CN110276762A - A kind of full-automatic bearing calibration of respiratory movement of the diffusion-weighted Abdominal MRI imaging of more b values - Google Patents
A kind of full-automatic bearing calibration of respiratory movement of the diffusion-weighted Abdominal MRI imaging of more b values Download PDFInfo
- Publication number
- CN110276762A CN110276762A CN201810212608.9A CN201810212608A CN110276762A CN 110276762 A CN110276762 A CN 110276762A CN 201810212608 A CN201810212608 A CN 201810212608A CN 110276762 A CN110276762 A CN 110276762A
- Authority
- CN
- China
- Prior art keywords
- image
- diffusion
- values
- value
- weighted
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/05—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves
- A61B5/055—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Surgery (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Veterinary Medicine (AREA)
- Theoretical Computer Science (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Public Health (AREA)
- Animal Behavior & Ethology (AREA)
- General Physics & Mathematics (AREA)
- High Energy & Nuclear Physics (AREA)
- Radiology & Medical Imaging (AREA)
- Artificial Intelligence (AREA)
- Physiology (AREA)
- Psychiatry (AREA)
- Signal Processing (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
Of the invention provides a kind of full-automatic bearing calibration of respiratory movement of diffusion-weighted Abdominal MRI imaging of more b values.This method correction is the renal image that is obtained by multiple b value Diffusion-weighted imaging (DWI) sequence acquisitions in the case where freely breathing.Kidney is split using U-NET network first, obtains kidney profile.Then, by golden (Lucas-Kanade) algorithm of pyramid Lucas be zero to first b value by multiple image registrations of high b value image on.Fitting after the graphical arrangement after registration is finally obtained into the parameter of inhomogenous movement (IVIM) model in vivo.This system is not necessarily to the manual participation of operator, does not reject any data.Have the advantages that high degree of automation, application cost are low etc..
Description
Technical field
The present invention relates to a kind of technical field of image processing more particularly to a kind of more diffusion-weighted Abdominal MRIs of b value at
The full-automatic bearing calibration of the respiratory movement of picture.
Background technique
Diffusion-weighted imaging (DWI) can be widely used in non-with the diffusion motion of quantitative response hydrone
Invasive renal functional evaluation.Based on the two fingers number computation model of inhomogenous movement (IVIM) theory in vivo, hypothesis is more nearly
Moisture movement state in living tissue is able to reflect the local diffusion of true property and microcirculatory perfusion.In actual imaging,
DWI for example breathes molecular motion, mass motion and physiological movement very sensitive.Wherein, kidney caused by respiratory movement
It deviates up to 5-10mm, this moving displacement occurred at different acquisition time point can generate huge mistake in estimation function parameter
Difference.
Currently, clinically solving the problems, such as that the main means of respiration artefacts are that patient is allowed to feel suffocated in scanning.This is to the elderly
And weak patient is greatly to challenge.Simultaneously as clinically used DWI imaging sequence be single-shot echo planar at
Picture needs to shorten echo train length as far as possible, this is just for magnetic susceptibility, the chemical shift etc. for inhibiting the acquisition strategies may cause
Cause clinical DWI image that there is lower resolution ratio.Another sweeping scheme be using respiration gate control in the case where freely breathing into
Row acquisition.However, gate method only carries out data acquisition under specific time window, scan efficiency is caused to decline.Importantly,
If patient respiration mode is irregular or breathing zone placement position is inaccurate, it will obtain the breath signal of mistake, lead to figure
Moving displacement occurs for specific imaging position as in, brings image quality decrease, influences to diagnose.
To solve the above-mentioned problems, forefathers propose the registration Algorithm of some post-processings.However, it may generally be desirable to manually participate in
Registration selecting kidney position or directly carrying out more b value images to entire image.Therefore, waste of manpower and time, and match
Quasi- imaging accuracy is low.Therefore, the good more diffusion-weighted figures of b value of full automatic kidney of a kind of strong robustness, registration effect are needed
As movement whole school's correction method.
Summary of the invention
It is an object of the invention to be directed to the prior art, a kind of more diffusion-weighted Abdominal MRIs of b value are proposed
The full-automatic bearing calibration of the respiratory movement of imaging.
To achieve the goals above, the method for the invention mainly comprises the steps of:
1) diffusion-weighted Abdominal MRI image I0 progress internal organs region corresponding to each b value is divided automatically is corresponded to
Image I1；Wherein b value needs to choose 3-10 different b values, and wherein needs b=0s/mm2；And region segmentation method
The U-NET neural network model obtained using supervised learning carries out abdominal organs region to the image I0 of all difference b values
Automatic segmentation, the image series I1 after obtaining region segmentation.
