CN105809656B - Medical image processing method and device - Google Patents
Medical image processing method and device Download PDFInfo
- Publication number
- CN105809656B CN105809656B CN201410840692.0A CN201410840692A CN105809656B CN 105809656 B CN105809656 B CN 105809656B CN 201410840692 A CN201410840692 A CN 201410840692A CN 105809656 B CN105809656 B CN 105809656B
- Authority
- CN
- China
- Prior art keywords
- pixel point
- region
- area
- iteration
- characteristic value
- 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.)
- Active
Links
Images
Abstract
The invention provides a medical segmentation image, which is filled; it is characterized by also comprising: extracting any one region of interest, dividing the region of interest, and respectively acquiring a first region, a second region and a third region; calculating the characteristic value T of any pixel point i in the third area by adopting a first judgment methodi(ii) a The characteristic value T of any pixel point i in the third areaiAnd performing second judgment to obtain the filled segmentation image. The invention can effectively fill the discrete points with unclosed boundaries and serious internal vacancy in the segmented images, flexibly adjust the boundaries, achieve different smooth filling effects, improve the filling accuracy and robustness and meet different diagnosis requirements.
Description
Technical Field
The invention relates to the field of medical information processing, in particular to a medical image processing method and a medical image processing device.
Background
The medical image reconstruction is a theory, a method and a technology which convert a two-dimensional medical image sequence into a three-dimensional image to be displayed on a screen through computer graphics, a digital image processing technology, computer visualization, human-computer interaction and other technologies and provide an interactive processing means for a user according to requirements.
In order to accurately distinguish between target tissue and normal tissue in a medical image, the medical image needs to be segmented. Image segmentation is an indispensable means for extracting quantitative information of target tissues in medical images, is a key step from image processing to image analysis, and the accuracy of image segmentation affects the accuracy of image reconstruction and is important for diagnosis of diseases and the like.
However, due to defects of image information or algorithm, a mask (e.g., a mask of a bone and an organ) obtained by image segmentation calculation has various defects, for example, as shown in fig. 1a-1b, a boundary in a bone segmentation result is missing or a mask is not filled with a material and has discrete points, etc. due to the algorithm defect in the segmentation process, a complementary filling process is usually performed on the mask with defects to achieve a desired effect.
In the prior art, in the reconstruction of medical images, the filling of band defect masks is supplemented by various methods.
Reference 1: cn200580003233.6 discloses a filling method based on conventional expansion corrosion. Firstly, a certain expansion is carried out on the mask to be processed, and then the expanded mask is correspondingly etched, so that the filled mask is obtained. The limitations of this type of approach are: 1) masks with severe internal defects are prone to over-expansion so that the final result is greatly different from the actually required result; 2) the strength of swelling and corrosion needs to be judged in advance and is only suitable for a simpler treatment environment.
Reference 2: chenyang, Xu., and Jerry, L.P., Snakes, Shapes, and GradientVector Flow, Image Processing, IEEE trans.on,7(3),359 369(march 1998) discloses a dynamic contour (ACM, Snake) based filling method. Firstly, a distance field or a vector field is calculated through boundary information of an original mask or an actual image, and then a final closed contour is obtained through the evolution of a dynamic contour in the field. The limitations of this approach are: the calculation time is long, complex logic is needed when a plurality of separated masks are filled, the sensitivity of parameters is strong, and the parameters need to be adjusted independently for different application scenes.
Reference 3: samuel, g.a., Maryellen, l.g., castbolt, j.m., James, t.b., Kunio, d., Heber, m., Computerized Detection of cellular nodule on CT Scans, Imaging & Therapeutic Technology, vol.19(5), 1303-cup 1311(1999) discloses a boundary-supplementing method based on the rolling ball method (RollingBall). The method fills and smoothes the boundary of the region of interest by setting a dynamically rolling sphere and then closes the region of interest. The limitations of this approach are: 1) radius parameters of rolling balls need to be predefined, and the problem of excessive smoothness or insufficient filling exists for processing masks with different boundary defects; 2) when padding is performed for a plurality of divided masks, a more complicated logic is required similarly to the case of reference 2, and the sensitivity of parameters is strong. The robustness of the parameters for different application scenarios is not strong.
