CN109785405A - The method for generating quasi- CT image using the multivariate regression of multiple groups magnetic resonance image - Google Patents
The method for generating quasi- CT image using the multivariate regression of multiple groups magnetic resonance image Download PDFInfo
- Publication number
- CN109785405A CN109785405A CN201910087064.2A CN201910087064A CN109785405A CN 109785405 A CN109785405 A CN 109785405A CN 201910087064 A CN201910087064 A CN 201910087064A CN 109785405 A CN109785405 A CN 109785405A
- Authority
- CN
- China
- Prior art keywords
- magnetic resonance
- image
- resonance image
- multiple groups
- target person
- 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.)
- Withdrawn
Links
Landscapes
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
A method of quasi- CT image is generated using the multivariate regression of multiple groups magnetic resonance image: obtaining multiple groups magnetic resonance image and corresponding one group of CT image respectively from several trainers, and the magnetic resonance image of the different sequence groups of each trainer is obtained using different magnetic resonance parameters；Magnetic resonance image in the multiple groups magnetic resonance image of different trainers positioned at same sequence group is obtained using identical magnetic resonance parameters；Each image in every group of magnetic resonance image of the same position of each trainer is aligned with corresponding CT image；Generate the mapping function that identical voxel CT value is corresponded to from the intensity value of the multiple groups magnetic resonance image of all trainers；The quasi- CT image of target person is generated from the multiple groups magnetic resonance image of target person.The present invention is not needed trainer's image and target person's image alignment, thus avoids error caused by alignment procedures in map set method.The present invention can be used for simulation calculating and the medical imaging of guided by magnetic resonance radiotherapy treatment planning.
Description
Technical field
The present invention relates to a kind of generations of CT image.It is returned more particularly to a kind of multivariable using multiple groups magnetic resonance image
Return the method for generating quasi- CT image.
Background technique
Computed tomography (CT) and magnetic resonance imaging (MRI) are two kinds of main imaging moulds for obtaining 3-D image
Formula.Computed tomography is conventionally used for the imaging pattern of creation radiotherapy treatment planning.CT image can be accurately
The geometry of target person is described, and the corresponding CT value of CT image can be directly changed into electron density for calculating target
The intracorporal Radiation dose distribution of person.However, CT image does not have good contrast for soft tissue, and CT also meets target person
By additional dose of radiation.Compared with CT image, magnetic resonance imaging mode also has extensively due to its excellent soft tissue contrast
General purposes.Magnetic resonance imaging is free of ionising radiation, and the functional informations such as metabolism that can provide target person.
At present magnetic resonance image be mainly used for supplement CT image to obtain more accurate anatomical structure profile and cancer target,
So needing for the magnetic resonance image of target person to be aligned with corresponding CT image.Since magnetic resonance image and CT image usually exist
It is obtained on different machines, therefore magnetic resonance image and CT image will not be aligned completely in overlapping；This is cancer target positioning
One important sources of error.Guided by magnetic resonance radiotherapy treatment planning utilizes magnetic resonance image without CT image, so swollen
Tumor targeting can be more acurrate.Since magnetic resonance intensity value cannot be converted directly into electron density, it is therefore desirable to there is method by magnetic resonance
Image is accurately converted to image corresponding with electron density value, i.e., quasi- CT image, also referred to as pCT image or derivative CT figure
Picture.
Journal of writings " the A Review of Substitute CT Generation for of Edmund and Nyholm
MRI-only Radiation Therapy ", RadiatOncol 12:28 (2017), doi:10.1186/s13014-016-
0747-y has commented on various methods used by creation pCT, including atlas (atlas) method and voxel (voxel) method.
Such as journal of writings " the An Atlas-based Electron Density Mapping Method for of Dowling et al.
