WO2017028519A1 - 一种肝脏血管的分类方法 - Google Patents
一种肝脏血管的分类方法 Download PDFInfo
- Publication number
- WO2017028519A1 WO2017028519A1 PCT/CN2016/074386 CN2016074386W WO2017028519A1 WO 2017028519 A1 WO2017028519 A1 WO 2017028519A1 CN 2016074386 W CN2016074386 W CN 2016074386W WO 2017028519 A1 WO2017028519 A1 WO 2017028519A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- blood vessel
- dimensional
- image
- segments
- hepatic
- Prior art date
Links
- 230000002792 vascular Effects 0.000 title claims abstract description 66
- 238000000034 method Methods 0.000 title claims abstract description 41
- 230000002440 hepatic effect Effects 0.000 title abstract description 4
- 210000004204 blood vessel Anatomy 0.000 claims abstract description 318
- 210000004185 liver Anatomy 0.000 claims abstract description 44
- 238000013507 mapping Methods 0.000 claims description 35
- 210000003240 portal vein Anatomy 0.000 claims description 32
- 210000002989 hepatic vein Anatomy 0.000 claims description 27
- 210000002767 hepatic artery Anatomy 0.000 claims description 23
- 238000012937 correction Methods 0.000 claims description 11
- 210000003462 vein Anatomy 0.000 claims description 5
- 238000012545 processing Methods 0.000 abstract description 3
- 238000011282 treatment Methods 0.000 abstract 1
- 230000011218 segmentation Effects 0.000 description 10
- 230000007547 defect Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 230000006872 improvement Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000003902 lesion Effects 0.000 description 3
- 210000000056 organ Anatomy 0.000 description 3
- 210000001835 viscera Anatomy 0.000 description 3
- 206010019695 Hepatic neoplasm Diseases 0.000 description 2
- 210000001367 artery Anatomy 0.000 description 2
- 239000003086 colorant Substances 0.000 description 2
- 238000002059 diagnostic imaging Methods 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 208000014018 liver neoplasm Diseases 0.000 description 2
- 238000002271 resection Methods 0.000 description 2
- 230000003187 abdominal effect Effects 0.000 description 1
- 210000003484 anatomy Anatomy 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000017531 blood circulation Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000002591 computed tomography Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000013506 data mapping Methods 0.000 description 1
- 238000003702 image correction Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/14—Vascular patterns
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2431—Multiple classes
Definitions
- the present invention relates to the field of medical image processing technologies, and in particular, to a method for classifying liver blood vessels.
- the conventional technique uses the conventional abdominal CT (English full name: Computed Tomography) to enhance the three-phase dynamics. Scanning, obtaining three images of arterial phase, portal vein phase and equilibrium phase, and then imaging the hepatic artery, portal vein and hepatic vein by three-dimensional reconstruction technique, analyzing the distribution structure and variation of the three in the liver, for liver segmentation, liver Tumor resection has important guiding significance.
- abdominal CT International full name: Computed Tomography
- an arterial model can be created using the arterial phase image
- a portal vein model can be created using the portal phase image
- a hepatic vein model can be created using the equilibrium image.
- the three-dimensional reconstruction technique is used to image the hepatic artery, portal vein and hepatic vein.
- Three-dimensional vascular images often have overlapping of hepatic artery, portal vein and hepatic vein, as well as other non-vascular tissue.
- medical personnel can only determine the root of the hepatic artery, hepatic vein and portal vein. Then, according to the thickness, connectivity, and shape of the blood vessels, the hepatic artery, hepatic vein, and portal vein are classified. It takes a lot of energy for the medical staff, and the above methods often lose many details of the blood vessels.
- Embodiments of the present invention provide a method for classifying liver blood vessels by re-dividing a plurality of blood vessel segments according to liver blood vessel types, and receiving a classification indication of a blood vessel type of a plurality of blood vessel segments by a user, It is enough to improve the accuracy of blood vessel classification in three-dimensional blood vessel images of liver organs, thereby improving the practical medical reference application value of three-dimensional blood vessel images.
- each blood vessel type is divided into a plurality of blood vessel segments according to a hepatic artery, a portal vein, and a hepatic vein blood vessel type;
- the three-dimensional blood vessel image is updated according to the updated blood vessel type of the plurality of blood vessel segments.
- Advantageous Effects [0013]
- the liver blood vessel classification method provided by the embodiment of the present invention, after acquiring the three-dimensional blood vessel image of the liver organ, separately divides the blood vessel segment for each blood vessel type according to the liver blood vessel type, and receives a plurality of input by the user.
- the classification indication of the vessel type to which the vascular branch belongs provides manual intervention for professional medical users, and classifies the vessel type twice, and can use its own rich professional knowledge to cause existing 3D modeling algorithms and source data.
- the defects can be corrected to improve the accuracy of liver blood vessel classification.
- the program divides multiple blood vessel segments according to three large blood vessel types, which is considered.
- the overall accuracy of the three-dimensional modeling is high, and the defect error usually occurs in the small branch vessel branch.
- the large blood vessel type it is convenient to check the classification of small branch vessels.
- the solution of the embodiment of the present invention further updates the three-dimensional blood vessel image according to the updated blood vessel type of the plurality of blood vessel segments, thereby updating the classification of the three major blood vessels of the liver, thereby improving the practical medical reference application value of the three-dimensional blood vessel image, thereby It is convenient for medical staff to make preoperative estimation and accurate judgment during operation.
- FIG. 1 is a schematic diagram of a method for classifying liver blood vessels according to a first embodiment of the present invention
- FIG. 2 is a schematic diagram of creating a three-dimensional model according to Embodiment 1 of the present invention.
- FIG. 3 is a schematic diagram of a method for correcting a three-dimensional blood vessel image according to Embodiment 2 of the present invention.
- FIG. 4 is a schematic diagram of a method for removing a non-vascular tissue of a three-dimensional blood vessel image according to a third embodiment of the present invention.
- FIG. 5 is a schematic diagram of a method for erroneously connecting a broken blood vessel according to a fourth embodiment of the present invention.
- FIG. 6 is a schematic diagram of mapping of a three-dimensional blood vessel image on a two-dimensional CT image according to Embodiment 2 of the present invention.
- Embodiments of the present invention provide a method for classifying liver blood vessels. Referring to FIG. 1, the method includes the following steps:
- Step 100 Obtain a three-dimensional blood vessel image of a liver organ.
- an arterial blood vessel model can be created using an arterial phase image
- a portal vein model can be created using a portal phase image
- a hepatic vein model can be created using a balanced phase image
- [0029] 100a Performing a blood vessel discrimination on each of the two-dimensional blood vessel images in the two-dimensional CT image according to the region growing algorithm, and obtaining a blood vessel segmentation sequence of each two-dimensional blood vessel image.
- the region growing algorithm (English name: region growing) is a combination of pixels having similar properties to form a region. First select a seed point (x, y) as the growth in the area that needs to be segmented. Point, then merge similar pixels in the field around the seed point into the region to which the seed point pixel belongs according to a predefined rule, and continue the above process as the new seed pixel until the pixel that does not satisfy the condition is included Come in, such an area is generated. After the blood vessel is segmented for each two-dimensional blood vessel image by the region growing algorithm, each two-dimensional blood vessel image can be divided into a plurality of regions, and thus, a blood vessel segmentation sequence of each two-dimensional blood vessel image is obtained. as shown in picture 2,
- the blood vessel segmentation sequence of all CT images in the arterial phase is three-dimensionally reconstructed according to the moving cube algorithm to obtain a three-dimensional blood vessel image of the liver blood vessel in the arterial phase; Reconstructing the image of the venous phase, according to the moving cube algorithm, three-dimensional reconstruction of the blood vessel segmentation sequence of all CT images in the venous phase, obtaining a three-dimensional blood vessel image of the vascular phase of the venous phase; if three-dimensional reconstruction of the portal vein image is required, according to the movement
- the cube algorithm performs three-dimensional reconstruction of the blood vessel segmentation sequence of all CT images in the portal vein phase to obtain a three-dimensional blood vessel image of the liver blood vessels in the portal vein phase.
- Step 102 Obtain three-dimensional images of a hepatic artery, a portal vein, and a hepatic vein in the three-dimensional blood vessel image, respectively
- each blood vessel type usually includes a plurality of blood vessels.
