Method, device, equipment and medium for determining bleeding area based on brain imaging

Abstract

The present application provides a method, device, equipment and medium for determining a bleeding area based on brain images. The method includes: determining the brain tissue area of ​​each image in a brain CT image sequence; determining the ventricular edge corresponding to each brain tissue area according to each brain tissue area; performing Hough transform straight line detection on each ventricular edge to obtain a target line segment of each ventricular edge; determining a target area according to the target line segment of each ventricular edge, performing parabola fitting on the pixel points of the target area to obtain a fitting result, and determining whether there is a bleeding area in the target area according to the fitting result. The present application realizes the automatic determination of bleeding areas based on brain CT images, thereby eliminating the need for manual participation, and not only has a fast recognition speed, but also a high accuracy rate.

Description

Technical Field

[0001] The embodiments of the present application relate to image processing technology, and in particular to a method, device, equipment and medium for determining a hemorrhage area based on brain images. Background Art

[0002] Intracerebral hemorrhage (ICH) is a type of stroke, accounting for about 13% of strokes. It is caused by rupture of cerebral arteries, leading to localized bleeding in surrounding tissues. ICH includes traumatic bleeding and primary cerebral hemorrhage. Due to its rapid onset, serious and complex condition, high mortality and disability rates, ICH has become the main cause of adult disability and death, seriously endangering the health of the people.

[0003] Whether the bleeding area invades the ventricle is an important indicator to measure the severity of cerebral hemorrhage. The ventricle is a pathway for the circulation of cerebrospinal fluid in the human body. When there is a hematoma in the ventricle, the hematoma will cause the cerebrospinal fluid circulation pathway to be blocked, thus causing acute hydrocephalus, which in turn will cause symptoms of acute increased intracranial pressure, seriously affecting the patient's health.

[0004] At present, determining whether the bleeding area has invaded the ventricle is basically achieved by observing the brain CT scan with the human eye. This is not only affected by subjective factors, but also has a low recognition rate when the bleeding area is not clearly visible, which will lead to large diagnostic errors. Summary of the invention

[0005] The embodiments of the present application provide a method, device, equipment and medium for determining a bleeding area based on brain images, which realizes the automatic determination of the bleeding area based on brain CT images, thereby eliminating the need for human intervention and achieving not only fast recognition speed but also high accuracy.

[0006] In a first aspect, an embodiment of the present application provides a method for determining a hemorrhage area based on brain images, the method comprising:

[0007] Determine the brain tissue region for each image in a brain CT image sequence;

[0008] According to each of the brain tissue regions, determining the ventricle edge corresponding to each of the brain tissue regions;

[0009] Performing Hough transform straight line detection on each ventricle edge to obtain a target line segment of each ventricle edge;

[0010] A target area is determined according to each target line segment of the ventricle edge, a parabola fitting is performed on the pixel points of the target area to obtain a fitting result, and whether there is a hemorrhage area in the target area is determined according to the fitting result.

[0011] In a second aspect, an embodiment of the present application provides a device for determining a bleeding area based on a brain image, the device comprising:

[0012] A first determination module is used to determine the brain tissue area of ​​each image in the brain CT image sequence;

[0013] A second determination module is used to determine the ventricle edge corresponding to each brain tissue region according to each brain tissue region;

[0014] A detection module, used for performing Hough transform straight line detection on each ventricle edge to obtain a target line segment of each ventricle edge;

[0015] The third determination module is used to determine the target area according to the target line segment of each ventricle edge, perform parabola fitting on the pixel points of the target area to obtain a fitting result, and determine whether there is a hemorrhage area in the target area according to the fitting result.

[0016] In a third aspect, an embodiment of the present application provides a device for determining a bleeding area based on brain images, comprising:

[0017] A processor and a memory, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to execute the method for determining the hemorrhage area based on brain imaging as described in the embodiment of the first aspect.

[0018] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium for storing a computer program, wherein the computer program enables a computer to execute the method for determining a hemorrhage area based on brain imaging described in the embodiment of the first aspect.

[0019] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the method for determining a hemorrhage area based on brain imaging described in the embodiment of the first aspect.

[0020] The technical solution disclosed in the embodiments of the present application has the following beneficial effects:

[0021] By determining the brain tissue area of ​​each image in the brain CT image sequence and determining the ventricle edge corresponding to each brain tissue area, Hough transform straight line detection is performed on each ventricle edge to obtain the target line segment of each ventricle edge, and then the target area is determined according to the target line segment of each ventricle edge, and a parabola fitting is performed on the pixel points of the target area, and it is determined whether there is a hemorrhage area in the target area according to the fitting result. In this way, the problem of large errors in diagnostic results caused by the influence of subjective factors on human eye observation and the low recognition rate when the hemorrhage area is not obvious is solved, and the hemorrhage area is automatically determined based on the brain CT image, so that no human participation is required, and the recognition speed is fast and the accuracy is high. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0023] Figure 1 It is a flowchart of a method for determining a hemorrhage area based on brain imaging provided in an embodiment of the present application;

[0024] Figure 2a is a schematic diagram of a brain CT image sequence having 30 brain CT images provided in an embodiment of the present application;

[0025] Figure 2b is a schematic diagram of the brain tissue region of each image in the brain CT image sequence provided in the embodiment of the present application;

[0026] Figure 3 is a flow chart of another method for determining a hemorrhage area based on brain images provided in an embodiment of the present application;

[0027] Figure 4a is a schematic diagram of constructing a grayscale histogram provided in an embodiment of the present application;

[0028] Figure 4b is a schematic diagram of the extracerebral space region of each image in the brain CT image sequence provided in the embodiment of the present application;

[0029] Figure 4c is a schematic diagram of multiple line segments in the extracerebral space region in the 20th brain CT image provided in an embodiment of the present application;

[0030] Figure 4d is a schematic diagram of the sulcus and gyrus region for determining the extracerebral gap region in the 20th brain CT image provided by an embodiment of the present application;

[0031] Figure 4e is a schematic diagram of the sulcus and gyrus region of the extracerebral space region in each brain CT image provided in the embodiment of the present application;

[0032] Figure 4f is a schematic diagram of the lateral ventricle region of the extracerebral space region in each brain CT image provided in the embodiments of the present application;

[0033] Figure 4g is a schematic diagram of the ventricle edge of each lateral ventricle region provided in an embodiment of the present application;

[0034] Figure 5 is a flow chart of another method for determining a hemorrhage area based on brain images provided in an embodiment of the present application;

[0035] Figure 6a is a schematic diagram of Hough transform of the 16th ventricle edge image provided in the embodiment of the present application;

[0036] Figure 6b is a schematic diagram of a straight line expression of a point obtained according to Hough transform provided in an embodiment of the present application;

[0037] Figure 6c is a schematic diagram of performing Hough transformation on each pixel point at the edge of the ventricle provided by an embodiment of the present application;

[0038] Figure 6d It is a schematic diagram of the corresponding relationship between the parameter area and the physical meaning provided in the embodiment of the present application;

[0039] Figure 6e is a schematic diagram of selecting a target line segment from candidate line segments of the 16th ventricle edge image provided by an embodiment of the present application;

[0040] Figure 7 is a flowchart of another method for determining a hemorrhage area based on brain images provided in an embodiment of the present application;

[0041] Figure 8a is a schematic diagram of determining a target area based on a target line segment of the 16th ventricle edge image provided in an embodiment of the present application;

[0042] Figure 8b is a schematic diagram of obtaining an initial clustering curve based on target area 4 in the 16th ventricle edge image provided by an embodiment of the present application;

[0043] Figure 8c It is a schematic diagram of performing binary classification clustering on the initial clustering curve provided in an embodiment of the present application to obtain a clustering result;

[0044] Fig. 9 It is a flowchart of another method for determining a hemorrhage area based on brain images provided in an embodiment of the present application;

[0045] Fig.10 is a schematic block diagram of a device for determining a hemorrhage area based on brain images provided in an embodiment of the present application;

[0046] Fig.11 It is a schematic block diagram of a device for determining a hemorrhage area based on brain imaging provided in an embodiment of the present application. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0048] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0049] The embodiments of the present invention are mainly aimed at the problem in the prior art that when determining whether the bleeding area has invaded the ventricle by observing the brain CT with the human eye, it is not only affected by the doctor's subjective factors, but also the recognition rate is low when the bleeding area is not obviously developed, resulting in large errors in the diagnosis results. A method, device, equipment and medium for determining the bleeding area based on brain images are proposed.

