CN105259525A - Dynamic contrast enhanced magnetic resonance fast imaging method based on neighborhood sharing compression sensing - Google Patents
Dynamic contrast enhanced magnetic resonance fast imaging method based on neighborhood sharing compression sensing Download PDFInfo
- Publication number
- CN105259525A CN105259525A CN201510711789.6A CN201510711789A CN105259525A CN 105259525 A CN105259525 A CN 105259525A CN 201510711789 A CN201510711789 A CN 201510711789A CN 105259525 A CN105259525 A CN 105259525A
- Authority
- CN
- China
- Prior art keywords
- space
- sampling
- neighborhood
- shared
- compressed sensing
- 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.)
- Granted
Links
Landscapes
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The present invention provides a K space data rapid acquisition and reconstruction method which integrates compressed sensing sampling technology to a neighborhood sharing sampling trajectory. The method is used for dynamic contrast enhanced magnetic resonance fast imaging. The method comprises the random division of a K space, compressed sensing sampling matrix design, neighborhood shared compressed sensing sampling matrix acquisition, data collection based on a neighborhood shared compressed sensing method, and data reconstruction based on the neighborhood shared compressed sensing method. According to a sequence, the scanning speed can be improved further based on neighborhood sharing, a higher imaging time resolution and a higher image signal-noise ratio are obtained, the high image quality is ensured, and the method is more suitable for clinical dynamic contrast enhanced magnetic resonance imaging.
Description
Technical field
The present invention is a kind of integrated compressed sensing (compressedsensing, CS) neighborhood of technology shares (viewsharing) K space sampling trajectories optimization acquisition method, belong to magnetic resonance medical imaging technology field, internal organs Dynamic constrasted enhancement (DCE-MRI) image of high time resolution and high s/n ratio can be provided, may be used for the medical diagnosis on disease of abdominal organs etc., functional parameter quantitative measurment and prognostic evaluation etc.
Background technology
Chronic kidney disease and liver cancer become major global public health problem, and in developed country, the chronic kidney disease morbidity rate of general population, up to 6.5% ~ 16%, brings heavy burden to global Health Investment.According to " Chinese chronic kidney disease epidemiology survey " result display in 2012, in China's Adult Groups, the morbidity rate of chronic kidney disease was up to 10.8%, estimates existing adult Patients with Chronic Kidney Disease 1.2 hundred million people in the whole nation.In addition, the onset of liver cancer rate of developed country general population is up to ASR2.7 ~ 8.6, and the incidence of disease is improving year by year, and mortality is up to ASR2.5 ~ 7.1.Liver cancer has become the lethal second largest disease of male cancer in world wide, only the newly-increased liver cancer patient 780,000 in the whole world in 2012, and dead 740,000, wherein, Chinese liver cancer number of patients and death toll account for 50%.In belly clinical diagnosis, the Dynamic constrasted enhancement inspection based on mr imaging technique is the effective means of diagnosis of kidney disease and liver diseases.Therefore, improve temporal resolution and the picture quality of belly Dynamic constrasted enhancement magnetic resonance imaging, to reach organ function quantitative measurment more accurately, the clinical diagnosis of abdominal organs disease and prognostic evaluation are had very important significance.
In belly Dynamic constrasted enhancement DCE-MRI imaging, because 3 Dimension Image Technique better can show the form of the organs such as whole kidney, liver, and provide more image information and perfusion curve, thus by wide clinical application.At present, in clinical, three-dimensional Fast spoiled gradient echo 3DFSPGR sequence and three-dimensional random trajectory time resolution TWIST sequence is mainly adopted to carry out the magnetic resonance imaging of belly Dynamic constrasted enhancement.
Three-dimensional Fast spoiled gradient echo 3DFSPGR sequence adopts small-angle radio frequency pulse excitation signal, then phase encoding collection signal is carried out, last adding on gradient encode direction disturbs phase gradient and to be fallen apart by residual Mxy phase, to shorten the sequence acquisition time.3DFSPGR sequence needs integrating parallel imaging technique usually, realizes belly Dynamic constrasted enhancement scanning faster, and clinically because belly image coverage is large, phase encoding step number is many, and imaging time resolution is very limited.