2) it is reference with the image of b=0s/mm2 after dividing automatically, the image I1 after other all b value image segmentations is pressed
According to being reference with the image of b=0s/mm2, using golden (Pyramid Lucas-Kanade) algorithm of pyramid Lucas to region
Image series I1 after segmentation is registrated, the image I2 after respectively obtaining respiratory movement correction.
3) it is based on I2, the true property diffusion coefficient (D) after being corrected by IVIM models fitting is perfused score (f) and false
Property diffusion coefficient (D*) distribution map.
In conclusion the invention has the following beneficial effects:
The present invention is that the method proposed has the advantages that high degree of automation, application cost are low etc., solve current doctor and exist
Needed when calculating IVIM model parameter manually circle take the profile of kidney, caused by the increase of human cost and the consuming of time.
Detailed description of the invention
Fig. 1 is overall flow figure of the invention；
Fig. 2 is the map figure after the present invention is fitted by IVIM model data；
Specific embodiment
The present invention will be described below by way of specific embodiments, but the present invention is not limited thereto.
The full-automatic bearing calibration of respiratory movement of the diffusion-weighted Abdominal MRI imaging of a kind of more b values as shown in Figure 1, specifically
Processing step is as follows:
1) data carried out to subject's renal tract under freely breathing acquire, using b value: 0,10,20,50,100,
The sequence of 180,300,420,550 and 700s/mm2.
2) kidney segmentation is carried out using U-NET.U-NET network is made of 23 convolutional layers altogether, shrinks net in the network
Network and expansion network are the relationships mutually mapped.The work that network is mainly responsible for down-sampling is shunk, extracts high dimensional feature information, often
Down-sampling includes the convolution operation of two 3x3 and the pondization operation of a 2x2, and using amendment linear unit as sharp
Function living.Expansion network is mainly responsible for the work of up-sampling, and up-sampling includes the convolution operation of two 3x3 each time, using repairing
Linear positive unit is as activation primitive.In up-sampling operation, by the contraction network of output feature and phase mapping each time
Feature merges, the boundary information lost among completion.Finally, the feature that the convolution operation that 1x1 is added will be obtained before
Classification belonging to being mapped to is above.According to the profile of the U-NET kidney being partitioned into, kidney is cut out from original image.
3) collected data are divided into 3 classes by the size of one-dimensional sequence according to value.Every matroid is resolved into multiple dimensioned
The sum of block low-rank matrix, and by the sum of multiple dimensioned piece of low-rank matrix minimum as constraint condition be used for it is every one kind data weight
It builds, the image after obtaining 3 reconstructions.
4) kidney obtained after dividing data is registrated using golden (Lucas-Kanade) algorithm of pyramid Lucas,
Search for the motion transform parameter between two images step by step using image pyramid.First pyramidal smallest dimension layer into
Row search, finds the optimal mapping of registration two images, and later each layer is searched centered on the result of the search of preceding layer
Rope, and constantly amendment preceding layer pyramid obtain as a result, until out to out layer.It may finally obtain all moving image (b
> 0) displacement relative to reference picture (b=0) and direction.It is later zero by 9 image registrations of high b value to first b value
On image.
5) by the graphical arrangement after registration, according to IVIM model calculation formula (wherein for when DWI signal strength, be phase
The DWI signal strength for answering b value to obtain, f and the occupation rate that (1-f) respectively represents microcirculatory perfusion and true property is spread, D* are to represent
The ADC value of microcirculatory perfusion effect, D are the ADC values for representing true property disperse) fitting obtain the functional parameter figure of IVIM model.
Fig. 2 gives the map figure after the present invention is fitted by IVIM model data.It is tested that this figure compares registration front and back two
The map of person D/D*/f/ADC schemes.The visual effect of four Parameter Maps has enhancing: the profile of kidney be more clear and failure
Match pixel point is reduced.After registration, the visual effect of four parameters is obviously improved.
Claims (4)
1. a kind of full-automatic bearing calibration of respiratory movement of the diffusion-weighted Abdominal MRI imaging of more b values, which is characterized in that described
Method the following steps are included:
1) diffusion-weighted Abdominal MRI image I0 progress internal organs region corresponding to each b value is divided automatically obtains each b value
Corresponding image I1；
2) with b=0s/mm after dividing automatically2Image be reference, the corresponding image I1 of other all b values is registrated respectively
The corresponding I2 of each b value after obtaining respiratory movement correction；
3) it is based on the corresponding I2 of all b values, the diffusion of the true property after being corrected is fitted by internal inhomogenous motion model IVIM
Score distribution figure and false diffusion coefficient distribution map is perfused in index profile.
