Method, device, equipment, storage medium and program product for segmentation of postoperative intracranial pneumothorax

Abstract

The present invention discloses a method, device, equipment, storage medium and program product for segmenting postoperative intracranial pneuma. The method comprises: acquiring image data of a patient's cranial brain, the image data at least comprising a postoperative computed tomography CT image and a preoperative nuclear magnetic resonance MR image; establishing a registration relationship between the computed tomography CT image and the nuclear magnetic resonance MR image; determining a first CT region below a preset density threshold from the computed tomography CT image; determining a first MR region in the nuclear magnetic resonance MR image corresponding to the first CT region according to the registration relationship; and generating a patient's intracranial pneuma segmentation file according to the first MR region. The method can accurately segment the pneumocranial boundary, accurately count the volume of the patient's intracranial pneuma, and improve the efficiency of statistics and analysis of the patient's postoperative intracranial pneuma.

Description

Technical Field

[0001] The present application relates to the field of medical analysis technology, and in particular to a method, device, equipment, storage medium and program product for segmenting postoperative intracranial pneumatosis. Background Art

[0002] Neurosurgery requires very high precision. However, any neurosurgery involving opening the dura mater faces the risk of cerebrospinal fluid loss and air entering the skull through the surgical incision, which will cause air accumulation in the patient's skull and form pneumocephalus.

[0003] Gas accumulation can cause brain tissue to shift or even nonlinear deformation, thereby reducing surgical accuracy. It can also cause problems for doctors in fusing pre- and post-operative images and judging the effectiveness of the surgery.

[0004] At present, there is a lack of accurate and reliable algorithms or software tools to realize the automatic processing and analysis of pneumocephalus after surgery, and it is impossible to accurately segment the pneumocephalus boundary and quantitatively calculate the volume of intracranial pneumocephalus. When necessary, the above purpose can only be achieved by manual outlining, but the manual outlining method has the problems of inaccurate segmentation of pneumocephalus boundary and inaccurate calculation of intracranial pneumocephalus volume, which seriously affects the efficiency of statistics and analysis of postoperative intracranial pneumocephalus in patients.

[0005] The present invention solves at least one of the above problems. Summary of the invention

[0006] The purpose of the present invention is to provide a method, device, equipment, storage medium and program product for segmenting postoperative intracranial pneuma. The method can accurately segment the pneumocranial boundary, accurately count the volume of intracranial pneuma in patients, and improve the efficiency of statistics and analysis of postoperative intracranial pneuma in patients.

[0007] The purpose of the present invention is achieved by the following technical solutions:

[0008] A first aspect of the present invention provides a method for segmenting postoperative intracranial pneumatosis, comprising:

[0009] Acquiring image data of the patient's brain, wherein the image data at least includes a postoperative computed tomography (CT) image and a preoperative magnetic resonance (MR) image;

[0010] Establishing a registration relationship between the computed tomography (CT) image and the magnetic resonance (MR) image;

[0011] Determine a first CT region below a preset density threshold from the computed tomography (CT) image;

[0012] Determining, according to the registration relationship, a first MR region in the nuclear magnetic resonance MR image corresponding to the first CT region;

[0013] An intracranial pneumatosis segmentation file of the patient is generated according to the first MR region.

[0014] In a possible implementation manner of the first aspect, determining, according to the registration relationship, a first MR region in the nuclear magnetic resonance MR image corresponding to the first CT region includes:

[0015] Extracting the frontal lobe region in the nuclear magnetic resonance (MR) image to obtain a frontal lobe mask image;

[0016] According to the registration relationship, a first MR region corresponding to the first CT region in the frontal lobe mask image is determined.

[0017] In a possible implementation manner of the first aspect, the first MR region is an intersection of a region corresponding to the frontal lobe mask image and the first CT region.

[0018] In a possible implementation manner of the first aspect, extracting the frontal lobe region in the nuclear magnetic resonance MR image to obtain a frontal lobe mask image includes:

[0019] Segmenting the components of the frontal lobe from the MRI image to obtain a sub-mask image of each component;

[0020] The intersection of all sub-mask images is obtained to merge all sub-mask images and obtain the prefrontal lobe mask image of the prefrontal lobe.

[0021] In a possible implementation manner of the first aspect, generating a segmentation file of intracranial pneumatosis of the patient according to the first MR region includes:

[0022] According to the first MR region, binarization processing is performed on the voxels in the prefrontal mask image, wherein the voxels in the first MR region are marked as 1, and the voxels in the prefrontal mask image other than the first MR region are marked as 0;

[0023] Determining the volume of intracranial pneumocephalus after surgery of the patient according to the prefrontal lobe mask image after binarization processing and the volume of a single voxel;

[0024] An intracranial pneumothorax segmentation file of the patient is generated according to the volume of the patient's postoperative intracranial pneumothorax and the mask image corresponding to the first MR region.

[0025] In a possible implementation manner of the first aspect, the volume of a single voxel is determined by a size parameter of the computed tomography (CT) image.

[0026] In a possible implementation manner of the first aspect, binarization is performed on voxel values ​​in the frontal lobe mask image according to the first MR region, and then the method further includes:

[0027] The binary frontal lobe mask image is expanded.