[0050] The following is a detailed description of a method, device, apparatus and medium for determining a bleeding area based on brain imaging provided by an embodiment of the present application in conjunction with the accompanying drawings. Figure 1 , a method for determining a hemorrhage area based on brain imaging provided in an embodiment of the present application is described. Figure 1: This is a flow chart of a method for determining a bleeding area based on brain images provided in an embodiment of the present application. The method for determining a bleeding area based on brain images provided in this embodiment can be executed by a bleeding area determination device based on brain images to control the process of automatically determining the bleeding area from the brain image. The bleeding area determination device based on brain images can be composed of hardware and / or software, and can be integrated into a bleeding area determination device based on brain images. Among them, the method for determining a bleeding area based on brain images includes the following steps:

[0051] S101, determining the brain tissue region of each image in a brain CT image sequence.

[0052] Among them, CT, the full English name is: Computed Tomography, and the Chinese name is: electronic computer tomography. In this embodiment, the brain CT image is a plain scan brain CT image. The number of images in the brain CT image sequence is determined based on the slice accuracy of the CT machine. The higher the slice accuracy of the CT machine, the more images there are in the corresponding brain CT image sequence. For example, the number of images in the corresponding brain CT image sequence acquired by an ordinary CT machine is 30; the number of images in the corresponding brain CT image sequence acquired by a high-precision CT machine can be 50 or 100. This embodiment is explained by taking the number of images as 30 as an example.

[0053] Exemplarily, a brain CT image sequence of a patient is acquired by a CT machine, and a brain tissue region extraction operation is performed on each image in the acquired brain CT image sequence to determine the brain tissue region of each brain CT image.

[0054] Generally, in brain CT images, the CT value of skull pixels is 150-1000HU, while the CT value of brain tissue pixels is much lower than 120HU. Therefore, when extracting the brain tissue region of each brain CT image in the brain CT image sequence, the skull pixels in each brain CT image can be removed based on the threshold method combined with three-dimensional morphology to obtain brain tissue pixels, thereby obtaining the brain tissue region based on the brain tissue pixels. The specific implementation process includes the following steps:

[0055] S11, determining whether the pixel CT value of each image in the brain CT image sequence is less than or equal to a screening threshold.

[0056] S12: If the CT value of any pixel point is less than or equal to the screening threshold, the pixel point is determined to be a candidate brain tissue pixel point, and three-dimensional morphological processing is performed on the candidate brain tissue pixel point to obtain a target brain tissue pixel point.

[0057] S13, determining the brain tissue region of each image in the brain CT image sequence according to the target brain tissue pixel point.

[0058] The screening threshold is used to extract the value of the brain tissue pixel in each brain CT image. In this embodiment, the screening threshold is preferably the minimum CT value of the skull pixel, 150HU.

[0059] Exemplarily, the CT value of the pixel of each brain CT image can be compared with the screening threshold value to determine which pixel points in each brain CT image have a CT value less than or equal to the screening threshold value. The pixel points in each brain CT image whose CT value is less than or equal to the screening threshold value are determined as candidate brain tissue pixel points. Then, three-dimensional morphological processing, such as corrosion, expansion, and filling, is performed on the candidate brain tissue pixel points of each brain CT image to obtain target brain tissue pixel points, and the area formed by the target brain tissue pixel points is determined as the brain tissue area of ​​each image.

[0060] Optionally, when the screening threshold is 150HU, the brain CT image sequence has 30 images, such as Figure 2a As shown, the brain tissue area of ​​each brain CT image can be determined based on the screening threshold 150HU and three-dimensional morphology, as shown in Figure 2b shown.

[0061] Since the method of determining the brain tissue area of ​​each brain CT image is similar, the following is a specific description of determining the brain tissue area of ​​the 6th brain CT image among 30 brain CT images. Specifically, the CT value of each pixel of the 6th brain CT image is compared with the screening threshold 150. If the CT values ​​of the 12th pixel to the 35th pixel are all less than the screening threshold 150, then the 12th pixel to the 35th pixel are determined to be candidate brain tissue pixels. Then, the 23 candidate brain tissue pixels are respectively subjected to three-dimensional morphological processing to perform corrosion and expansion on the candidate brain tissue pixels, and the target brain tissue pixels are determined to be the 12th pixel to the 30th pixel as the target brain tissue pixels. Furthermore, the area formed by the above 18 target pixels is used as the brain tissue area of ​​the 6th brain CT image.

[0062] S102: According to each of the brain tissue regions, determine the ventricle edge corresponding to each of the brain tissue regions.

[0063] After each brain CT image is subjected to bone removal processing to obtain the brain tissue region, only the brain tissue region and the extracerebral space region remain in each brain CT image, and the extracerebral space region accounts for a smaller proportion than the brain tissue region in the human brain. Therefore, in order to determine the ventricle edge corresponding to each brain tissue region, the present embodiment can first determine the extracerebral space region based on the brain tissue region. Then, the lateral ventricle region is screened out from the extracerebral space region, and the ventricle edge of the lateral ventricle region is extracted to obtain the ventricle edge corresponding to each brain tissue region. Among them, the lateral ventricle region is located in the left and right cerebral hemispheres of the brain, respectively, and extends to each lobe of the hemisphere. Among them, the extracerebral space is also known as the low-density area of ​​the brain.

[0064] In specific implementation, the grayscale values ​​of all brain tissue regions can be counted, and a grayscale histogram of the brain CT image sequence can be constructed based on the grayscale value statistics. Then, the segmentation threshold of the ventricle region can be determined based on the grayscale histogram, and the extracerebral space region can be screened out from each brain tissue region based on the segmentation threshold.

[0065] Further, since the extracerebral space region includes the sulci and lateral ventricles, and determining whether the brain hemorrhage region invades the ventricles specifically refers to whether the hemorrhage region breaks into the lateral ventricles, then the sulci in the extracerebral space region need to be removed to obtain the lateral ventricles. Therefore, this embodiment can determine the sulci region in each extracerebral space, and then remove the sulci region to obtain the lateral ventricle region of each extracerebral space. Among them, the sulci and lateral ventricles constitute the circulation pathway of cerebrospinal fluid, which is the way of brain cell function metabolism.

[0066] After obtaining the lateral ventricle region of each brain CT image, three-dimensional morphology can be used to extract the edge of each lateral ventricle region image to obtain the ventricle edge of each lateral ventricle region. Among them, using three-dimensional morphology to extract the edge of each lateral ventricle region image is a conventional technical means, which will not be described in detail here.

[0067] S103, performing Hough transform straight line detection on each ventricle edge to obtain a target line segment of each ventricle edge.

[0068] Exemplarily, the Hough transform line detection algorithm can be used to detect all the straight lines in each ventricle edge, and return the angle and direction of the straight line in a normal form. Then, according to the screening threshold, the target line segment of each ventricle edge is screened out from all the straight lines of each ventricle edge. The screening threshold is set according to the actual application.

[0069] In this embodiment, when all the straight lines at the edge of each ventricle are detected using the Hough transform line detection algorithm, all the straight lines at the edge of each ventricle are detected by using the Hough transform hough_lines_dir operator of the local gradient direction.

[0070] It should be noted that collinear points in the image space correspond to a point in the parameter space, and a point in the parameter space represents the set of all collinear points in the image space. The parameter space refers to the Hough space.