2. the method as described in claim 1, which is characterized in that more b values diffusion-weighted Abdominal MRI needs when being imaged
3-10 different b values are chosen, and wherein need b=0s/mm2。
3. the method as described in claim 1, which is characterized in that the region segmentation method are as follows: obtained using supervised learning
To U-NET neural network model I0 corresponding to all difference b values carry out the automatic segmentation in abdominal organs region, obtain region
Image series I1 after segmentation.
4. the method as described in claim 1, which is characterized in that the Registration of Measuring Data method are as follows: with b=0s/mm2Image
For reference, using golden (Pyramid Lucas-Kanade) algorithm of pyramid Lucas to the image series I1 after region segmentation into
Row registration, the image series I2 after obtaining motion correction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810212608.9A CN110276762A (en) | 2018-03-15 | 2018-03-15 | A kind of full-automatic bearing calibration of respiratory movement of the diffusion-weighted Abdominal MRI imaging of more b values |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810212608.9A CN110276762A (en) | 2018-03-15 | 2018-03-15 | A kind of full-automatic bearing calibration of respiratory movement of the diffusion-weighted Abdominal MRI imaging of more b values |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110276762A true CN110276762A (en) | 2019-09-24 |
Family
ID=67958494
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810212608.9A Pending CN110276762A (en) | 2018-03-15 | 2018-03-15 | A kind of full-automatic bearing calibration of respiratory movement of the diffusion-weighted Abdominal MRI imaging of more b values |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110276762A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110889897A (en) * | 2019-11-21 | 2020-03-17 | 厦门大学 | Method and system for reconstructing incoherent motion magnetic resonance imaging parameters in voxel |
| CN111445553A (en) * | 2020-03-31 | 2020-07-24 | 浙江大学 | Depth learning-based intra-voxel incoherent motion imaging acceleration method and device |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102428479A (en) * | 2009-04-17 | 2012-04-25 | 里弗兰医疗集团公司 | Chest X-ray registration, subtraction and display |
| CN104323777A (en) * | 2014-10-30 | 2015-02-04 | 西安交通大学医学院第一附属医院 | Diffusion magnetic resonance imaging motion artifact eliminating method |
| CN106022384A (en) * | 2016-05-27 | 2016-10-12 | 中国人民解放军信息工程大学 | Image attention semantic target segmentation method based on fMRI visual function data DeconvNet |
| CN106204587A (en) * | 2016-05-27 | 2016-12-07 | 孔德兴 | Multiple organ dividing method based on degree of depth convolutional neural networks and region-competitive model |
| CN106228583A (en) * | 2016-07-14 | 2016-12-14 | 北京大学 | The device that a kind of abdomen images is rebuild |
| CN107067398A (en) * | 2017-04-28 | 2017-08-18 | 青岛海信医疗设备股份有限公司 | Complementing method and device for lacking blood vessel in 3 D medical model |
| CN107274443A (en) * | 2017-06-05 | 2017-10-20 | 河北大学 | Prostate diffusion weighted images sequence method for registering |
| CN107301645A (en) * | 2017-06-20 | 2017-10-27 | 上海联影医疗科技有限公司 | A kind of data processing method and device |
| CN107705305A (en) * | 2017-10-20 | 2018-02-16 | 上海联影医疗科技有限公司 | A kind of method and device of Medical Image Processing |
-
2018
- 2018-03-15 CN CN201810212608.9A patent/CN110276762A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102428479A (en) * | 2009-04-17 | 2012-04-25 | 里弗兰医疗集团公司 | Chest X-ray registration, subtraction and display |
| CN104323777A (en) * | 2014-10-30 | 2015-02-04 | 西安交通大学医学院第一附属医院 | Diffusion magnetic resonance imaging motion artifact eliminating method |
| CN106022384A (en) * | 2016-05-27 | 2016-10-12 | 中国人民解放军信息工程大学 | Image attention semantic target segmentation method based on fMRI visual function data DeconvNet |
| CN106204587A (en) * | 2016-05-27 | 2016-12-07 | 孔德兴 | Multiple organ dividing method based on degree of depth convolutional neural networks and region-competitive model |
