CN104156960A - Full-automatic CT image kidney segmentation method - Google Patents
Full-automatic CT image kidney segmentation method Download PDFInfo
- Publication number
- CN104156960A CN104156960A CN201410392928.9A CN201410392928A CN104156960A CN 104156960 A CN104156960 A CN 104156960A CN 201410392928 A CN201410392928 A CN 201410392928A CN 104156960 A CN104156960 A CN 104156960A
- Authority
- CN
- China
- Prior art keywords
- image
- kidney
- resolution
- low resolution
- template
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
- Image Processing (AREA)
Abstract
The invention provides a full-automatic CT image kidney segmentation method. Multi-template-based image matching algorithm is used and a multi-resolution two-step algorithm strategy is adopted. A region of interest of the kidney in a low-resolution CT image is firstly positioned, and then kidney tissues in a region of interest in a high-resolution CT image are precisely segmented. The experimental result shows that in view of renal arteriography and urography CT images, accurate segmentation results can be provided, the workload of manual segmentation by a doctor can be greatly reduced, and the efficiency and effects of diagnosis and treatment can be effectively improved.
Description
Technical field
The present invention relates to technical field of image processing, particularly a kind of full-automatic CT image kidney dividing method.
Background technology
CT image has been widely used in the urological department illnesss such as clinical diagnosis kidney neoplasms, kidney stone at present.When carrying out CT image acquisition, by injection of contrast medium and adjusting image acquisition parameter, can show the tissues such as kidney, kidney neoplasms, the arteria renalis, cortex renis, kidney medulla, ureter, to obtain dissection and the physiologic parameters for medical diagnosis on disease.In addition,, along with extensively carrying out of the disease Wicresoft interventional therapy operations such as kidney, kidney stone, the preoperative plan that these information also can be Minimally Invasive Surgery provides reliable foundation.In order accurately to obtain these parameters, conventionally need to cut apart the kidney region in CT image, if only rely on the manual sketch outline line of doctor to require a great deal of time, also likely introduce human error, greatly the efficiency of impact diagnosis and treatment.Therefore, need badly and adopt the computer program of robotization to carry out auto Segmentation to the kidney region in CT image.Meanwhile, also consider that normal renal arteriography CT image and this two classes image of urography CT image of adopting carries out medical diagnosis on disease in actual clinical diagnosis, kidney automatic segmentation algorithm should be able to be applicable to the conventional CT image of this two class simultaneously.But because contrast preparation in this two classes image strengthens the position difference showing, being illustrated in figure 1 renal arteriography CT image, being illustrated in figure 2 urography CT image, there is larger difference in visible renal arteriography CT image and this two classes image of urography CT image.Especially in urography CT image, the border of kidney and Near tissue is usually difficult to distinguish, and this brings very large difficulty to the conventional partitioning algorithm based on boundary information.At present not yet there is relevant kidney partitioning algorithm can be applicable to this two class CT image simultaneously.
Summary of the invention
The technical matters solving: for the deficiencies in the prior art, the present invention proposes a kind of full-automatic CT image kidney dividing method, a kind of method of computing machine auto Segmentation kidney CT image is provided, avoid manually cutting apart consuming time and error, also solved the kidney dividing method of the prior art technical matters to renal arteriography CT image and these two kinds of Image Segmentation Usings of urography CT image simultaneously simultaneously.
Technical scheme: for solving the problems of the technologies described above, the present invention by the following technical solutions:
A full-automatic CT image kidney dividing method, comprises the following steps of carrying out in turn:
Step (1), set up experts database, N the low resolution template image and M the high resolving power template image that in experts database, comprise template CT image, be all furnished with expert's mark result in kidney region on described low resolution template image and high resolving power template image;
Step (2), CT image to be split is generated respectively to low resolution CT image and high-resolution ct image;
Step (3), the low resolution CT image that step (2) is generated and N low resolution template image in experts database carry out respectively image registration, obtain the preliminary segmentation result in kidney region in low resolution CT image;
Step (4), utilize the preliminary segmentation result in kidney region of step (3) gained, locate the approximate location of left and right kidney as area-of-interest, and the part of the area-of-interest of the left and right kidney of correspondence is intercepted out in the high-resolution ct image of step (2) gained;
Step (5), by the part of the area-of-interest of the left and right kidney intercepting in step (4) middle high-resolution CT image separately with experts database in M high resolving power template image carry out respectively image registration, obtain the Accurate Segmentation result in kidney region.
