A tumor test film checking apparatus and method

Abstract

The present application relates to the technical field of image processing, in particular to a kind of tumor test film checking equipment and method, comprising: based on the gray scale gradient value of each pixel point in the tumor test image of target site, clear area and fuzzy area are determined from the tumor test image, clear area includes edge region and uniform area;Based on the edge profile line of the edge region of the gray scale value of the pixel point in fuzzy area and surrounding, different tissues of fuzzy area are divided into multiple sub-regions;According to the gray scale gradient value and the gray scale value in fuzzy area, the gray scale gradient value on the edge profile line of the edge region around fuzzy area, the first quantity of clear area around fuzzy area, the second quantity of uniform area around fuzzy area and the gray scale value of the pixel point in each sub-region, the division accuracy of fuzzy area is determined;Based on the division accuracy, the division result of fuzzy area is determined.Such, the present application improves the region segmentation and marking accuracy of brain tumor MRI image.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, specifically to a tumor examination film examination device and method. Background Technology

[0002] Brain tumors are among the most common types of tumors, second only to tumors of the stomach, lungs, uterus, breast, and esophagus. Malignant tumors account for over 50% of these, with an extremely high probability of progressing to brain cancer. Early screening and diagnosis are crucial for improving treatment effectiveness and preventing disease progression and cancerous development. Currently, clinical diagnosis of brain tumors often relies on imaging techniques such as magnetic resonance imaging (MRI). Brain scans are used to obtain images to determine the presence and location of the tumor, providing a basis for subsequent treatment. However, the obtained MRI images may have issues such as occlusion and irrelevant information, directly affecting the efficiency of diagnosis. Therefore, preprocessing of the images using diagnostic equipment, including image enhancement and segmentation marking, is usually necessary to optimize image quality and facilitate analysis by doctors.

[0003] In the region segmentation and labeling of brain tumor MRI images, current technologies largely rely on image grayscale values, matching them with the grayscale features of tissues such as the cerebral cortex, gray matter, and cerebrospinal fluid to achieve segmentation and labeling of each tissue. However, the spatial structure of the human brain is extremely complex, with numerous folds and sulci, and different tissues easily overlap, leading to image blurring. In this situation, relying solely on grayscale values ​​for segmentation and labeling is highly susceptible to misjudgment and omission of blurred areas, resulting in incorrect tissue labeling and affecting the accuracy of subsequent diagnosis and treatment. Summary of the Invention

[0004] To address the technical problem of low accuracy in region segmentation and labeling of brain tumor MRI images, the present invention aims to provide a tumor examination film examination device and method.

[0005] To solve the above technical problems, the specific technical solution adopted is as follows: In a first aspect, embodiments of the present invention provide a method for examining tumor examination images, comprising: determining clear regions and blurred regions from a tumor examination image based on the grayscale gradient values ​​of each pixel in the tumor examination image of the target site, wherein the clear regions include edge regions and uniform regions; dividing different tissues in the blurred regions into multiple sub-regions based on the grayscale values ​​of pixels in the blurred regions and the edge contours of the surrounding edge regions; determining the accuracy of the blurred region division based on the grayscale gradient values ​​and grayscale values ​​in the blurred regions, the grayscale gradient values ​​on the edge contours of the surrounding edge regions, a first number of clear regions surrounding the blurred regions, a second number of uniform regions surrounding the blurred regions, and the grayscale values ​​of pixels in each sub-region; and determining the division result of the blurred regions based on the division accuracy.

[0006] Optionally, dividing the blurred region into multiple sub-regions based on the grayscale values ​​of pixels within the blurred region and the edge contours of surrounding edge regions includes: determining the degree of blending between the blurred region and surrounding tissues based on a first number of clear regions surrounding the blurred region, a number of uniform regions within the clear region, a first area of ​​the uniform regions, and a second area of ​​the blurred region; determining a first probability that multiple tissues exist within the blurred region based on the degree of blending between the blurred region and surrounding tissues, as well as the grayscale values ​​and spacing between any two pixels within the blurred region; determining that multiple tissues exist within the blurred region if the first probability is greater than a first threshold; determining the edge contours of the edge regions surrounding the blurred region, and dividing the blurred region into multiple sub-regions based on each edge contour.

[0007] Optionally, determining the degree of blending between the blurred region and the surrounding tissue based on a first number of clear regions around the blurred region, a number of uniform regions in the clear region, a first area of ​​the uniform region, and a second area of ​​the blurred region includes: using the ratio between the first number and the number of uniform regions in the clear region as the degree of circumference of the edge region around the blurred region; calculating a first ratio between the first area and the second area; and determining the degree of blending between the blurred region and the surrounding tissue based on the degree of circumference of the edge region and each of the first ratios.

[0008] Optionally, determining the first possibility of multiple tissues existing within a blurred region based on the degree of mixing between the blurred region and surrounding tissues, as well as the grayscale value and spacing between any two pixels within the blurred region, includes: calculating a second ratio between the degree of mixing between the blurred region and surrounding tissues and the average degree of mixing of surrounding tissues, and a third ratio between the grayscale value and spacing between any two pixels within the blurred region; selecting the maximum and minimum values ​​from each of the third ratios; and determining the first possibility based on the second ratio, the maximum value, and the minimum value.