| CN106228583A (en) * | 2016-07-14 | 2016-12-14 | 北京大学 | The device that a kind of abdomen images is rebuild |
| CN107067398A (en) * | 2017-04-28 | 2017-08-18 | 青岛海信医疗设备股份有限公司 | Complementing method and device for lacking blood vessel in 3 D medical model |
| CN107274443A (en) * | 2017-06-05 | 2017-10-20 | 河北大学 | Prostate diffusion weighted images sequence method for registering |
| CN107301645A (en) * | 2017-06-20 | 2017-10-27 | 上海联影医疗科技有限公司 | A kind of data processing method and device |
| CN107705305A (en) * | 2017-10-20 | 2018-02-16 | 上海联影医疗科技有限公司 | A kind of method and device of Medical Image Processing |
Non-Patent Citations (6)
| Title |
|---|
| 刘小华 等，: "DWI中基于单指数与IVIM双指数模型早期预测淋巴瘤疗效效能比较", 《中国医学影像技术》 * |
| 崔兴宇 等，: "IVIM-DWI在肝外胆管癌分级中的临床应用", 《临床放射学杂志》 * |
| 徐光柱 等，: "《实用性目标检测与跟踪算法原理及应用》", 30 April 2015, 国防工业出版社 * |
| 江建芹 等，: "体素内不相干运动模型扩散加权成像在肿瘤疗效评估中的进展", 《中国医学影像技术》 * |
| 白小晶 等，: "基于光流场模型的医学图像配准", 《江南大学学报（自然科学版）》 * |
| 贺大林，: "《前列腺癌贺大林2016观点》", 31 March 2017 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110889897A (en) * | 2019-11-21 | 2020-03-17 | 厦门大学 | Method and system for reconstructing incoherent motion magnetic resonance imaging parameters in voxel |
| CN111445553A (en) * | 2020-03-31 | 2020-07-24 | 浙江大学 | Depth learning-based intra-voxel incoherent motion imaging acceleration method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9361686B2 (en) | Method and apparatus for the assessment of medical images | |
| CN103249358B (en) | Medical image-processing apparatus | |
| EP3133993B1 (en) | Lean tissue volume quantification | |
| US20020147392A1 (en) | Technique for manipulating medical images | |
| Kohlmann et al. | Automatic lung segmentation method for MRI-based lung perfusion studies of patients with chronic obstructive pulmonary disease | |
| JP2009512527A (en) | Image registration method, algorithm for performing the image registration method, program for registering an image using the algorithm, and biomedical image handling method for reducing image artifacts due to object movement | |
| US9402562B2 (en) | Systems and methods for improved tractographic processing | |
| CN106780643A (en) | Magnetic resonance repeatedly excites diffusion imaging to move antidote | |
| Bullitt et al. | 3d graph description of the intracerebral vasculature from segmented mra and tests of accuracy by comparison with x-ray angiograms | |
| CN110060315B (en) | Image motion artifact eliminating method and system based on artificial intelligence | |
| Xue et al. | Unsupervised inline analysis of cardiac perfusion MRI | |
| US11544848B2 (en) | Deep learning based methods and systems for automated subject anatomy and orientation identification | |
| US20150363951A1 (en) | Systems and methods for objective tractographic processing using features of grid structures of the brain | |
| Lötjönen et al. | Correction of movement artifacts from 4-D cardiac short-and long-axis MR data | |
| CN110276762A (en) | A kind of full-automatic bearing calibration of respiratory movement of the diffusion-weighted Abdominal MRI imaging of more b values | |
| WO2020114632A1 (en) | Three-dimensional shape reconstruction from a topogram in medical imaging | |
| CN114821049A (en) | Method, device and equipment for segmenting image and storage medium | |
| CN110930394B (en) | Method and terminal equipment for measuring slope and pinnate angle of muscle fiber bundle line | |
| Shahzad et al. | Automated extraction and labelling of the arterial tree from whole-body MRA data | |
| KR102258070B1 (en) | Method for evaluating foot type and device evaluating foot type using the same | |
| CN115631232B (en) | Method for determining radial position of double-probe detector | |
| CN116958217B (en) | MRI and CT multi-mode 3D automatic registration method and device | |
| CN113781594A (en) | Magnetic resonance imaging image reconstruction method based on anatomical prior data | |
| Ens et al. | Automatic motion correction in cone-beam computed tomography | |
| CN117522978A (en) | Method and device for determining position information, storage medium and electronic device |
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 |