Method and system for removing halo artifacts in intravascular ultrasound image

Abstract

The invention discloses a method and system for removing halo artifacts in an intravascular ultrasound image, and the method comprises the steps: obtaining an original intravascular ultrasound image through an imaging catheter, and segmenting a blood vessel lumen to obtain a lumen region; dividing the lumen area into an imaging dead area A, a mixed area B and a blood flow area C; acquiring a circular halo artifact image by using the mixed region B and the blood flow region C; and obtaining an optimized intravascular ultrasound image based on the original intravascular ultrasound image and the circular halo artifact image. Through the image processing technology, the circular halo artifact information is recognized based on the blood flow signal and the blood vessel structure image, and then the circular halo artifact information is removed, so that the quality of the IVUS image is remarkably improved on the premise that key anatomical information is not modified or lost.

Description

Technical Field

[0001] The present invention relates to the field of intravascular ultrasound imaging technology, and particularly to a method and system for removing halo artifacts in intravascular ultrasound images. Background Art

[0002] During the intravascular ultrasound (IVUS) imaging process, since the imaging probe is embedded in the catheter protection sheath, the reflection signal of the internal structure of the blood vessel may interfere with the reflection signal of the protection sheath, resulting in impaired imaging quality in the area around the catheter. This phenomenon is called halo artifact. Halo artifacts appear as uneven bright rings in the area around the catheter in IVUS images, and their thickness is uncertain, making it difficult to identify the anatomical structure in this area.

[0003] Halo artifacts are common in IVUS images, and usually halo gain control is used to alleviate halo artifacts. This technology belongs to a form of time gain compensation, and by adjusting the pixel intensity in the area near the imaging catheter, the influence of halo artifacts is reduced. However, when the catheter is close to the blood vessel wall, the halo artifacts overlap with the blood vessel tissue structure, and at this time, halo gain control may cause the anatomical structure information to be blurred or even lost. Summary of the Invention

[0004] Therefore, the purpose of the present invention is to provide a method and system for removing halo artifacts in intravascular ultrasound images; aiming to effectively eliminate halo artifacts while ensuring the integrity and accuracy of blood vessel anatomical structure information. This method uses image processing technology to eliminate halo artifact information based on blood flow signals and blood vessel structure images, and then eliminates halo artifact information, thereby significantly improving the quality of IVUS images without modifying or losing key anatomical information.

[0005] To achieve the above purpose, a method for removing halo artifacts in intravascular ultrasound images provided by the present invention includes the following steps:

[0006] S1. Use the imaging catheter to obtain the original intravascular ultrasound image, and segment the blood vessel lumen to obtain the lumen area;

[0007] S2. Divide the lumen area into an imaging dead zone A, a mixed zone B, and a blood flow zone C;

[0008] S3. Use the mixed zone B and the blood flow zone C to obtain the halo artifact image;

[0009] S4. Obtain the optimized intravascular ultrasound image based on the original intravascular ultrasound image and the halo artifact image.

[0010] Further preferably, in S2, dividing the lumen area into an imaging dead zone A, a mixed zone B, and a blood flow zone C; includes the following process:

[0011] S21. Determine the imaging dead zone A based on the size of the imaging catheter;

[0012] S22. Binarize the original intravascular ultrasound image and then perform morphological processing. After removing small connected regions, use the Hough transform to identify the mixed region B;

[0013] S23. Exclude the dead zone A and the mixed region B from the lumen range in the original intravascular ultrasound image, and denote the remaining region as the blood region C.

[0014] Further preferably, in S22, it includes: binarize the lumen region of the original intravascular ultrasound image in polar coordinates, perform morphological processing on the binarized image, and remove small connected regions;

[0015] Use the Hough transform to identify the straight lines in the image. Determine the mixed region B containing the halo artifact according to the straight line that is farthest from the dead zone A and covers all A-lines among all the straight lines.

[0016] Further preferably, in S22, it includes: binarize the lumen region of the original intravascular ultrasound image in rectangular coordinates, perform morphological processing on the binarized image, and remove small connected regions; use the Hough transform to identify the circles in the image, and obtain the mixed region B containing the halo artifact according to the circle that contains the dead zone A and has the largest area among all the circles.