Disclosure of Invention
The invention provides a medical image processing method and a medical image processing device, which are used for solving the technical problem of filling a medical segmentation image with defects.
To solve the above problem, the present invention provides a medical image processing method, including: providing a medical segmentation image, and performing filling operation on the medical segmentation image; it is characterized by also comprising:
extracting any one region of interest, dividing the region of interest, and respectively acquiring a first region, a second region and a third region;
calculating the characteristic value T of any pixel point i in the third area by adopting a first judgment method and utilizing the thermal diffusion principlei；
The characteristic value T of any pixel point i in the third areaiAnd performing second judgment to obtain the filled segmentation image.
Optionally, the first region includes a point in the region of interest whose pixel point value is 1;
setting the characteristic values of all pixel points in the first area as Thigh；
The second area is a boundary area of the region of interest, and the characteristic values of all pixel points in the second area are Tlow；
The rest part is the third area;
wherein, Thigh-Tlow>100。
Optionally, the first determination method includes:
will be initialized to value Ti 1＝TlowSubstituting equation 1 for iterative calculation;
if Ti n+1-Ti n<Threshold A, then stop iteration, Ti n+1A characteristic value T obtained for the (n + 1) th iteration of the ith pixel pointi；
Otherwise, let Ti n＝Ti n+1Substituting equation 1 for iteration;
wherein α is a constant term;
Tithe characteristic value of the ith pixel point in the third area is represented, and i is a natural number;
Ti nobtaining a characteristic value for the nth iteration of the pixel point i;
xiand the spatial coordinates of the ith pixel point in the third area are obtained.
Optionally, the calculation method of equation 1 includes a conjugate gradient method, an incomplete cholesky conjugate gradient method, a strong implicit solution method, or a gaussian method.
Optionally, the value range of the threshold a is a value less than or equal to 10-6。
Optionally, the first determination method further includes:
will be initialized to value Ti 1＝TlowSubstituting equation 1 for calculation;
if the iteration number n +1 is more than 200, stopping the iteration, and Ti n+1A characteristic value T obtained for the (n + 1) th iteration of the ith pixel pointi；
Wherein α is a constant term;
Tithe characteristic value of the ith pixel point in the third area is represented, and i is a natural number;
Ti nobtaining a characteristic value for the nth iteration of the pixel point i;
xiand the spatial coordinates of the ith pixel point in the third area are obtained.
Optionally, the second determination includes:
obtaining a characteristic value T of the ith pixel point in the third areai，
If B is<Ti<ThighExtracting the pixel point i;
otherwise, the pixel point values are set to the pixel values of the background color of the medical segmented image.
Optionally, the value range of the threshold B is Tlow<B<Thigh。
In order to solve the above problem, the present invention also provides a medical image processing apparatus comprising:
an image acquisition unit adapted to provide a medical segmentation image;
an assignment unit adapted to acquire a first region, a second region and a third region, respectively, in the region of interest;
a first judging unit, adapted to calculate a characteristic value T of any pixel point i in the third region by using a thermal diffusion principlei；
A second judgment unit for judging the characteristic value T of any pixel point i in the third areaiAnd performing second judgment to obtain the filled segmentation image.
Optionally, the first determining unit further includes:
a convergence unit adapted to convert an initial value T of a feature value of any pixel point i in the third region into an initial value Ti 1＝TlowCalculated by substituting into said equation 1 if Ti n+1-Ti n<Threshold A, then stop iteration, Ti n+1Obtaining a characteristic value T for the (n + 1) th iteration of the ith pixel pointi；
Otherwise, let Ti n＝Ti n+1Substituting equation 1 for iteration;
an iteration unit adapted to convert the initial value Ti 1＝TlowSubstituting into the equation 1 to iterate, if the iteration number n +1 is more than 200, stopping the iteration, and Ti n+1A characteristic value T obtained for the (n + 1) th iteration of the ith pixel pointi；
Wherein the threshold A is less than or equal to 10-6。
Compared with the prior art, the technical scheme of the invention has the following advantages:
(1) the invention can well reserve or supplement the boundary information of the mask with the defects, and can not cause the situations of over-smoothness and over-filling.