[0028] In a possible implementation manner of the first aspect, generating an intracranial pneumothorax segmentation file of the patient according to the volume of the postoperative intracranial pneumothorax of the patient and a mask image corresponding to the first MR region includes:

[0029] When the volume of intracranial pneumocephalus after surgery is less than a preset pneumocephalus volume threshold, it is determined that the patient has no intracranial pneumocephalus;

[0030] When the postoperative intracranial pneumocephalus volume of the patient is greater than or equal to the preset pneumocephalus volume threshold, the seed growth method is used to perform diffusion processing on the pneumocephalus area corresponding to the intracranial pneumocephalus volume until the preset diffusion stop condition is reached, thereby obtaining the target intracranial pneumocephalus volume and the intracranial pneumocephalus segmentation file of the patient corresponding to the target intracranial pneumocephalus volume.

[0031] In a possible implementation manner of the first aspect, performing diffusion processing on the pneumocephalus region corresponding to the intracranial pneumocephalus volume by using a seed growth method includes:

[0032] Setting diffusion constraints, wherein the diffusion constraints at least include the range of the patient's brain cranial cavity;

[0033] Based on the diffusion constraint condition, taking the pneumocephalus region corresponding to the intracranial pneumonic volume as the center, diffusing to the first CT region at a preset diffusion speed, and updating the pneumocephalus region after each diffusion;

[0034] According to the number of voxels in the pneumocephalus area after each diffusion, the volume of intracranial pneumothorax after each diffusion is calculated to determine the volume increase after each diffusion;

[0035] Determining whether a preset diffusion stop condition is reached according to the volume growth after each diffusion and / or the number of diffusions;

[0036] If yes, the intracranial pneumothorax volume finally obtained by diffusion is determined as the target intracranial pneumothorax volume, and the intracranial pneumothorax segmentation file of the patient is determined according to the pneumocephalus region corresponding to the target intracranial pneumothorax volume.

[0037] In a possible implementation manner of the first aspect, the preset diffusion stop condition includes one of the following conditions:

[0038] The volume of intracranial pneumothorax increases less than the preset volume compared with the previous iteration; or

[0039] The number of diffusions reaches the preset number of diffusions.

[0040] Compared with the prior art, the beneficial effects of the present invention are as follows: by acquiring dual-modality imaging data of the patient, namely, the postoperative computed tomography CT image and the preoperative magnetic resonance MR image, and performing registration, the present invention can be well applied to the statistics and analysis of postoperative intracranial pneumothorax in the patient; by a preset density threshold, a first CT region representing the gas portion below the preset density threshold can be accurately located from the CT image; by the first CT region, a first MR region representing the pneumothorax region can be accurately located from the MR image, and by the first MR region, an intracranial pneumothorax segmentation file can be accurately segmented from the CT image, which can not only accurately segment the pneumothorax boundary, but also accurately count the volume of intracranial pneumothorax in the patient, thereby improving the efficiency of statistics and analysis of postoperative intracranial pneumothorax in the patient.

[0041] Specifically, the present application aligns the postoperative CT images with the patient's preoperative MR images through multimodal image data registration. The registered image data is combined with information about brain partitions in the MR images and information about pneumocephalus in the CT images, and multimodal image semantic information is integrated to enhance the reliability and accuracy of the algorithm.

[0042] Specifically, based on the recognition and utilization of the clinical characteristics of intracranial pneumocephalus in neurosurgery, the present application performs binarization and expansion operations on the frontal lobe mask image, thereby smoothing out the errors between CT images and MR images regarding the pneumocephalus area, improving the application effect of the mask file, and ensuring the integrity and accuracy of the boundaries of the frontal lobe mask image.

[0043] Specifically, the present application provides an efficient screening method based on HU values, which effectively distinguishes gas parts and non-gas parts in CT images, thereby improving screening accuracy.

[0044] Specifically, the present application creatively uses the seed growth method to iteratively expand the initial pneumocephalus region. By iteratively expanding the boundary of the initial pneumocephalus region in the highly stable prefrontal mask image, new pneumocephalus regions are continuously incorporated. In addition, the volume growth rate is defined as the main termination iteration indicator to improve the segmentation accuracy of the pneumocephalus region and ensure the robustness of the algorithm.

[0045] A second aspect of the present invention provides a postoperative intracranial pneumatosis segmentation device, comprising:

[0046] Image acquisition module: used to acquire image data of the patient's brain, the image data at least including postoperative computed tomography (CT) images and preoperative magnetic resonance (MR) images;

[0047] Image registration module: used to establish a registration relationship between the computer tomography (CT) image and the magnetic resonance (MR) image;

[0048] A CT region determination module: used to determine a first CT region below a preset density threshold from the computed tomography CT image;

[0049] An MR region determination module: used for determining a first MR region in the nuclear magnetic resonance MR image corresponding to the first CT region according to the registration relationship;

[0050] Intracranial pneuma segmentation module: used to generate an intracranial pneuma segmentation file of the patient according to the first MR region.

[0051] A third aspect of the present invention provides a medical device, comprising:

[0052] A first processor is configured to communicate with a medical acquisition device and execute the above method, wherein the medical acquisition device is used to acquire brain image data of a patient;

[0053] The display screen is configured to display a three-dimensional image of the patient's brain, wherein the three-dimensional image of the brain at least includes a three-dimensional image of intracranial pneumosis of the patient.