[0071] S104, determining a target area according to each target line segment of the ventricle edge, performing parabola fitting on the pixel points of the target area to obtain a fitting result, and determining whether there is a hemorrhage area in the target area according to the fitting result.

[0072] The number of the target area is at least one. In this embodiment, the parabola fitting is preferably a least squares parabola fitting.

[0073] Exemplarily, the target area may be first determined based on the target line segment of each ventricle edge, and then a parabola fitting may be performed on each pixel point in the target area to obtain a fitting result, and then the hemorrhage area in the target area may be determined according to the fitting result.

[0074] When determining the target area, the target line segment of each ventricle edge can be used as the length of the target area, and then the width of the target area is calculated according to the length of the target area according to the preset ratio, and then the target area is determined according to the length and width of the target area. In this embodiment, the preset ratio can be adaptively set according to the needs of disease diagnosis, for example, the preset ratio can be set to 3:1 or 5:1. That is, the width of the target area is one third of the length of the target area, or the width of the target area is one fifth of the length of the target area.

[0075] For example, if the preset ratio is 3:1, then when the length of any target line segment at the edge of the ventricle in the fifth brain CT image is

[0076] 1*6=2

[0077] When the length is 6 mm, the width of the target area determined based on the target line segment is 3 mm. Thus, it can be determined that the length of the target area is 6 mm and the width is 2 mm, that is, 6 mm*2 mm.

[0078] After the target area is determined based on the target line segment of each ventricle edge, a parabola fitting is performed on each pixel point in the target area according to the least squares parabola fitting algorithm to obtain a fitting result. The specific process is: by traversing the grayscale values ​​of all pixels in each target area, the row where the maximum grayscale value is located and the row where the minimum grayscale value is located are determined. Then, the two curves drawn by the pixel points in the row where the maximum grayscale value is located and the row where the minimum grayscale value is located are clustered as initial clustering curves, and the clustering results are curve fitted based on the least squares parabola fitting algorithm to obtain a fitting result. The fitting result includes two fitting curves, namely the first fitting curve and the second fitting curve.

[0079] Further, after obtaining the fitting result, it can be determined whether there is a bleeding area in the target area according to the first fitting parameter in the first fitting curve and / or the second fitting parameter in the second fitting curve in the fitting result, wherein the first fitting parameter is used to determine the opening direction of the first fitting curve, and the second fitting parameter is used to determine the opening direction of the second fitting curve.

[0080] The specific implementation process is to compare the first fitting parameter and the second fitting parameter with the threshold value, respectively, to determine whether the first fitting parameter is less than the threshold value, and whether the second fitting parameter is less than the threshold value. If the first fitting parameter and / or the second fitting parameter is less than the threshold value, it is determined that there is a bleeding area in the target area. If one of the first fitting parameter and the second fitting parameter is greater than or equal to the threshold value, it is determined that there is no bleeding area in the target area. In this embodiment, the threshold value is zero.

[0081] That is, when the first fitting parameter is less than zero, and / or the second fitting parameter is less than zero, it indicates that there is a bleeding area in the target area, otherwise there is no bleeding area in the target area.

[0082] The method for determining the hemorrhage area based on brain images provided in the embodiment of the present application determines the brain tissue area of ​​each image in the brain CT image sequence, and determines the ventricle edge corresponding to each brain tissue area, so as to perform Hough transform straight line detection on each ventricle edge, obtain the target line segment of each ventricle edge, and then determine the target area according to the target line segment of each ventricle edge, perform parabola fitting on the pixel points of the target area, and determine whether there is a hemorrhage area in the target area according to the fitting result. Thus, the problem of large errors in the diagnosis results caused by the influence of subjective factors on human eye observation and the low recognition rate when the hemorrhage area is not obvious is solved, and the hemorrhage area is automatically determined based on the brain CT image, so that no human participation is required, and not only the recognition speed is fast, but also the accuracy rate is high.

[0083] From the above description, it can be seen that the embodiment of the present application obtains the target line segment by performing Hough transform straight line detection on each acquired ventricular edge, and performs parabola fitting on the pixel points of the target area determined by the target line segment to determine whether there is a hemorrhage area in the target area according to the fitting result.

[0084] Based on the above embodiment, this embodiment further optimizes the determination of the ventricle edge corresponding to each brain tissue region according to each brain tissue region. Figure 3 , the above-mentioned optimization process of the method for determining the hemorrhage area based on brain imaging provided in an embodiment of the present application is specifically described.

[0085] like Figure 3As shown, the method for determining the hemorrhage area based on brain images provided in the embodiment of the present application may include the following steps:

[0086] S201, determining the brain tissue region of each image in the brain CT image sequence.

[0087] S202: Counting the grayscale values ​​of all brain tissue regions to construct a grayscale histogram of the brain CT image sequence.

[0088] Exemplarily, the grayscale histogram of the brain CT image sequence may be constructed in the following manner:

[0089] Method 1

[0090] The grayscale value of each image in the brain CT image sequence is counted, and the grayscale histogram of each image is constructed according to the statistical results. Then, all grayscale histograms are superimposed to obtain the grayscale histogram of the brain CT image sequence.

[0091] Method 2

[0092] The grayscale values ​​of all images in the brain CT image sequence are counted, and a grayscale histogram of the brain CT image sequence is constructed according to the statistical results.

[0093] The following example uses the second method to construct a grayscale histogram. Specifically, assuming that the number of brain tissue area images corresponding to the brain CT image is 30, Figure 2b As shown, by counting the gray values ​​of all images in the brain CT image sequence and constructing a gray histogram of the brain CT image sequence based on the statistical results, it can be shown as Figure 4a As shown. Among them, Figure 4a The horizontal axis represents the gray value, and the vertical axis represents the number of pixels.

[0094] S203, extracting the grayscale value corresponding to the second peak in the grayscale histogram, and using the grayscale value as a segmentation threshold.

[0095] Because the human brain is composed of brain tissue area and extracerebral space area, that is, in addition to the brain tissue area, the human brain is also the extracerebral space area. In other words, there are only brain tissue area and extracerebral space area in the brain CT image, and in the human brain, the extracerebral space area accounts for a smaller proportion than the brain tissue area.

[0096] Based on the above reasons, in this embodiment, the gray value corresponding to the second peak in the gray histogram can be used as the segmentation threshold for determining the extra-brain gap area. Then, based on the segmentation threshold, the extra-brain gap area is screened out from each brain tissue area. In other words, Figure 4a The grayscale value corresponding to the first peak is used to determine the brain tissue area, while the grayscale value corresponding to the second peak is used to determine the extra-brain space area.

[0097] Continue with Figure 4a Take this as an example to illustrate, based on Figure 4a It can be seen that the second peak is point P1 (x1, y1), so the horizontal coordinate x1 of point P1 is used as the segmentation threshold for separating the extra-brain gap area.

[0098] S204: If it is determined that the grayscale value of any pixel point in each of the brain tissue regions is less than the segmentation threshold, the pixel point is determined as an extra-brain gap pixel point, and the extra-brain gap region corresponding to each of the brain tissue regions is determined based on the extra-brain gap pixel point.

[0099] Exemplarily, according to the segmentation threshold, the extra-brain gap region is screened out from each brain tissue region, specifically: the gray value of each pixel point of each brain tissue region is compared with the segmentation threshold, so as to determine the pixel points with gray value less than the segmentation threshold as the extra-brain gap pixel points. Then, the region formed by the extra-brain gap pixel points is used as the extra-brain gap region to obtain the extra-brain gap region corresponding to each brain tissue region.