(2) The invention is based on the operation of physical significance, thus not needing complex logic to fuse, having strong robustness, and being capable of well distinguishing and filling the separated areas.
(3) The method can fill the discontinuous boundary of the mask on one hand, can fill and submerge a larger cavity in the mask on the other hand, has strong flexibility, and is suitable for filling requirements of various discrete masks such as filling with a defect boundary, clustering of scattered point areas and the like.
Drawings
FIGS. 1a-1b are schematic diagrams of medical image segmentation results in the prior art;
FIG. 2 is a flow chart of a medical image processing method according to an embodiment of the present invention;
FIG. 3 is a flow chart of a medical image processing method provided by an embodiment of the invention;
FIG. 4 is a schematic illustration of medical image processing results provided by an embodiment of the present invention;
fig. 5 is a schematic structural diagram of a medical image processing apparatus according to an embodiment of the present invention.
Detailed Description
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein, but rather construed as limited to the embodiments set forth herein.
Next, the present invention is described in detail by using schematic diagrams, and when the embodiments of the present invention are described in detail, the schematic diagrams are only examples for convenience of description, and the scope of the present invention should not be limited herein.
In medical image reconstruction, image segmentation is an important link, but the image segmentation result often has defects such as boundary missing or internal unfilled, and the accuracy of later-stage disease diagnosis is directly affected, so that the segmentation image needs to be filled. In the prior art, an unclosed area cannot be effectively filled, or a large cavity in the unclosed area cannot be completely filled, or the boundary corrosion causes over expansion, so that the subsequent diagnosis is influenced.
In order to solve the above technical problems, the present invention provides a method and an apparatus for processing a medical image.
Fig. 2 is a schematic flow chart of a medical image processing method according to the technical solution of the invention.
Step S1 is first performed to provide a medical segmented image. The medical image may be a binary image (mask) of a two-dimensional or three-dimensional image segmentation result, and in a specific implementation, the two-dimensional or three-dimensional medical image may be acquired by using CT, MRI, Positron Emission Tomography (PET), X-ray equipment, ultrasound equipment, or the like.
Step S2 is executed to extract any one of the regions of interest, divide the region of interest, and acquire the first region, the second region, and the third region, respectively. The first region, the second region and the third region constitute the extracted region of interest.
The first region comprises a target tissue region in the region of interest, and for a binary image (mask), the first region is a region formed by pixel points with pixel point values of 1; setting the characteristic values of all pixel points in the first area as Thigh。
The second area is a boundary area of the region of interest, and the characteristic values of all pixel points in the second area are TlowIs a relative ThighLow value of (c).
The rest part is the third area which is a characteristic value TiA region to be calculated;
it should be noted that the medical image processing method provided by the present invention can perform a filling operation when there are a plurality of regions of interest separated but need to be filled separately in the segmentation result, and the present invention can well distinguish the separated regions and fill them without performing complicated logic processing.
Continuing to execute step S3, calculating the characteristic value T of any pixel point i in the third area by adopting a first judgment method and utilizing the thermal diffusion principlei。
It should be noted that in the medical image, the pixel point values displayed on the image by the same target tissue fluctuate within a certain range. For example, in CT images, the concept of a quantity, the CT value, is used. The CT value represents the attenuation of the X-ray after passing through the tissues and being absorbed, and the CT values of different tissues are different and fluctuate within a certain range; if the CT value of the bone is up to +1000HU, the CT value of the soft tissue is 20-70 HU, and the CT value of the water is 0 (+ -10 HU), the target tissue can be determined according to the CT value range.
In the principle of thermal diffusion in the physical category, the temperature of a physical quantity diffuses from a high-temperature field to a low-temperature field to achieve thermal balance, the temperature diffuses fast in the high-temperature field, and the diffusion between the high-temperature field and the low-temperature field is slow; the heat diffusion inside the same object is fast, and the heat diffusion between different objects is slow, that is, whether the same object or different parts of the same object can be determined through the temperature change.