Enforcement of the present invention depends on low resolution template image and high resolving power template image and the expert's segmentation result separately in experts database, and consider accuracy and calculated amount, the quantitative range separately of low resolution template image and high resolving power template image is better between 6～10; After experts database forms, expert is made full use of the manual segmentation result of kidney in template image, first utilize low resolution CT image to carry out registration and obtain the preliminary segmentation result in kidney region in low resolution CT image, then mark off on this basis left and right kidney approximate region as the area-of-interest of left and right kidney, then in high-resolution ct image, intercept out this piece region and high resolving power template image and carry out registration, in the high-resolution ct image of intercepting, obtain net result; Said process is realized computing machine auto Segmentation completely, has avoided general medical worker's manual operations, and meanwhile, kidney is detected to conventional renal arteriography CT image in the present invention and urography CT image is all applicable.
Further, in the present invention, in step (1), the process of establishing of experts database is as follows:
First select template CT image, then the template CT image of choosing is reconstructed into X in three-dimensional cartesian coordinate system, Y, volume data that Z all directions resolution is equal by image linear interpolation algorithm; Then by expert, in the xsect in this volume data, by manual mode, describe kidney border, generate expert's mark result in kidney region; The kidney region finally this volume data and expert being marked is down-sampled generation low resolution template image respectively; Also according to expert's mark result, the area-of-interest that comprises kidney in this volume data is intercepted out simultaneously, generate high resolving power template image.
In experts database process of establishing, expert describes kidney border and carries out in the volume data without down-sampled, and sharpness is high, cuts apart accurately; And high resolving power template image does not directly rely on low resolution template image by obtaining after the volume data intercepting without down-sampled.In the high resolving power template image finally obtaining, only comprise kidney region, and in low resolution template image, comprise whole abdominal tissues; It is no matter equal visible expert's mark result in kidney region clearly in high resolving power template image or in low resolution template image, difference is only that expert's mark result in high resolving power template image is comparatively meticulous and only comprise single kidney region, and expert's mark result in low resolution template image because of through down-sampled processing not as good as down-sampled front meticulous.
Further, in the present invention, in step (2), low resolution CT image and high-resolution ct image obtain by image interpolation algorithm, specifically comprise the following steps:
First original CT image to be split is reconstructed into volume data that X in three-dimensional cartesian coordinate system, Y, Z all directions resolution are equal as high-resolution ct image by image interpolation algorithm, and then this volume data is down-sampled to low resolution, generate low resolution CT image.In the visible high-resolution ct image obtaining and low resolution CT image, all comprise whole abdominal tissues here.
Image interpolation algorithm is conventional a kind of disposal route during image scaling is processed, conventional have 3 kinds of concrete methods, be respectively arest neighbors interpolation, linear interpolation, spline interpolation, in the present invention, can utilize Tri linear interpolation algorithm, can realize good image scaling effect.
Further, in the present invention, there is some difference to consider anatomic form due to kidney in different patients' CT image, so in order to adapt to these differences, can adopt the two-step algorithm policy based on multi-template image matching algorithm and multiresolution, be partitioned into the kidney region in CT image.Based on multi-template image matching algorithm, be mainly to utilize low resolution template image in experts database and high resolving power template image and CT image to be split to carry out image registration, thereby obtain the segmentation result of respective regions in CT image to be split; In the present invention, used the thought of multiresolution, first with the image of low resolution, realized kidney region Primary Location, calculated amount is less, and then raising speed realize the accurate division of kidney edge and kidney inside with high-resolution image, and accuracy is high.