Magnetic Resonance Imaging(MRI)-Alone Treatment Planning and Adaptive MRI-
Based Prostate Radiation Therapy ", Int J RadiatOncolBiol Phys 83,5 (2012), doi:
Described in 10.1016/j.ijrobp.2011.11.056, map set method is using as the pre-existing atlas figure of reference
As helping to generate quasi- CT image.During generating quasi- CT image, atlas magnetic resonance image and atlas CT image are
The reference of quasi- CT image is generated from new target person's magnetic resonance image.Atlas magnetic resonance image needs the magnetic resonance with target person
Image alignment, and identical alignment transformation is applied to the atlas CT image co-registration of same position to scheme at the quasi- CT of target person
As during.However, alignment procedures of atlas image and from atlas magnetic resonance image to target person's magnetic resonance image
Alignment procedures can all have some errors.Voxel method is created using from magnetic resonance image intensity value to the conversion method of CT value
Quasi- CT image is built, without the magnetic resonance image of target person and trainer to be aligned.
Summary of the invention
The technical problem to be solved by the invention is to provide a kind of multivariate regression generations using multiple groups magnetic resonance image
The method of quasi- CT image.
The technical scheme adopted by the invention is that: it is a kind of to be schemed using the quasi- CT of multivariate regression generation of multiple groups magnetic resonance image
The method of picture, includes the following steps:
1) multiple groups magnetic resonance image (MRI) and corresponding one group of CT image are obtained respectively from several trainers, each
The magnetic resonance image of the different sequence groups of trainer is obtained using different magnetic resonance parameters；The multiple groups magnetic of different trainers is total
Being located at the magnetic resonance image of same sequence group in vibration image is obtained using identical magnetic resonance parameters；
2) by each image and the corresponding CT image pair in every group of magnetic resonance image of the same position of each trainer
It is quasi-；
3) the mapping letter that identical voxel CT value is corresponded to from the intensity value of the multiple groups magnetic resonance image of all trainers is generated
Number；
4) the quasi- CT image of target person is generated from the multiple groups magnetic resonance image of target person.
When the magnetic resonance image that trainers different in step 1) are located at same sequence group be using different scanning machines with/
Or different image-forming condition will be standardized the magnetic resonance image, make to be located at same row in all trainers when obtaining
The average intensity value of the magnetic resonance image of sequence group is identical.
Step 3) includes:
(3.1) human body contour outline is divided to each magnetic resonance image and CT image respectively；
(3.2) region segmentation mask is established to the magnetic resonance image of the first sequence group；
(3.3) magnetic for other sequence groups that obtained region segmentation mask is used for other than same position the first sequence group is total to
Region segmentation is carried out on vibration image and CT image；
(3.4) corresponding magnetic resonance image intensity value and CT value are extracted from the voxel in each region other than exclusion zone；
(3.5) voxel according to multiple groups magnetic resonance image intensity value and with identical multiple groups magnetic resonance image intensity value is flat
Equal CT value determines the mapping function of the higher polynomial in each region using multivariate regression method:
Wherein N is multinomial highest item number, CT (S1..., Sm) be mapping function be dependent variable, S1It is that first group of magnetic is total
Vibration image intensity value is independent variable, S in mapping functionmIt is m group magnetic resonance image intensity value, for certainly in mapping function
Variable, i1It is S1Index, imIt is SmIndex,It is fitting coefficient.
Division human body contour outline described in (3.1) step is to separate organization of human body in image and surrounding air.
In (3.2) step, when there was only bone or soft tissue in the magnetic resonance image of the first sequence group, the magnetic is total to
Image shake as a region；It, will be described when there is Bone and soft tissue simultaneously in the magnetic resonance image of the first sequence group
Magnetic Resonance Image Segmentation is bony areas, soft tissue area and line and staff control region, and the line and staff control region is phalanges
The boundary of uncertain part between bone and soft tissue, each region constitutes region segmentation mask.
In (3.2) step, by the multiple groups magnetic resonance image of same position and CT image because of organ daily routines or figure
The region of anatomical structure misalignment and the region other than human body contour outline is caused to be set as exclusion zone as being aligned insufficient.
