Side fissure extraction method, device, storage medium and equipment

Abstract

The present invention discloses a method for automatically locking the optimal sulcus segmentation threshold based on corrected images and identifying the sylvian fissure in combination with its anatomical location, belonging to the field of medical image processing technology. Based on threshold segmentation, this method extracts candidate regions based on the anatomical location of the sylvian fissure. The sylvian fissure is extracted within the corresponding anatomical region. For cases where lesions occupy the sylvian fissure, the abnormal side is fitted based on the sylvian fissure obtained from the normal side segmentation to obtain a sylvian fissure segmentation result for the entire brain. This method is applicable to CT plain scan (NCCT) and perfusion (CTP) images, and is suitable for batch, rapid, automated, and accurate identification of various brain shapes across multiple hospitals, multiple devices, and various brain shapes, and has important clinical significance.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical image processing, and particularly relates to a method for automatically locking an optimal sulcus segmentation threshold based on a corrected image and identifying the sylvian fissure in combination with the anatomical position of the sylvian fissure. Background Art

[0002] The Sylvian fissure is a groove located in the middle of the lateral surface of the cerebral hemisphere. It originates from the anterior perforated substance on the base of the brain and runs along the lateral side of the hemisphere. Its posterior portion marks the boundary between the frontal and temporal lobes, and it contains the insula. The Sylvian fissure channels the middle cerebral artery (MCA), the anterior perforating artery, the superficial middle cerebral vein, and the deep middle cerebral vein.

[0003] The transpterional lateral fissure approach is one of the most widely used surgical approaches in neurosurgery. Due to the presence of arterial and venous branches in the lateral fissure, surgery in this region is highly valued by neurosurgeons. Blood flow disturbances or bleeding caused by surgery can have serious consequences. While this approach may be relatively simple for experts, it is more challenging for inexperienced surgeons. This is particularly true in cerebral hemorrhage surgery, where hematoma compression can cause brain tissue displacement, occupying the space of the lateral fissure, making the fissure indistinguishable from imaging.

[0004] Clinical identification of the sylvian fissure is primarily based on physician experience, with some also relying on computer technology and manual annotation. However, due to the varying expertise of physicians, inexperienced physicians can easily misjudge the sylvian fissure and lead to cerebral hemorrhage. Furthermore, computer-assisted image labeling requires layer-by-layer labeling, which is time-consuming and laborious, and cannot be performed if the sylvian fissure is compressed. Summary of the Invention

[0005] To address the above problems, the present invention provides a method for automatically locking the optimal sulcus segmentation threshold on CT plain scan images based on corrected images, and identifying the lateral fissure in combination with the anatomical position of the lateral fissure. For the compressed part of the lateral fissure, the symmetry characteristics of the brain anatomical structure are utilized to fit and generate abnormal lateral fissures.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] The present invention provides a side fissure extraction method, which is characterized by comprising the following steps:

[0008] The first step is to use the existing patented technology CN114266789A method 1 to 4 to extract the extracerebral space and cerebral cortex area of the whole brain ROI , get the whole brain sulcus ventricular area csf ROI ;

[0009] In the second step, the ITK rigid registration method is used to achieve three-dimensional correction of the head;

[0010] The third step is to extract the side fissure;

[0011] The fourth step is to fit and generate abnormal sylvian fissure.

[0012] Furthermore, the third step specifically includes:

[0013] (1) Extract the candidate z-axis range: extract the 3 cm range of the 1.5 cm above and below the layer with the largest cross-sectional area as the candidate z-axis layer where the lateral fissure is located (slice 1, ..., slice M);

[0014] (2) Extract candidate ranges in the y-axis direction: traverse (slice1, ... sliceM), calculate the distance H between the upper and lower edges of each slice, H is divided into 4 parts, and the 2 / 4 area is the area where the sylvian fissure is located (slice1 y 、……sliceM y );

[0015] (3) Extract the candidate range of the x-axis: Calculate the distance W between the leftmost and rightmost sides of the fault, divide W into 3 equal parts, and take out (slice 1 y 、……sliceM y ) on both sides of the area (slice1 xy 、……sliceM xy );

[0016] (4) Extraction of side fissure: Extraction (slice 1 xy 、……sliceM xy ) The first step in the region csf ROI The middle part shows the bilateral Sylvian fissures obtained by segmentation.

[0017] Furthermore, the fourth step specifically includes:

[0018] (1) Determine the side of the lesion: Calculate the volume of the sylvian fissure on both sides, and the side with the smaller volume is the lesion side;

[0019] (2) Symmetrical fitting generation: The normal sylvian fissure is mirrored along the central axis and fitted into the sylvian fissure on the diseased side.

[0020] The present invention also provides a lateral fissure extraction model, characterized in that it is constructed by any of the lateral fissure extraction methods described above.

[0021] The present invention also provides a lateral fissure extraction device, characterized in that it comprises the lateral fissure extraction model described above.

[0022] The present invention also provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and wherein the processor implements the steps of any one of the above-mentioned side fissure extraction methods when executing the program.

[0023] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements any one of the above-mentioned side fissure extraction methods.

[0024] Compared with the prior art, the present invention has the following beneficial effects.

[0025] The inventor creatively proposed to extract candidate regions based on threshold segmentation and the anatomical position of the sylvian fissure, extract the sylvian fissure within the corresponding anatomical region, and in the case of lesion occupation, fit the abnormal side according to the sylvian fissure obtained by normal side segmentation to obtain the sylvian fissure segmentation result of the whole brain.

[0026] This method is applicable to both plain scan CT images (NCCT) and perfusion images (CTP), and is suitable for batch, rapid, automated, and accurate identification of various brain shapes across multiple hospitals, multiple devices, and various brain shapes. For thin-scan image data of sylvian fissure abnormalities, the processing time for a set of 160-slice NCCT data is an average of only 20 seconds, and the computational results are stable, with consistent results across multiple runs of the same data set. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 Overall flow chart.

[0028] Figure 2 The extracerebral space of the whole brain, where a is the cross-sectional view of the original scan data; b is the extracerebral space area of the whole brain obtained by segmentation.

[0029] Figure 3 Three-dimensional correction.

[0030] Figure 4 Triad axis definitions.

[0031] Figure 5 Candidate range in the Z-axis direction, where a is the sagittal view of the candidate lateral fissure area; b is the cross-sectional view of the candidate lateral fissure area.

[0032] Figure 6 Candidate range in the Y-axis direction.

[0033] Figure 7 Candidate range in the X-axis direction.

[0034] Figure 8 Detected side fissure.

[0035] Figure 9 Fitting the generated side fissure.

[0036] Figure 10 3D display of the segmented Sylvian fissure. DETAILED DESCRIPTION

[0037] The present invention will be further described in detail below by means of specific examples and accompanying drawings. The following examples are only provided to further illustrate the present invention and should not be construed as limiting the present invention.

[0038] The specific process of the present invention is as follows Figure 1 shown.

[0039] In the first step, the patent CN114266789A method is used to obtain the whole brain extracerebral space threshold (ThreshValue), and the image is segmented according to the threshold to obtain the whole brain sulcus ventricle (extracerebral space) area csf ROI ,like Figure 2 shown.

[0040] Step 2: Use ITK 3D rigid registration to correct the image (such as Figure 3 shown).

[0041] Step 3: Extract the side fissure. The specific steps are as follows:

[0042] (1) Extract the candidate z-axis range: Extract the 1.5 cm above and below the layer with the largest fault section area, a total of 3 cm, as the candidate z-axis layer where the lateral fissure is located (slice 1, ... slice M), as shown in the schematic diagram. Figure 4 and 5 .

[0043] (2) Extract candidate ranges in the y-axis direction: traverse (slice1, ... sliceM), calculate the distance H between the upper and lower edges of each slice, H is divided into 4 parts, and the 2 / 4 area is the area where the sylvian fissure is located (slice1 y 、……sliceM y ), the schematic diagram is as follows Figure 6 .

[0044] (3) Extract the candidate range of the x-axis: Calculate the distance W between the leftmost and rightmost sides of the fault, divide W into 3 equal parts, and take out (slice 1 y 、……sliceM y ) on both sides of the area (slice1 xy 、……sliceM xy ), the schematic diagram is as follows Figure 7 .

[0045] (4) Extraction of side fissure: Extraction (slice 1 xy 、……sliceM xy ) csf in the area ROI middle( Figure 2 ) is the bilateral sylvian fissure obtained by segmentation. Figure 8 .

[0046] Step 4: Fit and generate the whole brain lateral fissure.

[0047] (1) Determine the side of the lesion: Calculate the volume of the sylvian fissure on both sides, and the side with the smaller volume is the lesion side.

[0048] (2) Symmetrical fitting generation: The normal side fissure is mirrored with the central axis and fitted into the side fissure of the diseased side ( Figure 9 and 10 ).

Claims

1. A side crack extraction method, characterized in that: The following steps are involved: The first step is to obtain a brain CT image and extract the candidate ventricular region Csf through the following steps: ROI : (a) removing the skull region from the brain CT image to obtain the brain tissue region; (b) determining a cerebrospinal fluid grayscale threshold according to a grayscale histogram of the brain tissue region; (c) segmenting a cerebrospinal fluid region from the brain tissue region according to the cerebrospinal fluid grayscale threshold; (d) Extract the largest connected region within the cerebrospinal fluid region as the candidate ventricular region Csf ROI ; In the second step, the ITK rigid registration method is used to achieve three-dimensional correction of the head; The third step is to extract candidate regions in the z-axis, y-axis, and x-axis directions based on the anatomical position, and segment the sylvian fissure within the candidate regions. In the fourth step, the lesion side was determined based on the volume difference of the bilateral sylvian fissures, and the abnormal sylvian fissure was generated by mirror symmetry fitting.

2. A side crack extraction method according to claim 1, characterized in that: The third step specifically includes: (1) Extract the candidate z-axis range: extract the 3 cm range of the 1.5 cm above and below the layer with the largest cross-sectional area as the candidate z-axis layer where the lateral fissure is located (slice 1, ..., slice M); (2) Extract candidate ranges in the y-axis direction: traverse (slice1, ... sliceM), calculate the distance H between the upper and lower edges of each slice, H is divided into 4 parts, and the 2 / 4 area is the area where the sylvian fissure is located (slice1 y 、……sliceM y ); (3) Extract the candidate range of the x-axis: Calculate the distance W between the leftmost and rightmost sides of the fault, divide W into 3 equal parts, and take out (slice 1 y 、……sliceM y ) on both sides of the area (slice1 xy 、……sliceM xy ); (4) Extracting side fissures: In the candidate range (slice 1 xy 、……sliceM xy ), from the first step Csf ROI Bilateral sylvian fissures were divided medianly.

3. A side crack extraction method according to claim 1, characterized in that: The fourth step specifically includes: (1) Determine the side of the lesion: Calculate the volume of the sylvian fissure on both sides, and the side with the smaller volume is the lesion side; (2) Symmetrical fitting generation: The normal sylvian fissure is mirrored along the central axis and fitted into the sylvian fissure on the diseased side.

4. A side fissure extraction device, characterized in that: The invention comprises a module for executing the side fissure extraction method according to any one of claims 1 to 3.

5. An electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, characterized in that: When the processor executes the program, the steps of the side fissure extraction method according to any one of claims 1 to 3 are implemented.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the side fissure extraction method according to any one of claims 1 to 3 is implemented.

Images