TWIST sequence adopts neighborhood to share the three-dimensional pseudorandom track sets of (viewsharing), during data acquisition, intensive sampling is carried out and outer region (B) carries out sub-sampling in K space center region (A), and when data reconstruction, adjacent K space data is carried out merging shared (view-shared) reconstruction, thus shorten the acquisition time of each two field picture.In three-dimensional K space, TWIST Sampling Machine is built in phase encoding gradient k
yphase encoding gradient k is selected with layer
zupper execution, adopt polar coordinate mode to express, sample track is along with radiation radius k
rand azimuth angle theta (0< θ <2 π) increase is carried out to periphery gradually, the threshold value k of a setting radiation radius
c, at threshold value k
cwithin K space-intensive sampling, k
coutside K space pseudorandom sub-sampling.First periphery K area of space B is divided into the N number of subset B do not overlapped each other by periphery, K space sub-sampling mode at random
i(i=1 ... N), when carrying out data acquisition, first gathering the data set #0 that a K spatial point is full sampling, then gathering A successively, B
1, A, B
2, A, B
1number of regions strong point obtains #1, #2, #3, #4, #5, #6 ... data set.When rebuilding, by the #1 that odd-times gathers, former and later two even number K space data collection of #3 and #5 are merged into current data set, are combined into a complete full sampling K space and carry out image reconstruction, as shown in Figure 1.So the image frame per second that TWIST rebuilds image depends on the frame number that region A gathers.
But, clinically during belly three-dimensional imaging in order to realize covering on a large scale and obtain enough high-resolution enhancing image, the usually more number of plies of setting (>=32 layers), larger sampling matrix (phase encoding k
y>=128, frequency coding k
x>=128) and repeatedly excite (Nex>=1), sequence time resolution is between 10 ~ 60 seconds, make the time-intensity curves resolution that collects low, far can not the accurate measurement of content with funtion parameter, or to sacrifice spatial resolution to compensate, cause image spatial resolution to decline.Clinical 3DFSPGR sequence scanning is full sample mode, usually integrating parallel imaging technique (parallelimaging) is needed to carry out picture signal collection, but, parallel imaging technique causes signal noise ratio (snr) of image SNR to reduce, and acceleration multiple is limited within three times usually, accelerate speed higher, signal to noise ratio (S/N ratio) is lower, hinders the further raising of temporal resolution.On the other hand, TWIST sequence scanning mode is interval sampling K space center region and outer region, needs adjacent K space data collection to merge in process of reconstruction, ensure that the spatial resolution that image is higher, but image frame per second is reduced, limits the raising of its temporal resolution.So, study and a kind ofly not only improved temporal resolution but also can ensure that the magnetic resonance fast imaging method of sufficiently high spatial resolution improves the important prerequisite that medical diagnosis on disease and functional parameter accurately measure.
Summary of the invention
For the problems referred to above, the present invention proposes one and compressed sensing Integration ofTechnology is shared (view-sharedcompressedsensing in sample track to K spatial neighborhood, VCS) the quick sub-sampling method of K space data, improve the temporal resolution of belly Dynamic constrasted enhancement magnetic resonance imaging, ensure that higher signal noise ratio (snr) of image and picture quality simultaneously.
