A quality detection method for breast three-dimensional ultrasound images

Abstract

The present invention relates to the field of three-dimensional ultrasonic image processing, and in particular to a quality detection method for breast three-dimensional ultrasonic images. The present invention adopts the following technical scheme: a plurality of cross-sectional images are parsed from a breast three-dimensional ultrasonic image, all artifact regions are identified from each cross-sectional image frame and their sizes are marked, the artifact regions are classified according to the aspect ratio and average pixel value of the artifact regions, and grouped according to the frame number of the cross-sectional image where the artifact regions are located, and the quality detection results are output after the artifact results are counted. The beneficial effect of the present invention is that the size and classification of each artifact region can be obtained, and according to the frame number corresponding to the artifact region, the classification, number and width of the artifact regions in each cross-sectional image frame can be judged, so that the quality detection results of the breast three-dimensional ultrasonic image can be accurately output based on this, which is convenient for doctors to quickly and accurately judge the quality of ultrasonic detection images and improve detection efficiency.

Description

Technical Field

[0001] The present invention relates to the field of three-dimensional ultrasonic image processing, and in particular to a quality detection method for breast three-dimensional ultrasonic images. Background Art

[0002] In order to accurately detect the breast and determine whether there is a lesion, most of the breasts are currently tested by three-dimensional ultrasound imaging. In the breast three-dimensional ultrasound image, due to improper detection (such as insufficient fit between the probe and the detection site or the presence of bubbles or thin air layers between the probe and the detection site), artifacts such as empty areas, poor coupling or multi-layer reflections often appear. These artifacts have different degrees of influence on the detection of breast lesions. For the quality detection of breast three-dimensional ultrasound images, doctors are often required to identify and judge artifacts based on the cross-sectional images analyzed from the breast three-dimensional ultrasound images. Whether the quality of the breast three-dimensional ultrasound image is qualified also depends on the doctor's experience. Therefore, it is easy to have inaccurate judgments and the efficiency is low. At present, there is a trained artifact detection model for identifying and detecting the artifact area, but it can only identify the area range of the artifact, but cannot judge the type of the artifact. Different types of artifacts have different effects on the quality of the three-dimensional ultrasound image, so it is also impossible to accurately output the quality detection results. Summary of the invention

[0003] The purpose of the present invention is to provide a quality detection method for breast three-dimensional ultrasound images, specifically to provide a quality detection method that can automatically and quickly detect the size, classification, number, etc. of artifact areas existing in each frame of cross-sectional images in breast three-dimensional ultrasound images and output quality detection results.

[0004] To achieve the above object, the present invention adopts the following technical solution: a quality detection method for breast three-dimensional ultrasound images, characterized in that it includes the following steps:

[0005] S01. First, perform three-dimensional ultrasonic detection on the breast to obtain a three-dimensional ultrasonic breast image, and parse a plurality of frames of cross-sectional images from the three-dimensional ultrasonic breast image.

[0006] S02, use the trained artifact detection model to perform artifact region detection on a plurality of frames of cross-sectional sequence images, detect and segment all I artifact regions in the plurality of frames of cross-sectional images, and record the widths W1, W2, W3, ... W of the I artifact regions respectively. I and heights H1, H2, H3, ... H I , and the frame number of the cross-sectional image where the I artifact areas are located.

[0007] S03, classifying the I artifact regions, specifically including:

[0008] S31, first remove the artifact regions whose width is less than the set threshold W0 from among the I artifact regions;

[0009] S32, calculate the aspect ratios W1 / H1, W2 / H2, W3 / H3, ..., W of the I artifact regions I / H I The aspect ratios of the artifact regions removed in step S31 are arranged in ascending order, and an aspect ratio growth point with an obvious jump growth is found therein, and the artifact regions corresponding to the aspect ratio growth point and the subsequent aspect ratios are classified as "multi-layer reflection" artifacts, and the remaining artifact regions are classified as "poor coupling" or "idle area" artifacts;

[0010] S33. In the artifact areas classified as "poor coupling" or "idle area" artifacts in step S32, the average pixel value of each artifact area is calculated respectively, and the average pixel values ​​are arranged in ascending order, and the average pixel value growth point where an obvious jump-like increase occurs is found, and the artifact areas corresponding to the average pixel value growth point and the subsequent average pixel values ​​are classified as "poor coupling" artifacts, and the remaining artifact areas are classified as "idle area" artifacts.

[0011] S04. Group the artifact areas classified in step S32 and step S33 according to the frame numbers of the cross-sectional images where the artifact areas are located, group the artifact areas with the same frame numbers of the cross-sectional images into the same group, and count the classification, number and width of the artifact areas in the same group respectively.

[0012] S05. Outputting the quality detection result of the breast three-dimensional ultrasound image according to the statistical result obtained in step S04.