[0054] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a second processor, the steps of the above-mentioned postoperative intracranial pneumatosis segmentation method are implemented.

[0055] A fifth aspect of the present invention provides a computer program product, which includes instructions, and when the instructions are executed on an electronic device, the electronic device implements the steps of the above-mentioned postoperative intracranial pneumatosis segmentation method. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is a flow chart of the overall steps of the method for segmenting postoperative intracranial pneumatosis according to an embodiment of the present invention;

[0057] Figure 2 This is a flow chart of the steps of determining the first MR region according to an embodiment of the present invention;

[0058] Figure 3 This is a flow chart of the steps of extracting the frontal lobe mask image according to an embodiment of the present invention;

[0059] Figure 4 A flow chart of the steps for generating a segmentation file of intracranial pneumatosis in a patient according to an embodiment of the present invention;

[0060] Figure 5 A flow chart of steps for generating a pneumatosis segmentation file corresponding to a target pneumatosis volume according to an embodiment of the present invention;

[0061] Figure 6A flowchart of the steps of diffusion processing of pneumocranial regions corresponding to the volume of intracranial pneumosis according to an embodiment of the present invention;

[0062] Figure 7 This is a schematic diagram of the structure of a device for segmenting postoperative intracranial pneumatosis according to an embodiment of the present invention;

[0063] Figure 8 A schematic diagram of the structure of a medical device according to an embodiment of the present invention;

[0064] Fig. 9 It is a schematic diagram of a cross-sectional MR image of the brain of a patient according to an embodiment of the present invention;

[0065] Fig.10 A schematic diagram of a cross-sectional CT image of the brain of a patient according to an embodiment of the present invention;

[0066] Fig.11 A pneumocephalus model for three-dimensional visualization of the intracranial pneumocephalus segmentation file according to an embodiment of the present invention. DETAILED DESCRIPTION

[0067] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided to make the present invention more comprehensive and complete and to fully convey the concepts of example embodiments to those skilled in the art. The same reference numerals in the figures represent the same or similar structures, and thus their repeated description will be omitted.

[0068] See attached Figure 1 As shown, this embodiment provides a postoperative intracranial pneumatosis segmentation method, including steps S1 to S5.

[0069] Step S1: Acquire image data of the patient's brain, the image data at least including postoperative computed tomography (CT) images and preoperative magnetic resonance (MR) images.

[0070] In an embodiment of the present invention, MR images and CT images are two common diagnostic images in medicine, which are generated by MR (MEGNETIC RRESONANCE) imaging technology and CT (Computed Tomography) technology, respectively. In some other embodiments, the image data may also be other types of image data, such as X-ray imaging, ultrasonic imaging, optical coherence tomography, etc., and the specific type is not limited.

[0071] Step S2: Establishing a registration relationship between the computed tomography (CT) image and the magnetic resonance (MR) image.

[0072] The purpose of registration of CT images and MR images is to seek spatial transformation so that the corresponding points of the two images are consistent in space. The registration method can be the body surface marker method, the interactive automatic registration method, etc. The registration relationship between CT images and MR images can be established by the above methods. This is a conventional technical means and will not be described in detail here.

[0073] Furthermore, the registration effect can be measured by evaluating the off-axis curve matching and the deviation of the center position of the region of interest volume, which will not be elaborated here.

[0074] It should be further explained that the same set of body surface landmarks should be used when acquiring MR images and CT images to improve the accuracy of registration.

[0075] For patients with head and neck, both registration methods have good registration results with small deviations. For patients with chest tumors, the interactive automatic registration method is superior to the body surface marker method and has better off-axis curve matching.

[0076] By establishing the registration relationship between CT images and MR images through the above-mentioned registration method, the corresponding points between CT images and MR images of different imaging modes can be made consistent in space, thereby improving the accuracy of the analysis and statistics of postoperative intracranial pneumatosis.

[0077] Step S3: determining a first CT region whose density is lower than a preset density threshold from the computed tomography (CT) image.

[0078] According to the principle of CT imaging, materials of different densities have different HU (Hounsfield unit) values ​​in CT scan imaging results. HU is a unit of measurement for determining the density of a local tissue or organ in the human body. The HU value of brain tissue is usually 25~45, the HU value of cerebrospinal fluid is usually 0~12, the HU value of compact bone is usually 900~1500, the HU value of gas is usually less than or equal to -900, and the HU value of DBS electrodes and other metal parts is usually greater than or equal to 1600.

[0079] Therefore, the preset density threshold is the HU threshold. By dividing the HU values ​​at different positions in the CT image, an area roughly related to the corresponding type can be obtained. For example, taking gas as an example, the HU threshold in this embodiment can be between -800 and 1000, preferably, can be -900.

[0080] Because the HU value of gas is usually less than or equal to -900, the first CT region representing the gas portion in the CT image can be screened out according to the HU threshold.

[0081] Step S4: Determine a first MR region in the nuclear magnetic resonance MR image corresponding to the first CT region according to the registration relationship.

[0082] Since the above registration relationship makes the corresponding points between the CT image and the MR image consistent in space, the first MR region corresponding to the gas portion can be determined in the MR image.

[0083] In some specific embodiments, see the attached Figure 2 As shown, step S4 includes step S41-step S42.

