CN108520542B - Reconstruction method for time phase matching of PET/CT data - Google Patents
Reconstruction method for time phase matching of PET/CT data Download PDFInfo
- Publication number
- CN108520542B CN108520542B CN201810268898.9A CN201810268898A CN108520542B CN 108520542 B CN108520542 B CN 108520542B CN 201810268898 A CN201810268898 A CN 201810268898A CN 108520542 B CN108520542 B CN 108520542B
- Authority
- CN
- China
- Prior art keywords
- data
- image
- pet
- images
- attenuation
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/003—Reconstruction from projections, e.g. tomography
- G06T11/005—Specific pre-processing for tomographic reconstruction, e.g. calibration, source positioning, rebinning, scatter correction, retrospective gating
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2211/00—Image generation
- G06T2211/40—Computed tomography
- G06T2211/416—Exact reconstruction
Abstract
The invention discloses a reconstruction method for PET/CT data time phase matching, which directly carries out reconstruction containing attenuation coefficient on a total set of PET coincidence data to obtain a reconstructed image A, the reconstruction of the sub-coincidence data sets without attenuation coefficients yields N sets of images of the image data sets B1 to BN, normalizing the images while selecting a specific brightness for N sets of images B1-N, calculating demon estimation error formula through A, AT, Bn, BnT for any set of images N in B1-BN, taking the best matching PET gated image as a reference image, and performing matching operation on other images to obtain other images under the same gating, and performing image merging operation to obtain a merged image Bc, wherein the final result BF = BnAC/Bn Bc.
Description
Technical Field
The invention relates to the field of CT, in particular to a reconstruction method for PET/CT data time phase matching.
Background
PET/CT is a device that combines CT (X-ray computed tomography) and PET (positron emission tomography). CT imparts fundamental structural features to the body, and PET imparts potential pathological features to the body.
In PET respiratory reconstruction, the scanned coincidence data is divided into sub-data of a plurality of phases. Each sub-data corresponds to a particular breathing phase of the patient. There are two commonly used CT scanning methods: one is to directly utilize CT fast scanning, directly load the scanned CT result on PET, and the PET image and the CT image are easy to generate the problem of fusion dislocation. Another approach is to use a 4D scan mode in CT to acquire gating information. This approach can be matched by CT gating and PET gating to reduce motion artifacts, but increase the patient X-ray scan dose.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a PET/CT data time phase matching reconstruction method for improving the matching degree of a CT and a PET image.
In order to achieve the purpose, the invention provides the following technical scheme: a reconstruction method of PET/CT data time phase matching is characterized in that: comprises the following steps
(1) The CT image is converted into attenuation coefficient distribution in an image domain;
(2) converting the attenuation coefficient in the step (1) from an image domain to a projection domain through a projection algorithm, wherein the projection algorithm is obtained by performing path integration on attenuation values on paired detector paths according to PET detector coordinates and then performing exponential operation;
(3) the PET coincidence data is divided into a plurality of sub-coincidence data sets by gating;
(4) directly carrying out reconstruction containing attenuation coefficients on the total set of PET coincidence data to obtain a reconstructed image A;
(5) reconstructing the sub-coincidence data set without attenuation coefficients to obtain N groups of images of the image data sets B1 to BN;
(6) normalizing the image A, selecting specific brightness as a threshold point, normalizing the image, and selecting specific brightness for N groups of images B1-N to perform image brightness normalization;
(7) performing noise smoothing operation on B1-BN;
(8) B1-BN respectively carries out mutual information conversion on A to obtain N groups of B mutual information conversion images from B1T, B2T … to BNT and N groups of A mutual information conversion images from A1T … ANT;
for a particular luminance m and f on both sets of images, the mutual information is defined as
p (M, F) is the joint probability distribution of M and F, p (M) and p (F) are the distribution of images M and F in the gray scale domain, M and F are the original images, and after I (M, F) is calculated, the mutual information mapping images of M and F are obtained according to the calculated mutual information mapping images
Calculating;
(9) for any set of images n in B1-BN, computing demon estimation error formula through A, AT, Bn, BnT;
(10) extracting E1-EN multiple groups of error estimators, wherein the PET gate number where the minimum value is located is the corresponding optimal gate position in the CT image and PET multiple groups of data;
(11) matching other images by using the best matching PET gated image as a reference image to obtain other images under the same gate control, and performing image merging operation to obtain a merged image Bc;
(12) attenuation correction is carried out on the matched gating data in the PET data subset by the attenuation data in the step (2) to obtain BnAC, the optimal gating signal is reconstructed in the same way as in the step (4), attenuation correction is carried out, and the only difference is that the input PET coincidence data total set is replaced by the optimal gating coincidence signal;
(13) and the final result BF = BnAC/Bn Bc is used as a processing unit, and multiplication and division operations are carried out in the image domain in a one-to-one correspondence mode according to pixels, wherein the BnAC/Bn term is used as an attenuation factor, and certain smoothing operation is needed due to the difference of image noise.