[0009] Optionally, determining the accuracy of blurry region segmentation based on the gray-level gradient values ​​and gray values ​​within the blurry region, the gray-level gradient values ​​on the edge contours of the edge regions surrounding the blurry region, the first number of clear regions surrounding the blurry region, the second number of uniform regions surrounding the blurry region, and the gray values ​​of pixels in each sub-region includes: determining a second possibility that multiple tissues exist in each sub-region; determining a first tissue difference degree between sub-regions based on the maximum gray-level gradient value within the blurry region, the average gray-level gradient values ​​on the edge contours of the edge regions surrounding the blurry region, the first number, the second number, and the gray values ​​of pixels in each sub-region; determining a second tissue difference degree between the blurry region and the surrounding clear regions based on the maximum gray-level gradient value within the blurry region, the average gray-level gradient values ​​on the edge contours of the edge regions surrounding the blurry region, the first number, the second number, and the gray values ​​of pixels in the clear regions surrounding the blurry region; and determining the accuracy of blurry region segmentation based on the first possibility, the second possibility, the first tissue difference degree, and the second tissue difference degree.

[0010] Optionally, determining the first organizational difference degree between sub-regions based on the maximum gray-level gradient value within the blurred region, the average gray-level gradient value on the edge contour lines of the edge regions surrounding the blurred region, a first quantity, a second quantity, and the gray-level values ​​of pixels within each sub-region includes: calculating the variance of the mean gray-level values ​​of pixels within each sub-region, and the sum of the ranges of gray-level values ​​within each sub-region; calculating a fourth ratio between the average gray-level gradient value on the edge contour lines of the edge regions surrounding the blurred region and the maximum gray-level gradient value within the blurred region; and determining the first organizational difference degree between sub-regions based on the difference between the first quantity and the second quantity, the sum of the variance and the range, and the fourth ratio.

[0011] Optionally, determining the accuracy of the fuzzy region segmentation based on the first possibility, the second possibility, the first organizational difference, and the second organizational difference includes: calculating a fifth ratio between the first organizational difference and the second organizational difference, and a second difference between the first possibility and the second possibility; and determining the segmentation accuracy based on the fifth ratio and the second difference.

[0012] Optionally, determining the segmentation result of the fuzzy region based on the segmentation accuracy includes: adjusting the segmentation result of the fuzzy region based on the gradient values ​​of the surrounding pixels of the boundary lines of each sub-region after the fuzzy region is segmented, and selecting the segmentation result with the highest segmentation accuracy as the final segmentation result of the fuzzy region.

[0013] Optionally, determining clear and blurred regions from a tumor examination image based on the grayscale gradient values ​​of each pixel in the tumor examination image of the target site includes: calculating the grayscale gradient values ​​of each pixel in the tumor examination image and determining a gradient value histogram of the tumor examination image based on the grayscale gradient values ​​of each pixel; performing edge detection on the tumor examination image to determine the minimum grayscale gradient value of pixels in the edge contour; determining the target grayscale gradient value corresponding to the maximum number of occurrences of the same grayscale gradient value from the gradient value histogram; classifying pixels with grayscale gradient values ​​greater than or equal to the minimum value as edge regions, pixels with grayscale gradient values ​​less than or equal to the target grayscale gradient value as uniform regions, and pixels with grayscale gradient values ​​between the minimum value and the target grayscale gradient value as blurred regions.

[0014] In a second aspect, embodiments of the present invention provide an electronic device, including: a processor and a memory; wherein the memory is used to store a computer program that can run on the processor; the processor is used to execute the program stored in the memory to implement the steps of the tumor examination film examination method mentioned in the first aspect.

[0015] This invention offers the following advantages: It distinguishes between clear regions (including edge regions and uniform regions) and blurred regions based on pixel grayscale gradient values, overcoming the limitations of traditional image segmentation that relies solely on a single grayscale value and easily confuses similar grayscale regions. Grayscale gradient values ​​can more sensitively capture differences in grayscale changes between pixels, making the contour features of edge regions more prominent and the grayscale consistency of uniform regions more explicit. This allows for precise localization of the blurred region's range, avoiding misjudgment between clear and blurred regions and improving the accuracy of region segmentation and labeling in brain tumor MRI images. Furthermore, by combining pixel grayscale values ​​within the blurred region with the contour lines of surrounding edge regions, multiple sub-regions are divided, further refining the boundaries of different tissues within the blurred region, thus achieving precise segmentation of different tissues within the blurred region. By integrating the grayscale gradient values ​​of the blurred region itself, the grayscale gradient values ​​of the surrounding edge region contour lines, and the number of surrounding clear and uniform regions, the segmentation accuracy is determined. This directly reflects the reliability of the blurred region segmentation results, yielding more accurate segmentation results and improving the accuracy of subsequent diagnosis and treatment. Attached Figure Description