Step 4) includes:
(4.1) each group magnetic resonance figure of target person is obtained using magnetic resonance parameters identical with the same sequence group of trainer
Picture；
(4.2) the multiple groups magnetic resonance image of target person's same position is mutually aligned；
(4.3) when the magnetic resonance image of target person is using the scanning machine different from trainer and/or different imagings
When condition obtains, the magnetic resonance image of target person is standardized, make the same of target person and all trainer's same positions
The average intensity value of the magnetic resonance image of one sequence group is identical；
(4.4) use (3.2) in step 3), region segmentation mode described in (3.3) step to target person's the first sequence group
Magnetic resonance image carry out region segmentation and establishing region segmentation mask, and obtain target person's same position the first sequence group with
The region of the magnetic resonance image of other outer sequence groups, wherein the image of target person is not provided with exclusion zone；
(4.5) multiple groups magnetic resonance image intensity value is extracted to the voxel in target person's magnetic resonance image；
(4.6) intensity value of the multiple groups magnetic resonance image of each voxel of target person is obtained by step 3) (3.5) step
Corresponding region mapping function, obtain the CT value of the identical voxel of target person；
(4.7) set of the CT value of all voxels of target person is constituted to the quasi- CT image of target person.
The method that multivariate regression using multiple groups magnetic resonance image of the invention generates quasi- CT image, uses multivariable letter
Number is corresponding at quasi- CT image by the multiple groups magnetic resonance image of target person.The accuracy of the method is existing best with map set method
As a result similar, but it is more more convenient than map set method and quick.The present invention uses voxel method by the magnetic resonance figure of target person
As being converted directly into quasi- CT image.It is not needed in this approach by trainer's image and target person's image alignment, thus avoided
Error caused by alignment procedures in map set method.The present invention can be used for guided by magnetic resonance radiotherapy treatment planning simulation calculate and
Medical imaging.
Detailed description of the invention
Fig. 1 is the exemplary stream that the method for regression model is determined using the multiple groups magnetic resonance image and CT image of trainer
Cheng Tu；
Fig. 2 is the exemplary process diagram that quasi- CT image is generated using the multiple groups magnetic resonance image of target person.
Specific embodiment
Quasi- CT is generated to the multivariate regression of the invention using multiple groups magnetic resonance image below with reference to embodiment and attached drawing
The method of image is described in detail.
When describing illustrative examples, used attached drawing and concrete term merely to for the sake of clear, there is no restriction this
The purpose of description.
The method that multivariate regression using multiple groups magnetic resonance image of the invention generates quasi- CT image, including walk as follows
It is rapid:
1) multiple groups magnetic resonance image (MRI) and corresponding one group of CT image are obtained respectively from several trainers, each
The magnetic resonance image of the different sequence groups of trainer is obtained using different magnetic resonance parameters；The multiple groups magnetic of different trainers is total
Being located at the magnetic resonance image of same sequence group in vibration image is obtained using identical magnetic resonance parameters；
When the magnetic resonance image that different trainers are located at same sequence group is to use different scanning machine and/or different
When image-forming condition obtains, the magnetic resonance image is standardized, make the magnetic for being located at same sequence group in all trainers
The average intensity value of resonance image is identical.Standardization is in order to ensure magnetic resonance image intensity value collected is for different instructions
White silk person is in harmony certainly.It is a kind of to decide whether that standardized straightforward procedure is voxel in the same area to different trainers
Magnetic resonance image intensity value does histogram and is compared；If the histogram of different trainers shows similar shape but very not
Same peak value and average value, then need to standardize.During exemplary standardized, first finds each trainer and be located at same row
The average intensity value of the magnetic resonance image of sequence group, then determine the correction factor of each trainer come multiplied by corresponding trainer this
The magnetic resonance image intensity value of voxel in the magnetic resonance image of sequence group, to make the magnetic of this sequence group of all trainers total
The average intensity value of vibration image is the same.