[0017] Further preferably, in S3, it includes:

[0018] S31. For any A-line, perform Fourier transforms on the signals of the mixed region B and the blood region C on the A-line respectively to obtain the frequency spectra of the mixed region B signal and the blood region C signal;

[0019] S32. Subtract the frequency spectrum of the blood region C signal from the frequency spectrum of the mixed region B signal, and filter out the low-frequency noise signals through a low-pass filter to obtain the frequency domain characteristics of the halo artifact signal;

[0020] S33. Convert the obtained frequency domain characteristics of the halo artifact signal back to the time domain to obtain the halo artifact signal α corresponding to the A-line in the time domain;

[0021] S34. Judge the halo artifact signal α to determine whether it meets the condition for successful removal. If it meets the condition, retain the signal α for further processing; if it does not meet the condition, mark it as an unsatisfactory situation;

[0022] S35. Process all A-lines of the current frame in the order of data acquisition to obtain the halo artifact signals of all A-lines in the current frame, and then form the halo artifact image in the time domain of this frame of image, and store the halo artifact image of this frame of image;

[0023] S36. Loop the above steps to process all frames of the original intravascular ultrasound image to obtain the halo artifact image corresponding to the original image.

[0024] Further preferably, in S34, determine whether it meets the conditions for successful removal, including:

[0025] The signal β in the A-line mixing region B minus the halo artifact signal α to obtain the corrected region signal β′, and calculate the structural similarity SSIM(β,ε) between β and the signal ε in the blood region C, and SSIM(β′,ε) between β′ and the signal ε in the blood region C respectively,

[0026]

[0027] where, μ β represents the mean value of the signal β, μ β′ represents the mean value of the signal β′; i takes β′, β, ε; respectively represents the variance of the signals β′, β, ε, σ βε represents the covariance between the signal β and the signal ε, σ β′ε represents the covariance between the signal β′ and the signal ε, and C1 and C2 are constants;

[0028] If SSIM(β,ε) < SSIM(β′,ε), it means that the halo artifact removal is successful, and the halo artifact signal α is retained, otherwise the halo artifact signal α is marked as an unsatisfactory situation.

[0029] The present invention also provides a system for removing halo artifacts in intravascular ultrasound images, which is used to implement the steps of the method for removing halo artifacts in intravascular ultrasound images described above, including:

[0030] A data acquisition module, which uses an imaging catheter to acquire the original intravascular ultrasound image and segments the vascular lumen to obtain the lumen region;

[0031] A data processing module, which divides the lumen region into an imaging dead zone A, a mixing region B, and a blood flow region C;

[0032] A halo artifact acquisition module, which uses the mixing region B and the blood flow region C to acquire the halo artifact image;

[0033] An artifact removal module, which obtains the optimized intravascular ultrasound image based on the original intravascular ultrasound image and the halo artifact image.

[0034] Further preferably, the data processing module includes dividing the lumen area into an imaging dead zone A, a mixed area B and a blood flow area C; including the following process:

[0035] Determining an imaging dead zone A based on the size of the imaging catheter;

[0036] The original intravascular ultrasound image is binarized and then morphologically processed. After removing the small connected domain, the mixed area B is identified using Hough transform.

[0037] The dead area A and the mixed area B are removed from the lumen range of the original intravascular ultrasound image, and the remaining area is recorded as the blood area C.

[0038] The present invention also provides an electronic device, comprising: a memory storing computer program instructions; and a processor, which implements the steps of the above-mentioned method for removing halo artifacts in intravascular ultrasound images when the computer program instructions are executed by the processor.

[0039] The present invention also provides a computer-readable storage medium, which is used to store instructions. When the stored instructions are executed on a computer, the computer executes the steps of the above-mentioned method for removing halo artifacts in intravascular ultrasound images.

[0040] The present application discloses a method and system for removing halo artifacts in intravascular ultrasound images, which aims to effectively eliminate halo artifacts while ensuring the integrity and accuracy of vascular anatomical structure information. The method uses image processing technology to remove halo artifact information, thereby significantly improving the quality of IVUS images without modifying or losing key anatomical information. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A schematic flow chart of a method for removing halo artifacts in intravascular ultrasound images provided by the present invention.

[0042] Figure 2 This is a schematic diagram of the A-line intraluminal area division proposed by the present invention.

[0043] Figure 3 This is a schematic diagram of the IVUS image mixed area B in the polar coordinate system proposed by the present invention.