[0084] Step S41: extracting the frontal lobe region in the nuclear magnetic resonance MR image to obtain a frontal lobe mask image.

[0085] In some specific embodiments, see the attached Figure 3 As shown, step S41 includes steps S411 to S412.

[0086] Step S411: segmenting the components of the frontal lobe from the MR image to obtain a sub-mask image of each component.

[0087] In actual situations, after a patient undergoes craniotomy, air generally enters the frontal lobe, thereby forming a certain gas accumulation environment in the frontal lobe region. Therefore, in order to improve the accuracy and efficiency of intracranial gas segmentation, the patient's frontal lobe region can be segmented in advance, and gas accumulation can be identified and segmented only from this region. Alternatively, by segmenting the MR images, the various components of the frontal lobe can be more accurately identified and located, which helps to extract the frontal lobe region in the MR images and improves the image processing efficiency and the accuracy of identification and positioning.

[0088] Step S412: Obtain the intersection of all sub-mask images to merge all sub-mask images and obtain a prefrontal lobe mask image of the prefrontal lobe.

[0089] By merging the sub-mask images, the image integrity of the prefrontal lobe mask image is improved. By obtaining the intersection of the sub-mask images, it can be ensured that no component of the prefrontal lobe is missed during the merging process, thereby obtaining a complete prefrontal lobe mask image of the prefrontal lobe.

[0090] Furthermore, the merging process can reduce image noise, improve image clarity and contrast, and make the frontal lobe more clearly visible.

[0091] It should be noted that the applicant flexibly applied medical prior knowledge here. Since pneumocephalus generally only exists in the frontal lobe of the patient's brain, the applicant first extracted the frontal lobe area from the MR image to obtain a frontal lobe mask image.

[0092] Step S42: Determine the first MR region corresponding to the first CT region in the frontal lobe mask image according to the registration relationship.

[0093] According to the above registration relationship, the gas portion is defined from the frontal lobe mask image, that is, the pneumocephalus region in the patient's brain is determined from the frontal lobe region.

[0094] See attached Fig. 9 As shown, Image I is a top view cross-sectional view of the patient's cranial MR image, wherein A represents the cranial brain region, and a and b both represent the pneumocephalus region in the prefrontal lobe region; Image II is a front view cross-sectional view of the patient's cranial MR image, wherein B represents the cranial brain region, and c and d both represent the pneumocephalus region in the prefrontal lobe region; Image III is a side view cross-sectional view of the patient's cranial MR image, wherein C represents the cranial brain region, and e represents the pneumocephalus region in the prefrontal lobe region.

[0095] Because the frontal mask image is obtained, the pneumocephalus part can be limited in the frontal region. At the same time, because the non-frontal region is eliminated first, the interference of the gas part in the non-frontal region on the determination of the pneumocephalus region is avoided, making the subsequent pneumocephalus analysis statistics more accurate.

[0096] It should be further explained that the first MR region is the intersection of the region corresponding to the frontal lobe mask image and the first CT region, and the intersection is the location of the pneumocephalus region in the frontal lobe region.

[0097] Step S5: Generate a patient's intracranial pneumothorax segmentation file based on the first MR region.

[0098] In some specific embodiments, see the attached Figure 4 As shown, step S5 includes steps S51 to S53.

[0099] Step S51: binarizing the voxels in the prefrontal mask image according to the first MR region, wherein each voxel in the first MR region is marked as 1, and each voxel in the prefrontal mask image except the first MR region is marked as 0.

[0100] Therefore, in this embodiment, the voxel value of each voxel in the first MR region, that is, the pneumocephalus region, is 1, and the voxel value of each voxel in the prefrontal mask image except the first MR region, that is, the non-pneumocephalus region, is marked as 0.

[0101] Alternatively, each voxel in the first MR region is marked as 0, and each voxel in the prefrontal mask image except the first MR region is marked as 1, which is not limited in the present application.

[0102] Binarization can highlight features such as edges and textures in the image, which is beneficial for feature extraction and region recognition. The frontal lobe mask image can be clearly divided into two parts, the pneumocephalus area and the non-pneumocephalus area, through binarization processing, which is helpful for subsequent pneumocephalus analysis statistics.

[0103] In some implementations, the voxel values ​​in the prefrontal lobe mask image are binarized according to the first MR region, and then the method further includes: dilating the binarized prefrontal lobe mask image.

[0104] Due to the presence of pneumocephalus in the patient's brain after surgery, the MR image will have a certain offset after the postoperative CT image and MR image are registered, that is, it will offset toward the back of the brain, and the corresponding frontal mask image will also have a certain offset. In this way, the first MR area determined by the intersection of the area corresponding to the frontal mask image and the first CT area will be relatively small, and more iterations will be required in subsequent iterations. Therefore, in order to improve the efficiency of subsequent iterations, the binarized frontal mask image can be appropriately expanded and enlarged before iteration.

[0105] Step S52: Determine the volume of intracranial pneumothorax after surgery of the patient based on the binarized prefrontal mask image and the volume of a single voxel.

[0106] It should be noted that the volume of a single voxel is determined by the size parameters of the computed tomography (CT) image. In some embodiments, the volume of each voxel is the same. In other embodiments, the voxels may have different volumes.