- the three-dimensional images of the hepatic artery, portal vein, and hepatic vein in the three-dimensional blood vessel image were obtained and distinguished.
- the three large types of blood vessels can be distinguished by different colors.
- the hepatic artery is marked in red
- the hepatic vein is marked in green
- the portal vein is marked in blue.
- Step 104 Divide each blood vessel type into a plurality of blood vessel segments according to the hepatic artery, portal vein, and hepatic vein blood vessel types, respectively.
- the overall three-dimensional blood vessel image modeling such as spatial position, volume and the like, achieves a certain degree of accuracy, but there are often some misjudgments at small blood vessel branches. Therefore, according to the large blood vessel type, it is divided again to obtain a plurality of blood vessel segments, wherein the hepatic artery, the portal vein, and the hepatic artery respectively have a plurality of blood vessel segments.
- blood vessels of each blood vessel type are divided into a plurality of blood vessel segments, thereby facilitating observation of small blood vessel segments.
- the plurality of blood vessel segments are arranged separately, and the coloration is the same as the associated large blood vessel type.
- the three-dimensional blood vessel image and the multiple blood vessel segments are respectively displayed in the same interface.
- a three-dimensional blood vessel image is displayed on the left side of the interface, and three different toggle buttons are provided on the right side, each button pairing a plurality of blood vessel branch interfaces of a blood vessel type.
- a corresponding plurality of vessel segments under the divided vessel type are displayed, each vessel segment being in a box.
- Each vessel segment maintains its spatial position throughout the three-dimensional vessel image.
- the vascular branch is classified too much, the consideration according to the spatial position, the trend or the degree of the individual lesion of the patient becomes many and complicated, and it is not easy to divide, but it is easy to be misjudged, and if the blood vessel is The segment is too small, and it is easy to ignore the details, and the purpose of correcting the classification is not achieved.
- each large blood vessel type is divided into 4 to 8 vascular branches, which can It satisfies the purpose of correcting the three-dimensional modeling and meets the medical reference needs.
- the workload of reclassification is also less, and the probability of misjudging is also reduced.
- Step 106 Receive a classification indication of a blood vessel type of the plurality of blood vessel segments input by a user.
- the doctor or other medical staff can use the professional experience and knowledge to re-evaluate the blood vessel branch of the three-dimensional blood vessel image, thereby classifying it into the correct blood vessel type classification.
- a blood vessel segment of the hepatic vein may be close to the hepatic vein, and the vein may not be observed in each two-dimensional CT image due to blood flow, due to the compensation algorithm. For other reasons, it may be classified into the hepatic vein, but it should be classified into the portal vein classification.
- an attribute value may be set for a plurality of blood vessel segments, the attribute value including at least the blood vessel type of the blood vessel branch, and the user may select the blood vessel segment
- the attribute value of the vascular segment is modified, and the vascular type classification indication for the vascular branch is completed.
- the attribute value can be selected by right-clicking the selected vascular segment, and the user can select it through the drop-down list, or can be classified into the corresponding blood vessel by selecting the vascular branch and then holding the left mouse button to drag and drop. Under the type button, the attribute value changes accordingly.
- the corresponding coloration also changes to the color corresponding to the updated blood vessel type classification.
- Step 108 Update the three-dimensional blood vessel image according to the updated blood vessel type classification of the plurality of blood vessel segments.
- the background data of the peers also changes, and the entire three-dimensional blood vessel image model is updated, so that the user can perform other operations based on this, such as simulated resection, segmentation observation and the like.
- the blood vessel branch is divided again for each blood vessel type according to the liver blood vessel type, and the plurality of blood vessel branches input by the user are received.
- the classification indication of the vessel type to which the segment belongs provides manual intervention for professional medical personnel, secondary classification of blood vessel types, and the ability to use its own rich professional knowledge to cause defects in existing 3D modeling algorithms and source data. Can be corrected to improve the accuracy of liver vascular classification.
- the program divides multiple blood vessel segments according to three large blood vessel types, which is considered.
- the overall accuracy of the three-dimensional modeling is high, and the defect error usually occurs in the small branch vessel branch.
- the large blood vessel type it is convenient to check the classification of small branch vessels.
- the solution of the embodiment of the present invention further updates the three-dimensional blood vessel image according to the updated blood vessel type of the plurality of blood vessel segments, thereby updating the classification of the three major blood vessels of the liver, thereby improving the practical medical reference application value of the three-dimensional blood vessel image, thereby It is convenient for medical staff to make preoperative estimation and accurate judgment during operation.
- Embodiment 2 of the present invention is an improvement on the classification method of liver blood vessels based on the first embodiment.
- each blood vessel type is divided into a plurality of blood vessel segments, and further comprising step 103, performing the three-dimensional blood vessel image Two-dimensional image map correction.
- the blood vessel is divided according to the region growing algorithm for each two-dimensional blood vessel image in the first step 100a of the first embodiment
- similar pixels are also classified into the blood vessel area, so the three-dimensional blood vessel image is mapped in each two-dimensional blood vessel image, and the inaccurate position is adjusted to make the reconstructed
- the three-dimensional blood vessel image is more accurate to conform to the actual vascular system of the internal organs, and then proceeds to steps 103a-103d, as shown in FIG.
- Step 103a Acquire a mapping contour on the two-dimensional CT image of the corresponding sequence of the three-dimensional blood vessel image in the set of two-dimensional CT images.
- a set of two-dimensional CT images refers to a two-dimensional data source for the same period used to create a three-dimensional blood vessel image model of the liver.
- the liver three-dimensional blood vessel image model is mapped in order to compare the difference between the two-dimensional data of the three-dimensional blood vessel image and the original two-dimensional data source. Since the three-dimensional blood vessel image is volume data, and the two-dimensional CT image arranges a plurality of images of the same period in a sequence, it can be understood that the diurnal sequence characteristics of the two-dimensional CT image are expressed by three-dimensional volume data.
- a volume of data can be mapped back to a plurality of two-dimensional images. In principle, the mapping of the volume data is performed on each of the two-dimensional images. However, due to imaging and the like, the object organ tissue is not observed on each of the two-dimensional CT images.
- the mapping contour on the two-dimensional CT image of the corresponding sequence in the set of two-dimensional CT images obtained by acquiring the three-dimensional blood vessel image in step 103a may include:
- the three-dimensional blood vessel images are all obtained by three-dimensional reconstruction of a plurality of two-dimensional CT images
- the three-dimensional blood vessel images have corresponding data mappings on the two-dimensional CT images of the corresponding sequences, and the blood vessels in each two-dimensional blood vessel image
- the three-dimensional blood vessel image there are corresponding three-dimensional three-dimensional blood vessel parts, and each three-dimensional blood vessel has its own blood vessel edge line.
- the three-dimensional blood vessel image is mapped back to the two-dimensional CT image, and the map is represented by a closed curve in the figure. contour line.
- the blood vessels in the three-dimensional blood vessel image have respective positional information.
- the location information may be spatial location coordinates. Therefore, the three-dimensional blood vessel image has respective spatial position coordinates in the blood vessel contour line in the corresponding two-dimensional blood vessel image, and the spatial position coordinates of the corresponding blood vessel edge line in the three-dimensional blood vessel image and each two-dimensional blood vessel tissue image are acquired. [0065] 1033a.
- the blood vessel mapping contour of the three-dimensional blood vessel image on each two-dimensional blood vessel image is drawn according to the position relative information of the blood vessel edge line corresponding to each two-dimensional blood vessel image.
- Step 103b Determine whether the mapping contour is coincident with the boundary of the vascular tissue in the two-dimensional CT image of the corresponding sequence in the set of two-dimensional CT images.
- the boundary of the vascular tissue in each two-dimensional blood vessel image can be calculated according to the boundary region algorithm.
- the closed curve shown in Figure 6 maps the three-dimensional blood vessel image back to the mapping contour of a corresponding two-dimensional CT image.
- mapping contour of the three-dimensional model data on this two-dimensional CT image compares whether or not the original tissue boundary of the vascular tissue coincides.
- the judging method can compare the position information of the mapping contour of the three-dimensional model data with the position information of the tissue boundary by means of a computer algorithm, and specifically, can be a comparison of the coordinate position information.
- Step 103c If not coincident, adjust the mapping contour to coincide with the boundary of the blood vessel tissue in the two-dimensional CT image of the corresponding sequence.