Therefore, the first determination method of the present invention utilizes the principle of thermal diffusion to set the first region of the target tissue composition obtained by image segmentation, analogous to a high-temperature constant-temperature field, to have a constant characteristic value Thigh(ii) a The second region composed of the extracted boundary region of the region of interest is analogized to a low-temperature constant-temperature field and has a constant characteristic value Tlow(ii) a The T islowIs less than ThighLow value of (a), characteristic value T of the first and second regionshigh-Tlow>100, there is a temperature difference in the analog temperature field, which is shown as a difference in brightness, i.e., contrast, in the image. According to the principle of thermal diffusion, different spatial positions of all pixel points in the third area have different characteristic values TiCalculating the characteristic value T of any pixel point in the third areaiIf the image belongs to the same target organization, the characteristic value of the pixel point i approaches to ThighAnd the images are displayed with similar brightness, so that the aim of filling the same target tissue is fulfilled.
The invention carries out the filling operation by utilizing the thermal diffusion principle, firstly, the inside of the tissue can be absolutely filled, and secondly, the inside can also be completely filled under the condition that the boundaries have scattered non-closed positions, thereby meeting the filling requirement in the medical segmentation image. Compared with the prior art, the first judgment method is carried out simultaneously based on physical meanings, so that complex logic fusion is not needed, the filling result has robustness, and the separated areas can be well distinguished and filled.
In summary, the first determination method is similar to the thermal diffusion principle, effectively and accurately fills the mask region of interest with defects, and can fill each separated region independently without causing mutual influence or resetting of parameters when a plurality of separated regions exist in the mask, so that the first determination method has high accuracy and robustness.
Finally, step S4 is executed to determine the feature value T of any pixel point i in the third regioniPerforming second judgment, and if the judgment condition is met, extracting the pixel point i; otherwise, setting the pixel point values to the medical segmentation mapAnd obtaining the filled segmentation image according to the pixel value of the background color of the image.
For example, in the blood vessel extraction result, if a doctor needs to accurately measure the volume of a blood vessel, an accurate target tissue boundary is needed, and the method can effectively solve the problems that in the prior art, the boundary in the segmentation result is easily over-expanded when a filling method of expansion corrosion is utilized, and the boundary is over-smooth due to the adoption of a rolling ball boundary supplementing method, so that the accurate target tissue boundary requirement is met, and different diagnosis requirements of the doctor are met.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below.
In this embodiment, the three-dimensional CT mask is used as an example of a segmentation result of a bone image, and the image is processed to fill in the internal defect of the bone due to boundary loss.
Fig. 3 is a flowchart illustrating a medical image processing method according to an embodiment of the present invention.
Step S3001 is first executed to provide a binarized image (i.e., mask) of three-dimensional CT, which may be a segmentation result obtained by using various conventional image segmentation methods, and the present embodiment is not particularly limited to the type of image.
It should be noted that, the medical image processing method provided by this embodiment has a low requirement on the accuracy of the segmentation result obtained by the provided medical image segmentation method, and can still obtain an accurate and stable filling effect on the rough segmentation result, so that the medical image processing method is suitable for performing a filling operation on any medical segmented image.
Step S3002 is executed to extract any region of interest in the binarized image, for example, to extract a bone image lacking a partial boundary and having a missing internal portion as shown in fig. 1a, wherein the region of interest can be automatically obtained by an existing algorithm or manually extracted. And dividing the interested region, and respectively acquiring a first region, a second region and a third region.
The first region is a highlighted bone part in the region of interest, namely a region with a pixel value of 1; the second area is a boundary area of the region of interest and is automatically acquired when the region of interest is extracted; the third region is the remaining portion; the first region, the second region and the third region constitute the extracted region of interest.
Step S3003 is executed, in the third area, the initial value T is takeni 1＝TlowCalculated by substituting into equation 1.
In equation 1, α is a constant term, and since the actual diffusion effect does not need to be considered, α in this embodiment preferably takes a value of 1.