2) by each image and the corresponding CT image pair in every group of magnetic resonance image of the same position of each trainer
It is quasi-.Magnetic resonance image and CT image usually obtain on different machines, therefore magnetic resonance image and CT image are in overlapping
It will not be perfectly aligned.So needing to be aligned the magnetic resonance image and corresponding position of each trainer by image alignment technique
CT image, to strive for that each voxel of trainer's magnetic resonance image and CT image is allowed to correspond to each other.
3) the mapping letter that identical voxel CT value is corresponded to from the intensity value of the multiple groups magnetic resonance image of all trainers is generated
Number；Include:
(3.1) human body contour outline is divided to each magnetic resonance image and CT image respectively；The division human body contour outline is
Organization of human body in image and surrounding air are separated.
(3.2) region segmentation mask is established to the magnetic resonance image of the first sequence group；Wherein,
When there was only bone or soft tissue in the magnetic resonance image of the first sequence group, using the magnetic resonance image as one
A region；When there is Bone and soft tissue simultaneously in the magnetic resonance image of the first sequence group, by the magnetic resonance image point
Be segmented into bony areas, soft tissue area and line and staff control region, the line and staff control region refer to Bone and soft tissue it
Between uncertain part, the boundary in each region constitutes region segmentation mask；And by the multiple groups magnetic resonance image of same position
With in CT image because the daily routines of organ or image are aligned it is insufficient due to lead to region and the human body of anatomical structure misalignment
Region other than profile is set as exclusion zone.
(3.3) magnetic for other sequence groups that obtained region segmentation mask is used for other than same position the first sequence group is total to
Region segmentation is carried out on vibration image and CT image；
(3.4) corresponding magnetic resonance image intensity value and CT value are extracted from the voxel in each region other than exclusion zone；
(3.5) voxel according to multiple groups magnetic resonance image intensity value and with identical multiple groups magnetic resonance image intensity value is flat
Equal CT value determines the mapping function of the higher polynomial in each region using multivariate regression method:
Wherein N is multinomial highest item number, CT (S1..., Sm) be mapping function be dependent variable, S1It is that first group of magnetic is total
Vibration image intensity value is independent variable, S in mapping functionmIt is m group magnetic resonance image intensity value, for certainly in mapping function
Variable, i1It is S1Index, imIt is SmIndex,It is fitting coefficient.
4) the quasi- CT image of target person is generated from the multiple groups magnetic resonance image of target person.Include:
(4.1) each group magnetic resonance figure of target person is obtained using magnetic resonance parameters identical with the same sequence group of trainer
Picture.In the preferred case, the magnetic resonance image of target person and trainer are generated by identical magnetic resonance scanner；In other realities
In example, the magnetic resonance image of trainer and target person can be generated by different magnetic resonance scanners.
(4.2) the multiple groups magnetic resonance image of target person's same position is mutually aligned；
(4.3) when the magnetic resonance image of target person is using the scanning machine different from trainer and/or different imagings
When condition obtains, the magnetic resonance image of target person is standardized, make the same of target person and all trainer's same positions
The average intensity value of the magnetic resonance image of one sequence group is identical；
(4.4) use (3.2) in step 3), region segmentation mode described in (3.3) step to target person's the first sequence group
Magnetic resonance image carry out region segmentation and establishing region segmentation mask, and obtain target person's same position the first sequence group with
The region of the magnetic resonance image of other outer sequence groups, wherein the image of target person is not provided with exclusion zone；
(4.5) multiple groups magnetic resonance image intensity value is extracted to the voxel in target person's magnetic resonance image；
(4.6) intensity value of the multiple groups magnetic resonance image of each voxel of target person is obtained by step 3) (3.5) step
Corresponding region mapping function, obtain the CT value of the identical voxel of target person；
(4.7) set of the CT value of all voxels of target person is constituted to the quasi- CT image of target person.