The attenuation coefficient distribution in the further step (1) is obtained by subjecting the CT values to absorption coefficient conversion within the gamma photon frequency.
And (4) further, the gating area in the step (3) is divided into a plurality of subsets by finding out the corresponding respiration amplitude of each coincidence event through a respirator or directly through data analysis and then dividing the coincidence events through different amplitudes.
In the further step (4), reconstruction of the PET coincidence data aggregate containing attenuation coefficient is realized by adopting a filtering back projection algorithm or a fast maximum expectation iteration method, wherein the attenuation coefficient is realized by introducing the attenuation result in the step (2) into calculation.
And (5) reconstructing the sub-coincidence data set without attenuation coefficient by adopting a filtering back projection algorithm or a quick maximum expectation iteration method, wherein the step of not including the attenuation coefficient means that the attenuation result in the step (2) is not introduced into the calculation.
And (4) further normalizing the brightness of the image in the step (6) by taking the image mean value as the step of the threshold value, continuously increasing the threshold value until the threshold value is higher than 99% of non-zero points in the data and is used as the maximum value of the image, and then dividing the whole image by the threshold value, wherein the part higher than 1 is set as 1.
The noise smoothing operation on B1-BN in further step (7) uses gaussian convolution filtering, the size of the filtering kernel may vary with the pixel size of the image.
In a further step (8), any group of images M, F can generate mutual information maps MT, FT by formula, and since the gating generates N groups of B images in step 5, there is one group of images (Bn BnT a AnT) per group of images, resulting in N groups of BnT, AnT.
In the further step (9), (Bn BnT A Ant) corresponds to M, MT, F and FT in the formula, S in the formula is a motion deformation matrix, and U is an update matrix. The result of each set of data would be one estimate, so there are a total of N sets of estimates.
In a further step (11) an image domain registration method is used to register the other gated images to the optimal gating selected in step 10. And carrying out mean processing on the registered result to obtain Bc.
In conclusion, the beneficial effects of the invention are as follows:
1. rapid CT scanning is used without increasing the dose received by the patient.
2. The matching degree of the CT image and the PET image is improved, and the attenuation correction precision is also improved.
3. And the CT image is converted into a PET result, so that the analysis accuracy is improved.
4. And the mutual information is used for calculating data with different brightness, so that the reliability of analysis is improved.
Drawings
FIG. 1 is a flow chart of gated PET data selection;
FIG. 2 is a flow chart of phase matched reconstruction of PET/CT data.
Detailed Description
An embodiment of a reconstruction method for phase matching PET/CT data according to the present invention is further described with reference to fig. 1 and 2.