- the adjusting method may be that the computer expands or scales the mapping contour to the tissue boundary according to the comparison result of the position information of the two, and changes the range of the mapping contour.
- the calibration instruction that is, the movement information of a point on the contour line is mapped, the point position of the mapping contour is adjusted to the user-specified position.
- Step 103d Perform three-dimensional modeling according to the two-dimensional CT image data corresponding to the adjusted mapping contour to obtain a corrected three-dimensional blood vessel image.
- mapping contour means that the range of the two-dimensional source data is changed, so that the moving cube algorithm is re-executed according to the data range of the two-dimensional CT image corresponding to the adjusted mapping contour.
- Modeling to obtain a corrected three-dimensional blood vessel image It should be noted that the above is an example of three-dimensional blood vessel image correction by adjusting the mapping contour on the two-dimensional CT image mapped to a corresponding sequence, and the change of the mapping contour in each two-dimensional CT image is caused. Correction of the 3D source data range, and thus re-modeling, correcting the 3D vessel model
- multiple 2D CT images can be selected to observe the mapping contour of the 3D model, and the distribution is adjusted to correct the 3D blood vessel image more accurately.
- the embodiment of the present invention adds a step of performing two-dimensional image mapping correction on the three-dimensional blood vessel image based on the first embodiment, and can further improve the accuracy of the three-dimensional modeling based on the beneficial effects brought by the first embodiment.
- Sexuality which provides a more accurate basis for the division of the vascular branches in the next step, and ultimately affects the classification results of the three major types of blood vessels in the liver.
- the embodiment of the present invention is an improvement based on the combination of the first embodiment or the first embodiment and the second embodiment.
- each blood vessel type is divided into a plurality of blood vessel segments according to the hepatic artery, portal vein, and hepatic vein blood vessel type, respectively, or in step 103 of the second embodiment, After the two-dimensional blood vessel image is subjected to two-dimensional image mapping correction, a further step of correcting the three-dimensional blood vessel image is performed, and the non-vascular tissue in the three-dimensional blood vessel image is cleared.
- the embodiment of the present invention is mainly directed to a three-dimensional image of a blood vessel, the non-vascular tissue in the three-dimensional blood vessel image can be first cleared to obtain a relatively accurate three-dimensional blood vessel image.
- the marking instruction is used to indicate that the non-vascular tissue in the three-dimensional blood vessel image is marked out.
- the marking instruction may be a user clicking a mouse or an instruction code related to the marking input by the user.
- a user can identify obvious non-vascular tissues according to the anatomy of the internal organs and the shape of the blood vessels. For these non-vascular tissues, these non-vascular tissues are marked by clicking the mouse. , such as drawing a closed curve to circle these non-vascular tissues, further Clear the non-vascular tissue in the delineated area to obtain a more accurate three-dimensional blood vessel image.
- the step of clearing the non-vascular tissue in the three-dimensional blood vessel image since the step of clearing the non-vascular tissue in the three-dimensional blood vessel image is added, the unnecessary non-target object can be cleared, thereby making the three-dimensional blood vessel image more accurate, and preventing subsequent division of the blood vessel branch is unnecessary.
- the organization is divided, resulting in inaccurate division of the vascular branch, which in turn reduces the accuracy of liver vascular classification.
- the embodiment of the present invention is improved on the basis of the combination of the first embodiment or the first embodiment and the second embodiment. Therefore, the advantages of the first embodiment and the second embodiment are the same, and the details are not described herein again.
- the embodiment of the present invention is based on the foregoing embodiment 1, or the combination of the first embodiment and the second embodiment, or the combination of the first embodiment and the third embodiment, or the combination of the first, second and third embodiments. Improvements based on the basis.
- each blood vessel type is divided into a plurality of blood vessel segments according to the hepatic artery, the portal vein, and the hepatic vein blood vessel type, respectively, or in step 103 of the second embodiment, After the two-dimensional blood vessel image is subjected to two-dimensional image mapping correction, or before the second step 103 of the second embodiment, or before the third step 105, clearing the non-vascular tissue in the three-dimensional blood vessel image, or in the third step 105 of the third embodiment A further step of correcting the three-dimensional blood vessel image is further included, and the vascular connection error position in the three-dimensional blood vessel image is broken.
- step 107 is to optimize the effect of the above embodiment, there is no strict order limitation in the implementation step arrangement, as long as the accuracy of the three-dimensional blood vessel image can be improved, thereby improving the classification of the liver blood vessel.
- the purpose of sex can be.
- step 107 breaking the vascular connection error position in the three-dimensional blood vessel image includes:
- Step 107a Receive a second marking instruction of the user.
- the second marking instruction is used to mark a position of the blood vessel connection error in the three-dimensional blood vessel image.
- the second marking instruction may be a user clicking a mouse, or a user-entered instruction code associated with the marking.
- Step 107b Mark the position of the blood vessel connection error in the three-dimensional blood vessel image according to the second marking instruction.
- Step 107c Receive a second clearing instruction of the user.
- the second clear command is used to instruct to clear a position where the blood vessel marked in the three-dimensional blood vessel image is connected incorrectly.
- Step 108d Clear the position of the blood vessel connection error in the three-dimensional blood vessel image according to the second clearing instruction.
- the user determines the position of the three-dimensional blood vessel image with an incorrect blood vessel connection according to the anatomical knowledge, and identifies the spatial coordinate of the identified position by clicking a mouse, using a sphere knife or other shape, and storing the three-dimensional blood vessel.
- the data set of the image data clears the coordinate values of this part, thus breaking the incorrect connection position.
- the three-dimensional blood vessel images with clear vascular connection errors are classified to provide an accurate reference for subsequent medical operations.
- the embodiment of the present invention increases the processing steps of breaking the vascular connection error position in the three-dimensional blood vessel image, and improves the modeling accuracy of the three-dimensional blood vessel image.
- the vascular connection error may lead to the division of the vascular branch in the division ⁇ , which leads to the classification of the vascular segment reclassification ,, which leads to the inaccurate classification result of the three major vascular types of the liver.
- the increasing step provided by the present embodiment can also improve the accuracy of subsequent segmentation of blood vessel segments and classification of blood vessels.
- the above embodiments of the present invention provide a plurality of methods for classification of liver blood vessels.
- the defects caused by the existing 3D modeling algorithms and source data can be corrected to improve the accuracy of liver blood vessel classification.
- the program divides multiple blood vessel segments according to three large blood vessel types, which is considered.
- the overall accuracy of the three-dimensional modeling is high, and the defect error usually occurs in the small branch vessel branch.
- the large blood vessel type it is convenient to check the classification of small branch vessels.
- the solution of the embodiment of the present invention further updates the three-dimensional blood vessel image according to the updated blood vessel type of the plurality of blood vessel segments, thereby updating the classification of the three major blood vessels of the liver, thereby improving the practical medical reference application value of the three-dimensional blood vessel image, thereby It is convenient for medical staff to make preoperative estimation and accurate judgment during operation.
- the three-dimensional blood vessel image is corrected to make the three-dimensional blood vessel image more accurate, in order to conform to the actual liver blood vessel, and it is convenient for the medical staff to accurately classify the blood vessels in the liver, thereby facilitating the preoperative estimation and accurate judgment of the medical staff.
- Optimizing the three-dimensional blood vessel image can also improve the accuracy of the three-dimensionally modeled blood vessel image and improve the accuracy of the final blood vessel classification result.
- the above two-dimensional image mapping of the three-dimensional blood vessel image, the removal of the non-vascular tissue, or the disconnection of the vascular misconnection position can be regarded as a correction step for the three-dimensional blood vessel model, and can be based on the three-dimensional image.