A specific example is given below:
Fig. 1 is the exemplary stream that the method for regression model is determined using the multiple groups magnetic resonance image and CT image of trainer
Cheng Tu.
In step 110, training data is that the two groups of magnetic obtained using identical magnetic resonance scanner from multiple trainers are total to
The CT of vibration (referred to as MRI1 group and MRI2 group) image and the corresponding site obtained with identical CT scanner from these trainers
Image.Magnetic resonance image in the multiple groups magnetic resonance image of different trainers positioned at same sequence group is using identical magnetic resonance
Parameter and image-forming condition obtain.The magnetic resonance image of the different sequence groups of each trainer is obtained using different magnetic resonance parameters
It takes, for example obtained using the magnetic resonance image sequential parameter (T1 is weighted, T2 weighting) with different contrast attribute.Each instruction
The data of white silk person include two groups of magnetic resonance image and one group of corresponding CT image for use as basic fact.Due to different trainer positions
In the magnetic resonance image of same sequence group be to be obtained using identical magnetic resonance parameters and image-forming condition, so not needing to instruction
The magnetic resonance image of white silk person is standardized.
In step 120, magnetic resonance image and CT image usually obtain on different machines, thus magnetic resonance image and
CT image will not be perfectly aligned in overlapping.So needing to be aligned the magnetic resonance figure of each trainer by image alignment technique
The CT image of picture and corresponding position, to strive for that each voxel of trainer's magnetic resonance image and CT image is allowed to correspond to each other.
In step 130, human body contour outline is divided to each magnetic resonance image and CT image respectively first, by human body in image
Structure is separated with surrounding air, then carries out region segmentation for each image.First carried out for the magnetic resonance image of the first sequence group
Region segmentation.When there was only bone in the magnetic resonance image of the first sequence group or when soft tissue, using the magnetic resonance image as
One region；When there is Bone and soft tissue simultaneously in the magnetic resonance image of the first sequence group, by the magnetic resonance image
It is divided into bony areas, three regions of soft tissue area and line and staff control.Bony areas only includes determining skeleton dissection knot
Structure, including cortex and spongy (cancellous bone) skeletal structure.Soft tissue area only includes the non-instrument of skeleton anatomical structure of all determinations.
The remaining area retained between the region (i.e. uncertain region) that bony areas and soft tissue area contact is area, line and staff control
Domain.In addition, by the multiple groups magnetic resonance image of same position and CT image because of organ daily routines or image alignment it is insufficient
And the region of anatomical structure misalignment and the region other than human body contour outline is caused to be set as exclusion zone.Region segmentation can be with hand
Trainer's profile completion that is dynamic or being generated using computer.The boundary in each region in the magnetic resonance image of the first sequence group
Constitute region segmentation mask；Then other sequences obtained region segmentation mask being used for other than same position the first sequence group
Region segmentation is carried out in the magnetic resonance image and CT image of group.
In step 140, from the magnetic resonance image intensity value and CT value of each image zooming-out voxel of training data.Every individual
The data of element are the magnetic resonance image intensity values by the magnetic resonance image intensity value from MRI1 group, from MRI2 group, and corresponding
CT value composition triple.The data acquisition system extracted from the voxel of each trainer given area is together；But in exclusion zone
Voxel be excluded the numerical value extraction except, to guarantee the true association between magnetic resonance image intensity value and CT value.
Step 150 carries out multivariate regression to each region, i.e., is fitted with the function of two groups of magnetic resonance image intensity values
Corresponding mean CT-number (the i.e. voxel of training data the same area being averaged in two groups of magnetic resonance image intensity value sections
Value).Multivariate regression in this example uses the mapping function of following binary high-order moment:
Wherein multinomial highest item times N=30, CT (S1, S2) be mapping function be dependent variable, MRI1 group magnetic resonance image
Intensity value S1With MRI2 group magnetic resonance image intensity value S2It is independent variable,It is fitting coefficient.