For achieving the above object, the present invention takes following technical scheme: a kind of quick three-dimensional K space sub-sampling fill method of Dynamic constrasted enhancement magnetic resonance imaging, comprises the random division in K space, the design of CS sampling matrix, the acquisition of neighborhood share compressed perception sampling matrix, the data acquisition based on neighborhood share compressed cognitive method, data reconstruction five parts based on neighborhood share compressed cognitive method.Particular content is: first, adopt adjacent K space periphery data sharing sample track, gather central area and portion perimeter area data in a K space simultaneously, the mutual zero lap of periphery, K space sub-sampling track that neighborhood is shared, all adjacent sub-sampling K space data combinations can form a complete K space.Secondly, compressed sensing being accelerated Sampling techniques is embedded in the shared sample track in adjacent K space, each K space is filled track and is determined by compression sensing method design, wherein the central area reservation in K space is not shared, 2 ~ 5 shared subsets are divided at random, to realize the neighborhood share compressed perception sampling of continuous multiple K space sampling trajectories after the design of outer region compressed perception sampling matrix.Neighborhood share compressed perception sampling is when a K spatial sampling, and simultaneously subset is shared in the central area in collect and process perception K space and 1 periphery.K space data rebuild time, continuous multiple non-overlapping shared subset is merged into current K space, reassemble into meet compressed sensing sampling matrix track data set for further compressed sensing image reconstruction.Be integrated into neighborhood share sample track compressed sensing accelerate multiple can be set to 2 ~ 8 times.The design of compressed sensing sampling matrix adopts non uniformly varying density function to generate, and meets Gaussian distribution, and sparse conversion of compressed sensing can be Fourier transform, wavelet transformation, finite difference conversion etc.Finally, neighborhood is shared the K space sub-sampling data of restructuring through l
1the minimized nonlinear algorithm of normal form is rebuild and is obtained image, and reconstruction formula is as follows:
minimizeλ
1||ψS||
1+λ
2TV(S),
In formula, S is target image, λ
1and λ
2represent the weight between ψ sparse expression and finite difference sparse expression TV, F represents the Fourier transform corresponding with sub-sampling, d
irepresent VCS
ithe k-space data collected, ε is usually arranged on getting off of background noise level and ensures to rebuild the fidelity of image.And be rewritten as clinical criteria medical image according to DICOM form.
The present invention is owing to taking above technical scheme, tool has the following advantages: the present invention is a kind of belly quick three-dimensional Dynamic constrasted enhancement formation method of high time resolution, K space filling mode adopts neighborhood technology of sharing, each K space data collection gathers central area and portion perimeter number of regions strong point simultaneously, and image frame per second is remained unchanged; Each K spatial sampling that neighborhood is shared is integrated with compressed sensing Sampling techniques, and the neighborhood simultaneously achieving data is shared and compressed sensing collection, further shorten data acquisition time; When neighborhood share compressed cognitive method is rebuild, adjacent K space data collection is carried out merging to rebuild, image information is not lost; Neighborhood share compressed cognitive method image reconstruction algorithm adopts based on l
1normal form compressed sensing reconstruction algorithm, the data that sub-sampling obtains reliably can be reconstructed original image by CS reconstruction algorithm.The present invention can be widely used in the magnetic resonance imaging of abdominal organs Dynamic constrasted enhancement and clinical practice thereof.
Accompanying drawing explanation
Fig. 1 is that track schematic diagram is filled in TWIST sequence K space:
Fig. 2 is neighborhood share compressed perception VCS sample mode schematic diagram:
Fig. 3 is neighborhood share compressed cognitive method VCS image reconstruction schematic diagram:
Embodiment
Below in conjunction with accompanying drawing, the present invention is described in further detail, specific as follows:
Be embodiments of the invention as shown in Figures 2 and 3, comprise the random division in K space, the design of CS sampling matrix, the acquisition of neighborhood share compressed perception sampling matrix, the data acquisition based on neighborhood share compressed cognitive method, data reconstruction five parts based on neighborhood share compressed cognitive method, first neighborhood share compressed perception sampling matrix VCS will be generated during work, then gather continuously dynamic K space data, finally institute's image data is rebuild.