Concrete, described step (3) specifically comprises the following steps:
Step (31), low resolution CT image and N low resolution template image that step (1) is generated carry out respectively image registration, obtain No. one coordinate mapping parameters;
Step (32), utilize a coordinate mapping parameters of gained in step (31), expert's mark result in kidney region in N low resolution template image is all mapped on low resolution CT image after deformation;
Step (33), expert's mark result in N kidney region on the low resolution CT image after the resulting Means of Deformation Mapping Approach of step (32) is merged, generate the preliminary segmentation result in kidney region in low resolution CT image.
Similar with the processing mode of low resolution CT image, for high-resolution ct image, described step (5) specifically comprises the following steps:
Step (51), high resolving power area-of-interest CT image and M high resolving power template image that step (4) is intercepted carry out respectively image registration, obtain No. two coordinate mapping parameters;
Step (52), utilize No. two coordinate mapping parameters of gained in step (51), expert's mark result in kidney region in M high resolving power template image is all mapped on the high-resolution ct image of intercepting after deformation;
Step (53), expert's mark result in N kidney region on the high-resolution ct image after the resulting Means of Deformation Mapping Approach of step (52) is merged, in the high-resolution ct image of intercepting, generate the Accurate Segmentation result in kidney region.
As preferably, in the present invention, described merging adopts most Voting principles to carry out.On the CT image after Means of Deformation Mapping Approach, whether the number of times that is confirmed as kidney region according to same pixel is greater than half of total ballot number of times of carrying out for this pixel, if be greater than half of total ballot number of times of carrying out for this pixel, just this pixel be defined as to kidney region in final segmentation result.Most Voting principles are simple the most the most frequently used a kind of in merging method, in addition, also have a lot of other merging modes to select, and all belong to the routine techniques means in this area.
Beneficial effect:
Full-automatic CT image kidney dividing method provided by the present invention, the two-step algorithm policy of employing based on multi-template image matching algorithm and multiresolution, rely on existing expert's mark result in experts database, the approximate location of first locating kidney in low resolution CT image is as area-of-interest, be reflected in high-resolution ct image, and intercept corresponding part, and then utilize the renal tissue in the high-resolution ct image Accurate Segmentation area-of-interest being truncated to.Complete the accurate division to each patient renal region, be applicable to cut apart the renal tissue region in renal arteriography CT image and urography CT image simultaneously;
The present invention relies on computing machine to carry out completely, can significantly reduce the manual workload of delineating of doctor, avoids the manual error of bringing of delineating simultaneously;
The result of cutting apart obtains relevant disease diagnosis and required important dissection and the physiologic parameters of surgery planning by can be used in, thereby improves efficiency and the accuracy of medical diagnosis on disease and treatment.
Accompanying drawing explanation
Fig. 1 is renal arteriography CT image;
Fig. 2 is urography CT image;
Fig. 3 is process flow diagram of the present invention;
Fig. 4 is 4 groups of low resolution template images that comprise left and right kidney and corresponding expert's mark result example;
The high-resolution template image that comprises 4 groups of kidneys that Fig. 5 is intercepting and corresponding expert's mark result example;
Fig. 6 is kidney segmentation result and area-of-interest location example in low resolution CT image in renal arteriography CT image;
Fig. 7 is kidney segmentation result and area-of-interest location example in low resolution CT image in urography CT image;
Fig. 8 is high resolving power renal arteriography CT image kidney segmentation result example;
Fig. 9 is high resolving power urography CT image kidney segmentation result example.
Embodiment
Below in conjunction with accompanying drawing, the present invention is further described.