[0013] Specifically, in step S05, when outputting the quality detection result of the breast three-dimensional ultrasound image according to the statistical result obtained in step S04, the following steps are specifically included:

[0014] S51, for any artifact region in the same group, if there is an artifact region classified as a "multi-layer reflection" artifact, the breast three-dimensional ultrasound image quality detection result is directly determined to be unqualified, and the result of "quality detection unqualified, multi-layer reflection exists" is output; if there is no artifact region classified as a "multi-layer reflection" artifact for all artifact regions in the same group, enter S52;

[0015] S52, if the width of the artifact region in the same group is greater than the set width threshold W' or the number of the artifact regions in the same group is greater than the set number threshold N, the quality of the cross-sectional image of the frame number corresponding to the group of artifact regions is unqualified, and the number of frames of the cross-sectional image with unqualified quality is recorded;

[0016] S53. When the number of frames of the unqualified cross-sectional images recorded in step S52 is greater than the set frame number threshold M, the breast three-dimensional ultrasound image quality detection result is judged to be unqualified; if it is not greater than the set frame number threshold M, the breast three-dimensional ultrasound image quality detection result is judged to be qualified.

[0017] Specifically, in step S53, when the number of frames of unqualified cross-sectional images recorded in step S52 is greater than the set frame number threshold M, not only is the breast three-dimensional ultrasound image quality detection result determined to be unqualified, but the quality problems existing in the breast three-dimensional ultrasound image are also output based on the number of artifact areas classified as "poor coupling" artifacts and the number of artifact areas classified as "idle area" artifacts in step S33.

[0018] The beneficial effects of the present invention are as follows: by screening the aspect ratio and average pixel value of the artifact areas of each frame of the cross-sectional image of the breast three-dimensional ultrasound image, the size and classification of each artifact area can be obtained, and according to the frame number corresponding to the artifact area, the classification, number and width of the artifact area in each frame of the cross-sectional image can be judged, so that the quality detection result of the breast three-dimensional ultrasound image can be accurately output based on this, which is convenient for doctors to quickly and accurately judge the quality of the ultrasound detection image and improve the detection efficiency. DETAILED DESCRIPTION

[0019] Embodiment 1, a method for detecting quality of breast three-dimensional ultrasound images, characterized in that it comprises the following steps:

[0020] S01. First, perform three-dimensional ultrasonic detection on the breast to obtain a three-dimensional ultrasonic breast image, and parse a plurality of frames of cross-sectional images from the three-dimensional ultrasonic breast image.

[0021] S02, use the trained artifact detection model to perform artifact region detection on a plurality of frames of cross-sectional sequence images, detect and segment all I artifact regions in the plurality of frames of cross-sectional images, and record the widths W1, W2, W3, ... W of the I artifact regions respectively. I and heights H1, H2, H3, ... H I , and the frame number of the cross-sectional image where the I artifact areas are located.

[0022] S03, classifying the I artifact regions, specifically including:

[0023] S31, first remove the artifact regions whose width is less than the set threshold W0 from among the I artifact regions;

[0024] S32, calculate the aspect ratios W1 / H1, W2 / H2, W3 / H3, ..., W of the I artifact regions I / H IThe aspect ratios of the artifact regions removed in step S31 are arranged in ascending order, and an aspect ratio growth point with an obvious jump growth is found therein, and the artifact regions corresponding to the aspect ratio growth point and the subsequent aspect ratios are classified as "multi-layer reflection" artifacts, and the remaining artifact regions are classified as "poor coupling" or "idle area" artifacts;

[0025] S33. In the artifact areas classified as "poor coupling" or "idle area" artifacts in step S32, the average pixel value of each artifact area is calculated respectively, and the average pixel values ​​are arranged in ascending order, and the average pixel value growth point where an obvious jump-like increase occurs is found, and the artifact areas corresponding to the average pixel value growth point and the subsequent average pixel values ​​are classified as "poor coupling" artifacts, and the remaining artifact areas are classified as "idle area" artifacts.

[0026] S04. Group the artifact areas classified in step S32 and step S33 according to the frame numbers of the cross-sectional images where the artifact areas are located, group the artifact areas with the same frame numbers of the cross-sectional images into the same group, and count the classification, number and width of the artifact areas in the same group respectively.

[0027] S05. Outputting the quality detection result of the breast three-dimensional ultrasound image according to the statistical result obtained in step S04.

[0028] Specifically, in step S05, when outputting the quality detection result of the breast three-dimensional ultrasound image according to the statistical result obtained in step S04, the following steps are specifically included:

[0029] S51, for any artifact region in the same group, if there is an artifact region classified as a "multi-layer reflection" artifact, the breast three-dimensional ultrasound image quality detection result is directly determined to be unqualified, and the result of "quality detection unqualified, multi-layer reflection exists" is output; if there is no artifact region classified as a "multi-layer reflection" artifact for all artifact regions in the same group, enter S52;

[0030] S52, if the width of the artifact region in the same group is greater than the set width threshold W' or the number of the artifact regions in the same group is greater than the set number threshold N, the quality of the cross-sectional image of the frame number corresponding to the group of artifact regions is unqualified, and the number of frames of the cross-sectional image with unqualified quality is recorded;

[0031] S53. When the number of frames of the unqualified cross-sectional images recorded in step S52 is greater than the set frame number threshold M, the breast three-dimensional ultrasound image quality detection result is judged to be unqualified; if it is not greater than the set frame number threshold M, the breast three-dimensional ultrasound image quality detection result is judged to be qualified.

[0032] Specifically, in step S53, when the number of frames of cross-sectional images with unqualified quality recorded in step S52 is greater than the set frame number threshold M, not only is it determined that the breast three-dimensional ultrasound image quality detection result is unqualified, but also the quality problems existing in the breast three-dimensional ultrasound image are output according to the number of artifact areas classified as "poor coupling" artifacts and the number of artifact areas classified as "idle area" artifacts in step S33. Specifically, the quality problems existing in the breast three-dimensional ultrasound image can be output directly by outputting the number of artifact areas classified as "poor coupling" artifacts and the number of artifact areas classified as "idle area" artifacts, or the problem of "poor coupling" and / or "idle area" existing in the breast three-dimensional ultrasound image can be output according to the number of "poor coupling" and "idle area". For example, when the number of "poor coupling" exceeds the set value, it is output that the breast three-dimensional ultrasound image has the problem of "poor coupling", and when the number of "idle area" exceeds the set value, it is output that the breast three-dimensional ultrasound image has the problem of "idle area".

[0033] Of course, the above are only preferred embodiments of the present invention, and are not intended to limit the scope of use of the present invention. Therefore, any equivalent changes made to the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A quality detection method for breast three-dimensional ultrasound images, characterized in that: The steps include: S01, firstly perform three-dimensional ultrasound detection on the breast to obtain a three-dimensional ultrasound breast image, and parse a plurality of frames of cross-sectional images from the three-dimensional ultrasound breast image; S02, use the trained artifact detection model to perform artifact region detection on a plurality of frames of cross-sectional sequence images, detect and segment all I artifact regions in the plurality of frames of cross-sectional images, and record the widths W1, W2, W3, ... W of the I artifact regions respectively. I and heights H1, H2, H3, ... H I , and the frame number of the cross-sectional image where the I artifact area is located; S03, classifying the I artifact regions, specifically including: S31, first remove the artifact regions whose width is less than the set threshold W0 from among the I artifact regions; S32, calculate the aspect ratios W1 / H1, W2 / H2, W3 / H3, ..., W of the I artifact regions I / H I The aspect ratios of the artifact regions removed in step S31 are arranged in ascending order, and an aspect ratio growth point with an obvious jump growth is found among them, and the artifact regions corresponding to the aspect ratio growth point and the subsequent aspect ratios are classified as "multi-layer reflection" artifacts, and the remaining artifact regions are classified as "poor coupling" or "idle area" artifacts; S33, in the artifact areas classified as "poor coupling" or "idle area" artifacts in step S32, respectively calculate the average pixel value of each artifact area, and arrange the average pixel values ​​in ascending order, find the average pixel value growth point with obvious jump growth, and classify the artifact areas corresponding to the average pixel value growth point and the subsequent average pixel values ​​as "poor coupling" artifacts, and classify the remaining artifact areas as "idle area" artifacts; S04, grouping the artifact regions classified in step S32 and step S33 according to the frame numbers of the cross-sectional images where the artifact regions are located, grouping the artifact regions with the same frame numbers of the cross-sectional images into the same group, and counting the classification, number and width of the artifact regions in the same group respectively; S05, outputting the quality detection result of the breast three-dimensional ultrasound image according to the statistical result obtained in step S04, specifically comprising the following steps: S51, for any artifact region in the same group, if there is an artifact region classified as a "multi-layer reflection" artifact, the breast three-dimensional ultrasound image quality detection result is directly determined to be unqualified, and the result of "quality detection unqualified, multi-layer reflection exists" is output; if there is no artifact region classified as a "multi-layer reflection" artifact for all artifact regions in the same group, enter S52; S52, if the width of the artifact region in the same group is greater than the set width threshold W' or the number of the artifact regions in the same group is greater than the set number threshold N, the quality of the cross-sectional image of the frame number corresponding to the group of artifact regions is unqualified, and the number of frames of the cross-sectional image with unqualified quality is recorded; S53. When the number of frames of the unqualified cross-sectional images recorded in step S52 is greater than the set frame number threshold M, the breast three-dimensional ultrasound image quality detection result is judged to be unqualified; if it is not greater than the set frame number threshold M, the breast three-dimensional ultrasound image quality detection result is judged to be qualified.

2. A quality detection method for breast three-dimensional ultrasound images according to claim 1, characterized in that: In the step S53, when the frame number of the cross-sectional images with unqualified quality recorded in step S52 is greater than the set frame number threshold M, not only is the breast three-dimensional ultrasound image quality detection result determined to be unqualified, but the quality problems existing in the breast three-dimensional ultrasound image are also output based on the number of artifact areas classified as "poor coupling" artifacts and the number of artifact areas classified as "idle area" artifacts in step S33.