- the specific defect problems of the model are selected to improve the accuracy of the three-dimensional blood vessel image model, thus laying the foundation for the correct classification of liver blood vessels.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Human Computer Interaction (AREA)
- Vascular Medicine (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
本发明提供了一种肝脏血管的分类方法,通过根据肝脏血管类型分别重新划分多个血管支段,并接收用户对多个血管支段的血管类型的分类指示,根据多个血管支段更新的血管类型分类对三维血管图像进行更新,能够提高对肝脏器官的三维血管图像中血管分类的准确性,从而提高三维血管图像的实际医学参考应用价值。本发明应用于对医疗中的三维建模图像处理。
Description
发明名称:一种肝脏血管的分类方法
[0001] 相关申请的交叉引用
[0002] 本申请要求于 2015年 8月 18日提交中国专利局、 申请号为 201510506303.5、 发明 名称为"一种肝脏血管的分类方法"的中国专利申请的优先权, 其全部内容通过引 用结合在本申请中。
技术领域
[0003] 本发明涉及医疗图像处理技术领域, 尤其涉及一种肝脏血管的分类方法。
背景技术
[0004] 随着医疗技术水平的不断提高, 为了更加精准的获取病人的病变位置, 现有技 术中采用对人体进行常规腹部 CT (英文全称: Computed Tomography中文: 电子 计算机断层扫描) 增强三期动态扫描, 获得动脉期、 门静脉期及平衡期的三期 图像, 进而采用三维重建技术对肝动脉、 门静脉和肝静脉血管成像, 分析三者 在肝内的分布结构以及变异, 对于肝脏分段, 肝脏肿瘤的切除术具有重要的指 导意义。
[0005] 如果提供的 CT图像质量较好, 使用动脉期图像可以创建出动脉血管模型, 使 用门脉期图像可以创建出门静脉模型, 使用平衡期图像创建出肝静脉模型。 但 是, 由于医学成像自身以及病人个体肝脏肿瘤的位置和血管的变异等因素, 导 致实际 CT图像质量并不高, 因此, 使用三维重建技术对肝动脉、 门静脉和肝静 脉血管进行成像吋, 得到的三维血管图像经常会出现肝动脉、 门静脉和肝静脉 重叠, 以及其他非血管组织混杂的现象, 通常情况下医护人员只能根据肝动脉 、 肝静脉和门静脉的解剖结构的不同, 确定出它们的根部, 然后再根据血管的 粗细、 连通性以及走形分类出肝动脉、 肝静脉和门静脉, 需要医护人员花费较 多的精力, 而采样上述方法还会经常丢失许多血管的细节。
技术问题
[0006] 本发明的实施例提供一种肝脏血管的分类方法, 通过根据肝脏血管类型分别重 新划分多个血管支段, 并接收用户对多个血管支段的血管类型的分类指示, 能
够提高对肝脏器官的三维血管图像中血管分类的准确性, 从而提高三维血管图 像的实际医学参考应用价值。
问题的解决方案
技术解决方案
[0007] 为达到上述目的, 本发明的实施例采用如下技术方案:
[0008] 获取肝脏器官的三维血管图像;
[0009] 在三维血管图像中根据肝动脉、 门静脉、 肝静脉血管类型分别标记不同的颜色 以区分;
[0010] 分别根据肝动脉、 门静脉、 肝静脉血管类型将每种血管类型均划分为多个血管 支段;
[0011] 接收用户输入的对多个血管支段的血管类型的分类指示;
[0012] 根据多个血管支段更新后的血管类型分类, 更新三维血管图像。 有益效果 [0013] 本发明实施例提供的肝脏血管分类方法, 在获取肝脏器官的三维血管图像后, 根据肝脏血管类型分别对每种血管类型再次进行血管支段划分, 接收用户输入 的对多个血管支段的所属的血管类型的分类指示, 为专业医学人员用户提供了 人工干预手段, 对血管类型进行二次分类, 能够利用其自身丰富的专业知识对 现有三维建模算法和源数据造成的缺陷能够进行纠正, 从而提高肝脏血管分类 的准确性。
[0014] 同吋, 本方案根据三种大的血管类型分别再划分多个血管支段, 这是考虑, 一 方面三维建模整体准确性较高, 缺陷错误通常出现在小的分支血管支段方面, 因此将大的血管类型进行再次划分, 方便对小的分支血管归类进行检査。
[0015] 本发明实施例方案还根据多个血管支段更新后的血管类型分类更新三维血管图 像, 从而也更新了肝脏三大血管的分类, 提高了三维血管图像的实际医学参考 应用价值, 从而便于医护人员进行术前预估以及术中的精确判断。
发明的有益效果
对附图的简要说明
附图说明
[0016] 为了更清楚地说明本发明实施例或现有技术中的技术方案, 下面将对实施例或 现有技术描述中所需要使用的附图作简单地介绍, 显而易见地, 下面描述中的 附图仅仅是本发明的一些实施例, 对于本领域普通技术人员来讲, 在不付出创 造性劳动的前提下, 还可以根据这些附图获得其他的附图。
[0017] 图 1为本发明实施例一提供的一种肝脏血管的分类方法示意图;
[0018] 图 2为本发明实施例一提供的三维模型创建示意图;
[0019] 图 3为本发明实施例二提供的三维血管图像校正方法示意图;
[0020] 图 4为本发明实施例三提供的三维血管图像的非血管组织清除方法示意图; [0021] 图 5为本发明实施例四提供的断幵血管错误连接位置方法示意图;
[0022] 图 6为本发明实施例二提供的三维血管图像在二维 CT图像上的映射示意图。
本发明的实施方式
[0023] 下面将结合本发明实施例中的附图, 对本发明实施例中的技术方案进行清楚、 完整地描述, 显然, 所描述的实施例仅仅是本发明一部分实施例, 而不是全部 的实施例。 基于本发明中的实施例, 本领域普通技术人员在没有做出创造性劳 动前提下所获得的所有其他实施例, 都属于本发明保护的范围。
[0024] 实施例一、
[0025] 本发明的实施例提供一种肝脏血管的分类方法, 参照图 1, 该方法包括以下步 骤:
[0026] 步骤 100、 获取肝脏器官的三维血管图像。
[0027] 其中, 三维血管图像是根据一组处于同期的一组二维 CT图像进行建模得到的 。 比如背景技术中所述, 使用动脉期图像可以创建出动脉血管模型, 使用门脉 期图像可以创建出门静脉模型, 使用平衡期图像创建出肝静脉模型。
[0028] 具体的, 对于步骤 100中获取肝脏器官的三维血管图像具体包括如图 2所示,
[0029] 100a. 根据区域生长算法对一组二维 CT图像中的每个二维血管图像进行血管分 害 |J, 得到每个二维血管图像的血管分割序列。
[0030] 其中, 区域生长算法 (英文全称: region growing) 是将具有相似性质的像素集 合起来构成区域。 首先在需要分割的区域中选取一个种子点 (x, y) 作为生长
点, 然后根据预先定义的规则将种子点周围领域中的相似像素合并到种子点像 素所属的区域中, 将这些新像素作为新的种子像素继续进行上述过程, 直到没 有满足条件的像素点被包括进来, 这样一个区域就生成完成。 通过区域生长算 法对每个二维血管图像进行血管分割后, 可以将每个二维血管图像分割成多个 区域, 因而, 得到每个二维血管图像的血管分割序列。 如图 2所示,
[0031] 100b、 根据移动立方体 (英文全称: Marching Cubes, 简称 MC) 算法对所有二 维血管图像的血管分割序列进行三维重建, 得到三维血管图像。
[0032] 示例的, 若需要三维重建动脉期的图像, 则根据移动立方体算法, 对动脉期的 所有 CT图像的血管分割序列进行三维重建, 得到动脉期的肝脏血管的三维血管 图像; 若需要三维重建静脉期的图像, 则根据移动立方体算法, 对静脉期的所 有 CT图像的血管分割序列进行三维重建, 得到静脉期的肝脏血管的三维血管图 像; 若需要三维重建门静脉期的图像, 则根据移动立方体算法, 对门静脉期的 所有 CT图像的血管分割序列进行三维重建, 得到门静脉期的肝脏血管的三维血 管图像。
[0033] 步骤 102、 分别获取所述三维血管图像中肝动脉、 门静脉、 肝静脉的三维图像
[0034] 具体地, 在建模好的三维血管图像上, 根据现有的一些血管识别算法, 可以区 分出肝动脉、 门静脉、 肝静脉三大类型的血管, 每个血管类型通常包括多条血 管。 分别获取三维血管图像中肝动脉、 门静脉、 肝静脉的三维图形, 并加以区 分。 为了更加清楚的区分, 这三种大的种类的血管可以分别着不同的颜色以进 行区分。
[0035] 比如, 肝动脉用红色标记, 肝静脉用绿色标记, 门静脉用蓝色标记。
[0036] 步骤 104、 分别根据所述肝动脉、 门静脉、 肝静脉血管类型将每种血管类型均 划分为多个血管支段。