A reconstruction method of PET/CT data time phase matching is characterized in that: comprises the following steps
(1) The CT image is converted into attenuation coefficient distribution in an image domain;
(2) converting the attenuation coefficient in the step (1) from an image domain to a projection domain through a projection algorithm, wherein the projection algorithm is obtained by performing path integration on attenuation values on paired detector paths according to PET detector coordinates and then performing exponential operation;
(3) the PET coincidence data is divided into a plurality of sub-coincidence data sets by gating;
(4) directly carrying out reconstruction containing attenuation coefficients on the total set of PET coincidence data to obtain a reconstructed image A;
(5) reconstructing the sub-coincidence data set without attenuation coefficients to obtain N groups of images of the image data sets B1 to BN;
(6) normalizing the image A, selecting specific brightness as a threshold point, normalizing the image, and selecting specific brightness for N groups of images B1-N to perform image brightness normalization;
(7) performing noise smoothing operation on B1-BN;
(8) B1-BN respectively carries out mutual information conversion on A to obtain N groups of B mutual information conversion images from B1T, B2T … to BNT and N groups of A mutual information conversion images from A1T … ANT;
for a particular luminance m and f on both sets of images, the mutual information is defined as
p (M, F) is the joint probability distribution of M and F, p (M) and p (F) are the distribution of images M and F in the gray scale domain, M and F are the original images, and after I (M, F) is calculated, the mutual information mapping images of M and F are obtained according to the calculated mutual information mapping images
Calculating;
(9) for any set of images n in B1-BN, computing demon estimation error formula through A, AT, Bn, BnT;
(10) extracting E1-EN multiple groups of error estimators, wherein the PET gate number where the minimum value is located is the corresponding optimal gate position in the CT image and PET multiple groups of data;
(11) matching other images by using the best matching PET gated image as a reference image to obtain other images under the same gate control, and performing image merging operation to obtain a merged image Bc;
(12) attenuation correction is carried out on the matched gating data in the PET data subset by the attenuation data in the step (2) to obtain BnAC, the optimal gating signal is reconstructed in the same way as in the step (4), attenuation correction is carried out, and the only difference is that the input PET coincidence data total set is replaced by the optimal gating coincidence signal;
(13) and the final result BF = BnAC/Bn Bc is used as a processing unit, and multiplication and division operations are carried out in the image domain in a one-to-one correspondence mode according to pixels, wherein the BnAC/Bn term is used as an attenuation factor, and certain smoothing operation is needed due to the difference of image noise.
The attenuation coefficient distribution in the further step (1) is obtained by subjecting the CT values to absorption coefficient conversion within the gamma photon frequency.
The preferred gating area in step (3) of the present invention is to find the corresponding respiration amplitude of each coincidence event by a ventilator or directly by data analysis, and then divide the coincidence events by different amplitudes to form a plurality of subsets.
In the preferred step (4) of the invention, the reconstruction of the PET coincidence data aggregate containing attenuation coefficient is realized by adopting a filtering back projection algorithm or a fast maximum expectation iteration method, and the attenuation coefficient containing means introducing the attenuation result in the step (2) into the calculation.
In the preferred step (5) of the present invention, the reconstruction of the sub-coincidence data set without attenuation coefficient is realized by using a filtered back projection algorithm or a fast maximum expectation iteration method, and the non-attenuation coefficient means that the attenuation result in the step 2 is not introduced into the calculation.
The image brightness normalization in the step (6) preferably adopts the image mean value as the step of the threshold value, the threshold value is continuously raised until the threshold value is higher than 99% of non-zero points in the data and is used as the maximum value of the image, then the whole image is divided by the threshold value, and meanwhile, the part higher than 1 is set as 1.
The noise smoothing operation on B1-BN in the preferred step (7) of the present invention uses gaussian convolution filtering, and the size of the filtering kernel can vary with the pixel size of the image.
In the preferred step (8) of the present invention, any group of images M, F can generate mutual information mapping MT, FT by formula, and because the gating generates N groups of B images in step 5, each group of images has one group (Bn BnT a AnT), resulting in N groups of BnT, AnT.
In the preferred step (9) of the present invention, (Bn BnT A Ant) corresponds to M, MT, F, FT in the formula, S in the formula is the motion deformation matrix, and U is the update matrix. The result of each set of data would be one estimate, so there are a total of N sets of estimates.
In a preferred step (11) of the present invention, an image domain registration method is used to register other gated images to the optimal gating selected in step 10. And carrying out mean processing on the registered result to obtain Bc.