As shown in Figure 3, be process flow diagram of the present invention.Comprise the steps:
Method in step (1), the present invention depends on the foundation of experts database: first select template CT image, then the template CT image of choosing is reconstructed into X in three-dimensional cartesian coordinate system, Y, volume data that Z all directions resolution is equal by image linear interpolation algorithm; Then by expert, in this volume data, by manual mode, describe kidney border, mark kidney region, generate expert's mark result in kidney region; The kidney region finally this volume data and expert being marked is down-sampled respectively, reduce sampling frequency is set as 4, can generate like this resolution is the down-sampled low resolution template image of image size 1/64 before, select N=8 low resolution template image, in Fig. 4,4 low resolution template images that comprise whole abdominal tissues have been listed in left side, in Fig. 4 right side be with Fig. 4 on the left of expert's mark result example in corresponding kidney region; Also according to expert's mark result, the area-of-interest that comprises kidney in the volume data before not down-sampled is intercepted out simultaneously, generate high resolving power template image.Selected M=8 high resolving power template image, in Fig. 5,4 high resolving power template images that only comprise single kidney that intercepted out have been listed in left side, and in Fig. 5, the expert mark result example corresponding with Fig. 5 left side listed on right side.Due to similar in the kidney-shaped of left and right, in orientation, be only that mirror image is symmetrical, so all right side kidneys are transformed into the direction identical with left kidney by mirror transformation, to avoid that left and right kidney is set up respectively to high-resolution template image, can simplify like this process of establishing of template image.
Step (2), CT image to be split is generated respectively to low resolution CT image and high-resolution ct image; First original CT image to be split is reconstructed into volume data that X in three-dimensional cartesian coordinate system, Y, Z all directions resolution are equal as high-resolution ct image by image interpolation algorithm, and then this volume data is down-sampled, reduce sampling frequency is set as 4, and the low resolution CT image size obtaining after down-sampled is high-resolution ct image size 1/64.
Step (3), the low resolution CT image that step (2) is generated and N low resolution template image in experts database carry out respectively image registration, obtain the preliminary segmentation result in kidney region in low resolution CT image; Specifically comprise the steps:
Step (31), low resolution CT image and N low resolution template image that step (1) is generated carry out respectively image registration, obtain No. one coordinate mapping parameters;
Step (32), utilize a coordinate mapping parameters of gained in step (31), expert's mark result in kidney region in N low resolution template image is all mapped on low resolution CT image after deformation;
Step (33), the most Voting principles of expert's mark result utilization in N kidney region on the low resolution CT image after the resulting Means of Deformation Mapping Approach of step (32) are merged, generate the preliminary segmentation result in kidney region in low resolution CT image.As shown in Figure 6, Fig. 6 left side is the low resolution CT image obtaining after renal arteriography CT image is processed, and Fig. 6 right side is the preliminary segmentation result in kidney region obtaining after renal artery radiography CT image is processed; Fig. 7 left side is the low resolution CT image obtaining after urography CT image is processed, and Fig. 6 lower right is the preliminary segmentation result in kidney region obtaining after urography CT image is processed.
Step (4), utilize the preliminary segmentation result in kidney region of step (3) gained, locate the approximate location of left and right kidney as area-of-interest, as the region of rectangle frame choosing in Fig. 6 and Fig. 7, and the part of the area-of-interest of the left and right kidney of correspondence is intercepted out in the high-resolution ct image of step (2) gained.
Step (5), by the part of the area-of-interest of the left and right kidney intercepting in step (4) middle high-resolution CT image separately with experts database in M high resolving power template image carry out respectively image registration, obtain the Accurate Segmentation result in kidney region; Specifically comprise the following steps:
Step (51), high resolving power area-of-interest CT image and M high resolving power template image that step (4) is intercepted carry out respectively image registration, obtain No. two coordinate mapping parameters;
Step (52), utilize No. two coordinate mapping parameters of gained in step (51), expert's mark result in kidney region in M high resolving power template image is all mapped on the high-resolution ct image of intercepting after deformation;
Step (53), the most Voting principles of expert's mark result utilization in N kidney region on the high-resolution ct image after the resulting Means of Deformation Mapping Approach of step (52) are merged, in the high-resolution ct image of intercepting, generate the Accurate Segmentation result in kidney region.If Fig. 8 is high resolving power renal arteriography CT image kidney segmentation result example, wherein upside 2 width figure show with two-dimensional approach, and downside 2 width figure show with three dimensional constitution; Specifically comprise the high resolving power renal arteriography CT image that Fig. 8 left side is intercepting, Fig. 8 right side is kidney region Accurate Segmentation result.Similarly, Fig. 9 high resolving power urography CT image kidney segmentation result example, upside 2 width figure show with two-dimensional approach equally, downside 2 width figure show with three dimensional constitution; Specifically comprise the high resolving power urography CT image that Fig. 9 left side is intercepting, Fig. 9 right side is kidney region Accurate Segmentation result.