[0100] For example, Figure 2a The 30 brain tissue region images shown in the figure can be used to compare the grayscale value of the pixels in each brain tissue region with the segmentation threshold to extract all the pixels in each brain tissue region whose grayscale value is less than the segmentation threshold. Then, all the extracted pixels are used as extracerebral gap pixels, and the region formed by the extracerebral gap pixels is used as the extracerebral gap region, thereby obtaining 30 extracerebral gap regions corresponding to each brain tissue region. For specific results, see Figure 4b For ease of understanding, the extracerebral space areas in some brain CT images can be marked, for example Figure 4b The area marked as 41 in FIG. It should be noted that due to the accuracy of image display, some extracerebral space areas may not be clearly displayed in the figure.

[0101] S205, removing the gyrus region in each of the extracerebral space regions to obtain the lateral ventricle region, and extracting the ventricle edge of each of the lateral ventricle regions to obtain the ventricle edge corresponding to each of the brain tissues.

[0102] Because the extracerebral space region includes the cerebral sulci and lateral ventricles, determining whether the brain hemorrhage region invades the ventricles specifically refers to whether the brain hemorrhage region enters the lateral ventricles. For this reason, the present embodiment can process the extracerebral space region to remove the cerebral sulci in the extracerebral space region to obtain the lateral ventricle region of each extracerebral space region. Furthermore, an edge extraction operation is performed on each lateral ventricle region to obtain the ventricle edge of each lateral ventricle region.

[0103] In specific implementation, because the lateral ventricle is located in the central area of ​​the brain, and the sulci are located in the edge area of ​​the brain, the sulci in each extracerebral space area can be removed by the following method, which specifically includes the following steps:

[0104] S21, determining the central pixel point of each extra-brain gap region and the edge pixel points of the brain tissue region corresponding to each extra-brain gap region.

[0105] For example, to determine the central pixel point of each extra-brain gap region, a central point detector may be used to identify the central pixel point of each extra-brain gap region. Of course, other methods may also be used to determine the central pixel point, which is not limited here.

[0106] To determine the edge pixel points of the brain tissue region corresponding to each extra-brain gap region, an edge detector may be used to identify all edge pixel points on the edge of the brain tissue region corresponding to each extra-brain gap region; or, a three-dimensional 1-scale erosion operation may be performed on the brain tissue region corresponding to each extra-brain gap region to obtain an erosion result, and then the erosion result is removed in each brain tissue region to obtain all edge pixel points on the edge of each brain tissue region. Wherein, 1 scale refers to one pixel point.

[0107] For example, if any brain tissue region is vBrainmask, then the erosion operation of scale 1 is performed on the brain tissue region vBrainmask to obtain the erosion result: vBrainmask_out. Then, the brain tissue region vBrainmask minus the erosion result vBrainmask_out can obtain all the edge pixels of the brain tissue region.

[0108] S22, determining a plurality of line segments of each of the extra-brain gap regions according to a central pixel point of each of the extra-brain gap regions and all edge pixel points of the brain tissue region corresponding to each of the extra-brain gap regions.

[0109] For example, a line segment can be established based on the central pixel point of each extra-brain space region and each edge pixel point of the brain tissue region corresponding to each extra-brain space region, thereby determining multiple line segments of the brain CT image corresponding to each extra-brain space region, wherein the two endpoints of each line segment are the central pixel point and the edge pixel point, respectively.

[0110] Since the implementation principle of determining multiple line segments of each extra-brain gap region is the same, any one of the 30 extra-brain gap regions is used as an example for explanation. Figure 4bTake the 20th extracerebral space region image in the figure as an example to illustrate. Assuming that the central pixel point of the 20th extracerebral space region is P11, and the edge pixel points of the 20th extracerebral space region corresponding to the brain tissue region are: P12, P13, P14, P15, P16, P17, P18, P19 and P20, then a line segment can be established based on P11 and P12 in the 20th extracerebral space region image. Create a line segment based on P11 and P13 P11 and P14 create line segments P11 and P15 create line segments Create line segments based on P11 and P16 Create line segments based on P11 and P17 Create line segments based on P11 and P18 Create line segments based on P11 and P19 And establish line segments based on P11 and P20 like Figure 4c shown.

[0111] S23, calculating the gyrus sulcus region of each of the extracerebral gap regions according to the multiple line segments and calculation parameters of each of the extracerebral gap regions, and removing the gyrus sulcus region to obtain the lateral ventricle region.

[0112] Among them, the calculation parameter is a parameter used to participate in the calculation of the sulcus area, and the parameter can be an empirical value. Specifically, the size of the sulcus area varies in different age groups. For example, the sulcus area is smaller for older people, and vice versa. Therefore, the calculation parameters in this embodiment can be adaptively set according to the age of the patient, and no specific restrictions are made here.

[0113] Continue with the above Figure 4c Take this as an example to illustrate, assuming that the calculation parameter is α, and the number of line segments in the extracerebral gap area of ​​the 20th image is 9, which are: and Then we can start from the edge pixel P12 of the brain tissue area in the 20th extracerebral space area and follow the line segment The sliding length is Arrive at P12' and get the line segment Starting from P13 along the line segment The sliding length is Arrive at P13' and get the line segment Starting from P14 along the line segment The sliding length is Arrive at P14' and get the line segment Starting from P15 along the line segment The sliding length is Arrive at P15' and get the line segment Starting from P16 along the line segment The sliding length is Arrive at P16' and get the line segment Starting from P17 along the line The sliding length is Arrive at P17' and get the line segment Starting from P18 along the line The sliding length is Arrive at P18' and get the line segment Starting from P19 along the line The sliding length is Arrive at P19' and get the line segment And starting from P20 along the line segment The sliding length is Arrive at P20' and get the line segment Then, based on the line segment and The determined closed loop closed area is used as the cerebral sulcus region (not shown in the figure). Then, the cerebral sulcus region is eliminated to obtain the lateral ventricle region of the 20th extracerebral space region image, as shown in 4d.

[0114] In this embodiment, the above method can also be used to Figure 4b The images of other extracerebral space regions are processed by removing the sulci and gyri, so as to obtain the lateral ventricle region of each extracerebral space region. Figure 4e For ease of understanding, the excluded sulci and gyri in some brain CT images can be marked, for example Figure 4e It should be noted that due to display accuracy reasons, some brain sulci and gyri that are removed from the brain CT images may not be clearly displayed in the figure. Figure 4b The sulci and gyrus elimination process of other extracerebral space area images is the same as the process of the 20th extracerebral space area, which will not be described in detail here.

[0115] Since each person's brain structure is different, when the sulci and gyri in each extracerebral space area are removed based on the above method, part of the lateral ventricle area may be removed, or part of the sulci and gyri may remain and not be completely removed.

[0116] In order to avoid the above-mentioned situation of removing part of the lateral ventricle area, or not removing all the sulcus and gyrus areas, this embodiment can also remove the sulcus and gyrus areas in each extracerebral gap area, and then extract the connected domain of each processed extracerebral gap area to obtain the maximum connected domain, and then remove the remaining gyrus and gyrus areas according to the maximum connected domain, and at the same time retrieve part of the lateral ventricle area that was removed, so as to obtain the complete lateral ventricle area corresponding to each extracerebral gap area. The connected domain extraction in this embodiment is specifically a three-dimensional maximum connected domain extraction. The three-dimensional maximum connected domain extraction operation is a conventional technical means, which will not be described in detail here.

[0117] For example, Figure 4e The above processing can be performed on the 30 processed extracerebral space areas to obtain the lateral ventricle area corresponding to each extracerebral space area. Figure 4f For ease of understanding, the lateral ventricle area in some brain CT images can be marked, for example Figure 4f The area marked as 43 in the figure. It should be noted that due to display accuracy, some extracerebral space areas may not be clearly displayed in the figure.

[0118] Further, after obtaining the lateral ventricle region corresponding to each extracerebral space region, the present embodiment can use three-dimensional morphology to perform edge extraction on each lateral ventricle region to obtain the ventricle edge corresponding to each brain tissue. Among them, using three-dimensional morphology to perform edge extraction on each lateral ventricle region image is a conventional technical means, which will not be described in detail here.