Neighborhood share compressed cognitive method Part I is the random division in K space, at the k in three-dimensional K space
y-k
zin plane, be divided into central area A and outer region B, wherein region B is divided into N number of subset B do not overlapped each other at random
i(i=1,2 ... N), as shown in Figure 2, each subset Bi and A combines the K space tracking K that composition is used for neighborhood shared (view-shared)
i(i=1,2 ... N), i.e. set (the K of white point and its periphery white point in dashed circle
1, K
2, K
3) represent three K space trackings of random division.
Neighborhood share compressed cognitive method Part II is the design of CS sampling matrix, in three-dimensional K space, and phase encoding (k in layer
y) and layer select phase encoding (k
z) phase encoding of two dimensions all can carry out compression sampling, theoretical according to CS, adopts Monte Carlo method to obtain the two-dimensional observation matrix Φ meeting independent same distribution Gaussian distribution
cS, the sampling mechanism of this observing matrix meets non uniformly varying density function, and irrelevant with sparse matrix, at k
yand k
zdirection achieves sub-sampling, frequency coding direction (k
x) being subject to hardware condition restriction must for sampling entirely.As shown in Figure 2, during work, k
ydirection and k
zdirection is according to Φ
cSsampling matrix carries out random coded simultaneously, and namely in dashed circle, the set (CSmatrix) of white point and its periphery white point represents execution random coded.Phase encoding after optimizing according to this, K space center region point is intensive sampling, and outer region is Random sparseness sampling by the probability density function of variable density outward, realizes the compression sampling of K space data.Φ
cSthe number percent that the step number of the actual execution of phase encoding accounts for K space sum determines the temporal resolution of this imaging, and in this embodiment, K counts in space and is set to 50%, and the CS realizing 2 times accelerates sampling.
Neighborhood share compressed cognitive method Part III is the acquisition of neighborhood share compressed perception sampling matrix, as shown in Figure 2, first, and the K space tracking K that neighborhood is shared
i(i=1,2 ... N) illustrate adjacent K space outer region and perform different track fillings, then by the CS sampling matrix Φ of 2 times
cSsame K
ido mathematics and computing, obtain neighborhood share compressed perception sampling matrix VCS (view-sharedCSmatrix).VCS
i(i=1,2 ... N) the neighborhood share compressed perception sampling matrix of successive image frame is represented respectively.
Neighborhood share compressed cognitive method Part IV is the data acquisition based on neighborhood share compressed cognitive method, share sampling matrix VCS by the neighborhood of Embedded compression perception and carry out continuous print Dynamic Data Acquiring, obtain the original rawdata data d in K space of Dynamic constrasted enhancement scanning
i(i=1,2 ... N), as shown in Figure 3.
Neighborhood share compressed cognitive method Part V is the data reconstruction based on neighborhood share compressed cognitive method, as shown in Figure 3, according to VCS sampling mechanism, and the raw data d first will collected
ithat carries out adjacent data collection shares (view-shared), reassembles into the data reconstruction meeting CS sample track, then, carries out based on l the data reconstruction of these restructuring
1the minimized nonlinear algorithm of normal form is rebuild.Reconstruction formula is as follows:
minimizeλ
1||ψS||
1+λ
2TV(S),
S is target image, λ
1and λ
2represent the weight between ψ sparse expression and finite difference sparse expression TV, F represents the Fourier transform corresponding with sub-sampling, d
irepresent VCS
ithe k-space data collected, ε is usually arranged on getting off of background noise level and ensures to rebuild the fidelity of image.
Finally, reconstruction image is rewritten as clinical criteria medical image according to DICOM form.
Claims (6)
1. the present invention relates to a kind of quick three-dimensional K space sub-sampling fill method of Dynamic constrasted enhancement magnetic resonance imaging, it is characterized in that: described K space sub-sampling fill method adopts adjacent K space periphery data sharing sample track, gather central area and portion perimeter area data in a K space simultaneously, the mutual zero lap of periphery, K space sub-sampling track that neighborhood is shared, the K space that all adjacent sub-sampling K space data combination formations one are complete.