The above is only the preferred embodiment of the present invention; be noted that for those skilled in the art; under the premise without departing from the principles of the invention, can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.
Claims (6)
1. a full-automatic CT image kidney dividing method, is characterized in that: comprise the following steps of carrying out in turn:
Step (1), set up experts database, N the low resolution template image and M the high resolving power template image that in experts database, comprise template CT image, be all furnished with expert's mark result in kidney region on described low resolution template image and high resolving power template image;
Step (2), CT image to be split is generated respectively to low resolution CT image and high-resolution ct image;
Step (3), the low resolution CT image that step (2) is generated and N low resolution template image in experts database carry out respectively image registration, obtain the preliminary segmentation result in kidney region in low resolution CT image;
Step (4), utilize the preliminary segmentation result in kidney region of step (3) gained, locate the approximate location of left and right kidney as area-of-interest, and the part of the area-of-interest of the left and right kidney of correspondence is intercepted out in the high-resolution ct image of step (2) gained;
Step (5), by the part of the area-of-interest of the left and right kidney intercepting in step (4) middle high-resolution CT image separately with experts database in M high resolving power template image carry out respectively image registration, obtain the Accurate Segmentation result in kidney region.
2. a kind of full-automatic CT image kidney dividing method according to claim 1, is characterized in that: in step (1), the process of establishing of experts database is as follows:
First select template CT image, then the template CT image of choosing is reconstructed into X in three-dimensional cartesian coordinate system, Y, volume data that Z all directions resolution is equal by image linear interpolation algorithm; Then by expert, in the xsect in this volume data, by manual mode, describe kidney border, generate expert's mark result in kidney region; The kidney region finally this volume data and expert being marked is down-sampled generation low resolution template image respectively; Also according to expert's mark result, the area-of-interest that comprises kidney in this volume data is intercepted out simultaneously, generate high resolving power template image.
3. a kind of full-automatic CT image kidney dividing method according to claim 1, is characterized in that: in step (2), low resolution CT image and high-resolution ct image obtain by image interpolation algorithm, specifically comprise the following steps:
First original CT image to be split is reconstructed into volume data that X in three-dimensional cartesian coordinate system, Y, Z all directions resolution are equal as high-resolution ct image by image interpolation algorithm, and then this volume data is down-sampled to low resolution, generate low resolution CT image.
4. a kind of full-automatic CT image kidney dividing method according to claim 1, is characterized in that: described step (3) specifically comprises the following steps:
Step (31), low resolution CT image and N low resolution template image that step (1) is generated carry out respectively image registration, obtain No. one coordinate mapping parameters;
Step (32), utilize a coordinate mapping parameters of gained in step (31), expert's mark result in kidney region in N low resolution template image is all mapped on low resolution CT image after deformation;
Step (33), expert's mark result in N kidney region on the low resolution CT image after the resulting Means of Deformation Mapping Approach of step (32) is merged, generate the preliminary segmentation result in kidney region in low resolution CT image.