[0119] Continue with the above Figure 4f Take 30 lateral ventricle regions as an example, and extract the edges of them to obtain the ventricle edge results of 30 lateral ventricle regions. Figure 4g shown.

[0120] S206, performing Hough transform straight line detection on each ventricle edge to obtain a target line segment of each ventricle edge.

[0121] S207, determining a target area according to each target line segment of the ventricle edge, performing parabola fitting on the pixel points of the target area to obtain a fitting result, and determining whether there is a hemorrhage area in the target area according to the fitting result.

[0122] The method for determining the hemorrhage area based on brain images provided in the embodiment of the present application determines the brain tissue area of ​​each image in the brain CT image sequence, and determines the ventricle edge corresponding to each brain tissue area, so as to perform Hough transform straight line detection on each ventricle edge, obtain the target line segment of each ventricle edge, and then determine the target area according to the target line segment of each ventricle edge, perform parabola fitting on the pixel points of the target area, and determine whether there is a hemorrhage area in the target area according to the fitting result. Thus, the problem of large errors in the diagnosis results caused by the influence of subjective factors on human eye observation and the low recognition rate when the hemorrhage area is not obvious is solved, and the hemorrhage area is automatically determined based on the brain CT image, so that no human participation is required, and not only the recognition speed is fast, but also the accuracy rate is high.

[0123] On the basis of the above embodiment, in this embodiment, each ventricle edge is subjected to Hough transform straight line detection to obtain a target line segment of each ventricle edge for further optimization. Figure 5 , the above-mentioned optimization process of the method for determining the hemorrhage area based on brain imaging provided in an embodiment of the present application is specifically described.

[0124] like Figure 5 As shown, the method for determining the hemorrhage area based on brain images provided in the embodiment of the present application may include the following steps:

[0125] S301, determining the brain tissue region of each image in the brain CT image sequence.

[0126] S302: According to each of the brain tissue regions, determine the ventricle edge corresponding to each of the brain tissue regions.

[0127] S303, traversing all pixel points at the edge of each ventricle to convert each pixel point into a Hough space, and obtaining a Hough transform result of each pixel point.

[0128] Among them, Hough transform refers to taking the lower left corner of the image as the origin (0,0), for any straight line l in the image i , expressed as (γ,θ). Where γ is the straight line l i The vertical distance to the origin, θ is the straight line l i The angle between the x-axis and the x-axis, i is a positive integer. The method of using (γ, θ) to represent a straight line is called Hough transform.

[0129] In order to facilitate the understanding of the above Hough transform principle, the following Figure 6a Take this as an example to illustrate. Assume that the image is the 16th ventricle edge image. Then the lower left corner of the ventricle edge image is the origin (0,0). For any straight line l in the ventricle edge image i , expressed as (γ,θ). Where γ is the straight line l iThe vertical distance to the origin, θ is the straight line l i The angle with the X axis.

[0130] Since a point can emit countless straight lines, the straight line expression of the point obtained according to the Hough transformation can be as follows Figure 6b As shown. Figure 6b The midpoints include: Pm and Pn, then the straight line through point Pm is (r m ,θ m ) is represented by the straight line passing through point Pn (r n ,θ n ) is represented by two points. Since two points determine a straight line, the straight line passing through the two points Pm and Pn must have r m =r n and θ m =θ n In this embodiment, m and n are positive integers.

[0131] Exemplarily, this embodiment performs a traversal operation on all pixel points at the edge of each ventricle to convert each pixel point at the edge of each ventricle into the Hough space, i.e., the (r, θ) space, to obtain the Hough transform result. Since the collinear points on the image plane have the same (r, θ), then by counting the points with the same r and θ in the Hough space, the statistical result (i.e., the Hough transform result) is: (r 1 ,θ 1 )、(r 2 ,θ 2 )、(r 3 ,θ 3 )……(r R ,θ R ). Based on the statistical results, it can be seen that there are R straight lines at the edge of each ventricle, where R is an integer.

[0132] In this embodiment, each pixel point at the edge of each ventricle is converted to the Hough space and the obtained Hough result is as follows: Figure 6c As shown. Among them, θ∈(0,180°), γ∈(r min ,r max ), γ min Represents the minimum distance from the origin to the left border of the brain tissue area, γ max Represents the farthest distance from the origin to the right border of the brain tissue area. For details, the corresponding relationship between the above parameter areas and their physical meanings can be found in Figure 6d shown.

[0133] S304: Filter out candidate line segments from the Hough transform result according to a first screening threshold.

[0134] Since the physiological structure of the ventricle contour is a smooth curved surface, and red blood cells are heavier than cerebrospinal fluid, that is, the horizontal line of the junction of cerebrospinal fluid and hemorrhage is the longest line on the edge of the ventricle. Therefore, the first screening threshold is used to filter out the value of a certain length of line segment from the Hough transform result corresponding to each ventricle edge. In this embodiment, the first screening threshold is specifically the number of pixels, and the number of pixels can be selected as 100. Of course, the number of pixels can be adaptively set according to the actual application needs, and is not limited here.

[0135] For example, assuming that the first screening threshold is 100 pixels, line segments with a length greater than or equal to 100 pixels are screened out from the Hough transform results corresponding to each ventricle edge, and the screened line segments are determined as candidate line segments.

[0136] S305: Filter out a target line segment of each ventricle edge from the candidate line segments according to a second screening threshold.

[0137] The second screening threshold is used to screen the target line segment from the candidate line segments. The length of time the patient lies flat and the amount of bleeding have an impact on the horizontal line of the boundary between cerebrospinal fluid and bleeding. Therefore, in order to reduce the above impact, the second screening threshold can be set to θ∈(-6°,6°), γ∈(γ min ,γ max ). In addition, in order to meet the detection accuracy of different bleeding areas entering the ventricle, the second screening threshold of this embodiment also includes a length threshold, which can be set based on the structure of the human brain. For example, the length threshold is 10 millimeters (mm), etc.

[0138] by Figure 6a The 16th ventricle edge image shown in the figure is used as an example to illustrate that if the second screening threshold is set to θ∈(-6°,6°), γ∈(γ min ,γ max ), and the length threshold is 10, then the candidate line segments in the 16th ventricle edge image are screened twice to obtain the target line segment of the 16th ventricle edge, as shown in the following figure: Figure 6e 61, 62, 63 and 64 shown.

[0139] S306, determining a target area according to each target line segment of the ventricle edge, performing parabola fitting on the pixel points of the target area to obtain a fitting result, and determining whether there is a hemorrhage area in the target area according to the fitting result.

[0140] The method for determining the hemorrhage area based on brain images provided in the embodiment of the present application determines the brain tissue area of ​​each image in the brain CT image sequence, and determines the ventricle edge corresponding to each brain tissue area, so as to perform Hough transform straight line detection on each ventricle edge, obtain the target line segment of each ventricle edge, and then determine the target area according to the target line segment of each ventricle edge, perform parabola fitting on the pixel points of the target area, and determine whether there is a hemorrhage area in the target area according to the fitting result. Thus, the problem of large errors in the diagnosis results caused by the influence of subjective factors on human eye observation and the low recognition rate when the hemorrhage area is not obvious is solved, and the hemorrhage area is automatically determined based on the brain CT image, so that no human participation is required, and not only the recognition speed is fast, but also the accuracy rate is high.

[0141] Based on the above embodiment, this embodiment further optimizes the target area determined according to the target line segment of each ventricle edge in this embodiment, and the parabola fitting of the pixel points in the target area to obtain the fitting result. Figure 7 , the above-mentioned optimization process of the method for determining the hemorrhage area based on brain imaging provided in an embodiment of the present application is specifically described.

[0142] like Figure 7 As shown, the method for determining the hemorrhage area based on brain images provided in the embodiment of the present application may include the following steps:

[0143] S401, determining the brain tissue region of each image in the brain CT image sequence.

