CN106897993A - The construction method of probability collection of illustrative plates is rolled into a ball based on quantitative susceptibility imaging human brain gray matter core - Google Patents
The construction method of probability collection of illustrative plates is rolled into a ball based on quantitative susceptibility imaging human brain gray matter core Download PDFInfo
- Publication number
- CN106897993A CN106897993A CN201710020265.1A CN201710020265A CN106897993A CN 106897993 A CN106897993 A CN 106897993A CN 201710020265 A CN201710020265 A CN 201710020265A CN 106897993 A CN106897993 A CN 106897993A
- Authority
- CN
- China
- Prior art keywords
- illustrative plates
- collection
- tested
- magnetic susceptibility
- space
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
- G06T7/0014—Biomedical image inspection using an image reference approach
-
- 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
- G06T2207/10088—Magnetic resonance imaging [MRI]
-
- 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
- G06T2207/30016—Brain
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The invention discloses a kind of construction method that probability collection of illustrative plates is rolled into a ball based on quantitative susceptibility imaging human brain gray matter core, comprise the following steps：Use MR imaging apparatus gathered data；In being MNI spaces to normed space by tested quantitative magnetic susceptibility image registration；Brain deep layer kankar group region is sketched out on the magnetic susceptibility figure of normed space after registration by hand, and different probability collection of illustrative plates is made according to different overlap proportion；In for the magnetic susceptibility image evaluated, the measurement of similarity and coverage value is carried out with the goldstandard delineated by hand to the automatic segmentation result of different probability collection of illustrative plates, take the final probability collection of illustrative plates of the map construction of overlap proportion when similarity reaches peak value.The probability collection of illustrative plates that the present invention builds can avoid delineating by hand the artificial difference of introducing to the automatic segmentation that brain deep layer kankar is rolled into a ball, and core group segmentation result accuracy is higher, better than existing AAL and JH collection of illustrative plates；And the method than delineating by hand saves the time, improves the efficiency of graphical analysis work.
Description
Technical field
The invention belongs to mr imaging technique field, the human brain deep layer kankar of quantitative susceptibility imaging is based especially on
The construction method of the probability collection of illustrative plates of group.
Background technology
3 D stereo brain map can provide useful anatomical reference for the analysis of brain imaging data, by tested and brain template
Between autoregistration, tested brain effectively can be divided into corresponding region of interest by brain map.The 26S Proteasome Structure and Function of human brain is answered
It is miscellaneous various, there is its specificity per part.Spatial normalization can reduce the difference of brain anatomy between individuality, therefore it is
An important step in human brain atlas research.At present, the widely used big brain map of MR investigation comes from t1 weighted image,
Such as Talairach and Tournoux collection of illustrative plates, automatic anatomical landmarks (AAL) collection of illustrative plates based on Colin27 templates.
With the development of High field strenghth MRI technology, quantitative susceptibility imaging provides a kind of novel contrast mechanism,
It obtains local magnetic field variation characteristic using the phase information that general mr imaging technique is given up, then by complicated field to source
Inversion Calculation, can directly obtain quantitative magnetic susceptibility distribution map.The contrast mechanism of brain magnetic susceptibility image is mainly derived from iron content
Deep kankar group and the white matter containing myelin, its contrast is very good for deep nuclei.Johns is come from recently
The research team of Hopkins universities has made one based on the big brain map (JH collection of illustrative plates) for quantifying magnetic susceptibility figure, but its collection of illustrative plates
It is the brain based on single-subject, it is impossible to fully reflect the diversity of the anatomical structure of a large amount of Normal brains.
The content of the invention
The one kind proposed the invention aims to solve to split brain deep layer kankar clique problem automatically is based on fixed
Amount susceptibility imaging technology creates a kind of construction method of the automatic segmentation measurement collection of illustrative plates of deep layer kankar group, and the method uses one group
The new probability brain deep layer kankar group segmentation collection of illustrative plates of the quantitative magnetic susceptibility image creation one of tested high-resolution, the collection of illustrative plates
Can as a useful template carry out automatically brain deep layer core group identification and carry out big data group analysis measurement.