2. K space sub-sampling fill method according to claim 1, it is characterized in that: compressed sensing is accelerated Sampling techniques and is embedded in the shared sample track in adjacent K space, each K space is filled track and is determined by compression sensing method design, wherein the central area reservation in K space is not shared, be divided into 2 ~ 5 shared subsets at random after the design of outer region compressed perception sampling matrix, realize the neighborhood share compressed perception sampling of continuous multiple K space sampling trajectories.
3. K space sub-sampling fill method according to claim 1, it is characterized in that: compressed sensing is accelerated Sampling techniques and is embedded in the shared sample track in adjacent K space, neighborhood share compressed perception sampling is when a K spatial sampling, and simultaneously subset is shared in the central area in collect and process perception K space and 1 periphery.
4. K space sub-sampling fill method according to Claims 2 or 3, it is characterized in that: K space data is when rebuilding, continuous multiple non-overlapping shared subset is merged into current K space, reassemble into meet compressed sensing sampling matrix track data set for further compressed sensing image reconstruction.
5. K space sub-sampling fill method according to Claims 2 or 3, is characterized in that: be integrated into neighborhood share sample track compressed sensing accelerate multiple can be set to 2 ~ 8 times.
6. K space sub-sampling fill method according to Claims 2 or 3, it is characterized in that: the design of compressed sensing sampling matrix adopts non uniformly varying density function to generate, and meeting Gaussian distribution, the sparse conversion of compressed sensing can be Fourier transform, wavelet transformation, finite difference conversion etc.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510711789.6A CN105259525B (en) | 2015-10-28 | 2015-10-28 | Dynamic constrasted enhancement magnetic resonance fast imaging method based on the perception of neighborhood share compressed |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510711789.6A CN105259525B (en) | 2015-10-28 | 2015-10-28 | Dynamic constrasted enhancement magnetic resonance fast imaging method based on the perception of neighborhood share compressed |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105259525A true CN105259525A (en) | 2016-01-20 |
CN105259525B CN105259525B (en) | 2019-08-13 |
Family
ID=55099286
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510711789.6A Active CN105259525B (en) | 2015-10-28 | 2015-10-28 | Dynamic constrasted enhancement magnetic resonance fast imaging method based on the perception of neighborhood share compressed |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105259525B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106526514A (en) * | 2016-11-21 | 2017-03-22 | 清华大学 | Magnetic resonance diffusion imaging data compression and accelerated reconstruction method |
CN106880356A (en) * | 2017-03-11 | 2017-06-23 | 潍坊新力超导磁电科技有限公司 | A kind of magnetic resonance time delay imaging method |
CN107993271A (en) * | 2017-12-26 | 2018-05-04 | 上海交通大学 | A kind of magnetic resonance dynamic imaging method of sampling and image rebuilding method |
CN108567422A (en) * | 2018-03-02 | 2018-09-25 | 上海东软医疗科技有限公司 | A kind of blood vessel imaging method and apparatus |
CN110213467A (en) * | 2019-05-27 | 2019-09-06 | 中国科学院国家空间科学中心 | The multiple dimensioned modulation compressed sensing imaging system of one kind and its imaging method |
CN111080738A (en) * | 2019-11-28 | 2020-04-28 | 上海联影智能医疗科技有限公司 | System and method for image reconstruction |