[0044] Figure 4 This is a schematic diagram of the spectrum of the mixed area B and the blood area C proposed by the present invention. DETAILED DESCRIPTION

[0045] The present invention is further described in detail below through the accompanying drawings and specific embodiments.

[0046] like Figure 1As shown in the figure, the method for removing the halo artifact in the intravascular ultrasound image provided by an embodiment of the present invention on the one hand includes the following steps:

[0047] S1. Use an imaging catheter to obtain the original intravascular ultrasound image, and segment the vascular lumen to obtain the lumen area; this step can segment the vascular lumen in the intravascular ultrasound image by using the existing technology.

[0048] S2. Divide the lumen area into an imaging dead zone A, a mixed zone B, and a blood flow zone C;

[0049] As Figure 2 and Figure 3 shown, the lumen area is divided into three areas. Among them, the first area is defined as area A, that is, the area inside the outer sheath of the imaging catheter, which is the imaging dead zone and the signal is completely set to zero; the second area is defined as area B, which is the area near the outside of the outer sheath, representing the mixed area, containing halo artifact information and blood flow information; area C is the pure blood area, only containing blood signals. The above areas A, B, and C together constitute the lumen area. For the same catheter, area A is determined and confirmed by converting the real physical size of the imaging catheter. In step S1, the lumen area has been identified. Therefore, as long as the mixed area B is identified, area C can be obtained.

[0050] Specifically, S21. Determine the imaging dead zone A based on the size of the imaging catheter;

[0051] S22. Binarize the original intravascular ultrasound image and then perform morphological processing. After removing small connected components, use the Hough transform to identify the mixed area B; the following will explain the identification of the mixed area B in detail:

[0052] In one embodiment, the lumen area of the IVUS image can be binarized in polar coordinates, and the binarized image is subjected to morphological processing to remove small connected components;

[0053] Use the Hough transform to identify the straight lines in the image, and determine the mixed area B containing the halo artifact according to the straight line that is the farthest from the dead zone A and covers all A-lines among all the straight lines;

[0054] In another embodiment, the lumen area of the original intravascular ultrasound image (i.e., the pie chart obtained after processing) is binarized in rectangular coordinates, and the binarized image is subjected to morphological processing to remove small connected components;

[0055] Use the Hough transform to identify the circles in the image, and obtain the mixed area B containing the halo artifact according to the circle that contains the dead zone A and has the largest area among all the circles.

[0056] S23. Determine the blood area C based on the lumen area, the imaging dead zone A, and the mixed area B;

[0057] Specifically, in the IVUS image, the area remaining after removing the dead zone A and the mixed area B within the lumen range is the blood area C.

[0058] Step S2 divides the intracoronary lumen area into three areas, namely area A, the area within the outer sheath of the catheter; area B, the area containing the halo artifact component and the blood flow component; and area C, the pure blood flow area.

[0059] S3. Obtain the halo artifact image by using the mixed area B and the blood area C;

[0060] S31. For any A-line, perform Fourier transforms on the signals of the mixed area B and the blood area C on the A-line respectively to obtain the signal spectrum of the mixed area B and the signal spectrum of the blood area C;

[0061] S32. Subtract the signal spectrum of the blood area C from the signal spectrum of the mixed area B, and filter out the low-frequency noise signal through a low-pass filter to obtain the frequency-domain characteristics of the halo artifact signal;

[0062] As Figure 4 shown, usually, the blood flow signal shows the characteristics of high frequency and low amplitude in the frequency domain, while the halo artifact signal shows the characteristics of low frequency and high amplitude. After determining the mixed area B, combined with the blood flow signal in area C, perform frequency-domain decomposition on the signal in area B to separate the blood flow signal and the halo artifact signal, so as to eliminate the influence of the halo artifact.

[0063] Considering that in practical applications, the imaging probe and the outer sheath are not completely concentric, which may lead to different artifacts in the circumferential direction. Therefore, calculate separately for each A-line. For example, perform Fourier transforms on the signal ε of the mixed area B and the signal ε of the blood area C of a certain A-line respectively to obtain their respective spectrograms, as Figure 4 shown. The amplitude of the signal in the mixed area B in the low-frequency band is significantly higher than that of the signal in the blood area C, which is consistent with the frequency-domain characteristics of the halo artifact signal. By finding the difference between the two spectrograms, the frequency-domain characteristics of the halo artifact signal can be obtained.