The above-mentioned
At the time t of n +1 iterations, the characteristic value of the pixel point i;
for example, the initial value Ti 1＝TlowFor the time t of the first iteration1When the characteristic value of the pixel point i is Tlow。
TiThe characteristic value of the ith pixel point in the third area is represented, and i is a natural number;
Ti nobtaining a characteristic value for the nth iteration of the pixel point i;
xithe spatial coordinates of the ith pixel point in the third area,
in this embodiment, T is preferredi 1＝TlowThe 1 st iteration value of the pixel point i is Tlow。
As detailed in the foregoing, the present invention determines whether the missing part belongs to the target tissue by comparing the characteristics of the pixel points of the target tissue in the medical image with the thermal diffusion principle, so as to achieve the purpose of completely filling the inside of the target tissue even if the boundary is not closed.
Therefore, we define saidThe pixel points in a region have constant characteristic value Thigh200 in this embodiment, the pixel point values in the second region have a constant characteristic value TlowIs lower than ThighLow value of (a), 0 in this example; the remaining pixel points in the third region are regions to be calculated, and the characteristic value T of any pixel point i in the third region is calculated and obtainedi. Wherein, ThighAnd TlowThe difference of (A) is required to achieve a certain contrast ratio, i.e. Thigh-Tlow>100, analogous to the high and low temperature fields, appear as contrasts in this example, which is summarized in medical images.
Will Ti 1＝TlowAnd substituting the initial value into equation 1 to perform iterative calculation.
Executing S3004, judging if Tn+1-Tn<Threshold A, i.e. convergence determination, said threshold A ≦ 10-6In this embodiment, the threshold A is 10 according to the general convergence principle-6In specific implementation, the threshold a may be set according to experience or actual needs; if yes, go to step S007, Ti n+1A characteristic value T obtained for the (n + 1) th iteration of the ith pixel pointi(ii) a Otherwise, go to S3005 to let Ti n＝Ti n+1Substituting into equation 1 for iterative computation. The calculation method for solving equation 1 includes a conjugate gradient method, an incomplete cholesky conjugate gradient method, a strong implicit solution method or a gaussian method, and the present invention is not limited to this method.
Executing step S3006, if the iteration number is greater than 200, executing step S3007, stopping iteration, and obtaining the characteristic value T of the ith pixel point in the third regioni. In this embodiment, when the number of iterations is greater than 200, the iteration result of calculating equation 1 may be considered to have converged, and the iteration may be stopped. The above embodiments respectively describe how to obtain the characteristic value T of the pixel point i by convergence judgment and iteration judgmenti. It should be noted that, in order to improve the efficiency of filling the divided image, the convergence determination in step S3004 and the iteration number determination in step S3006 may be adopted to perform the feature value calculationPerforming comprehensive judgment, judging convergence and judging the number of iterations to be an OR calculation relation, stopping iteration if any step of the comprehensive judgment, the convergence judgment and the iteration number judgment meets the judgment condition, and obtaining the characteristic value T of the ith pixel point in the third areai。
Executing step S3008, obtaining a feature value T of the ith pixel point in the third areaiExecuting the judgment of step S3009, if B is true<Ti<ThighExtracting the pixel point i, namely the pixel point i is the filled target organization required by us; otherwise, step S3010 is executed to set the pixel point to the background color of the medical segmented image, and in the binary image in this embodiment (mask), the pixel point value is set to 0. By controlling the threshold B, the boundary of the segmented target tissue can be flexibly adjusted, the problem of over expansion or corrosion is prevented, and different medical image diagnosis requirements are met. E.g. infinite approximation T of threshold BhighThe boundaries of the target tissue region tend to be accurate, suitable for accurate calculation of the target tissue volume.
Step S3011 is executed to obtain a filled medical segmented image, as shown in the bone image after filling the defect in fig. 4, compared with fig. 1a, by the medical image processing method provided in this embodiment, firstly, the interior of the tissue can be absolutely filled, and secondly, the interior can also be filled in the case of a non-closed position where the boundary has scatter, the image boundary is flexibly adjusted according to the medical requirement, and the filling result has strong robustness.
In the present embodiment, a single bone is used as a region of interest to perform a filling operation, and it can be understood that, in a specific implementation, the same filling operation can be performed on a segmentation effect of other organ tissues or on a plurality of organ tissues existing simultaneously to achieve the same effect. The invention also provides a medical image processing device corresponding to the medical image processing method.