5. a kind of full-automatic CT image kidney dividing method according to claim 1, is characterized in that: described step (5) specifically comprises the following steps:
Step (51), area-of-interest and M the high resolving power template image of the left and right kidney in the high-resolution ct image of step (4) intercepting are carried out respectively to image registration, obtain No. two coordinate mapping parameters;
Step (52), utilize No. two coordinate mapping parameters of gained in step (51), expert's mark result in kidney region in M high resolving power template image is all mapped on the high-resolution ct image of intercepting after deformation;
Step (53), expert's mark result in N kidney region on the high-resolution ct image after the resulting Means of Deformation Mapping Approach of step (52) is merged, in the high-resolution ct image of intercepting, generate the Accurate Segmentation result in kidney region.
6. according to a kind of full-automatic CT image kidney dividing method described in claim 4 or 5, it is characterized in that: described merging adopts most Voting principles to carry out.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410392928.9A CN104156960A (en) | 2014-08-11 | 2014-08-11 | Full-automatic CT image kidney segmentation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410392928.9A CN104156960A (en) | 2014-08-11 | 2014-08-11 | Full-automatic CT image kidney segmentation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104156960A true CN104156960A (en) | 2014-11-19 |
Family
ID=51882450
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410392928.9A Pending CN104156960A (en) | 2014-08-11 | 2014-08-11 | Full-automatic CT image kidney segmentation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104156960A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106228198A (en) * | 2016-08-17 | 2016-12-14 | 广东工业大学 | A kind of super-resolution recognition methods of medical treatment CT image |
| CN107492104A (en) * | 2017-07-24 | 2017-12-19 | 广东顺德中山大学卡内基梅隆大学国际联合研究院 | The automatic division method of the subcortical structure subregion of downfield MRI |
| CN110363774A (en) * | 2019-06-17 | 2019-10-22 | 上海联影智能医疗科技有限公司 | Image partition method, device, computer equipment and storage medium |
| WO2020125806A1 (en) * | 2018-12-17 | 2020-06-25 | Shanghai United Imaging Intelligence Co., Ltd. | Systems and methods for image segmentation |
| CN112016394A (en) * | 2020-07-21 | 2020-12-01 | 影石创新科技股份有限公司 | Obstacle information acquisition method, obstacle avoidance method, mobile device, and computer-readable storage medium |
| CN112241656A (en) * | 2019-07-17 | 2021-01-19 | 上海肇观电子科技有限公司 | Image detection method and apparatus, processor chip circuit, and storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130223704A1 (en) * | 2012-02-28 | 2013-08-29 | Siemens Aktiengesellschaft | Method and System for Joint Multi-Organ Segmentation in Medical Image Data Using Local and Global Context |
| CN103854286A (en) * | 2014-03-05 | 2014-06-11 | 东南大学 | Renal artery blood-supply area segmenting method based on CT contrastographic picture |
-
2014
- 2014-08-11 CN CN201410392928.9A patent/CN104156960A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130223704A1 (en) * | 2012-02-28 | 2013-08-29 | Siemens Aktiengesellschaft | Method and System for Joint Multi-Organ Segmentation in Medical Image Data Using Local and Global Context |
| CN103854286A (en) * | 2014-03-05 | 2014-06-11 | 东南大学 | Renal artery blood-supply area segmenting method based on CT contrastographic picture |
Non-Patent Citations (4)
| Title |
|---|
| REMI CUINGNET等: "Automatic Detection and Segmentation of Kidneys in 3D CT Images Using Random Forests", 《PROC.MICCAI 2012》 * |
| 智少丹: "LED自动固晶机视觉定位系统的图像处理研究", 《中国优秀硕士学位论文全文数据库》 * |
| 李昕 等: "三维超声图像肾脏位置自动定位算法", 《华中科技大学学报（自然科学版）》 * |
| 陈雯艳: "基于ROI多图谱配准的海马磁共振图像分割", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106228198A (en) * | 2016-08-17 | 2016-12-14 | 广东工业大学 | A kind of super-resolution recognition methods of medical treatment CT image |