[0037] 基于目前计算机三维建模的技术, 整体三维血管图像的建模, 比如在空间位置 , 体积等方面准确性达到了一定程度, 但在小的血管分支处往往出现一些误判 。 因此根据大的血管类型进行再次划分, 得到多个血管支段, 其中, 肝动脉、 门静脉、 肝动脉分别具有多个血管支段。
[0038] 具体地, 将每种血管类型的血管划分为多个血管支段, 从而便于观察小的血管 支段。 该多个血管支段分别排列显示, 着色与所属的大的血管类型相同。 为方 便对比观察, 提高互动性, 三维血管图像和多个血管支段分别显示在同一个界 面中。 比如, 三维血管图像显示在界面左侧, 右侧设置三个不同的切换按钮, 每个按钮对已一种血管类型的多个血管支段界面。 选中其中一个血管类型按钮 后, 就显示对应的已划分好的该血管类型下的多个血管支段, 每个血管支段处 于一个方框中。 每个血管支段保持其在整个三维血管图像中的空间位置。
[0039] 进一步地, 如果对血管支段分类过多, 根据其空间位置, 走向或者病人个体病 变程度的考虑因素就会变得很多且复杂, 不容易划分, 反而容易错判, 而如果 对血管支段分段过少, 又容易忽略细节, 达不到校正分类的目的。
[0040] 因此根据肝脏血管的空间位置, 医生对重要血管支段的关注程度, 结合解剖学 原理, 在实际操作中对每个大的血管类型种类划分为 4到 8条血管支段, 既能满 足对三维建模校正的目的, 满足医学上的参考需要, 同吋, 对于医生来说, 重 新分类的工作量也较少, 判断失误的概率也降低。
[0041] 步骤 106、 接收用户输入的对所述多个血管支段的血管类型的分类指示。
[0042] 由于医学领域的专业性较强, 病人的个体差异性也比较大, 医学成像不具有绝 对的准确性, 因此在现实应用吋, 如果能够同吋借助医护人员的丰富的专业知 识和经验对三维模型进行校正, 可以达到建模更加准确的目的。
[0043] 在本发明实施例步骤中, 医生或其他医护人员可以借助专业经验和知识对三维 血管图像的血管分支进行再次判断, 从而将其归入正确的血管类型分类中。
[0044] 具体地, 比如肝静脉的一个血管支段可能距离肝静脉血管很近, 而静脉由于血 液流动原因, 在扫描成像吋不一定能够在每张二维 CT图像中都能观察到, 由于 补偿算法等原因, 可能被归入了肝静脉, 而实际应当分到门静脉分类中。
[0045] 在本发明实施例方案中, 针对这种情况, 可以对多个血管支段设置属性值, 该 属性值至少包括该血管支段所述的血管类型, 用户可以选中该血管支段后对该 血管支段的属性值进行更改, 于是完成对该血管支段的血管类型分类指示。
[0046] 属性值可以通过选中血管支段后点击右键出现, 用户通过下拉列表选择, 也可 以通过选中血管支段后按住鼠标左键进行拖拽的方式将其归类到对应的血管类
型按钮下, 进而属性值也随之改变。
[0047] 当血管支段的属性值改变后, 其对应的着色也改变为更新后的血管类型分类对 应的颜色。
[0048] 步骤 108、 根据所述多个血管支段更新后的血管类型分类, 更新所述三维血管 图像。
[0049] 当用户对血管支段的分类进行改变后, 血管支段的着色发生变化, 同吋该血管 支段在三维血管图像界面中同吋发生改变, 着色变为更新分类后血管类型对应 的颜色, 方便用户实吋观察判断。
[0050] 同吋后台数据也随之改变, 将整个三维血管图像模型进行了更新, 从而便于用 户以此为依据进行其他的操作, 比如模拟切除, 分割观察等手术辅助措施。
[0051] 本发明实施例提供的肝脏血管分类方法, 在获取肝脏器官的三维血管图像后, 根据肝脏血管类型分别对每种血管类型再次进行血管支段划分, 接收用户输入 的对多个血管支段的所属的血管类型的分类指示, 为专业医学人员用户提供了 人工干预手段, 对血管类型进行二次分类, 能够利用其自身丰富的专业知识对 现有三维建模算法和源数据造成的缺陷能够进行纠正, 从而提高肝脏血管分类 的准确性。
[0052] 同吋, 本方案根据三种大的血管类型分别再划分多个血管支段, 这是考虑, 一 方面三维建模整体准确性较高, 缺陷错误通常出现在小的分支血管支段方面, 因此将大的血管类型进行再次划分, 方便对小的分支血管归类进行检査。
[0053] 本发明实施例方案还根据多个血管支段更新后的血管类型分类更新三维血管图 像, 从而也更新了肝脏三大血管的分类, 提高了三维血管图像的实际医学参考 应用价值, 从而便于医护人员进行术前预估以及术中的精确判断。
[0054] 实施例二
[0055] 本发明实施例二是在实施例一基础上对肝脏血管的分类方法的改进。
[0056] 具体地, 在步骤 104、 分别根据所述肝动脉、 门静脉、 肝静脉血管类型将每种 血管类型均划分为多个血管支段之前, 还包括步骤 103、 将所述三维血管图像进 行二维图像映射校正。
[0057] 由于在实施例一步骤 100a中对每个二维血管图像根据区域生长算法进行血管分
割吋, 有吋会把不属于血管区域的相似像素点也归类到血管区域, 因此, 将三 维血管图像在每个二维血管图像进行映射, 对不准确的位置进行调整, 使得重 建后的三维血管图像更加准确, 以符合实际的内脏器官的血管系统, 进而继续 执行步骤 103a-103d, 如图 3所示。
[0058] 步骤 103a、 获取所述三维血管图像在所述一组二维 CT图像中对应序列的二维 C T图像上的映射轮廓线。
[0059] 其中, 一组二维 CT图像是指用于创建肝脏三维血管图像模型所使用的同一期 的二维数据源。 将肝脏三维血管图像模型进行映射, 目的是为了比对生成三维 血管图像的二维数据与原始的二维数据源的差别。 由于三维血管图像是体数据 , 而二维 CT图像将同一期的多张图像按照序列排列, 也可以理解成, 将二维 CT 图像的吋间序列特性用三维体数据表现出来。 从而一个体数据可以映射回这一 组多张二维图像上, 原则上每张二维图像上都会有这个体数据的映射, 但由于 成像等原因, 不是每张二维 CT图像上都能观察到对象器官组织。
[0060] 对于步骤 103a中获取所述三维血管图像在所述一组二维 CT图像中对应序列的二 维 CT图像上的映射轮廓线可以包括:
[0061] 1031a. 获取三维血管图像在每个二维 CT图像中对应的血管边缘线。
[0062] 由于三维血管图像都是由多张二维 CT图像进行三维重建得到的, 因此, 三维 血管图像在对应序列的二维 CT图像上会具有相应的数据映射, 每个二维血管图 像中的血管在三维血管图像都有对应的立体三维血管部分, 每个三维血管都有 各自的血管边缘线, 如图 6所示, 三维血管图像映射回二维 CT图像, 图中以闭合 曲线的方式表示映射轮廓线。
[0063] 1032a. 获取与三维血管图像对应序列的每个二维 CT图像中血管组织对应的血 管边缘线的位置信息。
[0064] 由于三维血管图像是一个三维立体的图像, 因此, 三维血管图像中血管都有各 自的位置信息。 在具体的实现中, 该位置信息可以为空间位置坐标。 因此, 三 维血管图像在对应的二维血管图像中的血管轮廓线都有各自的空间位置坐标, 获取三维血管图像与每个二维血管组织图像中对应的血管边缘线的空间位置坐 标。
[0065] 1033a. 根据与每个二维血管图像对应的血管边缘线的位置相对信息, 绘制得 到三维血管图像在每个二维血管图像上的血管映射轮廓线。
[0066] 步骤 103b、 判断所述映射轮廓线是否与所述一组二维 CT图像中的对应序列的 二维 CT图像中血管组织的边界是否重合。
[0067] 也就是比较三维血管图像映射回对应序列的二维 CT图像的数据范围与原始二 维 CT图像的数据范围的差别。
[0068] 每个二维血管图像中血管组织的边界可以根据边界区域算法计算得到。 如图 6 所示的闭合曲线为三维血管图像映射回某一对应序列的二维 CT图像的映射轮廓 线。
[0069] 判断比较在这张二维 CT图像上三维模型数据的映射轮廓线是否血管组织的原 始组织边界与是否重合。
[0070] 判断方法可以借助于计算机算法, 对三维模型数据的映射轮廓线的位置信息和 组织边界的位置信息进行比较, 具体地, 可以为坐标位置信息的比较。
[0071] 以及, 也可以是用户人工的观察。
[0072] 步骤 103c、 若不重合, 调整所述映射轮廓线与所述对应序列的二维 CT图像中血 管组织的边界重合。
[0073] 具体地, 调整方法可以是计算机根据两者位置信息的比较结果, 将映射轮廓线 外扩或者缩放至组织边界处, 改变映射轮廓线的范围。
[0074] 或者, 也可以借助人工干预的方法。 根据用户的操作, 获取校准指令,
[0075] 根据校准指令, 即映射轮廓线上某一点的移动信息, 将映射轮廓线的该点位置 调整至用户指定位置。
[0076] 重复上述过程, 将映射轮廓线调整与对应的二维血管图像中血管组织的边界重 合。
[0077] 步骤 103d、 根据所述调整后的映射轮廓线对应的二维 CT图像数据重新三维建 模, 得到校正后的三维血管图像。
[0078] 具体的, 映射轮廓线被调整后就意味着二维源数据的范围发生了变化, 从而根 据调整后的映射轮廓线所对应的二维 CT图像的数据范围重新进行移动立方体算 法进行三维建模, 得到校正后的三维血管图像。
[0079] 需要说明的是, 上述是以映射回一张对应序列的二维 CT图像上映射轮廓线的 调整进行三维血管图像校正的举例, 每一张二维 CT图像中的映射轮廓线的变化 都会引起二维源数据范围的改变, 从而重新建模后, 对三维血管模型进行校正
。 也就是, 可以选取多张二维 CT图像, 观察三维模型的映射轮廓线, 分布进行 调整, 从而比较精确的对三维血管图像进行校正。
[0080] 本发明实施例方案在实施例一方案基础上增加了将三维血管图像进行二维图像 映射校正的步骤, 在实施例一带来的有益效果基础上, 能够进一步地提高三维 建模的准确性, 从而为下一步血管支段的划分提供更加准确的依据, 并最终影 响肝脏中三大类血管的分类结果。
[0081] 实施例三
[0082] 本发明实施例是在实施例一或者实施例一和实施例二结合的基础上的改进。
[0083] 具体地, 在实施例一步骤 104、 分别根据所述肝动脉、 门静脉、 肝静脉血管类 型将每种血管类型均划分为多个血管支段之前, 或者在实施例二步骤 103、 将所 述三维血管图像进行二维图像映射校正之后, 还包括另一对三维血管图像的校 正步骤 105、 清除所述三维血管图像中的非血管组织。
[0084] 由于本发明实施例主要针对血管的三维图像, 因此, 可以将三维血管图像中的 非血管组织先进行清除, 以便得到较为准确的三维血管图像。
[0085] 对于如何清除三维血管图像中的非血管组织, 如图 4所示, 具体包括:
[0086] 105a. 接收用户的标记指令。
[0087] 其中, 标记指令用于指示将三维血管图像中的非血管组织标记出来。 其中, 标 记指令可以是用户点击鼠标, 也可以是用户输入的与标记相关的指令代码。
[0088] 105b. 根据标记指令, 标记出三维血管图像中的非血管组织。
[0089] 105c 接收用户的清除指令。
[0090] 其中, 清除指令用于指示将三维血管图像中标记出的非血管组织清除。
[0091] 105d、 根据清除指令, 清除三维血管图像中的非血管组织。
[0092] 示例的, 用户特别是专业的医护人员可以根据内脏器官的解剖结构以及血管的 走形, 确认出明显的非血管组织, 对于这些非血管组织, 通过点击鼠标, 标记 出这些非血管组织, 如绘制闭合的曲线将这些非血管组织圈定出来, 进一步的
, 对于圈定区域的非血管组织进行清除, 得到较为准确的三维血管图像。
[0093] 从而本发明实施例由于增加了清除三维血管图像中非血管组织的步骤, 能够将 多余非目标对象清除, 从而使得三维血管图像更加准确, 防止后续对血管支段 划分吋将不必要的组织划入, 造成血管支段划分的不准确, 进而降低肝脏血管 分类的准确性。
[0094] 本发明实施例是在实施例一或者实施例一和二结合的基础上进行改进的, 因此 同吋兼具实施例一和实施例二的有益效果, 在此不再赘述。
[0095] 实施例四
[0096] 本发明实施例是在上述实施例一基础上, 或者实施例一与实施例二结合的基础 上, 或者实施例一、 三结合的基础上, 或者实施例一、 二、 三结合的基础上所 做的改进。
[0097] 具体地, 在实施例一步骤 104、 分别根据所述肝动脉, 门静脉, 肝静脉血管类 型将每种血管类型均划分为多个血管支段之前, 或者在实施例二步骤 103、 将所 述三维血管图像进行二维图像映射校正之后, 或者在实施例二步骤 103之前, 或 者在实施例三步骤 105、 清除所述三维血管图像中的非血管组织之前, 或者在实 施例三步骤 105之后还包括另一对三维血管图像的校正步骤 107、 断幵所述三维 血管图像中血管连接错误位置。
[0098] 在本发明实施例中, 由于步骤 107是对上述实施例效果的优化, 在实施步骤排 列上没有严格的顺序限制, 只要能够达到提高三维血管图像的准确性, 进而提 高肝脏血管分类准确性的目的即可。
[0099] 在三维建模中, 由于所有组织器官在二维 CT源图像中并不一定每张都可以观 察到, 有的静脉扫描图像会吋隐吋现, 因此三维血管图像中的血管也并不是完 全连通的, 会出现几个分支, 本发明实施例利用这些不连通的分支进行动脉、 静脉和门脉的分类, 分类后的血管分支保持了原来血管的解剖学位置, 能清晰 的看出三者在内脏器官的分布结构、 变异情况以及与病灶的毗邻关系。 在实际 情况中, 由于三维血管建模过程中算法的原因, 可能对二维 CT源图像中组织边 界的误判, 或者补偿算法的不确定性等多方面的因素, 用户在进行分类之前或 者分类之后或者分类过程中, 根据解剖学知识, 发现一些动脉、 静脉和门脉相
互混杂在一起, 因此, 就需要对这些连接错误的位置进行清除。
[0100] 具体的, 如图 5所示, 步骤 107、 断幵所述三维血管图像中血管连接错误位置包 括:
[0101] 步骤 107a、 接收用户的第二标记指令。
[0102] 其中, 第二标记指令用于将三维血管图像中血管连接错误的位置进行标记。 第 二标记指令可以是用户点击鼠标, 也可以是用户输入的与标记相关的指令代码
[0103] 步骤 107b、 根据第二标记指令, 标记出三维血管图像中血管连接错误的位置。
[0104] 步骤 107c、 接收用户的第二清除指令。
[0105] 其中, 第二清除指令用于指示将三维血管图像中标记出的血管连接错误的位置 进行清除。
[0106] 步骤 108d、 根据第二清除指令, 清除三维血管图像中血管连接错误的位置。
[0107] 从而实现血管错误连接位置的断幵。
[0108] 示例的, 用户根据解剖学知识确定三维血管图像中有血管连接不正确的位置, 通过点击鼠标, 用球体刀或者其他形状进行标识, 获取标识出的位置的空间坐 标, 在存储三维血管图像数据的数据集中把这部分的坐标值清空, 这样就断幵 了不正确的连接位置。 进而对清除血管连接错误的三维血管图像进行分类, 为 后续的医学手术提供准确的参考。
[0109] 在上述实施例一、 二、 三的基础上, 本发明实施例增加了断幵所述三维血管图 像中血管连接错误位置的处理步骤, 也提高了三维血管图像的建模准确性, 如 果血管连接错误, 有可能导致血管支段在划分吋划分错误, 从而导致血管支段 重新分类吋的分类错误, 进而导致最终的肝脏三大血管类型的分类结果不准确 。 从而本实施例提供的增加步骤也能够提高后续血管支段划分和血管分类的准 确度。
[0110] 本发明实施例是在实施例一或者实施例一和二结合的基础上进行改进得到的, 因此同吋兼具实施例一、 实施例二、 实施例三的有益效果, 在此不再赘述。
[0111] 综上, 本发明上述实施例提供了多种肝脏血管分类的方法, 首先在获取肝脏器 官的三维血管图像后, 根据肝脏血管类型分别对每种血管类型再次进行血管支
段划分, 接收用户输入的对多个血管支段的所属的血管类型的分类指示, 为专 业医学人员用户提供了人工干预手段, 对血管类型进行二次分类, 能够利用其 自身丰富的专业知识对现有三维建模算法和源数据造成的缺陷能够进行纠正, 从而提高肝脏血管分类的准确性。
[0112] 同吋, 本方案根据三种大的血管类型分别再划分多个血管支段, 这是考虑, 一 方面三维建模整体准确性较高, 缺陷错误通常出现在小的分支血管支段方面, 因此将大的血管类型进行再次划分, 方便对小的分支血管归类进行检査。
[0113] 本发明实施例方案还根据多个血管支段更新后的血管类型分类更新三维血管图 像, 从而也更新了肝脏三大血管的分类, 提高了三维血管图像的实际医学参考 应用价值, 从而便于医护人员进行术前预估以及术中的精确判断。
[0114] 以及, 在进行血管支段的划分之前还包括对三维血管图像进行二维图像映射校 正, 通过比较三维血管图像的在对应序列的二维 CT图像上的映射轮廓线与二维 血管图像中血管组织的边界线是否重合, 若不重合, 则对血管映射轮廓线进行 调整, 从而改变了三维重建模型所使用的二维数据范畴, 使更接近实际的血管 组织源数据, 进行重建后起到了三维血管图像进行校正的作用, 使得三维血管 图像更加准确, 以符合实际的肝脏血管, 便于医护人员对肝脏内的血管进行准 确分类, 进而便于医护人员进行术前预估以及术中的精确判断。
[0115] 以及, 在进行血管支段的划分之前或者在三维血管图像进行二维图像映射校正 之后, 还包括对三维血管图像中的非血管组织进行清除, 从而能够刪除一些非 必要的组织部分, 优化三维血管图像, 也能够提高三维建模的血管图像的准确 性, 能够提高最终的血管分类结果的准确性。
[0116] 以及, 在进行血管支段的划分之前, 或者在三维血管图像进行二维图像映射校 正之后, 或者之后, 或者在对三维血管图像中的非血管组织进行清除之前, 或 者之后, 还包括断幵三维血管图像中血管连接的错误位置, 从而能够对错误的 三维建模部位进行纠正, 也优化了三维血管图像, 从而提高血管分类结果的准 确性。
[0117] 上述无论是三维血管图像的二维图像映射, 非血管组织的清除, 还是血管错误 连接位置的断幵, 都可以视为对三维血管模型的校正步骤, 可以根据三维图像