The invention uses rapid CT scanning, does not increase the dosage received by a patient, improves the matching degree of the CT and PET images, also improves the precision of attenuation correction, converts the CT image into the PET result, improves the accuracy of analysis, uses mutual information to calculate data with different brightness, and improves the reliability of analysis.
The above description is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments, and all technical solutions belonging to the idea of the present invention belong to the protection scope of the present invention. It should be noted that modifications and embellishments within the scope of the invention may occur to those skilled in the art without departing from the principle of the invention, and are considered to be within the scope of the invention.
Claims (9)
1. A reconstruction method of PET/CT data time phase matching is characterized in that: comprises the following steps
(1) The CT image is converted into attenuation coefficient distribution in an image domain;
(2) converting the attenuation coefficient in the step (1) from an image domain to a projection domain through a projection algorithm, wherein the projection algorithm is obtained by performing path integration on attenuation values on paired detector paths according to PET detector coordinates and then performing exponential operation;
(3) the PET coincidence data is divided into a plurality of sub-coincidence data sets by gating;
(4) directly carrying out reconstruction containing attenuation coefficients on the total set of PET coincidence data to obtain a reconstructed image A; the reconstruction of the total set of the PET coincidence data containing the attenuation coefficient is realized by adopting a filtering back projection algorithm or a rapid maximum expectation iteration method, wherein the attenuation coefficient is obtained by introducing the attenuation result in the step (2) into calculation;
(5) reconstructing the sub-coincidence data set without attenuation coefficients to obtain N groups of images of the image data sets B1 to BN;
(6) selecting specific brightness as a threshold value point, and carrying out normalization operation on the image A; then, respectively carrying out normalization operation on N groups of B1-BN images by selecting specific brightness as a threshold value point;
(7) performing noise smoothing operation on B1-BN;
(8) B1-BN respectively carries out mutual information conversion on A to obtain N groups of B mutual information conversion images from B1T, B2T … to BNT and N groups of A mutual information conversion images from A1T … ANT;
for a particular luminance m and f on both sets of images, the mutual information is defined as
p (M, F) is the joint probability distribution of M and F, p (M) and p (F) are the distribution of images M and F in the gray scale domain, M and F are the original images, and after I (M, F) is calculated, the mutual information mapping images of M and F are obtained according to the mapping conversion formula
Calculating;
(9) for any group of images n in B1-BN, substituting A into a mapping conversion formula to obtain AT, substituting Bn into a mapping conversion formula to obtain BnT, substituting A, AT, Bn and BnT into demons estimation error formula to obtain an estimation value;
(10) extracting E1-EN multiple groups of error estimators, wherein the PET gate number where the minimum value is located is the corresponding optimal gate position in the CT image and PET multiple groups of data;
(11) matching other images by using the best matching PET gated image as a reference image to obtain other images under the same gate control, and performing image merging operation to obtain a merged image Bc;
(12) attenuation correction is carried out on the matched gating data in the PET data subset by the attenuation data in the step (2) to obtain BnAC, the optimal gating signal is reconstructed in the same way as in the step (4), attenuation correction is carried out, and the only difference is that the input PET coincidence data total set is replaced by the optimal gating coincidence signal;
(13) and the final result BF = BnAC/Bn Bc is used as a processing unit, and multiplication and division operations are carried out in the image domain in a one-to-one correspondence mode according to pixels, wherein the BnAC/Bn term is used as an attenuation factor, and certain smoothing operation is needed due to the difference of image noise.
2. A method of time-matched reconstruction of PET/CT data as claimed in claim 1, characterized by: the attenuation coefficient distribution in step (1) is obtained by performing absorption coefficient conversion within the gamma photon frequency on the CT value.
3. A method of time-matched reconstruction of PET/CT data as claimed in claim 1, characterized by: and (4) the gating area in the step (3) is divided into a plurality of subsets by finding out the corresponding respiration amplitude of each coincidence event through a respirator or directly through data analysis and then dividing the coincidence events through different amplitudes.