Fig. 5 is a schematic structural diagram of a medical image apparatus provided in the technical solution of the present invention.
As shown in fig. 5, the apparatus includes an image acquisition unit U11, a distribution unit U21, a first determination unit U31, and a second determination unit U41.
The image obtaining unit U11 is adapted to provide various types of medical segmented images, such as a binary image (mask) of a two-dimensional or three-dimensional image as a segmentation result or various pixel value images, which are not specifically limited by the present invention.
The assigning unit U21, adapted to acquire a first region, a second region and a third region in the region of interest, respectively;
the first determination unit U31 is adapted to calculate the characteristic value T of any pixel point i in the third region by using the principle of thermal diffusioniAnd the complete filling effect inside the target tissue is realized.
The first determination unit U31 further includes:
the convergence unit U311 is adapted to apply the feature value T of any pixel point i in the third regioniSubstituting into the image equation 1 to calculate if Ti n+1-Ti n<Stopping iteration if the threshold value A is reached, and obtaining the characteristic value T of the pixel point ii(ii) a Otherwise, let Ti n＝Ti n+1Substituting equation 1 for iteration;
the iteration unit U312 is adapted to generate an initial value T1＝TlowSubstituting the equation 1 for iterative calculation, if the iteration number n is more than 200, stopping iteration, and obtaining the characteristic value T of the ith pixel point in the third areaiAnd so on, obtaining the characteristic values T of all the pixel points in the third areaiAnd displaying the brightness of different pixel points in the image.
It should be noted that, when any one of the convergence unit U311 and the iteration unit U312 completes the calculation, the first determination unit U31 completes the determination to obtain the characteristic value T of the ith pixel point in the third areaiAnd enters the second determination unit U41.
The second decision unit U41 flexibly adjusts the boundary of the segmented target tissue by setting a threshold B to obtain a filled segmented image, wherein,the value range of the threshold B is Tlow<B<Thigh。
Although the present invention is disclosed above, the present invention is not limited thereto. Various changes and modifications may be effected therein by one skilled in the art without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (7)
1. A medical image processing method, comprising: providing a medical segmentation image, and performing filling operation on the medical segmentation image; it is characterized by also comprising:
extracting any one region of interest, dividing the region of interest, and respectively acquiring a first region, a second region and a third region;
calculating the characteristic value T of any pixel point i in the third area by adopting a first judgment method and utilizing the thermal diffusion principlei；
The characteristic value T of any pixel point i in the third areaiPerforming second judgment to obtain a filled segmentation image;
the first region includes a point in the region of interest whose pixel point value is 1; setting the characteristic values of all pixel points in the first area as Thigh；
The second area is a boundary area of the region of interest, and the characteristic values of all pixel points in the second area are Tlow；
The rest part is the third area;
wherein, Thigh-Tlow>100；
The first determination method includes:
will be initialized to value Ti 1＝TlowSubstituting equation 1 for iterative calculation;
if Ti n+1-Ti n<Threshold A, stop the stackGeneration, said Ti n+1Is the characteristic value T of the ith pixel pointi；
Otherwise, let Ti n＝Ti n+1Substituting equation 1 for iteration;
wherein α is a constant term;
Tithe characteristic value of the ith pixel point in the third area is represented, and i is a natural number;
Ti nobtaining a characteristic value for the nth iteration of the pixel point i;
xiand the spatial coordinates of the ith pixel point in the third area are obtained.
2. The medical image processing method according to claim 1, wherein the equation 1 calculation method includes a conjugate gradient method, an incomplete cholesky conjugate gradient method, a strong implicit solution method, or a gaussian method.
3. A medical image processing method according to claim 1, wherein the threshold value a is in a range of a ≦ 10-6。
4. The medical image processing method according to claim 1, wherein the first determination method further comprises:
will be initialized to value Ti 1＝TlowSubstituting into equation 1 to perform iterative calculation;
if the iteration number n +1 is more than 200, stopping the iteration, and Ti n+1A characteristic value T obtained for the (n + 1) th iteration of the ith pixel pointi；
Otherwise, let Ti n＝Ti n+1Substituting equation 1 for iteration;
wherein α is a constant term;
Tithe characteristic value of the ith pixel point in the third area is represented, and i is a natural number;
Ti nobtaining a characteristic value for the nth iteration of the pixel point i;
xiand the spatial coordinates of the ith pixel point in the third area are obtained.
5. The medical image processing method according to claim 1, wherein the second determination includes:
obtaining a characteristic value T of the ith pixel point in the third areai，
If the threshold value B<Ti<ThighExtracting the pixel point i;
otherwise, the pixel point values are set to the pixel values of the background color of the medical segmented image.
6. The medical image processing method according to claim 5, wherein the threshold B has a value in a range Tlow<B<Thigh。
7. A medical image processing apparatus, characterized by comprising:
an image acquisition unit adapted to provide a medical segmentation image;
an assigning unit adapted to acquire a first region, a second region and a third region in the extracted arbitrary one of the regions of interest, respectively;
a first judging unit adapted to calculate a characteristic of any one pixel point i in the third region using a thermal diffusion principleCharacteristic value Ti；
A second judgment unit for judging the characteristic value T of any pixel point i in the third areaiPerforming a second determination adapted to obtain the filled segmented image; the first determination unit further includes:
a convergence unit adapted to convert an initial value T of a feature value of any pixel point i in the third region into an initial value Ti 1＝TlowCalculated by substituting into said equation 1 if Ti n+1-Ti n<Threshold A, then stop iteration, Ti n+1A characteristic value T obtained for the (n + 1) th iteration of the ith pixel pointi；
Otherwise, let Ti n＝Ti n+1Substituting equation 1 for iteration;
an iteration unit adapted to convert the initial value Ti 1＝TlowSubstituting the obtained product into the equation 1 to carry out iteration, stopping the iteration if the iteration number n +1 is more than 200,
wherein α is a constant term;
Ti nobtaining a characteristic value for the nth iteration of the pixel point i;
xithe spatial coordinates of the ith pixel point in the third area are obtained;
the T isi n+1The characteristic value T of the ith pixel point in the third areai；
Wherein the threshold A is less than or equal to 10-6。
Priority Applications (7)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410840692.0A CN105809656B (en) | 2014-12-29 | 2014-12-29 | Medical image processing method and device |
| US15/323,035 US10181191B2 (en) | 2014-12-02 | 2015-10-31 | Methods and systems for identifying spine or bone regions in computed tomography image sequence |
| GB1719333.5A GB2559013B (en) | 2014-12-02 | 2015-10-31 | A method and system for image processing |
| PCT/CN2015/093506 WO2016086744A1 (en) | 2014-12-02 | 2015-10-31 | A method and system for image processing |
| EP15865201.6A EP3213296B1 (en) | 2014-12-02 | 2015-10-31 | A method and system for image processing |
| GB1709225.5A GB2547399B (en) | 2014-12-02 | 2015-10-31 | A method and system for image processing |
| US16/247,080 US11094067B2 (en) | 2014-12-02 | 2019-01-14 | Method and system for image processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410840692.0A CN105809656B (en) | 2014-12-29 | 2014-12-29 | Medical image processing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105809656A CN105809656A (en) | 2016-07-27 |
| CN105809656B true CN105809656B (en) | 2020-03-31 |
Family
ID=56980117
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410840692.0A Active CN105809656B (en) | 2014-12-02 | 2014-12-29 | Medical image processing method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105809656B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108765289B (en) * | 2018-05-25 | 2022-02-18 | 李锐 | Digital image extracting, splicing, restoring and filling method |
| CN111260642A (en) * | 2020-02-12 | 2020-06-09 | 上海联影医疗科技有限公司 | Image positioning method, device, equipment and storage medium |
| CN112102284A (en) * | 2020-09-14 | 2020-12-18 | 推想医疗科技股份有限公司 | Marking method, training method and device of training sample of image segmentation model |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102034232A (en) * | 2009-09-25 | 2011-04-27 | 上海电机学院 | Method for segmenting medical image graph |