[0144] S402: According to each of the brain tissue regions, determine the ventricle edge corresponding to each of the brain tissue regions.

[0145] S403, performing Hough transform straight line detection on each ventricle edge to obtain a target line segment of each ventricle edge.

[0146] S404: taking each target line segment of the ventricle edge as the length of the target area, and determining the width of the target area based on the length of the target line segment according to a preset ratio.

[0147] S405: Determine the target area according to the length and width of the target area.

[0148] The preset ratio can be set according to actual application requirements and is not limited here. For example, it can be set to 3:1 or 5:1.

[0149] For example, if the preset ratio is 3:1, then when the length of any target line segment at the edge of the sixth ventricle is 9 mm, the width of the target area determined based on the target line segment is Millimeters. Therefore, it can be determined that the length of the target area is 9mm and the width is 3mm, that is, 9mm*3mm.

[0150] For example, Figure 8a As shown, based on the four target line segments in the 16th ventricular edge image, four target areas are determined, namely area 1, area 2, area 3 and area 4.

[0151] S406, determining a row where the maximum grayscale value and a row where the minimum grayscale value of the pixels in the target area are located.

[0152] S407 , obtaining the row curve where the maximum gray value is located and the row curve where the minimum gray value is located as two types of initial clustering curves, and clustering the two types of initial clustering curves to obtain a clustering result.

[0153] The clustering is specifically binary clustering of Kmeans clustering.

[0154] Specifically, the maximum grayscale value and the minimum grayscale value of the pixels in each target area in each ventricle edge image are traversed to determine the maximum grayscale value and the minimum grayscale value. In addition, the row where the maximum grayscale value is located and the row where the minimum grayscale value is located are determined. Then, the curve drawn by the pixel points in the row where the maximum grayscale value is located and the row where the minimum grayscale value is located are used as the initial clustering curve, and the two initial clustering curves corresponding to each target area are subjected to binary clustering to obtain the clustering result.

[0155] Because the processing process for each target area is the same, target area 4 in the 16th ventricle edge image is used as an example for explanation. Specifically, if the line with the maximum grayscale value of the pixel points in target area 4 is: lineMax, and the line with the minimum grayscale value of the pixel points is: lineMin, then the curve drawn by the pixel points of lineMax is used as the first type of initial clustering curve, and the curve drawn by the pixel points of lineMin is used as the second type of initial clustering curve, such as Figure 8b As shown. Then, the above two types of initial clustering curves are subjected to binary clustering to obtain clustering results. For details, see Figure 8c .

[0156] S408, performing parabola fitting on the two types of clustering results respectively to obtain fitting results, and determining whether there is a bleeding area in the target area according to the fitting results.

[0157] In this embodiment, the parabola fitting is a least squares parabola fitting.

[0158] Specifically, this embodiment uses the least squares parabola fitting algorithm to perform parabola fitting on the two types of clustering results to obtain fitting results. The fitting results include a first fitting curve for the row with the maximum gray value and a second fitting curve for the row with the minimum gray value, and the first fitting curve and the second fitting curve can be respectively obtained by the parabola equation y=a*x 2 +b*x+c. For example, the known parabola equation is: y=a*x 2 +b*x+c, then the first fitting curve equation of the kth target area is: k1 =a k1 *x k1 2 +b k1 *x k1 +c k1 , the corresponding second fitting curve equation of the kth target area is: k2 =a k2 *x k2 2 +b k2 *x k2 +c k2 . Wherein k is a positive integer greater than or equal to 1.

[0159] Due to the physiological characteristics that red blood cells are heavier than cerebrospinal fluid, when there is blood in the ventricles, the blood will settle in the occipital corner of the ventricles, that is, the center of the target area, and the grayscale distribution of the fitting curve of the corresponding target area will present a parabola-like feature of opening downward with a high middle and low sides. Therefore, in this embodiment, the fitting parameter a that determines the opening direction of the fitting curve (i.e., parabola) can be obtained from the first fitting curve equation and the second fitting curve equation respectively. Among them, the positive and negative values ​​of the fitting parameter a determine the opening direction of the fitting curve (i.e., parabola), a>0 indicates that the opening is upward, and a<0 indicates that the opening is downward. Then, it is determined whether the fitting parameter a in the first fitting curve equation and the fitting parameter a in the second fitting curve equation are both less than the threshold value 0. If the fitting parameter a in the first fitting curve equation and the fitting parameter a in the second fitting curve equation are both greater than 0, it means that there is no bleeding area in the target area. If the fitting parameter a in the first fitting curve equation is less than 0, and / or the fitting parameter a in the second fitting curve equation is less than 0, it means that there is a bleeding area in the target area. In addition, when it is determined that there is a hemorrhage area in the target area, the target line segment used to determine the target area is also determined as the boundary between the ventricle and the hemorrhage area. The threshold is used to determine the positive or negative value of the fitting parameter a.

[0160] Continuing with the above example, from the first fitting curve equation y k1 =a k1 *x k1 2 +b k1 *x k1 +ck1 Get the first fitting parameter a k1 , and from the parabola equation of the second fitting curve y k2 =a k2 *x k2 2 +b k2 *x k2 +c k2 Get the second fitting parameter a k2 Then, determine a k1 Is it less than the threshold 0, and judge a k2 Is it less than the threshold 0? k1 <0 or a k2 <0, it is determined that there is a bleeding area in the kth target area.

[0161] The method for determining the hemorrhage area based on brain images provided in the embodiment of the present application determines the brain tissue area of ​​each image in the brain CT image sequence, and determines the ventricle edge corresponding to each brain tissue area, so as to perform Hough transform straight line detection on each ventricle edge, obtain the target line segment of each ventricle edge, and then determine the target area according to the target line segment of each ventricle edge, perform parabola fitting on the pixel points of the target area, and determine whether there is a hemorrhage area in the target area according to the fitting result. Thus, the problem of large errors in the diagnosis results caused by the influence of subjective factors on human eye observation and the low recognition rate when the hemorrhage area is not obvious is solved, and the hemorrhage area is automatically determined based on the brain CT image, so that no human participation is required, and not only the recognition speed is fast, but also the accuracy rate is high.

[0162] In another implementation scenario of the present application, after the bleeding area is determined from the target area based on the fitting result, the bleeding volume of the bleeding area can also be determined. Fig. 9 , the above-mentioned situation of determining the bleeding volume of the bleeding area in the method for determining the bleeding area based on brain images provided in an embodiment of the present application is explained.

[0163] like Fig. 9 As shown, the method for determining the hemorrhage area based on brain images provided in the embodiment of the present application may include the following steps:

[0164] S501, determining the brain tissue region of each image in the brain CT image sequence.

[0165] S502: According to each of the brain tissue regions, determine the ventricle edge corresponding to each of the brain tissue regions.

[0166] S503, performing Hough transform straight line detection on each ventricle edge to obtain a target line segment of each ventricle edge.

[0167] S504, determining a target area according to each target line segment of the ventricle edge, performing parabola fitting on the pixel points of the target area to obtain a fitting result, and determining whether there is a hemorrhage area in the target area according to the fitting result.

[0168] S505: If it is determined that the target area has a bleeding area, segment the bleeding area from the target area.

[0169] Optionally, segmenting the bleeding area from the target area can be achieved by the following formula (1):

[0170] Bloodmask=CTValue>medianCTValue……………………(1)

[0171] Wherein, Bloodmask represents the bleeding area, CTValue represents the CT value of each pixel in the target area, and medianCTValue represents the blood volume threshold of the bleeding volume. In this embodiment, medianCTValue can be determined based on the median value of the CT values ​​of all pixels in the target area, or based on the average value of the CT values ​​of all pixels in the target area, etc. There is no specific limitation on this herein.

[0172] S506: Calculate the bleeding volume of the bleeding area according to the number of pixels in the bleeding area and the volume of each pixel.