模型的具体缺陷问题选择适用, 均是为了提高三维血管图像模型的准确性, 从 而奠定肝脏血管正确分类的基础。
[0118] 本领域普通技术人员可以理解: 实现上述方法实施例的全部或部分步骤可以通 过程序指令相关的硬件来完成, 前述的程序可以存储于一计算机可读取存储介 质中, 该程序在执行吋, 执行包括上述方法实施例的步骤; 而前述的存储介质 包括: ROM、 RAM. 磁碟或者光盘等各种可以存储程序代码的介质。
[0119] 以上所述, 仅为本发明的具体实施方式, 但本发明的保护范围并不局限于此, 任何熟悉本技术领域的技术人员在本发明揭露的技术范围内, 可轻易想到变化 或替换, 都应涵盖在本发明的保护范围之内。 因此, 本发明的保护范围应以所 述权利要求的保护范围为准。
Claims
[权利要求 1] 一种肝脏血管的分类方法, 其特征在于, 包括:
获取肝脏器官的三维血管图像;
分别获取所述三维血管图像中肝动脉、 门静脉、 肝静脉的三维图像; 分别根据所述肝动脉、 门静脉、 肝静脉血管类型将每种血管类型均划 分为多个血管支段;
接收用户输入的对所述多个血管支段的血管类型的分类指示; 根据所述多个血管支段更新后的血管类型分类, 更新所述三维血管图
[权利要求 2] 根据权利要求 1所述的方法, 其特征在于,
所述三维血管图像和所述多个血管支段分别显示在同一界面中。
[权利要求 3] 根据权利要求 1所述的方法, 其特征在于, 在所述分别根据所述肝动 脉、 门静脉、 肝静脉血管类型将每种血管类型均划分为多个血管支段 之前, 还包括:
将所述三维血管图像进行二维图像映射校正。
[权利要求 4] 根据权利要求 3所述的方法, 其特征在于, 所述将所述三维血管图像 进行二维图像映射校正包括:
获取所述三维血管图像在所述一组二维 CT图像中对应序列的二维 CT 图像上的映射轮廓线;
判断所述映射轮廓线是否与所述一组二维 CT图像中的对应序列的二 维 CT图像中血管组织的边界重合;
若不重合, 则调整所述映射轮廓线与所述对应序列的二维 CT图像中 血管组织的边界重合;
根据所述调整后的映射轮廓线对应的二维 CT图像数据重新三维建模 , 得到校正后的三维血管图像。
[权利要求 5] 根据权利要求 1所述的方法, 其特征在于, 在所述分别根据所述肝动 脉、 门静脉、 肝静脉血管类型将每种血管类型均划分为多个血管支段 之前, 还包括:
清除所述三维血管图像中的非血管组织。
根据权利要求 5所述的方法, 其特征在于, 所述清除所述三维血管图 像中的非血管组织包括:
接收用户的标记指令;
根据所述标记指令, 标记出所述三维血管图像中的所述非血管组织; 接收所述用户的清除指令;
根据所述清除指令, 清除所述三维血管图像中的非血管组织。
根据权利要求 1所述的方法, 其特征在于, 在所述分别根据所述肝动 脉、 门静脉、 肝静脉血管类型将每种血管类型均划分为多个血管支段 之前, 还包括:
断幵所述三维血管图像中血管连接错误位置。
根据权利要求 7所述的方法, 其特征在于, 所述断幵所述三维血管图 像中血管连接错误位置包括:
接收所述用户的第二标记指令;
根据所述第二标记指令, 标记出所述三维血管图像中血管连接错误的 位置;
接收所述用户的第二清除指令;
根据所述第二清除指令, 清除所述三维血管图像中血管连接错误的位 置。
根据权利要求 1所述的方法, 其特征在于, 所述分别根据所述肝动脉 、 门静脉、 肝静脉血管类型将每种血管类型均划分为多个血管支段具 体包括: 分别根据所述肝动脉、 门静脉、 肝静脉血管类型将每种血管 类型均划分为 4到 8个血管支段。
根据权利要求 2或 3所述的方法, 其特征在于, 所述接收用户输入的对 所述多个血管支段的血管类型的分类指示包括:
所述多个血管支段具有属性值, 所述属性值至少包括所述每个血管支 段所属的血管类型;
接收用户输入的对所述多个血管支段的属性值的更新信息以获取对所
述多个血管支段的血管类型的分类指示。
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510506303.5A CN105139030A (zh) | 2015-08-18 | 2015-08-18 | 一种肝脏血管的分类方法 |
CN201510506303.5 | 2015-08-18 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2017028519A1 true WO2017028519A1 (zh) | 2017-02-23 |
Family
ID=54724373
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2016/074386 WO2017028519A1 (zh) | 2015-08-18 | 2016-02-23 | 一种肝脏血管的分类方法 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN105139030A (zh) |
WO (1) | WO2017028519A1 (zh) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111369662A (zh) * | 2018-12-25 | 2020-07-03 | 罗园明 | Ct图像中血管的三维模型重建方法及系统 |
CN112489051A (zh) * | 2020-11-13 | 2021-03-12 | 深圳市旭东数字医学影像技术有限公司 | 基于血管及病变区域的肝脏裁切方法及系统 |
CN112842371A (zh) * | 2021-01-29 | 2021-05-28 | 上海商汤智能科技有限公司 | 图像处理方法、装置、电子设备及存储介质 |
CN117635950A (zh) * | 2024-01-04 | 2024-03-01 | 广东欧谱曼迪科技股份有限公司 | 血管分割矫正处理的方法、装置、电子设备及存储介质 |
Families Citing this family (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105139030A (zh) * | 2015-08-18 | 2015-12-09 | 青岛海信医疗设备股份有限公司 | 一种肝脏血管的分类方法 |
CN107784038B (zh) * | 2016-08-31 | 2021-03-19 | 法法汽车(中国)有限公司 | 一种传感器数据的标注方法 |
CN106780728B (zh) * | 2017-02-28 | 2020-05-01 | 成都金盘电子科大多媒体技术有限公司 | 一种基于医学影像的单个器官拆分方法及系统 |
CN106897564B (zh) * | 2017-02-28 | 2019-12-03 | 成都金盘电子科大多媒体技术有限公司 | 一种基于医学影像的病案图像建模系统 |
CN108937978A (zh) * | 2018-04-20 | 2018-12-07 | 重庆大学 | 一种血流全循环成像算法的分层显示方法 |
CN109758136A (zh) * | 2019-02-28 | 2019-05-17 | 四川大学华西医院 | 基于门脉血管特征的肝静脉压力梯度测量方法 |
CN109949322B (zh) * | 2019-03-27 | 2023-06-23 | 中山大学 | 一种基于磁共振t1增强图像的脑血管图像分割方法 |
CN111311583B (zh) * | 2020-02-24 | 2021-03-12 | 广州柏视医疗科技有限公司 | 肺气管和血管的分段命名方法 |
CN111666972A (zh) * | 2020-04-28 | 2020-09-15 | 清华大学 | 基于深度神经网络的肝部病例图像分类方法及系统 |
CN112826535B (zh) * | 2020-12-31 | 2022-09-09 | 青岛海信医疗设备股份有限公司 | 一种超声成像中自动定位血管的方法和装置及设备 |
CN113096066B (zh) * | 2021-02-26 | 2023-02-10 | 杭州臻合健康科技有限公司 | 一种基于智能肝门静脉的分型方法 |
CN113763331A (zh) * | 2021-08-17 | 2021-12-07 | 北京医准智能科技有限公司 | 冠状动脉优势型判定方法、装置、存储介质及电子设备 |
CN115760961B (zh) * | 2021-09-01 | 2023-08-29 | 数坤(北京)网络科技股份有限公司 | 血管图像处理方法、装置、服务器及后处理图像生成系统 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101203890A (zh) * | 2005-06-24 | 2008-06-18 | 美国医软科技公司 | 交互式肝病诊断方法 |
CN101896939A (zh) * | 2007-12-14 | 2010-11-24 | 皇家飞利浦电子股份有限公司 | 标记被分割对象 |