[0064] S33. Convert the obtained frequency-domain characteristics of the halo artifact signal back to the time domain to obtain the halo artifact signal α corresponding to the A-line in the time domain; judge the halo artifact signal α to determine whether it meets the condition for successful removal. If it meets the condition, retain the signal α for further processing; if it does not meet the condition, mark it as an unsatisfactory situation;

[0065] Specifically, the signal β in the A-line mixing region B minus the halo artifact signal α to obtain the corrected region signal β'. The structural similarity SSIM(β,ε) and SSIM(β',ε) between β, β' and the blood region C signal ε are calculated respectively.

[0066]

[0067] Among them, μ β represents the mean value of the signal β, and μ β′ represents the mean value of the signal β'. i takes β', β, ε; respectively represents the variance of the signals β', β, ε, and σ βε represents the covariance between the signal β and the signal ε, and σ β′ε represents the covariance between the signal β' and the signal ε, and C1 and C2 are constants.

[0068] S34. Judge the halo artifact signal α to determine whether it meets the condition for successful removal. If it meets the condition, retain the signal α for further processing; if it does not meet the condition, mark it as an unsatisfactory situation.

[0069] Based on S33, judge whether the halo artifact signal α is successfully removed. If SSIM(β,ε) < SSIM(β',ε), it means that the halo artifact is successfully removed, and retain the halo artifact signal α; otherwise, mark the halo artifact signal α as an unsatisfactory situation.

[0070] Among them, the halo artifact signal marked as an unsatisfactory situation is replaced by the weighted average value of the halo artifact signals near it.

[0071] S35. Process all the A-lines of the current frame in the order of data acquisition to obtain the halo artifact signals of all the A-lines of the current frame, and then form the halo artifact image in the time domain of this frame of image, and store the halo artifact image of this frame of image.

[0072] It should be noted that when the catheter is eccentric and close to the blood vessel wall or the blood vessel is stenotic, there is no blood region C in some A-lines, and the mixing region B contains halo artifact components and blood vessel tissue or plaque components. Since the blood vessel tissue or plaque components are non-uniformly distributed, it is impossible to effectively separate them from the halo artifact signals. Therefore, the halo artifact signal of the nearest A-line with a blood region C stored can be used as the halo artifact signal of the current A-line. Further, if there is no blood region C in all the A-lines of the current frame, the halo artifact image of the nearest frame of image stored is used as the halo artifact image of the current frame.

[0073] S36. Loop through steps S31 - S33 to process all frames of the original intravascular ultrasound image to obtain the halo artifact image corresponding to the original image.

[0074] S4. Obtain the optimized intravascular ultrasound image based on the original intravascular ultrasound image and the halo artifact image;

[0075] Specifically, subtract the halo artifact image obtained in step S3 from the mixed region B of the original IVUS image to obtain the optimized IVUS image.

[0076] Obviously, the above embodiments are only examples for clear illustration and not limitations on the implementation manners. For those of ordinary skill in the art, other different forms of changes or variations can be made based on the above description. It is not necessary and impossible to list all implementation manners here. And the obvious changes or variations derived therefrom are still within the protection scope of the present invention.

Claims

1. A method for removing halo artifacts in intravascular ultrasound images, characterized in that: The following steps are involved: S1, using an imaging catheter to obtain an original intravascular ultrasound image, and segmenting the vascular lumen to obtain the lumen area; S2, dividing the lumen area into an imaging dead zone A, a mixed zone B, and a blood flow zone C; S3, using the mixed area B and the blood flow area C to obtain a halo artifact image; S4. Obtain an optimized intravascular ultrasound image based on the original intravascular ultrasound image and the halo artifact image.

2. The method for removing halo artifacts in intravascular ultrasound images according to claim 1, characterized in that: In S2, the lumen area is divided into an imaging dead zone A, a mixed zone B, and a blood flow zone C; the following process is included: S21, determining an imaging dead zone A based on the size of the imaging catheter; S22, binarizing the original intravascular ultrasound image and then performing morphological processing, removing small connected domains, and using Hough transform to identify mixed region B; S23. Remove the dead area A and the mixed area B within the lumen range of the original intravascular ultrasound image, and record the remaining area as the blood area C.