Fig. 2 is the exemplary process diagram that quasi- CT image is generated using the multiple groups magnetic resonance image of target person.
In step 210, the magnetic resonance scanner and use and the same sequence group of trainer with trainer's same model are used
Identical magnetic resonance parameters and image-forming condition obtain two groups of magnetic resonance image (MRI1 group and MRI2 group) of target person.Due to target
The magnetic resonance image of person is obtained using the magnetic resonance scanner and image-forming condition with trainer's same model, so not needing
The magnetic resonance image of target person is standardized.
In step 220, will be mutually aligned in two groups of magnetic resonance image of the same position of target person.
In step 230, using the region segmentation mode described in step 130 to the magnetic resonance image of target person's the first sequence group
It carries out region segmentation and establishes region segmentation mask, and obtain the magnetic resonance image of target person's same position the second sequence group
Region；The image of target person is not provided with exclusion zone.
In step 240, the magnetic resonance image intensity value of each voxel of target person is extracted, i.e., from the magnetic resonance figure of MRI1 group
As intensity value and from the magnetic resonance image intensity value of MRI2 group.
In step 250, the multi-variable function in each region determined in step 150 is applied to the correspondence area of target person
Domain.It first uses two groups of magnetic resonance image intensity values of each voxel of target person as independent variable, then utilizes in step 150
The binary high-order moment of this determining voxel region obtains CT value of the dependent variable as this voxel.
In step 260, the whole of quasi- CT value corresponding to each voxel of target person constitutes quasi- CT image.The image
It can be used for simulation calculating and the medical imaging of target person's guided by magnetic resonance radiotherapy treatment planning.
Claims (7)
1. a kind of method that multivariate regression using multiple groups magnetic resonance image generates quasi- CT image, which is characterized in that including such as
Lower step:
1) multiple groups magnetic resonance image (MRI) and corresponding one group of CT image, each training are obtained respectively from several trainers
The magnetic resonance image of the different sequence groups of person is obtained using different magnetic resonance parameters；The multiple groups magnetic resonance figure of different trainers
Being located at the magnetic resonance image of same sequence group as in is obtained using identical magnetic resonance parameters；
2) each image in every group of magnetic resonance image of the same position of each trainer is aligned with corresponding CT image；
3) mapping function that identical voxel CT value is corresponded to from the intensity value of the multiple groups magnetic resonance image of all trainers is generated；
4) the quasi- CT image of target person is generated from the multiple groups magnetic resonance image of target person.
2. the method that the multivariate regression according to claim 1 using multiple groups magnetic resonance image generates quasi- CT image,
It is characterized in that, is using different scanning machines when trainers different in step 1) are located at the magnetic resonance image of same sequence group
And/or different image-forming condition will be standardized the magnetic resonance image, make to be located in all trainers same when obtaining
The average intensity value of the magnetic resonance image of sequence group is identical.
3. the method that the multivariate regression according to claim 1 using multiple groups magnetic resonance image generates quasi- CT image,
It is characterized in that, step 3) includes:
(3.1) human body contour outline is divided to each magnetic resonance image and CT image respectively；
(3.2) region segmentation mask is established to the magnetic resonance image of the first sequence group；
(3.3) obtained region segmentation mask is used for the magnetic resonance figure of other sequence groups other than same position the first sequence group
Region segmentation is carried out on picture and CT image；
(3.4) corresponding magnetic resonance image intensity value and CT value are extracted from the voxel in each region other than exclusion zone；
(3.5) the average CT of the voxel according to multiple groups magnetic resonance image intensity value and with identical multiple groups magnetic resonance image intensity value
Value determines the mapping function of the higher polynomial in each region using multivariate regression method:
Wherein N is multinomial highest item number, CT (S1..., Sm) be mapping function be dependent variable, S1It is first group of magnetic resonance figure
It is independent variable, S in mapping function as intensity valuemIt is m group magnetic resonance image intensity value, is independent variable in mapping function,
i1It is S1Index, imIt is SmIndex,It is fitting coefficient.
4. the method that the multivariate regression according to claim 2 using multiple groups magnetic resonance image generates quasi- CT image,
It is characterized in that, division human body contour outline described in (3.1) step, is to separate organization of human body in image and surrounding air.
5. the method that the multivariate regression according to claim 3 using multiple groups magnetic resonance image generates quasi- CT image,
It is characterized in that, in (3.2) step, when there was only bone or soft tissue in the magnetic resonance image of the first sequence group, by the magnetic
Resonance image is as a region；It, will be described when there is Bone and soft tissue simultaneously in the magnetic resonance image of the first sequence group
Magnetic Resonance Image Segmentation be bony areas, soft tissue area and line and staff control region, the line and staff control region refers to
The boundary of uncertain part between Bone and soft tissue, each region constitutes region segmentation mask.
6. the method that the multivariate regression according to claim 3 using multiple groups magnetic resonance image generates quasi- CT image,
Be characterized in that, in (3.2) step, by the multiple groups magnetic resonance image of same position and CT image because of organ daily routines or figure
The region of anatomical structure misalignment and the region other than human body contour outline is caused to be set as exclusion zone as being aligned insufficient.
7. the method that the multivariate regression according to claim 1 using multiple groups magnetic resonance image generates quasi- CT image,
It is characterized in that, step 4) includes:
(4.1) each group magnetic resonance image of target person is obtained using magnetic resonance parameters identical with the same sequence group of trainer；
(4.2) the multiple groups magnetic resonance image of target person's same position is mutually aligned；
(4.3) when the magnetic resonance image of target person is using the scanning machine different from trainer and/or different image-forming conditions
When acquisition, the magnetic resonance image of target person is standardized, make the same row of target person Yu all trainer's same positions
The average intensity value of the magnetic resonance image of sequence group is identical；
(4.4) using region segmentation mode described in (3.2) in step 3), (3.3) step to the magnetic of target person's the first sequence group
Resonance image carries out region segmentation and establishes region segmentation mask, and obtains other than target person's same position the first sequence group
The region of the magnetic resonance image of other sequence groups, wherein the image of target person is not provided with exclusion zone；
(4.5) multiple groups magnetic resonance image intensity value is extracted to the voxel in target person's magnetic resonance image；
(4.6) phase for obtaining the intensity value of the multiple groups magnetic resonance image of each voxel of target person by step 3) (3.5) step
The mapping function of corresponding region obtains the CT value of the identical voxel of target person；
(4.7) set of the CT value of all voxels of target person is constituted to the quasi- CT image of target person.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910087064.2A CN109785405A (en) | 2019-01-29 | 2019-01-29 | The method for generating quasi- CT image using the multivariate regression of multiple groups magnetic resonance image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910087064.2A CN109785405A (en) | 2019-01-29 | 2019-01-29 | The method for generating quasi- CT image using the multivariate regression of multiple groups magnetic resonance image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109785405A true CN109785405A (en) | 2019-05-21 |
Family
ID=66503634
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910087064.2A Withdrawn CN109785405A (en) | 2019-01-29 | 2019-01-29 | The method for generating quasi- CT image using the multivariate regression of multiple groups magnetic resonance image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109785405A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104794739A (en) * | 2015-05-03 | 2015-07-22 | 南方医科大学 | Method for predicting CT (computerized tomography) image from MR (magnetic resonance) image on the basis of combination of corresponding partial sparse points |