4. A method of time-matched reconstruction of PET/CT data as claimed in claim 1, characterized by: and (5) reconstructing the sub-coincidence data set without attenuation coefficients by adopting a filtering back projection algorithm or a rapid maximum expectation iteration method, wherein the step of reconstructing the sub-coincidence data set without attenuation coefficients means that attenuation results in the step 2 are not introduced into calculation.
5. A method of time-matched reconstruction of PET/CT data as claimed in claim 1, characterized by: and (4) in the image brightness normalization in the step (6), the image mean value is used as the step of the threshold value, the threshold value is continuously increased until the threshold value is higher than 99% of non-zero points in the data and is used as the maximum value of the image, then the whole image is divided by the threshold value, and meanwhile, the part higher than 1 is set as 1.
6. A method of time-matched reconstruction of PET/CT data as claimed in claim 1, characterized by: the noise smoothing operation on B1-BN in step (7) uses gaussian convolution filtering, with the filter kernel size varying with the pixel size of the image.
7. A method of time-matched reconstruction of PET/CT data as claimed in claim 1, characterized by: in step (8), any group of images M, F can generate mutual information maps MT, FT by using a mapping transformation formula, and since the gating generates N groups of B images in step 5, there is one group of images (Bn BnT) in each group, resulting in N groups BnT.
8. A method of time-matched reconstruction of PET/CT data as claimed in claim 1, characterized by: in the step (9), (Bn BnT a AT) corresponds to M, MT, F, FT in the mapping conversion formula and demons error estimation formula, and the calculation result of each group of data is an estimation value through the rigid motion deformation matrix S and the updated elastic motion matrix U in the demons error estimation formula, so that there are N groups of estimation values.
9. A method of time-matched reconstruction of PET/CT data as claimed in claim 1, characterized by: and (3) adopting an image domain registration method in the step (11), registering other gated images to the optimal gate selected in the step (10), and carrying out mean processing on the registered result to obtain Bc.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810268898.9A CN108520542B (en) | 2018-03-29 | 2018-03-29 | Reconstruction method for time phase matching of PET/CT data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810268898.9A CN108520542B (en) | 2018-03-29 | 2018-03-29 | Reconstruction method for time phase matching of PET/CT data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108520542A CN108520542A (en) | 2018-09-11 |
| CN108520542B true CN108520542B (en) | 2020-10-09 |
Family
ID=63431193
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810268898.9A Active CN108520542B (en) | 2018-03-29 | 2018-03-29 | Reconstruction method for time phase matching of PET/CT data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108520542B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109242925B (en) * | 2018-09-13 | 2023-05-23 | 明峰医疗系统股份有限公司 | Respiratory artifact correction method based on gating data |
| CN109961419B (en) * | 2019-03-26 | 2021-05-07 | 江苏赛诺格兰医疗科技有限公司 | Correction information acquisition method for attenuation correction of PET activity distribution image |
| CN110223247B (en) * | 2019-05-20 | 2022-06-24 | 上海联影医疗科技股份有限公司 | Image attenuation correction method, device, computer equipment and storage medium |
| CN110192886B (en) * | 2019-05-29 | 2022-12-23 | 明峰医疗系统股份有限公司 | Reconstruction method for dynamically calling respiratory gating |
| CN110660111B (en) * | 2019-09-18 | 2023-05-30 | 东软医疗系统股份有限公司 | PET scattering correction and image reconstruction method, device and equipment |
| CN110687585B (en) * | 2019-09-23 | 2021-05-11 | 上海联影医疗科技股份有限公司 | Method, device, computer equipment and storage medium for obtaining crystal efficiency |
| CN112258642B (en) * | 2020-12-21 | 2021-04-09 | 之江实验室 | Low-dose PET data three-dimensional iterative updating reconstruction method based on deep learning |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103845073A (en) * | 2012-12-04 | 2014-06-11 | 美国西门子医疗解决公司 | MR scan selection for PET attenuation correction |