3. The method for removing halo artifacts in intravascular ultrasound images according to claim 2, characterized in that: In S22, it includes: performing binarization processing on the lumen area of ​​the original intravascular ultrasound image in polar coordinates, performing morphological processing on the binarized image, and removing small connected domains; Hough transform is used to identify straight lines in the image, and the mixed area B containing the halo artifact is determined according to the straight line that is farthest from the dead zone A among all the straight lines and covers all A-lines.

4. The method for removing halo artifacts in intravascular ultrasound images according to claim 2, characterized in that: In S22, it includes: binarizing the lumen area in the original intravascular ultrasound image in rectangular coordinates, performing morphological processing on the binarized image to remove small connected domains; using Hough transform to identify circles in the image, and obtaining a mixed area B containing halo artifacts based on the circle with the largest area that contains the dead zone A among all circles.

5. The method for removing halo artifacts in intravascular ultrasound images according to claim 1, characterized in that: In S3, this includes: S31, for any A-line, performing Fourier transform on the mixed region B signal and the blood region C signal on the A-line to obtain a mixed region B signal spectrum and a blood region C signal spectrum respectively; S32, subtracting the signal spectrum of the blood region C from the signal spectrum of the mixed region B, and filtering the low-frequency noise signal through a low-pass filter to obtain the frequency domain characteristics of the halo artifact signal; S33, converting the obtained frequency domain features of the halo artifact signal back to the time domain to obtain the halo artifact signal α in the time domain corresponding to the A-line; S34, judging whether the halo artifact signal α satisfies the conditions for successful removal, and if so, retaining the signal α for further processing; if not, marking it as an undesirable situation; S35, processing all A-lines of the current frame in the order of data acquisition to obtain halo artifact signals of all A-lines of the current frame, thereby forming a halo artifact image of the frame image in the time domain, and storing the halo artifact image of the frame image; S36, looping the above steps to process all frames of the original intravascular ultrasound image to obtain a halo artifact image corresponding to the original image.

6. The method for removing halo artifacts in intravascular ultrasound images according to claim 5, characterized in that: In S34, it is determined whether the conditions for successful removal are met, including: The signal β in the A-line mixing region B minus the halo artifact signal α to obtain a corrected region signal β'. The structural similarity SSIM(β,ε) between β and the signal ε in the blood region C, and the structural similarity SSIM(β',ε) between β' and the signal ε in the blood region C are calculated respectively. Among them, μ β represents the mean of the signal β, μ β′ represents the mean of the signal β′; i takes β′, β, ε; they represent the variance of signals β′, β, ε respectively, σ βε represents the covariance of signal β and signal ε, σ β′ε represents the covariance of signal β′ and signal ε, C1 and C2 are constants; If SSIM(β,ε) < SSIM(β',ε), it indicates that the removal of the halo artifact is successful, and the halo artifact signal α is retained; otherwise, the halo artifact signal α is marked as an unsatisfactory situation.

7. The method for removing halo artifacts in intravascular ultrasound images according to claim 5, characterized in that: When the halo artifact signal α is marked as an unsatisfactory situation, the weighted average of the nearby halo artifact signals is used as the halo artifact signal of the A-line.

8. A system for removing halo artifacts in intravascular ultrasound images, characterized in that: The steps for implementing the method for removing halo artifacts in the intravascular ultrasound image according to any one of claims 1-7 above include: A data acquisition module that uses an imaging catheter to acquire the original intravascular ultrasound image and segments the vascular lumen to obtain the lumen region. A data processing module that divides the lumen region into an imaging dead zone A, a mixing region B, and a blood flow region C. A halo artifact acquisition module that uses the mixing region B and the blood flow region C to acquire the halo artifact image. An artifact removal module that obtains an optimized intravascular ultrasound image based on the original intravascular ultrasound image and the halo artifact image.

9. An electronic device, characterized in that: Including: A memory that stores computer program instructions. A processor that, when the computer program instructions are executed by the processor, implements the steps of the method for removing halo artifacts in the intravascular ultrasound image according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store instructions, and when the stored instructions are run on a computer, the computer is caused to execute the steps of the method for removing halo artifacts in the intravascular ultrasound image according to any one of claims 1 to 7.

Images