[0107] Therefore, the volume of intracranial pneumocephalus after surgery = the sum of the number of voxel values ​​1 or 0 in the pneumocephalus area * the volume of a single voxel.

[0108] For example, the number of voxels with a voxel value of 1 in the pneumocephalus region is 100, and the volume of a single voxel is 20, so the volume of the intracranial pneumocephalus is 2000.

[0109] For example, the number of voxels with a voxel value of 1 in the pneumocephalus region is 80, and the volume of a single voxel is 40, so the volume of the intracranial pneumocephalus is 3200.

[0110] For example, the number of voxels with a voxel value of 1 in the pneumocephalus region is 60, and the volume of a single voxel is 60, so the volume of the intracranial pneumocephalus is 3600.

[0111] Or, for example, the number of voxels with a voxel value of 0 in the pneumocephalus region is 60, and the volume of a single voxel is 60, so the volume of the intracranial pneumocephalus volume is 3600.

[0112] For example, the number of voxels with a voxel value of 0 in the pneumocephalus region is 80, and the volume of a single voxel is 40, so the volume of the intracranial pneumocephalus is 3200.

[0113] For example, the number of voxels with a voxel value of 0 in the pneumocephalus region is 100, and the volume of a single voxel is 20, so the volume of the intracranial pneumocephalus volume is 2000.

[0114] Step S53: Generate a segmentation file of the patient's intracranial pneuma according to the patient's postoperative intracranial pneuma volume and the mask image corresponding to the first MR region.

[0115] In some specific embodiments, see the attached Figure 5 As shown, step S53 includes steps S531 and S532.

[0116] Step S531: When the volume of intracranial pneumocephalus after surgery is less than a preset pneumocephalus volume threshold, it is determined that there is no intracranial pneumocephalus in the patient.

[0117] Because a small amount of air enters the brain, it generally does not cause adverse reactions, the patient is conscious, has no obvious neurological deficits, and the intracranial pressure is not high. In this case, no special treatment is required, just bed rest for a period of time and use some antibiotics to prevent infection, and the gas will usually be absorbed by itself.

[0118] A large amount of intracranial gas may lead to symptoms such as intracranial infection and increased intracranial pressure, and may even be life-threatening. At this time, timely surgical treatment is required, such as cerebrospinal fluid leak repair, drilling and exhaust drainage, etc.

[0119] To this end, it is necessary to set a preset pneumocephalus volume threshold and compare the volume of intracranial gas with the preset pneumocephalus volume threshold to determine whether the patient has intracranial gas.

[0120] For example, the preset pneumocephalus volume threshold is 3000, and the intracranial gas volume is 2000. At this time, since the intracranial gas volume is less than the preset pneumocephalus volume threshold, the intracranial gas volume is too small, and it is determined that the patient has no intracranial gas, and no subsequent analysis and statistics are required.

[0121] Step S532: When the postoperative intracranial pneumocephalus volume of the patient is greater than or equal to the preset pneumocephalus volume threshold, the pneumocephalus area corresponding to the intracranial pneumocephalus volume is diffused using the seed growth method until the preset diffusion stop condition is reached, thereby obtaining the target intracranial pneumocephalus volume and the intracranial pneumocephalus segmentation file of the patient corresponding to the target intracranial pneumocephalus volume.

[0122] For example, the preset pneumocephalus volume threshold is 3000, and the intracranial gas volume is 3200. At this time, since the intracranial gas volume is greater than the preset pneumocephalus volume threshold, the intracranial gas volume is too large, and it is determined that the patient has intracranial gas, which requires further analysis and statistics.

[0123] In some embodiments, see Appendix Figure 6 As shown, the diffusion process of the pneumocranium region corresponding to the volume of intracranial pneumosis using the seed growth method includes steps S100 to S400.

[0124] Step S100: setting diffusion constraints, where the diffusion constraints at least include the range of the patient's cranial cavity.

[0125] Most of the pneumocephalus in the brain can be screened out through the above-mentioned registration and density screening methods. However, as mentioned above, due to the existence of pneumocephalus, there will be a certain offset between the postoperative CT image and the preoperative MR image during registration, that is, the MR image will shift backward, which will cause the pneumocephalus after screening to be missing. Therefore, in order to screen out the pneumocephalus in the postoperative CT image as much as possible, it is also necessary to perform edge diffusion processing based on the screened pneumocephalus, so as to expand the boundary of the pneumocephalus until the corresponding requirements are met, and a more accurate and realistic pneumocephalus area is obtained.

[0126] By setting diffusion constraints, the diffusion of the pneumocephalus area corresponding to the volume of intracranial gas accumulation cannot exceed the range of the patient's cranial cavity. Because the outer area outside the human brain in CT images does not belong to the research object, the diffusion range of the pneumocephalus area must be limited to prevent interference from external areas and improve the accuracy of pneumocephalus analysis statistics.

[0127] Furthermore, the diffusion range can be narrowed to the prefrontal region, further improving the accuracy of pneumocephalus analysis statistics.

[0128] Step S200: Based on the diffusion constraint condition, the pneumocephalus region corresponding to the intracranial gas volume is taken as the center, and diffuses toward the first CT region at a preset diffusion speed, and updates the pneumocephalus region after each diffusion.