CN102117378A (zh) * | 2009-12-31 | 2011-07-06 | 苏州瑞派宁科技有限公司 | 一种基于三维多模影像的肝肿瘤综合手术规划模拟方法及其系统 |
CN102165465A (zh) * | 2008-09-23 | 2011-08-24 | 美国医软科技公司 | 用于在医学影像中交互式标记管状结构的方法 |
CN105139030A (zh) * | 2015-08-18 | 2015-12-09 | 青岛海信医疗设备股份有限公司 | 一种肝脏血管的分类方法 |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102048550B (zh) * | 2009-11-02 | 2013-07-17 | 上海交通大学医学院附属仁济医院 | 一种自动生成肝脏3d图像并准确定位肝脏血管支配区域的方法 |
CN106355582B (zh) * | 2012-11-30 | 2019-02-15 | 河北大学 | 一种基于形状先验的肝脏医学图像分割的方法 |
CN104463860A (zh) * | 2014-12-01 | 2015-03-25 | 冯炳和 | 一种肝脏分段的方法 |
-
2015
- 2015-08-18 CN CN201510506303.5A patent/CN105139030A/zh active Pending
-
2016
- 2016-02-23 WO PCT/CN2016/074386 patent/WO2017028519A1/zh active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101203890A (zh) * | 2005-06-24 | 2008-06-18 | 美国医软科技公司 | 交互式肝病诊断方法 |
CN101896939A (zh) * | 2007-12-14 | 2010-11-24 | 皇家飞利浦电子股份有限公司 | 标记被分割对象 |
CN102165465A (zh) * | 2008-09-23 | 2011-08-24 | 美国医软科技公司 | 用于在医学影像中交互式标记管状结构的方法 |
CN102117378A (zh) * | 2009-12-31 | 2011-07-06 | 苏州瑞派宁科技有限公司 | 一种基于三维多模影像的肝肿瘤综合手术规划模拟方法及其系统 |
CN105139030A (zh) * | 2015-08-18 | 2015-12-09 | 青岛海信医疗设备股份有限公司 | 一种肝脏血管的分类方法 |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111369662A (zh) * | 2018-12-25 | 2020-07-03 | 罗园明 | Ct图像中血管的三维模型重建方法及系统 |
CN112489051A (zh) * | 2020-11-13 | 2021-03-12 | 深圳市旭东数字医学影像技术有限公司 | 基于血管及病变区域的肝脏裁切方法及系统 |
CN112489051B (zh) * | 2020-11-13 | 2024-04-05 | 深圳市旭东数字医学影像技术有限公司 | 基于血管及病变区域的肝脏裁切方法及系统 |
CN112842371A (zh) * | 2021-01-29 | 2021-05-28 | 上海商汤智能科技有限公司 | 图像处理方法、装置、电子设备及存储介质 |
CN117635950A (zh) * | 2024-01-04 | 2024-03-01 | 广东欧谱曼迪科技股份有限公司 | 血管分割矫正处理的方法、装置、电子设备及存储介质 |
CN117635950B (zh) * | 2024-01-04 | 2024-04-09 | 广东欧谱曼迪科技股份有限公司 | 血管分割矫正处理的方法、装置、电子设备及存储介质 |
Also Published As
Publication number | Publication date |
---|---|
CN105139030A (zh) | 2015-12-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2017028519A1 (zh) | 一种肝脏血管的分类方法 | |
Tobon-Gomez et al. | Benchmark for algorithms segmenting the left atrium from 3D CT and MRI datasets | |
CN108133478B (zh) | 一种提取冠状动脉血管中心线的方法 | |
JP4824146B1 (ja) | 医療機器および医療機器の制御方法 | |
US7840044B2 (en) | Method and system for liver lobe segmentation and pre-operative surgical planning | |
CN106097305B (zh) | 双行程区域生长结合形态学重建的肺部气管树分割方法 | |
US9811912B2 (en) | Image processing apparatus, image processing method and medical imaging device | |
EP2157905B1 (en) | A method for tracking 3d anatomical and pathological changes in tubular-shaped anatomical structures | |
US10235753B2 (en) | Automatic recognition of anatomical landmarks | |
CN103810752B (zh) | 基于医学图像的肝脏分段方法及其肝脏分段系统 | |
JP2011098195A (ja) | 構造物検出装置および方法ならびにプログラム | |
US20110052026A1 (en) | Method and Apparatus for Determining Angulation of C-Arm Image Acquisition System for Aortic Valve Implantation | |
CN114332013B (zh) | 基于肺动脉树分级的ct图像靶肺段识别方法 | |
US20210142470A1 (en) | System and method for identification of pulmonary arteries and veins depicted on chest ct scans | |
CN104463860A (zh) | 一种肝脏分段的方法 | |
WO2017028516A1 (zh) | 一种三维图像的校准方法、装置和系统 | |
US20220335690A1 (en) | System and method for linking a segmentation graph to volumetric data | |
US20220139029A1 (en) | System and method for annotation of anatomical tree structures in 3d images | |
Qiao et al. | Automatic liver segmentation method based on improved region growing algorithm | |
Tobon-Gomez et al. | Left atrial segmentation challenge: a unified benchmarking framework | |
CN105574862A (zh) | 一种肝脏半自动分段的方法 | |
JP5992853B2 (ja) | 手術支援装置、方法およびプログラム | |
CN111566699A (zh) | 静态流程前规划数据到动态流程内分割数据的配准 | |
CN113628346A (zh) | 一种自由浏览vb的方法以及进行标记的方法和系统 | |
WO2022194910A1 (en) | Method for visualizing at least a zone of an object in at least one interface |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 16836384 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 16836384 Country of ref document: EP Kind code of ref document: A1 |