| CN108351395A (en) * | 2015-10-27 | 2018-07-31 | 皇家飞利浦有限公司 | Virtual CT images from magnetic resonance image |
| CN108770373A (en) * | 2015-10-13 | 2018-11-06 | 医科达有限公司 | It is generated according to the pseudo- CT of MR data using feature regression model |
| CN108778416A (en) * | 2015-10-13 | 2018-11-09 | 医科达有限公司 | It is generated according to the pseudo- CT of MR data using organizational parameter estimation |
| CN109272486A (en) * | 2018-08-14 | 2019-01-25 | 中国科学院深圳先进技术研究院 | Training method, device, equipment and the storage medium of MR image prediction model |
-
2019
- 2019-01-29 CN CN201910087064.2A patent/CN109785405A/en not_active Withdrawn
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104794739A (en) * | 2015-05-03 | 2015-07-22 | 南方医科大学 | Method for predicting CT (computerized tomography) image from MR (magnetic resonance) image on the basis of combination of corresponding partial sparse points |
| CN108770373A (en) * | 2015-10-13 | 2018-11-06 | 医科达有限公司 | It is generated according to the pseudo- CT of MR data using feature regression model |
| CN108778416A (en) * | 2015-10-13 | 2018-11-09 | 医科达有限公司 | It is generated according to the pseudo- CT of MR data using organizational parameter estimation |
| CN108351395A (en) * | 2015-10-27 | 2018-07-31 | 皇家飞利浦有限公司 | Virtual CT images from magnetic resonance image |
| CN109272486A (en) * | 2018-08-14 | 2019-01-25 | 中国科学院深圳先进技术研究院 | Training method, device, equipment and the storage medium of MR image prediction model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | Evaluation of a deep learning-based pelvic synthetic CT generation technique for MRI-based prostate proton treatment planning | |
| Liu et al. | MRI-based treatment planning for liver stereotactic body radiotherapy: validation of a deep learning-based synthetic CT generation method | |
| US11455732B2 (en) | Knowledge-based automatic image segmentation | |
| US8787648B2 (en) | CT surrogate by auto-segmentation of magnetic resonance images | |
| Hill et al. | A strategy for automated multimodality image registration incorporating anatomical knowledge and imager characteristics | |
| CN108770373B (en) | pseudo-CT generation from MR data using a feature regression model | |
| US7817836B2 (en) | Methods for volumetric contouring with expert guidance | |
| US10223794B2 (en) | Method and device for generating one or more computer tomography images based on magnetic resonance images with the help of tissue class separation | |
| US9082169B2 (en) | Longitudinal monitoring of pathology | |
| Whitfield et al. | Automated delineation of radiotherapy volumes: are we going in the right direction? | |
| Wu et al. | AAR-RT–a system for auto-contouring organs at risk on CT images for radiation therapy planning: principles, design, and large-scale evaluation on head-and-neck and thoracic cancer cases | |
| CN107865658A (en) | Method and apparatus for correcting synthesis electron-density map | |
| CN110946652A (en) | Method and device for planning screw path of bone screw | |
| Cheng et al. | Airway segmentation and measurement in CT images | |
| US10453224B1 (en) | Pseudo-CT generation with multi-variable regression of multiple MRI scans | |
| Alam et al. | Medical image registration: Classification, applications and issues | |
| Wang et al. | Deformable torso phantoms of Chinese adults for personalized anatomy modelling | |
| CN105339804B (en) | Divided according to the cortex bone of MR Dixon data | |
| Vaassen et al. | Real-world analysis of manual editing of deep learning contouring in the thorax region | |
| CN114881848A (en) | Method for converting multi-sequence MR into CT | |
| CN109785405A (en) | The method for generating quasi- CT image using the multivariate regression of multiple groups magnetic resonance image | |
| CN116168097A (en) | Method, device, equipment and medium for constructing CBCT sketching model and sketching CBCT image | |
| WO2022185222A1 (en) | Methods and apparatuses for visualization of tumor segmentation | |
| Chourak et al. | Voxel-wise analysis for spatial characterisation of Pseudo-CT errors in MRI-only radiotherapy planning | |
| Rachmawati et al. | Bone Scan Image Segmentation based on Active Shape Model for Cancer Metastasis Detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190521 |
|
| WW01 | Invention patent application withdrawn after publication |