[0129] Step S300: Calculate the volume of intracranial gas after each diffusion according to the number of voxels in the pneumocephalus region after each diffusion to determine the volume increase after each diffusion.

[0130] Step S400: judging whether a preset diffusion stop condition is reached according to the volume growth after each diffusion and / or the diffusion times.

[0131] In some embodiments, the preset diffusion stop condition further includes one of the following conditions: condition 1: the increase in the volume of intracranial pneumothorax compared with the previous iteration is less than the preset volume; or condition 2: the number of diffusions reaches the preset number of diffusions.

[0132] If yes, the final diffused pneumocephalus volume is determined as the target pneumocephalus volume, and the pneumocephalus segmentation file of the patient is determined according to the pneumocephalus region corresponding to the target pneumocephalus volume.

[0133] See attached Fig.10As shown, image IV is a top-view cross-sectional view of the patient's cranial CT image, wherein D represents the cranial brain region, and g and f both represent the pneumocephalus region corresponding to the target intracranial pneuma volume; image V is a front-view cross-sectional view of the patient's cranial CT image, wherein E represents the cranial brain region, and h, i, j and k both represent the pneumocephalus region corresponding to the target intracranial pneuma volume; image VI is a side-view cross-sectional view of the patient's cranial CT image, wherein F represents the cranial brain region. It should be noted that the pneumocephalus region corresponding to the target intracranial pneuma volume is not shown in the image in the side-view direction.

[0134] See attached Fig.11 As shown in the figure, the pneumocephalus model is displayed in three dimensions in the visualization imaging system by segmenting the intracranial pneumocephalus file.

[0135] In the brain cranial cavity range, the pneumocephalus region corresponding to the current intracranial pneumothorax volume is used as the seed point for growth and diffusion. According to certain similarity criteria, adjacent pixels are gradually merged into the growth and diffusion region to improve the efficiency and accuracy of the segmentation of the pneumocephalus region corresponding to the target intracranial pneumothorax volume.

[0136] At the same time, the preset diffusion stop condition is set to make the growth of seed points more stable, further improving the accuracy and efficiency of intracranial pneumothorax segmentation files.

[0137] In this way, the target intracranial pneumocephalus volume can be accurately counted from the CT image, and the pneumocephalus area corresponding to the target intracranial pneumocephalus volume can be accurately segmented, thereby improving the accuracy of the analysis and statistics of the patient's postoperative pneumocephalus.

[0138] It should be further explained that the method of the present application has a wide range of application scenarios and is applicable to all neurosurgery operations involving cerebrospinal fluid loss and air entering the skull, including but not limited to: deep brain stimulation (DBS), ventricular puncture, hematoma removal, etc.

[0139] The present application provides another embodiment, which provides a postoperative intracranial pneumatosis segmentation device, and applies the postoperative intracranial pneumatosis segmentation method described above, see the attached Figure 7 As shown, it includes an image acquisition module 100, an image registration module 200, a CT region determination module 300, an MR region determination module 400 and an intracranial pneumatosis segmentation module 500.

[0140] The image acquisition module 100 is used to acquire image data of the patient's brain, and the image data at least includes postoperative computed tomography (CT) images and preoperative magnetic resonance (MR) images.

[0141] The image registration module 200 is used to establish a registration relationship between the computed tomography (CT) image and the magnetic resonance (MR) image.

[0142] The CT region determination module 300 is used to determine a first CT region with a density lower than a preset density threshold from a computed tomography (CT) image.

[0143] The MR region determination module 400 is used to determine a first MR region in the nuclear magnetic resonance MR image corresponding to the first CT region according to the registration relationship.

[0144] The pneumatosis segmentation module 500 is used to generate a pneumatosis segmentation file of the patient according to the first MR region.

[0145] The device can accurately count the target intracranial pneumocephalus volume from CT images and accurately segment the pneumocephalus area corresponding to the target intracranial pneumocephalus volume, thereby improving the accuracy of the analysis and statistics of postoperative pneumocephalus in patients.

[0146] This application provides another embodiment, see the attached Figure 8 As shown, this embodiment provides a medical device 10 , which includes a first processor 11 and a display screen 12 .

[0147] The first processor 11 is configured to communicate with the medical acquisition device 20 and execute the above method. The medical acquisition device 20 is used to acquire brain image data of the patient.

[0148] The display screen 12 is configured to display a three-dimensional image of the patient's brain, wherein the three-dimensional image of the brain at least includes a three-dimensional image of intracranial pneumosis of the patient.

[0149] The medical device 10 can not only accurately count the target intracranial pneumocephalus volume from the CT image, accurately segment the pneumocephalus area corresponding to the target intracranial pneumocephalus volume, thereby improving the accuracy of the analysis and statistics of the patient's postoperative pneumocephalus, but also display the three-dimensional morphology of the image during transmission and processing in real time through the display screen, which is conducive to doctors to better analyze the patient's postoperative pneumocephalus.

[0150] The present application provides another embodiment, which provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a second processor, the steps of the above-mentioned postoperative intracranial pneumatosis segmentation method are implemented.

[0151] The present application provides another embodiment, which provides a computer program product. The computer program product includes instructions. When the instructions are executed on an electronic device, the electronic device implements the steps of the above-mentioned postoperative intracranial pneumatosis segmentation method.