[0016] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of a tumor examination film examination method disclosed in one embodiment of the present invention; Figure 2 A series of tumor images of a user's brain provided as an embodiment of the present invention; Figure 3 This invention provides a tumor examination image and its corresponding edge detection image as an embodiment of the present invention. Figure 4 This is a schematic diagram of a gradient value histogram provided in one embodiment of the present invention; Figure 5 This is a schematic diagram illustrating the intermingling relationship between a blurred region and its surrounding tissue, provided in one embodiment of the present invention. Figure 6 This is a schematic diagram of the structure of an electronic device disclosed in one embodiment of the present invention. Detailed Implementation

[0018] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a tumor examination film examination device and method proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0020] The specific scheme of the tumor examination film examination method disclosed in this invention will be described in detail below with reference to the accompanying drawings.

[0021] Example 1: Please see Figure 1 The document illustrates a flowchart of a tumor examination film examination method provided by an embodiment of the present invention, comprising: Step S101: Based on the grayscale gradient values ​​of each pixel in the tumor examination image of the target site, determine the clear area and the blurred area from the tumor examination image.

[0022] The clear regions include edge regions and uniform regions.

[0023] Specifically, the patient lies in the correct position on the MRI scanner. Taking the brain as the target area, MRI scans are performed on the brain to obtain a series of consecutive, adjacent brain images from the top of the patient's head to their chin, serving as a tumor detection method. For example, as shown... Figure 2 As shown, Figure 2 This invention provides a series of tumor images of a user's brain, as one embodiment of the present invention. Figure 2 The images include a series of consecutive adjacent brain MRI scans showing tumors from the top of the head to the chin.

[0024] Furthermore, in this embodiment of the invention, the grayscale gradient values ​​at each pixel of the tumor examination image are calculated, and a gradient value histogram is obtained based on the grayscale gradient values ​​of each pixel. An edge detection operation is then performed on the image to obtain an edge-detected image. For example, as shown... Figure 3 As shown, Figure 3 This invention provides a tumor examination image and its corresponding edge detection image as an embodiment of the present invention. Figure 3 The left side shows the original image of the tumor examination, and the right side shows the edge detection image of that original image. For example... Figure 4 As shown, Figure 4 This is a schematic diagram of a gradient value histogram provided in an embodiment of the present invention. In this gradient value histogram, the horizontal axis represents the gray-level gradient value, and the vertical axis represents the number of occurrences of that gray-level gradient value. Since the gray-level gradient value represents the degree of change in the gray-level value of a pixel, a smaller degree of change indicates that its gray-level value is more similar to that of surrounding pixels, suggesting a higher probability that it belongs to the same tissue, and can be considered a clear area of ​​the image. Furthermore, the area occupied by the same tissue within the image is larger, resulting in a relatively higher proportion of smaller gradient values. Conversely, a larger degree of change indicates that its gray-level value is significantly different from that of surrounding pixels, allowing for direct differentiation (detected by edge detection algorithms), suggesting a higher probability that it belongs to a different tissue, and can also be considered a clear area of ​​the image. Therefore, as an optional embodiment of the present invention, determining clear and blurred regions from a tumor examination image based on the gray-level gradient values ​​of each pixel in the tumor examination image of the target site includes: calculating the gray-level gradient values ​​of each pixel in the tumor examination image, and determining a gradient value histogram of the tumor examination image based on the gray-level gradient values ​​of each pixel; performing edge detection on the tumor examination image to determine the minimum gray-level gradient value of the pixels in the edge contour; determining the target gray-level gradient value corresponding to the maximum number of occurrences of the same gray-level gradient value from the gradient value histogram; classifying pixels whose gray-level gradient values ​​are greater than or equal to the minimum value as edge regions, pixels whose gray-level gradient values ​​are less than or equal to the target gray-level gradient value as uniform regions, and pixels whose gray-level gradient values ​​are between the minimum value and the target gray-level gradient value as blurred regions.

[0025] Specifically, embodiments of the present invention can obtain the minimum gray-level gradient value of pixels in the edge contour based on the edge detection image obtained from edge detection, denoted as: The target gray-level gradient value corresponding to the maximum number of occurrences in the gradient histogram is denoted as... When the grayscale gradient value of a pixel is greater than or equal to When the pixel's grayscale gradient value is less than or equal to the edge region that can be detected by the edge detection algorithm, it is considered a clear region; when the grayscale gradient value of a pixel is less than or equal to the edge region, it is considered a clear region. At that time, it is divided into uniform regions belonging to the same organization, which are clear regions. Finally, the grayscale gradient value is set to... The pixels between them are divided into blurred regions, and these blurred regions are marked for further subdivision.

[0026] Step S102: Based on the gray values ​​of pixels within the blurred area and the edge contours of the surrounding edge areas, the blurred area is divided into multiple sub-regions.