| CN106228198B (en) * | 2016-08-17 | 2019-05-07 | 广东工业大学 | A kind of super-resolution recognition methods of medical treatment CT image |
| CN107492104A (en) * | 2017-07-24 | 2017-12-19 | 广东顺德中山大学卡内基梅隆大学国际联合研究院 | The automatic division method of the subcortical structure subregion of downfield MRI |
| WO2020125806A1 (en) * | 2018-12-17 | 2020-06-25 | Shanghai United Imaging Intelligence Co., Ltd. | Systems and methods for image segmentation |
| US11341734B2 (en) * | 2018-12-17 | 2022-05-24 | Shanghai United Imaging Intelligence Co., Ltd. | Systems and methods for image segmentation |
| US20220284687A1 (en) * | 2018-12-17 | 2022-09-08 | Shanghai United Imaging Intelligence Co., Ltd. | Systems and methods for image segmentation |
| US11836925B2 (en) * | 2018-12-17 | 2023-12-05 | Shanghai United Imaging Intelligence Co., Ltd. | Systems and methods for image segmentation |
| CN110363774A (en) * | 2019-06-17 | 2019-10-22 | 上海联影智能医疗科技有限公司 | Image partition method, device, computer equipment and storage medium |
| CN110363774B (en) * | 2019-06-17 | 2021-12-21 | 上海联影智能医疗科技有限公司 | Image segmentation method and device, computer equipment and storage medium |
| CN112241656A (en) * | 2019-07-17 | 2021-01-19 | 上海肇观电子科技有限公司 | Image detection method and apparatus, processor chip circuit, and storage medium |
| CN112016394A (en) * | 2020-07-21 | 2020-12-01 | 影石创新科技股份有限公司 | Obstacle information acquisition method, obstacle avoidance method, mobile device, and computer-readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104156960A (en) | Full-automatic CT image kidney segmentation method | |
| CN106056537B (en) | A kind of medical image joining method and device | |
| CN106902475B (en) | Radiation therapy method, computer storage medium and computer system | |
| CN103679810B (en) | The three-dimensional rebuilding method of liver's CT image | |
| CN103325143B (en) | Labelling point automatic registration method based on Model Matching | |
| US7773786B2 (en) | Method and apparatus for three-dimensional interactive tools for semi-automatic segmentation and editing of image objects | |
| CN101971213A (en) | A method and system for anatomy structure segmentation and modeling in an image | |
| WO2015161728A1 (en) | Three-dimensional model construction method and device, and image monitoring method and device | |
| US10311585B2 (en) | Mechanism for advanced structure generation and editing | |
| US10667786B2 (en) | Ultrasound imaging apparatus and method for segmenting anatomical objects | |
| CN101082983A (en) | Self-adapting medicine sequence image values inserting method based on interested region | |
| EP4349265A1 (en) | Ultrasonic puncture guidance planning system based on multi-modal medical image registration | |
| CN103886621A (en) | Method for automatically extracting bed plate | |
| CN108182687A (en) | A kind of interactive three-dimensional dividing method based on brain tumor medical image | |
| van den Elsen et al. | Image fusion using geometrical features | |
| Hacihaliloglu et al. | Statistical shape model to 3D ultrasound registration for spine interventions using enhanced local phase features | |
| Alam et al. | Evaluation of medical image registration techniques based on nature and domain of the transformation | |
| EP2584970A2 (en) | Mechanism for advanced structure generation and editing | |
| US10032316B2 (en) | System for dynamically propagating real-time alterations of medical images | |
| Papalazarou et al. | Surgical needle reconstruction using small-angle multi-view x-ray | |
| CN107292351B (en) | A kind of matching process and device of tubercle | |
| CN113256754B (en) | Stacking projection reconstruction method for segmented small-area tumor mass | |
| Chen et al. | Image segmentation and registration techniques for MR-Guided Liver Cancer Surgery | |
| Wang et al. | Automatic hybrid registration for 2-dimensional CT abdominal images | |
| Kim et al. | SLIMMER: SLIce MRI motion estimation and reconstruction tool for studies of fetal anatomy |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20141119 |
|
| RJ01 | Rejection of invention patent application after publication |