[0173] Optionally, the bleeding volume of the bleeding area can be calculated by the following formula (2):

[0174] BloodVolum=Bloodmask*Voxel………………(2)

[0175] Wherein, BloodVolum represents the bleeding volume of the bleeding area, Bloodmask represents the bleeding area, and Voxel represents the volume of each pixel. In this embodiment, the volume of a pixel specifically refers to the spatial volume of the pixel.

[0176] The method for determining the bleeding area based on brain images provided in the embodiment of the present application determines the brain tissue area of ​​each image in the brain CT image sequence, and determines the ventricle edge corresponding to each brain tissue area, so as to perform Hough transform straight line detection on each ventricle edge, obtain the target line segment of each ventricle edge, and then determine the target area according to the target line segment of each ventricle edge, perform parabola fitting on the pixel points of the target area, and determine whether there is a bleeding area in the target area according to the fitting result. Thus, the problem of large errors in the diagnosis results caused by the influence of subjective factors on human eye observation and the low recognition rate when the bleeding area is not obvious is solved, and the bleeding area is automatically determined based on the brain CT image, so that no manual participation is required, and not only the recognition speed is fast, but also the accuracy is high. In addition, when it is determined that the brain bleeding area enters the ventricle, the amount of bleeding in the bleeding area can be quantitatively analyzed, so as to determine the severity of the brain bleeding condition, and provide a reference basis for doctors to diagnose and treat.

[0177] Please refer to the attached Fig.10 , a device for determining a bleeding area based on brain imaging proposed in an embodiment of the present application is described. Fig.10 It is a schematic block diagram of a device for determining a hemorrhage area based on brain images provided in an embodiment of the present application.

[0178] The brain image-based hemorrhage region determination device 600 includes: a first determination module 610 , a second determination module 620 , a detection module 630 and a third determination module 640 .

[0179] Wherein, the first determination module 610 is used to determine the brain tissue region of each image in the brain CT image sequence;

[0180] A second determination module 620 is used to determine, according to each of the brain tissue regions, a ventricle edge corresponding to each of the brain tissue regions;

[0181] A detection module 630, configured to perform Hough transform straight line detection on each ventricle edge to obtain a target line segment of each ventricle edge;

[0182] The third determination module 640 is used to determine a target area according to each target line segment of the ventricle edge, perform parabola fitting on the pixel points of the target area to obtain a fitting result, and determine whether there is a hemorrhage area in the target area according to the fitting result.

[0183] In an optional implementation of the embodiment of the present application, the fitting result includes a first fitting curve and a second fitting curve;

[0184] Accordingly, the third determining module 640 is specifically configured to:

[0185] If the first fitting parameter in the first fitting curve is less than a threshold value, and / or the second fitting parameter in the second fitting curve is less than the threshold value, it is determined that there is a bleeding area in the target area, otherwise there is no bleeding area;

[0186] The first fitting parameter is used to determine the opening direction of the first fitting curve, and the second fitting parameter is used to determine the opening direction of the second fitting curve.

[0187] An optional implementation of the embodiment of the present application further includes:

[0188] a fourth determining module, configured to segment the bleeding area from the target area if it is determined that the target area has a bleeding area;

[0189] The calculation module is used to calculate the bleeding volume of the bleeding area according to the number of pixels in the bleeding area and the volume of each pixel.

[0190] In an optional implementation of the embodiment of the present application, the second determining module 620 includes:

[0191] A construction unit, used for counting the grayscale values ​​of all brain tissue regions to construct a grayscale histogram of the brain CT image sequence;

[0192] A first extraction unit, used for extracting the grayscale value corresponding to the second peak in the grayscale histogram, and using the grayscale value as a segmentation threshold;

[0193] a determining unit, configured to determine the pixel as an extra-brain gap pixel if the grayscale value of any pixel of each of the brain tissue regions is less than the segmentation threshold, and determine the extra-brain gap region corresponding to each of the brain tissue regions according to the extra-brain gap pixel;

[0194] The second extraction unit is used to remove the gyrus and sulcus regions in each of the extracerebral space regions to obtain the lateral ventricle regions, and to extract the ventricle edges of each of the lateral ventricle regions to obtain the ventricle edges corresponding to each of the brain tissues.

[0195] In an optional implementation of the embodiment of the present application, the second extraction unit is specifically configured to:

[0196] Determine the central pixel point of each of the extra-brain gap regions and the edge pixel points of the brain tissue region corresponding to each of the extra-brain gap regions;

[0197] Determine a plurality of line segments of each of the extra-brain gap regions according to a central pixel point of each of the extra-brain gap regions and all edge pixel points of a brain tissue region corresponding to each of the extra-brain gap regions;

[0198] According to the multiple line segments and calculation parameters of each extracerebral space region, the gyrus sulcus region of each extracerebral space region is calculated, and the gyrus sulcus region is eliminated to obtain the lateral ventricle region.

[0199] In an optional implementation of the embodiment of the present application, the second determining module 620 further includes:

[0200] A third extraction unit is used to extract a connected domain from each of the processed extra-brain gap regions to obtain a maximum connected domain;

[0201] The elimination unit is used to eliminate the remaining gyrus and sulcus regions according to the maximum connected domain.

[0202] In an optional implementation of the embodiment of the present application, the detection module 630 is specifically configured to:

[0203] Traversing all pixel points at the edge of each ventricle to convert each pixel point into Hough space, and obtaining a Hough transform result of each pixel point;

[0204] Filtering candidate line segments from the Hough transform result according to a first screening threshold;

[0205] According to a second screening threshold, a target line segment of each ventricle edge is screened out from the candidate line segments.

[0206] In an optional implementation of the embodiment of the present application, the third determining module 640 is specifically configured to:

[0207] Taking each target line segment of the ventricle edge as the length of the target area, and determining the width of the target area based on the length of the target line segment according to a preset ratio;

[0208] The target area is determined according to the length and width of the target area.

[0209] In an optional implementation of the embodiment of the present application, the third determining module 640 is specifically configured to:

[0210] Determine the row where the maximum grayscale value of the pixel points in the target area is located and the row where the minimum grayscale value is located;

[0211] Obtaining the row curve where the maximum gray value is located and the row curve where the minimum gray value is located as two types of initial clustering curves, and clustering the two types of initial clustering curves to obtain a clustering result;

[0212] Parabolic fitting is performed on the two types of clustering results respectively to obtain fitting results.

[0213] It should be understood that the embodiment of the device for determining the hemorrhage region based on brain images and the embodiment of the method for determining the hemorrhage region based on brain images may correspond to each other, and similar descriptions may refer to the method embodiment. To avoid repetition, they will not be described here. Specifically, Fig.10 The brain image-based bleeding area determination device 600 can be implemented Figure 1 The corresponding method embodiment, and the aforementioned and other operations and / or functions of each module in the brain image-based hemorrhage area determination device 600 are respectively to achieve Figure 1 For the sake of brevity, the corresponding processes in each method are not repeated here.

[0214] In the above, the bleeding area determination device 600 based on brain images of the embodiment of the present application is described from the perspective of functional modules in conjunction with the accompanying drawings. It should be understood that the functional module can be implemented in hardware form, can be implemented by instructions in software form, and can also be implemented by a combination of hardware and software modules. Specifically, the steps of the bleeding area determination method based on brain images in the embodiment of the present application can be completed by hardware integrated logic circuits and / or software instructions in the processor, and the steps of the bleeding area determination method based on brain images disclosed in the embodiment of the present application can be directly embodied as a hardware decoding processor to perform, or a combination of hardware and software modules in the decoding processor to perform. Optionally, the software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory, and completes the steps in the above method embodiment in conjunction with its hardware.

[0215] Fig.11 It is a schematic block diagram of a brain imaging-based hemorrhage area determination device 700 provided in an embodiment of the present application.