CN113379860A (en) * | 2021-04-25 | 2021-09-10 | 上海索骥信息科技有限公司 | Medical image data generation method and system with marker based on digital object |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101263398A (en) * | 2005-07-08 | 2008-09-10 | 威斯康星校友研究基金会 | Backprojection reconstruction method for undersampled MRI |
CN101975936A (en) * | 2010-09-03 | 2011-02-16 | 杭州电子科技大学 | Rapid magnetic resonance imaging (MRI) method based on CS ( compressed sensing ) technique |
CN102488519A (en) * | 2011-11-29 | 2012-06-13 | 中国科学院深圳先进技术研究院 | Diffusion tensor imaging method and system |
CN102934995A (en) * | 2012-10-18 | 2013-02-20 | 中国科学院深圳先进技术研究院 | Magnetic resonance imaging method based on low-rank and sparse characteristics of image |
CN103142228A (en) * | 2012-12-14 | 2013-06-12 | 中国科学院深圳先进技术研究院 | Compressed sensing magnetic resonance fast imaging method |
CN103163496A (en) * | 2011-12-12 | 2013-06-19 | 中国科学院深圳先进技术研究院 | Plane echo imaging method and system |
CN103323805A (en) * | 2013-05-29 | 2013-09-25 | 杭州电子科技大学 | Method for SPEED rapid magnetic resonance imaging based on wavelet domain sparse representation |
WO2014147518A2 (en) * | 2013-03-21 | 2014-09-25 | Koninklijke Philips N.V. | Mr image reconstruction using compressed sensing |
CN104101849A (en) * | 2013-04-02 | 2014-10-15 | 西门子公司 | Establishing a Magnetic Resonance System Actuation Sequence |
-
2015
- 2015-10-28 CN CN201510711789.6A patent/CN105259525B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101263398A (en) * | 2005-07-08 | 2008-09-10 | 威斯康星校友研究基金会 | Backprojection reconstruction method for undersampled MRI |
CN101975936A (en) * | 2010-09-03 | 2011-02-16 | 杭州电子科技大学 | Rapid magnetic resonance imaging (MRI) method based on CS ( compressed sensing ) technique |
CN102488519A (en) * | 2011-11-29 | 2012-06-13 | 中国科学院深圳先进技术研究院 | Diffusion tensor imaging method and system |
CN103163496A (en) * | 2011-12-12 | 2013-06-19 | 中国科学院深圳先进技术研究院 | Plane echo imaging method and system |
CN102934995A (en) * | 2012-10-18 | 2013-02-20 | 中国科学院深圳先进技术研究院 | Magnetic resonance imaging method based on low-rank and sparse characteristics of image |
CN103142228A (en) * | 2012-12-14 | 2013-06-12 | 中国科学院深圳先进技术研究院 | Compressed sensing magnetic resonance fast imaging method |
WO2014147518A2 (en) * | 2013-03-21 | 2014-09-25 | Koninklijke Philips N.V. | Mr image reconstruction using compressed sensing |
CN104101849A (en) * | 2013-04-02 | 2014-10-15 | 西门子公司 | Establishing a Magnetic Resonance System Actuation Sequence |
CN103323805A (en) * | 2013-05-29 | 2013-09-25 | 杭州电子科技大学 | Method for SPEED rapid magnetic resonance imaging based on wavelet domain sparse representation |
Non-Patent Citations (1)
Title |
---|
STANISLAS RAPACCHI,等: "Reducing View-Sharing Using Compressed Sensing in Time-Resolved Contrast-Enhanced Magnetic Resonance Angiography", 《MAGNETIC RESONANCE IN MEDICINE》 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106526514A (en) * | 2016-11-21 | 2017-03-22 | 清华大学 | Magnetic resonance diffusion imaging data compression and accelerated reconstruction method |
CN106526514B (en) * | 2016-11-21 | 2019-05-31 | 清华大学 | Magnetic resonance diffusion imaging data compression accelerated reconstruction method |
CN106880356A (en) * | 2017-03-11 | 2017-06-23 | 潍坊新力超导磁电科技有限公司 | A kind of magnetic resonance time delay imaging method |
CN107993271A (en) * | 2017-12-26 | 2018-05-04 | 上海交通大学 | A kind of magnetic resonance dynamic imaging method of sampling and image rebuilding method |
CN108567422A (en) * | 2018-03-02 | 2018-09-25 | 上海东软医疗科技有限公司 | A kind of blood vessel imaging method and apparatus |
CN108567422B (en) * | 2018-03-02 | 2021-03-05 | 上海东软医疗科技有限公司 | Blood vessel imaging method and device |
CN110213467A (en) * | 2019-05-27 | 2019-09-06 | 中国科学院国家空间科学中心 | The multiple dimensioned modulation compressed sensing imaging system of one kind and its imaging method |