[0152] It is to be understood that the structure illustrated in the embodiment of the present invention does not constitute a specific limitation on the electronic device. In other embodiments of the present application, the electronic device may include more or fewer components, or combine certain components, or split certain components, or arrange the components differently. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0153] The first processor 11 and / or the second processor may include one or more processing units, for example, a central processing unit CPU (Central Processing Unit), a graphics processor GPU (Graphics Processing Unit), a digital signal processor DSP, a microprocessor MCU (Micro-programmed Control Unit), an AI (Artificial Intelligence, artificial intelligence) processor or a programmable logic device FPGA (Field Programmable Gate Array) and the like processing module or processing circuit. Among them, different processing units can be independent devices or integrated in one or more processors. A storage unit may be set in the first processor 11 and / or the second processor for storing instructions and data. In some embodiments, the storage unit in the first processor 11 and / or the second processor is a cache memory.

[0154] The various embodiments of the mechanism disclosed in the present application can be implemented in hardware, software, firmware or a combination of these implementation methods. The embodiments of the present application can be implemented as a computer program or program code executed on a programmable system, and the programmable system includes at least one third processor, a storage system (including volatile and non-volatile memory and / or storage elements), at least one input device and at least one output device.

[0155] Program code can be applied to input instructions to perform the functions described in this application and generate output information. The output information can be applied to one or more output devices in a known manner. For the purposes of this application, a processing system includes any system having a processor such as, for example, a digital signal processor (DSP), a microcontroller, an application specific integrated circuit (ASIC), or a microprocessor.

[0156] Program code can be implemented with high-level programming language or object-oriented programming language to communicate with the processing system. When necessary, program code can also be implemented with assembly language or machine language. In fact, the mechanism described in this application is not limited to the scope of any specific programming language. In either case, the language can be a compiled language or an interpreted language.

[0157] In some cases, the disclosed embodiments may be implemented in hardware, firmware, software, or any combination thereof. The disclosed embodiments may also be implemented as instructions carried or stored on one or more temporary or non-temporary machine-readable (e.g., computer-readable) storage media, which may be read and executed by one or more processors. For example, instructions may be distributed over a network or through other computer-readable media. Therefore, machine-readable media may include any mechanism for storing or transmitting information in a machine (e.g., computer) readable form, including, but not limited to, floppy disks, optical disks, optical disks, read-only memories (CD-ROMs), magneto-optical disks, read-only memories (ROMs), random access memories (RAMs), erasable programmable read-only memories (EPROMs), electrically erasable programmable read-only memories (EEPROMs), magnetic or optical cards, flash memory, or a tangible machine-readable memory for transmitting information (e.g., carrier waves, infrared signals, digital signals, etc.) using the Internet in electrical, optical, acoustic, or other forms of propagation signals. Therefore, machine-readable media include any type of machine-readable media suitable for storing or transmitting electronic instructions or information in a machine (e.g., computer) readable form.

[0158] In the accompanying drawings, some structural or method features may be shown in a specific arrangement and / or order. However, it should be understood that such a specific arrangement and / or order may not be required. Instead, in some embodiments, these features may be arranged in a manner and / or order different from that shown in the illustrative drawings. In addition, the inclusion of structural or method features in a particular figure does not mean that such features are required in all embodiments, and in some embodiments, these features may not be included or may be combined with other features.

[0159] It should be noted that the units / modules mentioned in the various device embodiments of the present application are all logical units / modules. Physically, a logical unit / module can be a physical unit / module, or a part of a physical unit / module, or can be implemented as a combination of multiple physical units / modules. The physical implementation method of these logical units / modules themselves is not the most important. The combination of functions implemented by these logical units / modules is the key to solving the technical problems proposed by the present application. In addition, in order to highlight the innovative part of the present application, the above-mentioned device embodiments of the present application do not introduce units / modules that are not closely related to solving the technical problems proposed by the present application, which does not mean that there are no other units / modules in the above-mentioned device embodiments.

[0160] It should be noted that in the examples and description of this patent, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "including one" do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0161] Although the present application has been illustrated and described with reference to certain preferred embodiments thereof, it will be apparent to those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the present application.

Claims

1. A method for segmenting postoperative intracranial pneumothorax, characterized in that: include: Acquiring image data of the patient's brain, wherein the image data at least includes a postoperative computed tomography (CT) image and a preoperative magnetic resonance (MR) image; Establishing a registration relationship between the computed tomography (CT) image and the magnetic resonance (MR) image; Determine a first CT region below a preset density threshold from the computed tomography (CT) image; the first CT region represents a gas portion in the CT image; Determining, according to the registration relationship, a first MR region in the nuclear magnetic resonance MR image corresponding to the first CT region, comprising: extracting a frontal lobe region in the nuclear magnetic resonance MR image to obtain a frontal lobe mask image; determining, according to the registration relationship, a first MR region in the frontal lobe mask image corresponding to the first CT region; An intracranial pneumatosis segmentation file of the patient is generated according to the first MR region.

2. The method for segmenting postoperative intracranial pneumothorax according to claim 1, characterized in that: The first MR region is an intersection of a region corresponding to the frontal lobe mask image and the first CT region.