[0216] like Fig.11 As shown, the brain image-based bleeding area determination device 700 may include:

[0217] The memory 710 and the processor 720, the memory 710 is used to store the computer program and transmit the program code to the processor 720. In other words, the processor 720 can call and run the computer program from the memory 710 to implement the method for determining the hemorrhage area based on brain imaging in the embodiment of the present application.

[0218] For example, the processor 720 may be configured to execute the above-mentioned embodiment of the method for determining a hemorrhage area based on brain images according to the instructions in the computer program.

[0219] In some embodiments of the present application, the processor 720 may include but is not limited to:

[0220] General-purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components, etc.

[0221] In some embodiments of the present application, the memory 710 includes but is not limited to:

[0222] Volatile memory and / or non-volatile memory. Among them, the non-volatile memory can be read-only memory (ROM), programmable ROM (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM) or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM bus random access memory (DR RAM).

[0223] In some embodiments of the present application, the computer program may be determined as one or more modules, which are stored in the memory 710 and executed by the processor 720 to complete the method provided by the present application. The one or more modules may be a series of computer program instruction segments capable of completing specific functions, and the instruction segments are used to describe the execution process of the computer program in the device for determining the hemorrhage area based on brain imaging.

[0224] like Fig.11As shown, the brain image-based bleeding area determination device 700 may also include:

[0225] The transceiver 730 may be connected to the processor 720 or the memory 710 .

[0226] The processor 720 may control the transceiver 730 to communicate with other devices, specifically, to send information or data to other devices, or to receive information or data sent by other devices. The transceiver 730 may include a transmitter and a receiver. The transceiver 730 may further include an antenna, and the number of antennas may be one or more.

[0227] It should be understood that the various components in the brain imaging-based bleeding area determination device 700 are connected via a bus system, wherein the bus system includes not only a data bus but also a power bus, a control bus and a status signal bus.

[0228] The present application also provides a computer storage medium on which a computer program is stored, and when the computer program is executed by a computer, the computer can execute the method for determining the hemorrhage area based on brain images of the above embodiment. In other words, the present application embodiment also provides a computer program product containing instructions, and when the instructions are executed by a computer, the computer can execute the method of the above method embodiment.

[0229] When software is used for implementation, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function according to the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (digital subscriber line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integration. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a digital video disc (digital video disc, DVD)), or a semiconductor medium (e.g., a solid state drive (solid state disk, SSD)), etc.

[0230] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0231] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the module is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0232] The modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. For example, each functional module in each embodiment of the present application may be integrated into a processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0233] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for determining a hemorrhage area based on brain images, characterized in that: include: Determine the brain tissue region for each image in a brain CT image sequence; According to each of the brain tissue regions, determining the ventricle edge corresponding to each of the brain tissue regions; Performing Hough transform straight line detection on each ventricle edge to obtain a target line segment of each ventricle edge; Determine a target area according to each target line segment of the ventricle edge, perform parabola fitting on the pixels in the target area to obtain a fitting result, and determine whether there is a hemorrhage area in the target area according to the fitting result, wherein performing parabola fitting on the pixels in the target area to obtain the fitting result includes: determining a row where the maximum grayscale value and the minimum grayscale value of the pixels in the target area are located; The row curve where the maximum gray value is located and the row curve where the minimum gray value is located are obtained as two types of initial clustering curves, and the two types of initial clustering curves are clustered to obtain clustering results; and parabola fitting is performed on the two types of clustering results respectively to obtain fitting results.

2. The method according to claim 1, characterized in that The fitting result includes a first fitting curve and a second fitting curve; Accordingly, determining whether there is a bleeding area in the target area according to the fitting result includes: If the first fitting parameter in the first fitting curve is less than a threshold value, and / or the second fitting parameter in the second fitting curve is less than the threshold value, it is determined that there is a bleeding area in the target area, otherwise there is no bleeding area; The first fitting parameter is used to determine the opening direction of the first fitting curve, and the second fitting parameter is used to determine the opening direction of the second fitting curve.

3. The method according to claim 1 or 2, characterized in that: The method further comprises: If it is determined that the target area has a bleeding area, segmenting the bleeding area from the target area; The bleeding volume of the bleeding area is calculated according to the number of pixels in the bleeding area and the volume of each pixel.

4. The method according to claim 1, characterized in that: According to each of the brain tissue regions, determining the ventricle edge corresponding to each of the brain tissue regions comprises: Counting the grayscale values ​​of all brain tissue regions to construct a grayscale histogram of the brain CT image sequence; Extracting the grayscale value corresponding to the second peak in the grayscale histogram, and using the grayscale value as a segmentation threshold; If it is determined that the grayscale value of any pixel point of each of the brain tissue regions is less than the segmentation threshold, the pixel point is determined as an extra-brain gap pixel point, and the extra-brain gap region corresponding to each of the brain tissue regions is determined based on the extra-brain gap pixel point; The gyrus and sulcus regions in each of the extracerebral space regions are eliminated to obtain the lateral ventricle regions, and the ventricle edges of each of the lateral ventricle regions are extracted to obtain the ventricle edges corresponding to each of the brain tissues.

5. The method according to claim 4, characterized in that The gyrus region in each of the extracerebral space regions is eliminated to obtain the lateral ventricle region, including: Determine the central pixel point of each of the extra-brain gap regions and the edge pixel points of the brain tissue region corresponding to each of the extra-brain gap regions; Determine a plurality of line segments of each of the extra-brain gap regions according to a central pixel point of each of the extra-brain gap regions and all edge pixel points of a brain tissue region corresponding to each of the extra-brain gap regions; According to the multiple line segments and calculation parameters of each extracerebral space region, the gyrus sulcus region of each extracerebral space region is calculated, and the gyrus sulcus region is eliminated to obtain the lateral ventricle region.

6. The method according to claim 5, characterized in that After the gyrus region is removed, the method further includes: Performing connected domain extraction on each of the processed extra-brain gap regions to obtain a maximum connected domain; According to the maximum connected domain, the remaining gyrus and sulcus regions are removed.

7. The method according to claim 1, characterized in that Performing Hough transform straight line detection on each of the ventricle edges to obtain a target line segment of each of the ventricle edges includes: Traversing all pixel points at the edge of each ventricle to convert each pixel point into Hough space, and obtaining a Hough transform result of each pixel point; Filtering candidate line segments from the Hough transform result according to a first screening threshold; According to a second screening threshold, a target line segment of each ventricle edge is screened out from the candidate line segments.

8. The method according to claim 1, characterized in that Determining a target area according to each target line segment of the ventricle edge includes: Taking each target line segment of the ventricle edge as the length of the target area, and determining the width of the target area based on the length of the target line segment according to a preset ratio; The target area is determined according to the length and width of the target area.

9. A device for determining a bleeding area based on brain images, characterized in that: include: A first determination module is used to determine the brain tissue area of ​​each image in the brain CT image sequence; A second determination module is used to determine the ventricle edge corresponding to each brain tissue region according to each brain tissue region; A detection module, used for performing Hough transform straight line detection on each ventricle edge to obtain a target line segment of each ventricle edge; A third determination module is used to determine a target area according to each target line segment of the ventricle edge, perform parabola fitting on the pixels in the target area to obtain a fitting result, and determine whether there is a hemorrhage area in the target area according to the fitting result, wherein the parabola fitting on the pixels in the target area to obtain the fitting result includes: determining a row where the maximum grayscale value and the minimum grayscale value of the pixels in the target area are located; The row curve where the maximum gray value is located and the row curve where the minimum gray value is located are obtained as two types of initial clustering curves, and the two types of initial clustering curves are clustered to obtain clustering results; and parabola fitting is performed on the two types of clustering results respectively to obtain fitting results.

10. A device for determining a bleeding area based on brain images, characterized in that: include: A processor and a memory, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to execute the method for determining a hemorrhage area based on brain imaging according to any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that: Used to store a computer program, wherein the computer program enables a computer to execute the method for determining a hemorrhage area based on brain images according to any one of claims 1 to 8.

Images