CN110213467B (en) * | 2019-05-27 | 2020-10-23 | 中国科学院国家空间科学中心 | Multi-scale modulation compressed sensing imaging system and imaging method thereof |
CN111080738A (en) * | 2019-11-28 | 2020-04-28 | 上海联影智能医疗科技有限公司 | System and method for image reconstruction |
CN111080738B (en) * | 2019-11-28 | 2024-02-27 | 上海联影智能医疗科技有限公司 | System and method for image reconstruction |
CN113379860A (en) * | 2021-04-25 | 2021-09-10 | 上海索骥信息科技有限公司 | Medical image data generation method and system with marker based on digital object |
Also Published As
Publication number | Publication date |
---|---|
CN105259525B (en) | 2019-08-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105259525A (en) | Dynamic contrast enhanced magnetic resonance fast imaging method based on neighborhood sharing compression sensing | |
CN104077791B (en) | A kind of several dynamic contrast enhancement magnetic resonance images joint method for reconstructing | |
CN105005012B (en) | Abdomen organ's Dynamic constrasted enhancement MR imaging method based on compressed sensing | |
WO2018014475A1 (en) | System and method for segmenting medical image | |
Altazi et al. | Investigating multi-radiomic models for enhancing prediction power of cervical cancer treatment outcomes | |
CN101626727A (en) | Additional automatic diagnosis and the aligning that the PET/MR flow estimation is arranged | |
CN103323805B (en) | Method for SPEED rapid magnetic resonance imaging based on wavelet domain sparse representation | |
US12016701B2 (en) | Quantitative differentiation of tumor heterogeneity using diffusion MR imaging data | |
CN101975936A (en) | Rapid magnetic resonance imaging (MRI) method based on CS ( compressed sensing ) technique | |
CN110223285A (en) | Imaging results prediction technique and system based on animal migration and neural network model | |
CN104267361A (en) | Composite reconstruction method for self-adaptation quantitative magnetisability distribution diagram based on structural feature | |
Han et al. | Temporal/spatial resolution improvement of in vivo DCE-MRI with compressed sensing-optimized FLASH | |
Hou et al. | Bag-of-features-based radiomics for differentiation of ocular adnexal lymphoma and idiopathic orbital inflammation from contrast-enhanced MRI | |
CN110619635A (en) | Hepatocellular carcinoma magnetic resonance image segmentation system and method based on deep learning | |
WO2019241266A1 (en) | Black-box assessment of disease and response-to-therapy by t1 or t2* weighted dce-mri | |
Zhang et al. | Comparison of CT and MRI images for the prediction of soft-tissue sarcoma grading and lung metastasis via a convolutional neural networks model | |
Vardoulis et al. | Single breath-hold 3D measurement of left atrial volume using compressed sensing cardiovascular magnetic resonance and a non-model-based reconstruction approach | |
WO2020113148A1 (en) | Single or a few views computed tomography imaging with deep neural network | |
CN116784820A (en) | Brain function network construction method and system based on seed point connection | |
Liu et al. | High‐temporospatial‐resolution dynamic contrast‐enhanced (DCE) wrist MRI with variable‐density pseudo‐random circular Cartesian undersampling (CIRCUS) acquisition: evaluation of perfusion in rheumatoid arthritis patients | |
CN1689510A (en) | Digitalized method for magnetic resonance perfusion imaging | |
CN103048632B (en) | A kind of quick fine angular resolution diffusion imaging method based on compressed sensing | |
CN111681184B (en) | Neural melanin image reconstruction method, device, equipment and storage medium | |
CN104305998A (en) | Three-dimensional thickness measurement and quantitative analysis method for bladder walls | |
CA3109456A1 (en) | System, method and computer-accessible medium for tissue fingerprinting |
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 |