3. The method for segmenting postoperative intracranial pneumothorax according to claim 1, characterized in that: Extracting the frontal lobe region in the nuclear magnetic resonance MR image to obtain a frontal lobe mask image includes: Segmenting the components of the frontal lobe from the MRI image to obtain a sub-mask image of each component; The intersection of all sub-mask images is obtained to merge all sub-mask images and obtain the prefrontal lobe mask image of the prefrontal lobe.

4. The method for segmenting postoperative intracranial pneumothorax according to claim 1, characterized in that: Generating a segmentation file of intracranial pneumatosis of the patient according to the first MR region includes: According to the first MR region, binarization processing is performed on the voxels in the prefrontal mask image, wherein the voxels in the first MR region are marked as 1, and the voxels in the prefrontal mask image other than the first MR region are marked as 0; Determining the volume of intracranial pneumocephalus after surgery of the patient according to the prefrontal lobe mask image after binarization processing and the volume of a single voxel; An intracranial pneumothorax segmentation file of the patient is generated according to the volume of the patient's postoperative intracranial pneumothorax and the mask image corresponding to the first MR region.

5. The method for segmenting postoperative intracranial pneumothorax according to claim 4, characterized in that: The volume of a single voxel is determined by the size parameters of the computed tomography (CT) image.

6. The method for segmenting postoperative intracranial pneumothorax according to claim 4, characterized in that: According to the first MR region, voxels in the frontal lobe mask image are binarized, and then the method further includes: The binary frontal lobe mask image is expanded.

7. The method for segmenting postoperative intracranial pneumothorax according to claim 4, characterized in that: Generating a segmentation file of the patient's intracranial pneuma according to the patient's postoperative intracranial pneuma volume and the mask image corresponding to the first MR region, including: When the volume of intracranial pneumocephalus after surgery is less than a preset pneumocephalus volume threshold, it is determined that the patient has no intracranial pneumocephalus; When the postoperative intracranial pneumocephalus volume of the patient is greater than or equal to the preset pneumocephalus volume threshold, the seed growth method is used to perform diffusion processing on the pneumocephalus area corresponding to the intracranial pneumocephalus volume until the preset diffusion stop condition is reached, thereby obtaining the target intracranial pneumocephalus volume and the intracranial pneumocephalus segmentation file of the patient corresponding to the target intracranial pneumocephalus volume.

8. The method for segmenting postoperative intracranial pneumothorax according to claim 7, characterized in that: The pneumocephalus region corresponding to the intracranial pneumocephalus volume is diffused using a seed growth method, including: Setting diffusion constraints, wherein the diffusion constraints at least include the range of the patient's brain cranial cavity; Based on the diffusion constraint condition, taking the pneumocephalus region corresponding to the intracranial pneumonic volume as the center, diffusing to the first CT region at a preset diffusion speed, and updating the pneumocephalus region after each diffusion; According to the number of voxels in the pneumocephalus area after each diffusion, the volume of intracranial pneumothorax after each diffusion is calculated to determine the volume increase after each diffusion; Determining whether a preset diffusion stop condition is reached according to the volume growth after each diffusion and / or the number of diffusions; If yes, the intracranial pneumothorax volume finally obtained by diffusion is determined as the target intracranial pneumothorax volume, and the intracranial pneumothorax segmentation file of the patient is determined according to the pneumocephalus region corresponding to the target intracranial pneumothorax volume.

9. The method for segmenting postoperative intracranial pneumothorax according to claim 8, characterized in that: The preset diffusion stop condition includes one of the following conditions: The volume of intracranial pneumothorax increases less than the preset volume compared with the previous iteration; or The number of diffusions reaches the preset number of diffusions.

10. A postoperative intracranial pneumatosis segmentation device, characterized in that: include: Image acquisition module: used to acquire image data of the patient's brain, the image data at least including postoperative computed tomography (CT) images and preoperative magnetic resonance (MR) images; Image registration module: used to establish a registration relationship between the computer tomography (CT) image and the magnetic resonance (MR) image; CT region determination module: used to determine a first CT region below a preset density threshold from the computed tomography CT image; the first CT region represents a gas portion in the CT image; The MR region determination module is used to determine the first MR region in the nuclear magnetic resonance MR image corresponding to the first CT region according to the registration relationship, including: extracting the frontal lobe region in the nuclear magnetic resonance MR image to obtain a frontal lobe mask image; determining the first MR region in the frontal lobe mask image corresponding to the first CT region according to the registration relationship; Intracranial pneuma segmentation module: used to generate an intracranial pneuma segmentation file of the patient according to the first MR region.

11. A medical device, characterized in that The medical equipment includes: A first processor, configured to communicate with a medical acquisition device and execute the method according to any one of claims 1 to 9, wherein the medical acquisition device is used to acquire brain image data of a patient; The display screen is configured to display a three-dimensional image of the patient's brain, wherein the three-dimensional image of the brain at least includes a three-dimensional image of intracranial pneumosis of the patient.

12. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by the second processor, the steps of the postoperative intracranial pneumatosis segmentation method according to any one of claims 1 to 9 are implemented.

13. A computer program product, characterized in that The computer program product comprises instructions, which, when executed on an electronic device, enable the electronic device to implement the steps of the postoperative intracranial pneumatosis segmentation method according to any one of claims 1 to 9.

Images