[0027] Specifically, the proton density varies among different tissues in the human brain (such as gray matter, cerebral cortex, and cerebrospinal fluid), resulting in significant differences in grayscale values ​​in MRI images. In the clearly defined regions obtained through the above steps, uniform regions exhibit small variations in grayscale values ​​and high similarity, likely representing the same tissue region; edge regions show significant variations in grayscale values, likely representing the boundary between different tissues; while the remaining blurred regions exhibit large variations in grayscale values ​​but are insufficient for clear division, potentially containing the same tissue region or multiple tissues coexisting. Therefore, it is necessary to analyze and further segment these blurred regions where multiple tissues coexist.

[0028] Furthermore, as an optional embodiment of the present invention, dividing the different tissues of the blurred region into multiple sub-regions based on the gray values ​​of pixels within the blurred region and the edge contours of the surrounding edge regions includes: determining the degree of mixing between the blurred region and the surrounding tissues based on a first number of clear regions surrounding the blurred region, a number of uniform regions within the clear region, a first area of ​​the uniform regions, and a second area of ​​the blurred region; determining a first probability that multiple tissues exist within the blurred region based on the degree of mixing between the blurred region and the surrounding tissues, as well as the gray values ​​and spacing between any two pixels within the blurred region; determining that multiple tissues exist within the blurred region if the first probability is greater than a first threshold; determining the edge contours of the edge regions surrounding the blurred region, and dividing the blurred region into multiple sub-regions based on each edge contour.

[0029] Specifically, brain tissues are often intertwined. If multiple tissues exist within a blurred region, there is a high probability that multiple tissues and their boundaries also exist in the surrounding area. In this embodiment of the invention, the smallest circumcircle of the blurred region is defined, and the number of distinct clear regions adjacent to the blurred region within this smallest circumcircle is denoted as . Furthermore, such as Figure 5 As shown, Figure 5This is a schematic diagram illustrating the intermingling relationship between a blurred region and its surrounding tissues, according to an embodiment of the present invention. The brain has numerous sulci and grooves, leading to frequent interpenetration and fusion between tissues. The blurred region exhibits intermingling with other surrounding tissues; the higher the degree of interpenetration, the greater the likelihood of that tissue existing within it. Let the number of other uniform regions existing within the aforementioned minimum circumcircle be denoted as . The area of ​​each uniform region within the smallest circumcircle is denoted as . The area of ​​the fuzzy region is .

[0030] Furthermore, as an optional embodiment of the present invention, determining the degree of blending between the blurred region and the surrounding tissue based on the first number of clear regions around the blurred region, the number of uniform regions in the clear region, the first area of ​​the uniform region, and the second area of ​​the blurred region includes: taking the ratio between the first number and the number of uniform regions in the clear region as the degree of edge region surrounding the blurred region, calculating the first ratio between the first area and the second area; and determining the degree of blending between the blurred region and the surrounding tissue based on the degree of edge region surrounding and each first ratio.

[0031] Specifically, the embodiments of the present invention use the following formula to calculate the degree of mixing between the blurred region and the surrounding tissue: In the above formula, It indicates the degree of blending between the blurred area and the surrounding tissue. This represents the first number of clear regions surrounding the blurred region. The number of uniform regions (different human tissues) in a clear area. This represents the number of marginal regions (inter-organizational boundaries) in a clear region. The larger this value, the more inter-organizational boundaries the ambiguous region has, the higher the probability that organizational boundaries exist within it, and the higher its degree of mixing with its surrounding organizations. Indicates the first The first area of ​​a uniform region within the smallest circumcircle. The second area represents the fuzzy region. This indicates the degree of encirclement of the edge region around the blurred area. This represents the degree to which uniform regions surrounding the blurred region penetrate into the blurred region. The larger this value, the higher the degree of mutual penetration between the blurred region and other surrounding tissues. When the degree of encirclement of the blurred region by the edge region and the degree of penetration by the uniform regions surrounding the blurred region are both greater, the degree of mixing between the blurred region and its surrounding tissues is higher.

[0032] In summary, the degree of mixing between the tissue within this fuzzy region and its surrounding tissue is denoted as . The larger this value, the higher its degree of mixing with surrounding tissues, and the greater the likelihood that it contains different tissues. The average degree of mixing with its adjacent surrounding tissues is calculated using the same method and denoted as . .

[0033] Furthermore, if the blurred region contains various different human brain tissues, the grayscale values ​​will differ significantly between these tissues, while remaining uniform within the same tissue. Therefore, the distance between pixels with significantly different grayscale values ​​within the blurred region may also be relatively large. Let the ratio of the difference in grayscale values ​​between any two pixels within this blurred region to the distance between them be denoted as... Further, obtain the results for any two pixels. Values, and record the resulting series The maximum value in the value is The minimum value is Based on the data obtained above, the first probability of the existence of multiple organizations within this ambiguous region is denoted as: The spacing between pixels can be represented by Euclidean distance or Manhattan distance. In this embodiment, Euclidean distance is used as the spacing between pixels.

[0034] Furthermore, as an optional embodiment of the present invention, determining the first possibility of the existence of multiple tissues within the blurred region based on the degree of mixing between the blurred region and surrounding tissues, as well as the grayscale value and spacing between any two pixels within the blurred region, includes: calculating a second ratio between the degree of mixing between the blurred region and surrounding tissues and the average degree of mixing of surrounding tissues, and a third ratio between the grayscale value and spacing between any two pixels within the blurred region; selecting the maximum and minimum values ​​from each of the third ratios; and determining the first possibility based on the second ratio, the maximum value, and the minimum value.

[0035] Specifically, the embodiments of the present invention use the following formula to calculate the first possibility: In the above formula, This indicates the first possibility that multiple organizations exist within the ambiguous region. This indicates the degree of blending between the fuzzy region and its surrounding tissues. The ratio is the average of the mixing degree between the fuzzy region and other regions surrounding it and its surrounding organizations. The larger the ratio, the higher the relative mixing degree within its surrounding organizations, and the higher the probability that it contains its surrounding organizations. This represents the maximum value among the ratios of the difference in grayscale values ​​between two pixels within a blurred region to the distance between the two pixels. It represents the minimum value of the ratio of the difference in grayscale value between two pixels within a blurred region to the distance between the two pixels. This represents the ratio of the minimum to the maximum value. The larger this value is and the closer it is to the maximum value, the better. The stronger the correlation between the grayscale value difference of each pixel within the blurred region and its distance, the more orderly the change of grayscale value with distance, and the higher the possibility of multiple structures contained within it. Used for Perform normalization, mapping its range to Inside.

[0036] In summary, it is possible that multiple organizations exist within this ambiguous region. The larger this value, the more likely it is that multiple different human brain tissues exist within the ambiguous region. This embodiment of the invention sets a first threshold. (This can be customized by the user; this example is set to...) ),like They believe that multiple organizations exist within this ambiguous region, requiring further subdivision.

[0037] Furthermore, the above steps revealed fuzzy regions with varying brain sizes. While no clearly high-gradient edge regions were identified within these regions, grayscale differences still existed between different tissues. Moreover, the boundary curves of the surrounding tissues could, to some extent, reflect the location of the boundaries between these tissues. The fuzzy regions could be pre-divided based on the trends of the surrounding edge regions, and the division results evaluated. Taking the fuzzy regions with varying tissues as an example, the edge regions adjacent to these fuzzy regions are likely areas where it is difficult to identify larger gradient values ​​within the fuzzy region. The trends of the edge regions could be further simulated, and the fuzzy regions could be divided by combining the simulated curves and the grayscale gradient values ​​of the surrounding areas. First, the tumor examination image was divided into image coordinate systems, and the coordinates of each point on a certain boundary region adjacent to the fuzzy region were obtained. The edge contour lines of this edge region were simulated using the least squares method, and the total number of edge contour lines was calculated. If multiple edge contour lines intersect within a fuzzy region, only the edge contour line before the intersection point is taken. This divides the fuzzy region into [number of lines]. Block region.

[0038] Step S103: Determine the accuracy of the blurry region division based on the gray-level gradient value and gray value within the blurry region, the gray-level gradient value on the edge contour line of the edge region surrounding the blurry region, the first number of clear regions surrounding the blurry region, the second number of uniform regions surrounding the blurry region, and the gray value of the pixels in each sub-region.

[0039] Specifically, the division Each sub-region of the block region should contain only the same tissue region. Based on the calculation method of the first possibility described above, the second possibility of multiple tissues existing in each sub-region is calculated, which is denoted as follows in this embodiment of the invention: After segmentation, each sub-region exhibits different tissue regions with significant differences in grayscale values, and the grayscale gradient values ​​of pixels along the edge contours of these tissue regions should be substantial. In this embodiment of the invention, the maximum grayscale gradient value within the blurred region is recorded as... The mean of the grayscale gradient values ​​on each edge contour line is denoted as . The range of gray values ​​in each sub-region is obtained based on the division results and denoted as . Further calculate the variance between the mean gray values ​​of pixels in each sub-region, denoted as . As an optional embodiment of the present invention, determining the first organizational difference degree between sub-regions based on the maximum gray-level gradient value within the blurred region, the average gray-level gradient values ​​on the edge contour lines of the edge regions surrounding the blurred region, a first quantity, a second quantity, and the gray-level values ​​of pixels in each sub-region includes: calculating the variance of the mean gray-level values ​​of pixels in each sub-region, and the superposition of the ranges of gray-level values ​​in each sub-region; calculating a fourth ratio between the average gray-level gradient values ​​on the edge contour lines of the edge regions surrounding the blurred region and the maximum gray-level gradient value within the blurred region; and determining the first organizational difference degree between sub-regions based on the difference between the first quantity and the second quantity, the superposition of the variance and the range, and the fourth ratio.

[0040] Specifically, the embodiments of the present invention use the following formula to calculate the first tissue difference degree between each sub-region: In the above formula, This indicates the first organizational difference between the sub-regions. This represents the variance between the mean grayscale values ​​of pixels in each sub-region. The larger this value, the greater the difference in the mean grayscale values ​​between the sub-regions. This is the sum of the ranges of gray values ​​within each sub-region. The smaller this value, the smaller the difference in gray values ​​within each sub-region. The larger the ratio of the two, the higher the probability that the sub-regions belong to the same organization and different sub-regions belong to different organizations, and the greater the organizational difference between the sub-regions. This represents the first number of clear regions surrounding the blurred region. The second number is the uniform region (different human tissues) in the clear region. This represents the number of peripheral regions (inter-organizational boundaries) in a clear region. The larger this value, the more inter-organizational boundaries surround the blurry region, and the higher the probability that organizational boundaries exist within it. For the fuzzy region, the first The average grayscale gradient value along the edge contour line, and its value relative to the maximum grayscale gradient value within the blurred region. The higher the ratio, the more likely the edge contour line is to be the boundary between different tissues within the ambiguous region. The mean value is the ratio of the mean gray-level gradient values ​​of all edge contours to the maximum gray-level gradient value within the sub-region. The larger this value is, the larger the gray-level gradient value of consecutive pixels on the divided edge contours within the sub-region, the more obvious the gray-level value change is relative to other pixels within the sub-region, and the greater the organizational difference between the divided sub-regions.

[0041] In summary, the first organizational difference among the sub-regions in this classification result is obtained. The larger this value, the higher the probability that the sub-regions obtained from this segmentation belong to the same tissue and that different sub-regions belong to different tissues. Further, based on the image before segmentation, the second tissue difference degree between the blurred region and its adjacent clear regions is calculated. .

[0042] It is understood that the embodiments of this application calculate the second tissue difference degree The specific method is the same as the above calculation of the first tissue difference. The specific methods are the same or similar, and will not be elaborated here.

[0043] Furthermore, as an optional embodiment of the present invention, determining the accuracy of the fuzzy region segmentation based on the first possibility, the second possibility, the first organizational difference degree, and the second organizational difference degree includes: calculating a fifth ratio between the first organizational difference degree and the second organizational difference degree, and a second difference between the first possibility and the second possibility; and determining the segmentation accuracy based on the fifth ratio and the second difference.

[0044] Specifically, the embodiments of the present invention use the following formula to calculate the division accuracy: In the above formula, Indicates the accuracy of the division. This represents the first organizational difference among the sub-regions obtained from this partitioning result. The second organizational difference degree is the ratio between the fuzzy region before division and its adjacent clear region. The ratio of the two indicates the consistency between the difference performance of each region in the fuzzy region after division and the difference performance of each organization around the fuzzy region. The larger the value, the closer the difference performance of each sub-region after division is to the organizational difference. The higher the probability that different sub-regions belong to different organizations, the higher the accuracy of this division. This indicates the first possibility that multiple organizations exist within the ambiguous region. This indicates the second possibility that multiple organizations exist within the i-th subregion. The difference between the possibility of multiple organizations existing in the fuzzy region before partitioning and the possibility of multiple organizations existing within each sub-region after partitioning is calculated. Indicates the number of sub-regions. This is the sum of the differences between the first possibility and each of the second possibilities. The larger this value is, the higher the probability that each sub-region after the division belongs to the same organization compared to the entire fuzzy region before the division, and the higher the accuracy of this division.

[0045] In summary, the accuracy of this division is obtained. The larger this value, the higher the accuracy of the division.

[0046] Step S104: Determine the segmentation result of the fuzzy region based on the segmentation accuracy.

[0047] Specifically, in this embodiment of the invention, a second threshold can be set. If the accuracy of the division is higher than the second threshold, it indicates that the accuracy of the division result is high, and it can be used as the final division result for subsequent steps. If it is lower than the second threshold, the above process is repeated to re-divide the data.

[0048] Furthermore, as an optional embodiment of the present invention, determining the segmentation result of the fuzzy region based on the segmentation accuracy includes: adjusting the segmentation result of the fuzzy region based on the surrounding pixel gradient values ​​of the pixel points of the boundary lines of each sub-region after the fuzzy region is segmented, and selecting the segmentation result with the highest segmentation accuracy as the final segmentation result of the fuzzy region.

[0049] Specifically, the segmentation result obtained solely from the edge contour line may not be the optimal result. Adjustments need to be made based on the performance of pixels surrounding the edge contour line, and the accuracy changes before and after the adjustment should be compared to obtain a better segmentation result. In this embodiment, each pixel along the edge contour line is evaluated. It compares whether there are other pixels with a higher grayscale gradient value than the pixel in its eight-neighborhood that are not simulated as part of the edge contour line. These pixels are then included in the edge contour line to replace the original pixels. This process is repeated for all pixels on the edge contour line to obtain the adjusted edge contour line. Based on the adjusted edge contour line, the blurred region is re-segmented using the method described in the above embodiment, and the segmentation accuracy of the re-segmented blurred region is re-evaluated. ,Compare and The larger value is selected as the partitioning result, and this step is repeated until the accuracy of the adjusted partitioning no longer increases. Until then, the partitioning result corresponding to the maximum partitioning accuracy is taken as the final partitioning result.

[0050] This invention distinguishes between clear regions (including edge regions and uniform regions) and blurred regions based on pixel grayscale gradient values. This overcomes the limitations of traditional image segmentation, which relies solely on a single grayscale value and is prone to confusing similar grayscale regions. Grayscale gradient values ​​can more sensitively capture differences in grayscale changes between pixels, making the contour features of edge regions more prominent and the grayscale consistency of uniform regions more explicit. This allows for precise localization of the blurred region's range, avoiding misjudgment between clear and blurred regions and improving the accuracy of region segmentation and labeling in brain tumor MRI images. Furthermore, by combining pixel grayscale values ​​within the blurred region with the contour lines of surrounding edge regions, multiple sub-regions are divided, further refining the boundaries of different tissues within the blurred region, thus achieving precise segmentation of different tissues within the blurred region. By integrating the grayscale gradient values ​​of the blurred region itself, the grayscale gradient values ​​of the surrounding edge region contour lines, and the number of surrounding clear and uniform regions, the segmentation accuracy is determined. This directly reflects the reliability of the blurred region segmentation results, yielding more accurate segmentation results and improving the accuracy of subsequent diagnosis and treatment.

[0051] Example 2: Corresponding to the tumor examination film examination method provided in the above embodiments, based on the same technical concept, this embodiment of the invention also provides an electronic device for performing the above-described tumor examination film examination method. Figure 6 This is a schematic diagram of the structure of an electronic device provided in one embodiment of the present invention, as shown below. Figure 6 As shown. Electronic devices can vary considerably due to differences in configuration or performance, and may include one or more processors 601 and memories 602. The memory 602 stores computer programs that can run on the processor 601, and the processor 601 executes the programs stored in the memory 602 to achieve the above. Figure 1 The various steps in the method embodiment are described. The memory 602 can be temporary or persistent storage. The application stored in the memory 602 may include one or more modules (not shown), each module may include a series of computer-executable instructions for the electronic device.

[0052] Furthermore, the processor 601 may be configured to communicate with the memory 602 and execute a series of computer-executable instructions stored in the memory 602 on the electronic device. The electronic device may also include one or more power supplies 603, one or more wired or wireless network interfaces 604, one or more input / output interfaces 605, and one or more keyboards 606.

[0053] Specifically, in this embodiment, the electronic device includes a processor, a communication interface, a memory, and a communication bus; wherein, the processor, the communication interface, and the memory communicate with each other via the bus; the memory is used to store computer programs; and the processor is used to execute the programs stored in the memory to achieve the above. Figure 1 The various steps in the method embodiments are the same as those in the above method embodiments, and have the same beneficial effects. To avoid repetition, the embodiments of the present invention will not be described again here.

[0054] It should be noted that the electronic device provided in this embodiment of the invention and the tumor examination film examination method provided in this embodiment of the invention are based on the same application concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned tumor examination film examination method, and has the same or similar beneficial effects. Repeated parts will not be repeated.

[0055] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0056] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0057] This invention also provides a computer-readable storage medium that stores one or more programs, which, when executed by an electronic device including multiple applications, cause the electronic device to perform... Figure 1 The methods disclosed in the embodiments shown achieve the functions and beneficial effects of the methods in the preceding method embodiments, and will not be repeated here.

[0058] The computer-readable storage media include read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for examining tumor imaging films, characterized in that, include: Based on the grayscale gradient values ​​of each pixel in the tumor examination image of the target site, clear regions and blurred regions are determined from the tumor examination image, and the clear regions include edge regions and uniform regions. Based on the gray values ​​of pixels within the blurred region and the edge contours of the surrounding edge regions, the different structures of the blurred region are divided into multiple sub-regions. The accuracy of the blurry region segmentation is determined based on the gray-level gradient value and gray value within the blurry region, the gray-level gradient value on the edge contour line of the edge region surrounding the blurry region, the first number of clear regions surrounding the blurry region, the second number of uniform regions in the clear regions surrounding the blurry region, and the gray-level value of the pixels in each sub-region. The division result of the fuzzy region is determined based on the division accuracy.

2. The method for examining tumor imaging films according to claim 1, characterized in that, The step of dividing the blurred region into multiple sub-regions based on the grayscale values ​​of pixels within the blurred region and the edge contours of surrounding edge regions includes: The degree of blending between the blurred region and the surrounding tissue is determined based on a first number of clear regions surrounding the blurred region, a second number of uniform regions within the clear regions, a first area of ​​the uniform regions, and a second area of ​​the blurred region. Based on the degree of mixing between the blurred region and the surrounding tissues, as well as the grayscale value and spacing of any two pixels within the blurred region, the first possibility of the existence of multiple tissues within the blurred region is determined. If the first probability is greater than the first threshold, it is determined that multiple tissues exist within the fuzzy region; The edge contour lines of the edge regions surrounding the blurred region are determined, and the blurred region is divided into multiple sub-regions based on each edge contour line.

3. The method for examining tumor imaging films according to claim 2, characterized in that, The step of determining the degree of blending between the blurred region and the surrounding tissue based on a first number of clear regions surrounding the blurred region, a second number of uniform regions within the clear regions, a first area of ​​the uniform regions, and a second area of ​​the blurred region includes: The ratio between the first quantity and the second quantity of uniform areas in the clear region is used as the degree of edge region encirclement around the blurred region; the first ratio between the first area and the second area is calculated. The degree of blending between the blurred region and the surrounding tissue is determined based on the degree of encirclement of the edge region and each of the first ratios.

4. The method for examining tumor imaging films according to claim 3, characterized in that, The determination of the first possibility of the presence of multiple tissues within the blurred region, based on the degree of blending between the blurred region and surrounding tissues, and the grayscale value and spacing between any two pixels within the blurred region, includes: Calculate a second ratio between the degree of blending of the blurred region with the surrounding tissue and the average degree of blending of the surrounding tissue, and a third ratio between the gray value of any two pixels within the blurred region and the distance between them; Select the maximum and minimum values ​​from each of the aforementioned third ratios; The first probability is determined based on the second ratio, the maximum value, and the minimum value.

5. The method for examining tumor images according to claim 4, characterized in that, The step of determining the accuracy of blurry region segmentation based on the grayscale gradient value and grayscale value within the blurry region, the grayscale gradient value on the edge contour line of the edge region surrounding the blurry region, the first number of clear regions surrounding the blurry region, the second number of uniform regions in the clear regions surrounding the blurry region, and the grayscale value of pixels in each sub-region includes: To determine the second possibility of multiple organizations existing in each subregion; The first organization difference degree between each sub-region is determined based on the maximum gray-level gradient value within the blurred region, the average gray-level gradient value on the edge contour line of the edge region surrounding the blurred region, the first quantity, the second quantity, and the gray-level value of the pixel in each sub-region. A second tissue difference degree between the blurred region and the surrounding clear region is determined based on the maximum gray-level gradient value within the blurred region, the average gray-level gradient value on the edge contour line of the edge region surrounding the blurred region, the first quantity, the second quantity, and the gray-level value of the pixels in the clear region surrounding the blurred region. The accuracy of the fuzzy region segmentation is determined based on the first probability, the second probability, the first organizational difference, and the second organizational difference.

6. The method for examining tumor imaging films according to claim 5, characterized in that, The step of determining the first organizational difference degree between each sub-region based on the maximum grayscale gradient value within the blurred region, the average grayscale gradient value on the edge contour line of the edge region surrounding the blurred region, the first quantity, the second quantity, and the grayscale value of the pixels in each sub-region includes: Calculate the variance of the mean grayscale value of pixels in each sub-region, and the sum of the ranges of grayscale values ​​in each sub-region; Calculate the fourth ratio between the average gray-level gradient value on the edge contour line of the edge region surrounding the blurred region and the maximum gray-level gradient value within the blurred region; Based on the difference between the first quantity and the second quantity, the sum of the variance and the range, and the fourth ratio, a first organizational difference degree between each of the sub-regions is determined.

7. The method for examining tumor imaging films according to claim 5, characterized in that, Determining the accuracy of the fuzzy region segmentation based on the first probability, the second probability, the first organizational difference, and the second organizational difference includes: Calculate a fifth ratio between the first tissue difference and the second tissue difference, and a second difference between the first probability and the second probability; The accuracy of the division is determined based on the fifth ratio and the second difference.

8. The method for examining tumor imaging films according to claim 1, characterized in that, The determination of the fuzzy region segmentation result based on the segmentation accuracy includes: The fuzzy region division result is adjusted based on the surrounding pixel gradient values ​​of the pixels at the boundary lines of each sub-region after the fuzzy region is divided, and the division result with the highest accuracy is selected as the final division result of the fuzzy region.

9. The method for examining tumor imaging films according to claim 1, characterized in that, The determination of clear and blurred regions from the tumor examination image based on the grayscale gradient values ​​of each pixel in the tumor examination image at the target location includes: Calculate the grayscale gradient value of each pixel in the tumor examination image, and determine the gradient value histogram of the tumor examination image based on the grayscale gradient value of each pixel; Edge detection is performed on the tumor examination image to determine the minimum gray-level gradient value of the pixels in the edge contour; Determine the target gray-level gradient value corresponding to the maximum number of occurrences of the same gray-level gradient value from the gradient value histogram; Pixels with grayscale gradient values ​​greater than or equal to the minimum value are classified as edge regions, pixels with grayscale gradient values ​​less than or equal to the target grayscale gradient value are classified as uniform regions, and pixels with grayscale gradient values ​​between the minimum value and the target grayscale gradient value are classified as blurred regions.

10. An electronic device, characterized in that, include: Processor and memory; wherein the memory is used to store computer programs that can run on the processor; A processor is configured to execute a program stored in memory to implement the steps of the tumor examination film examination method as described in any one of claims 1-9.

Images