| CN106456098A (en) * | 2016-07-08 | 2017-02-22 | 上海联影医疗科技有限公司 | A generation method and system for attenuation images |
| CN206075356U (en) * | 2016-06-28 | 2017-04-05 | 明峰医疗系统股份有限公司 | The device of PET linear attenuation coefficients is obtained based on CT images |
| CN107638188A (en) * | 2017-09-28 | 2018-01-30 | 江苏赛诺格兰医疗科技有限公司 | Image attenuation bearing calibration and device |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR102111626B1 (en) * | 2013-09-10 | 2020-05-15 | 삼성전자주식회사 | Image processing apparatus and image processing method |
-
2018
- 2018-03-29 CN CN201810268898.9A patent/CN108520542B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103845073A (en) * | 2012-12-04 | 2014-06-11 | 美国西门子医疗解决公司 | MR scan selection for PET attenuation correction |
| CN206075356U (en) * | 2016-06-28 | 2017-04-05 | 明峰医疗系统股份有限公司 | The device of PET linear attenuation coefficients is obtained based on CT images |
| CN106456098A (en) * | 2016-07-08 | 2017-02-22 | 上海联影医疗科技有限公司 | A generation method and system for attenuation images |
| CN107638188A (en) * | 2017-09-28 | 2018-01-30 | 江苏赛诺格兰医疗科技有限公司 | Image attenuation bearing calibration and device |
Non-Patent Citations (1)
| Title |
|---|
| 基于互信息法的多模医学图像配准研究;朱祥胜;《中国优秀硕士学位论文全文数据库 信息科技辑》;20050915(第5期);I138-303 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108520542A (en) | 2018-09-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108520542B (en) | Reconstruction method for time phase matching of PET/CT data | |
| Dong et al. | Deep learning-based attenuation correction in the absence of structural information for whole-body positron emission tomography imaging | |
| US8774481B2 (en) | Atlas-assisted synthetic computed tomography using deformable image registration | |
| US9582906B2 (en) | Method for subjecting PET image to motion compensation and attenuation correction by using small number of low-radiation-dose CT images | |
| CN106846430B (en) | Image reconstruction method | |
| JP2017094097A (en) | Medical image processing device, x-ray computer tomographic imaging device, and medical image processing method | |
| CN113256753B (en) | PET image region-of-interest enhancement reconstruction method based on multitask learning constraint | |
| Cheng et al. | Applications of artificial intelligence in nuclear medicine image generation | |
| CN111436958B (en) | CT image generation method for PET image attenuation correction | |
| Zhi et al. | High‐quality initial image‐guided 4D CBCT reconstruction | |
| CN111161182B (en) | MR structure information constrained non-local mean guided PET image partial volume correction method | |
| US20220130079A1 (en) | Systems and methods for simultaneous attenuation correction, scatter correction, and de-noising of low-dose pet images with a neural network | |
| Ote et al. | Deep-learning-based fast TOF-PET image reconstruction using direction information | |
| CN117813055A (en) | Multi-modality and multi-scale feature aggregation for synthesis of SPECT images from fast SPECT scans and CT images | |
| US7242004B2 (en) | Image correction method, image correction apparatus, and image correction program | |
| CN112529977B (en) | PET image reconstruction method and system | |
| CN105184741A (en) | Three-dimensional CBCT (cone-beam computed tomography) image denoising method on the basis of improved nonlocal means | |
| CN115311258A (en) | Method and system for automatically segmenting organs in SPECT (single photon emission computed tomography) plane image | |
| Chen et al. | Low dose cone-beam computed tomography reconstruction via hybrid prior contour based total variation regularization (hybrid-PCTV) | |
| De Moor et al. | Non-rigid registration with position dependent rigidity for whole body PET follow-up studies | |
| US11854126B2 (en) | Methods and apparatus for deep learning based image attenuation correction | |
| US20230056685A1 (en) | Methods and apparatus for deep learning based image attenuation correction | |
| WO2023131061A1 (en) | Systems and methods for positron emission computed tomography image reconstruction | |
| CN109242925B (en) | Respiratory artifact correction method based on gating data | |
| Lindström Söraas | Respiratory Motion Correction in PET Imaging: Comparative Analysis of External Device and Data-driven Gating Approaches |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |