a child, the lung area is small, so the doctor can reduce the scanning area to ensure that the scanning part includes the complete lungs. This flexible range selection method can minimize the radiation exposed to the patient and reduce the harm of CT scanning to the patient.

[0051] The image acquisition unit includes a CT scanner and a computer device connected to the CT scanner. The image acquisition unit is used to obtain a whole-body CT image of the patient's lungs, and transmit the obtained whole-body CT image of the lungs to the computer device connected to the CT scanner for subsequent processing.

[0052] It should be noted that since the pulmonary blood vessels are spread throughout the lungs, the lungs will move when the patient breathes. Directly photographing the lungs of a patient who is breathing may cause excessive artifacts. Therefore, the present invention has a built-in respiratory gating module in the CT scanner. The respiratory gating module can prevent the CT scanner from scanning the patient's lungs while the patient is breathing and obtaining an image with heavy artifacts.

[0053] The respiratory gating module includes a strap-type respiratory sensor, which is started synchronously with the CT scanner and wrapped around the patient's abdomen. When the CT scanner is turned on, the strap-type respiratory sensor is turned on synchronously to detect the patient's respiratory signal by measuring the volume change caused by respiratory movement. The strap-type respiratory sensor is used to obtain the respiratory signal of the currently scanned patient and analyze the patient's respiratory signal to obtain the respiratory cycle.

[0054] Specifically, the respiratory information transmitted by the strap-type respiratory sensor is analyzed, and the respiratory cycle obtained through the analysis includes: inhalation phase, end of inhalation, exhalation phase, and end of exhalation; the CT scanner uses the end of exhalation as the gating window, and the scanner stops scanning at the inhalation phase, end of inhalation, and exhalation phase, and performs a rapid scan when it detects that this phase is the end of exhalation.

[0055] It should be noted that a 320-row CT scanner is usually used for lung scanning. The end of exhalation is usually the most tranquil period of the human body, when the range of human activity is the smallest and the exhalation time is about 1-2 seconds. Therefore, two end-of-exhalation cycles are usually sufficient to complete a CT scan of the lungs. The present invention does not limit the number of rows of the CT scanner, because the respiratory gating technology is used, and even if multiple end-of-exhalation cycles are scanned, the accuracy will not be reduced.

[0056] The image processing unit is arranged in a computer device connected to the CT scanner. The image processing unit is used to enhance the whole lung CT image transmitted by the image acquisition unit, and obtain the lung vascular image by processing the whole lung CT image. The image processing unit includes an image enhancement unit and an image segmentation unit. The image enhancement unit improves the quality and visualization effect of the whole lung CT image through a series of image processing technologies; the image segmentation unit is used to separately segment and extract the vascular image in the whole lung CT image to obtain the lung vascular image.

[0057] The image enhancement unit processes the whole lung CT image by contrast stretching and gamma correction; wherein, contrast stretching maps the grayscale value of the whole lung CT image to a wider range through linear transformation; and gamma correction adjusts the brightness of the whole lung CT image through nonlinear changes. It should be noted that, since the present invention uses grayscale values ​​for segmentation, the image enhancement method of the present invention does not involve changing the specific value of the grayscale value, and does not perform image enhancement by classifying the closer grayscale values ​​to make the image clearer, such as using the whole lung CT image for histogram averaging. Although this method can make the image clearer, it changes the grayscale value of the original whole lung CT image, which will affect the subsequent segmentation.

[0058] After the whole lung CT image is enhanced by the image enhancement unit, the blood vessels of the enhanced whole lung CT image are segmented by the image segmentation unit. The image segmentation unit segments the whole lung CT image by optimizing the region growing method. The optimized region growing method selects seed points and gradually merges neighborhoods similar to the seed points in the whole lung CT image.

[0059] The basic idea of ​​the region growing method is to group pixels with similar properties to form a region. Specifically, it is to find a pixel in the area to be segmented as the seed point for growth, that is, the starting point, and then merge the pixels in the vicinity of the seed pixel that are the same or similar to the seed pixel into the area where the seed pixel is located according to the predetermined growth criteria. These new pixels are used as new seed pixels to continue the above process until there are no pixels that meet the growth conditions and stop growing. The growth criteria can be selected according to system needs, and the threshold is usually determined based on the grayscale characteristics of the target object in the image. There are three key points in the implementation of the region growing method: 1) Selection of seed points in the target area; 2) Determination of the region growth criteria; 3) Criteria or conditions for stopping region growth.

[0060] The present invention adopts an optimized regional growth method, which is more accurate than the regional growth method. The specific steps of the optimized regional growth method are:

[0061] A1, construct a seed point set and select the seed point, set the initial gray value threshold Tthreshold, and initialize the region R;

[0062] A2, check the grayscale of the neighboring points of the seed point; if the grayscale value of the neighboring point meets the merging condition, merge it into the region R, and add the pixel to the seed point set; if the grayscale value of the neighboring point does not meet the merging condition, ignore the point; if the termination condition is met, do not continue to execute downward;

[0063] A3. Select a random point from the seed point set and execute S2.

[0064] Termination condition: The grayscale values ​​of all neighborhood pixels do not meet the merging conditions.

[0065] The seed point is selected by the operator or doctor through the control unit. Due to the special structure of the pulmonary blood vessels, the pulmonary aorta is the clearest and easiest to select, so a point in the pulmonary aorta is selected as the seed point.

[0066] It should be noted that in the present invention, the seed point set avoids excessive random selection. If the same point is often selected during random selection, the segmentation speed will be too slow, especially when the image as a whole tends to be segmented. Therefore, in the present invention, when selecting random points from the seed point set, if a point is selected more than the MAX value, the point will be deleted from the seed point set.

[0067] The MAX value is determined by the number of image pixels. Accordingly, when the number of image pixels is large, the MAX value is set to a larger value, and when the number of image pixels is small, the MAX value is set to a smaller value.

[0068] The merging condition for stepwise merging in the optimized region growing method is grayscale similarity. Grayscale value similarity is to determine whether it meets the merging criteria by calculating the grayscale difference between the neighborhood pixels and the seed point. The merging criteria is whether the difference between the current grayscale value and the seed point grayscale value is less than the grayscale similarity threshold. The merging criteria calculation formula is:

[0069] |I neighborhood - I seed |≤T threshold

[0070] Among them, Ineighborhood is the gray value of the neighborhood pixels, Iseed is the gray value of the seed point, and Tthreshold is the gray similarity threshold.

[0071] It should be noted that, usually the region growing method will set the similarity threshold to a fixed value. This method can actually achieve the segmentation of pulmonary blood vessels to a certain extent. However, in the actual segmentation process, since the pulmonary blood vessels need to be segmented in the present invention, the pulmonary blood vessels have particularity, and the grayscale value of the blood vessels gradually decreases from the root to the end. Therefore, the use of a single fixed grayscale similarity threshold will lead to poor segmentation effect, and the segmentation difficulty increases at the end. Therefore, the grayscale similarity threshold in the present invention changes the judgment threshold at any time according to the grayscale value in the R area, so that the extraction effect can be better and the information of the pulmonary blood vessels can be more complete.

[0072] The grayscale similarity threshold is set to an unfixed value, and the grayscale similarity threshold is dynamically optimized and adjusted. By real-time monitoring of the grayscale value of the merged area, the grayscale similarity threshold is dynamically adjusted to obtain a more accurate segmentation solution.

[0073] The grayscale similarity threshold is dynamically adjusted according to the average grayscale value of all pixels currently merged. The grayscale value average is calculated as follows:

[0074]

[0075] Among them, μ is the mean gray value of the pixel, R is the current area, N is the number of pixels in the area, P is the current point, T is the gray value, and PT is the T value at the P point position.

[0076] The image output unit includes a display device, which is connected to the CT scanner and a computer device connected to the CT scanner. The display device is used to represent auxiliary information, the whole lung CT image, and the lung vascular image separately segmented and extracted from the whole lung CT image. The auxiliary information represented by the image output unit through the display device includes the patient's name and age. Through the image output unit, doctors and operators can intuitively observe the patient's whole lung CT image and lung vascular image, thereby realizing the judgment of the patient's physical information.

[0077] Embodiment 2: A method for using a pulmonary blood vessel accurate identification system, according to the above-mentioned pulmonary blood vessel accurate identification system, comprises the following steps:

[0078] S1. The doctor sends a control command through the control system to set the scanning position according to the patient's body shape;

[0079] S2, the image acquisition unit scans the patient's lungs through a CT scanner to obtain an overall CT image of the patient's lungs, and transmits the obtained overall CT image of the lungs to a computer device connected to the CT scanner for subsequent processing;

[0080] S3, the image processing unit receives the whole CT image of the patient's lungs;

[0081] The specific steps of S3 are:

[0082] S31, the image enhancement unit performs contrast stretching and gamma correction on the whole lung CT image for processing, and inputs the enhanced whole lung CT image into the image segmentation unit for pulmonary blood vessel segmentation;

[0083] S32, the image segmentation unit performs pulmonary vascular segmentation on the whole lung CT image by optimizing the region growing method to obtain a pulmonary vascular image;

[0084] S4. The image output unit represents the auxiliary information, the whole lung CT image, and the lung blood vessel image.

[0085] The device embodiments described above are merely illustrative and are not all embodiments. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some communication interfaces, indirect coupling or communication connection of devices or units, which may be electrical, mechanical or other forms. All other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present invention.

Claims

1. A pulmonary blood vessel accurate identification system, characterized by: It includes a control unit, an image acquisition unit, an image processing unit and an image output unit; The control unit is used to realize the interaction between the doctor and the pulmonary blood vessel accurate identification system. The doctor can limit the scanning area through the control system so that the image acquisition unit only obtains the minimum image range including the lungs; The image acquisition unit includes a CT scanner and a computer device connected to the CT scanner, and the image acquisition unit is used to acquire a whole lung CT image of the patient, and transmit the acquired whole lung CT image to the computer device connected to the CT scanner for subsequent processing; The image processing unit is arranged in a computer device connected to the CT scanner, and is used to perform enhancement processing on the whole lung CT image transmitted by the image acquisition unit, and obtain a lung blood vessel image by processing the whole lung CT image; The image processing unit includes an image enhancement unit and an image segmentation unit. The image enhancement unit improves the quality and visualization effect of the whole lung CT image through a series of image processing technologies. The image segmentation unit is used to separately segment and extract the blood vessel image in the whole lung CT image to obtain the lung blood vessel image. The image output unit includes a display device, which is connected to the CT scanner and a computer device connected to the CT scanner. The display device is used to represent auxiliary information, an overall lung CT image, and a lung vascular image separately segmented and extracted from the overall lung CT image.

2. The pulmonary blood vessel accurate identification system according to claim 1, characterized in that: The CT scanner has a built-in respiratory gating module, which includes a strap-type respiratory sensor. The strap-type respiratory sensor is started synchronously with the CT scanner and wrapped around the patient's abdomen. The patient's respiratory signal is detected by measuring the volume change caused by respiratory movement. The strap-type respiratory sensor is used to obtain the respiratory signal of the currently scanned patient and analyze the patient's respiratory signal to obtain the respiratory cycle.

3. The pulmonary blood vessel accurate identification system according to claim 2, characterized in that: The respiratory information transmitted by the strap-type respiratory sensor is analyzed, and the respiratory cycle obtained through the analysis includes: inhalation phase, end of inhalation, exhalation phase, and end of exhalation; the CT scanner uses the end of exhalation as a gating window, and the scanner stops scanning at the inhalation phase, end of inhalation, and exhalation phase, and performs a rapid scan when it is detected that the phase is the end of exhalation.