| CN102903103A (en) * | 2012-09-11 | 2013-01-30 | 西安电子科技大学 | Migratory active contour model based stomach CT (computerized tomography) sequence image segmentation method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8526697B2 (en) * | 2008-02-15 | 2013-09-03 | Koninklijke Philips N.V. | Apparatus for segmenting an object comprising sub-objects |
-
2014
- 2014-12-29 CN CN201410840692.0A patent/CN105809656B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102034232A (en) * | 2009-09-25 | 2011-04-27 | 上海电机学院 | Method for segmenting medical image graph |
| CN102903103A (en) * | 2012-09-11 | 2013-01-30 | 西安电子科技大学 | Migratory active contour model based stomach CT (computerized tomography) sequence image segmentation method |
Non-Patent Citations (2)
| Title |
|---|
| The Combination of Local Fuzzy-Entropy-Based Transition Region Extraction with Otsu Thresholding Method for Image Segmentation;Aiying Lin等;《2009 2nd International Congress on Image and Signal Processing》;20091019;第1-4页 * |
| 基于可分解马尔科夫网的图像分割方法研究;曹建农;《中国博士学位论文全文数据库信息科技辑》;20060515(第05期);第37-107页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105809656A (en) | 2016-07-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wolterink et al. | Generative adversarial networks for noise reduction in low-dose CT | |
| Rossi et al. | Combination of low level processing and active contour techniques for semi-automated volumetric lung lesion segmentation from thoracic CT images | |
| WO2018120644A1 (en) | Blood vessel extraction method and system | |
| US9547894B2 (en) | Apparatus for, and method of, processing volumetric medical image data | |
| US9710915B2 (en) | Automatic background region selection for lesion delineation in medical images | |
| Schreier et al. | Clinical evaluation of a full-image deep segmentation algorithm for the male pelvis on cone-beam CT and CT | |
| del Fresno et al. | A combined region growing and deformable model method for extraction of closed surfaces in 3D CT and MRI scans | |
| CN113240719A (en) | Method and system for characterizing anatomical features in medical images | |
| Davis et al. | Assessment of 18F PET signals for automatic target volume definition in radiotherapy treatment planning | |
| CN106573150B (en) | The covering of image medium vessels structure | |
| Prasad et al. | Automatic segmentation of lung parenchyma in the presence of diseases based on curvature of ribs | |
| EP1761897A1 (en) | Displaying a tracheobronchial tree | |
| CN105809656B (en) | Medical image processing method and device | |
| Yim et al. | Correction of segmented lung boundary for inclusion of pleural nodules and pulmonary vessels in chest CT images | |
| Yasmin et al. | Brain image analysis: a survey | |
| Tan et al. | An approach for pulmonary vascular extraction from chest CT images | |
| US10062167B2 (en) | Estimated local rigid regions from dense deformation in subtraction | |
| Woźniak et al. | 3D vascular tree segmentation using a multiscale vesselness function and a level set approach | |
| Tobon-Gomez et al. | Automatic construction of 3D-ASM intensity models by simulating image acquisition: Application to myocardial gated SPECT studies | |
| Brabant et al. | EDART, a discrete algebraic reconstructing technique for experimental data obtained with high resolution computed tomography | |
| WO2009035559A1 (en) | Correction of intensity inhomogeneity in breast mri | |
| Mahdavi et al. | A computer-based analysis for identification and quantification of small airway disease in lung computed tomography images: a comprehensive review for radiologists | |
| Leube et al. | PSMA-PET improves deep learning-based automated CT kidney segmentation | |
| Sghaier et al. | A new approach for microcalcification enhancement in digital breast tomosynthesis reconstruction | |
| Oliveira et al. | Automatic Couinaud Liver and Veins Segmentation from CT Images. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP01 | Change in the name or title of a patent holder | ||
| CP01 | Change in the name or title of a patent holder |
Address after: 201807 No. 2258 Chengbei Road, Jiading Industrial Zone, Jiading District, Shanghai. Patentee after: Shanghai Lianying Medical Technology Co., Ltd Address before: 201807 No. 2258 Chengbei Road, Jiading Industrial Zone, Jiading District, Shanghai. Patentee before: SHANGHAI UNITED IMAGING HEALTHCARE Co.,Ltd. |