4. The pulmonary blood vessel accurate identification system according to claim 1, characterized in that: The image enhancement unit processes the whole lung CT image by contrast stretching and gamma correction; Among them, contrast stretching maps the grayscale value of the whole lung CT image to a wider range through linear transformation; gamma correction adjusts the brightness of the whole lung CT image through nonlinear changes.

5. The pulmonary blood vessel accurate identification system according to claim 1, characterized in that: The image segmentation unit segments the whole lung CT image by optimizing the region growing method, wherein the optimized region growing method selects seed points and gradually merges neighborhoods similar to the seed points in the whole lung CT image; The steps of optimizing the region growing method are: A1, construct a seed point set and select the seed point, set the initial gray value threshold Tthreshold, and initialize the region R; A2. Check the grayscale of all points in the neighborhood of the seed point; if the grayscale value of the neighborhood meets the merging condition, merge it into region R and add the pixel to the seed point set; If the neighborhood grayscale value does not meet the merging condition, the point is ignored; If the termination condition is met, the execution will not continue; A3, select a random point from the seed point set and execute S2; Termination condition: The grayscale values ​​of all neighboring pixels do not meet the merging condition; The seed point is selected by the operator or doctor through the control unit. Due to the special structure of the pulmonary blood vessels, a point in the pulmonary aorta is selected as the seed point.

6. The pulmonary blood vessel accurate identification system according to claim 5, characterized in that: The merging condition for the stepwise merging in the optimized region growing method is grayscale similarity. The grayscale value similarity is determined by calculating the grayscale difference between the neighborhood pixels and the seed point to determine whether it meets the merging standard. The merging standard is whether the difference between the current grayscale value and the grayscale value of the seed point is less than the grayscale similarity threshold. The merging standard calculation formula is: |I neighborhood - I seed |≤T threshold Among them, Ineighborhood is the gray value of the neighborhood pixels, Iseed is the gray value of the seed point, and Tthreshold is the gray similarity threshold.

7. The pulmonary blood vessel accurate identification system according to claim 6, characterized in that: The grayscale similarity threshold is set to an unfixed value, and the grayscale similarity threshold is dynamically optimized and adjusted. By real-time monitoring of the grayscale value of the merged area, the grayscale similarity threshold is dynamically adjusted to obtain a more accurate segmentation solution.

8. The pulmonary blood vessel accurate identification system according to claim 7, characterized in that: The grayscale similarity threshold is dynamically adjusted according to the average grayscale value of all pixels currently merged. The grayscale value average is calculated as follows: Among them, μ is the mean gray value of the pixel, R is the current area, N is the number of pixels in the area, P is the current point, T is the gray value, and PT is the T value at the P point position.

9. The pulmonary blood vessel accurate identification system according to claim 1, characterized in that: The auxiliary information represented by the image output unit through the display device includes the patient's name and age.

10. A method for using a pulmonary blood vessel accurate identification system, using a pulmonary blood vessel accurate identification system as claimed in any one of claims 1 to 9, characterized in that: The following steps are involved: S1. The doctor sends a control command through the control system to set the scanning position according to the patient's body shape; S2, the image acquisition unit scans the patient's lungs through a CT scanner to obtain an overall CT image of the patient's lungs, and transmits the obtained overall CT image of the lungs to a computer device connected to the CT scanner for subsequent processing; S3, the image processing unit receives the whole CT image of the patient's lungs; S31, the image enhancement unit performs contrast stretching and gamma correction on the whole lung CT image for processing, and inputs the enhanced whole lung CT image into the image segmentation unit for pulmonary blood vessel segmentation; S32, the image segmentation unit performs pulmonary vascular segmentation on the whole lung CT image by optimizing the region growing method to obtain a pulmonary vascular image; S4. The image output unit represents the auxiliary information, the whole lung CT image, and the lung blood vessel image.

Images