The object of the present invention is achieved like this：
A kind of construction method that probability collection of illustrative plates is rolled into a ball based on quantitative susceptibility imaging human brain deep layer kankar, the method is included
Step in detail below：
Step 1：A collection of Healthy subjects are recruited, a portion is randomly selected tested as collection of illustrative plates making object, remaining quilt
Study the object evaluated for later stage inspection collection of illustrative plates；
Step 2：Using MR imaging apparatus gathered data and data rebuild obtain quantitative magnetic susceptibility image；
Step 3：In being MNI spaces to normed space by all tested quantitative magnetic susceptibility image registrations；
Step 4：On the magnetic susceptibility figure of the normed space after all collection of illustrative plates make the tested registration of object, sketch out by hand big
Brain deep layer kankar rolls into a ball region, and makes different probability collection of illustrative plates according to different overlap proportion；
Step 5：In the magnetic susceptibility image of the normed space after for the tested registration of collection of illustrative plates evaluation object, by two researchs
Person delineates obtain brain deep layer kankar group region by hand；
Step 6：The different probability collection of illustrative plates obtained using step 4 is entered to the magnetic susceptibility image being tested for collection of illustrative plates evaluation object
The automatic segmentation of row, and will split two researchers in the core group that obtains and step 5 automatically and delineate that the core group for obtaining is i.e. golden to mark by hand
Standard, carries out the measurement of similarity and coverage value, takes final general of map construction of overlap proportion when similarity reaches peak value
Rate collection of illustrative plates.
In the method for the present invention, by tested quantitative magnetic susceptibility image registration to normed space described in step 3, specifically wrap
Include following steps：
Step a1：Tested sagittal plain high-resolution T1 structure pictures are redeveloped into cross-section position, and remove scalp, cranium, extracted
Brain tissue part；
Step a2：Mould figure to being tested is removed scalp, cranium, extracts brain tissue part；
Step a3：Tested T1 structures picture is carried out with tested mould figure using linear registration Algorithm registering, be tested
T1 figures in mould map space；
Step a4：Used what step a3 was obtained in linear registration Algorithm and normed space by the T1 figures in die trial map space
ICBM T1 figures carry out registration, obtain tested T1 structure charts in normed space and turn from by die trial map space to normed space
The matrix for changing；
Step a5：The transition matrix obtained using step a4 is transformed to tested magnetic susceptibility figure in normed space, is obtained
Tested magnetic susceptibility image in normed space.
In the method for the present invention, described in step 6 in for the magnetic susceptibility image evaluated, different probability collection of illustrative plates is divided automatically
Cut the core group for obtaining and delineate the core group i.e. goldstandard for obtaining by hand with two researchers, carry out the survey of similarity and coverage value
Amount, using following evaluating：
Kappa coefficients：
Dice coefficients：
Coverage rate:
Wherein, similarity Dice coefficients refer to correct segmentation result number of pixels account for whole cut zone (comprising by hand
All regions of segmentation and collection of illustrative plates segmentation automatically) ratio, its difference to two area sizes and position is very sensitive, value model
It is [0,1] to enclose, and 1 represents completely the same.It is infinity to roll into a ball number of pixels relative to target core due to TN number of pixels, so that
Kappa coefficients are equal with Dice coefficients, and had research to point out that Dice coefficients are a kind of special cases of Kappa coefficients.Covering
Rate refers to that the correct number of pixels being partitioned into accounts for manual segmentation result and collection of illustrative plates segmentation result jointly comprises the ratio in region.
The probability collection of illustrative plates that the present invention makes can avoid delineating introducing by hand to the automatic segmentation that brain deep layer kankar is rolled into a ball
Artificial difference, core group segmentation result accuracy it is higher, better than existing AAL collection of illustrative plates and JH collection of illustrative plates；In addition, the present invention makes
Probability collection of illustrative plates than by hand delineate region of interest conventional method save the time, can effectively improve graphical analysis work effect
Rate.
Brief description of the drawings
Fig. 1 is flow chart of the present invention；
Fig. 2 is the flow chart that the present invention builds probability collection of illustrative plates embodiment；
Fig. 3 evaluates principle schematic for the present invention；
Fig. 4 is that the different probability collection of illustrative plates that the embodiment of the present invention makes splits each core group in the Basal ganglia region for obtaining automatically
Dice coefficients and coverage rate distribution trend curve map；
Fig. 5 is that the different probability collection of illustrative plates that the embodiment of the present invention makes splits each core group in the basis cranii and cerebellum for obtaining automatically
Dice coefficients and coverage rate distribution trend curve map.
Specific embodiment
With reference to specific examples below and accompanying drawing, the present invention is described in further detail.Implement process of the invention, bar
Part, experimental program method etc., in addition to the following special content for referring to, are the universal knowledege and common knowledge of this area, this
Invention is not particularly limited content.
The quantitative susceptibility imaging human brain deep layer kankar of the present invention rolls into a ball the construction method of probability collection of illustrative plates, based on to brain
On the basis of the extraordinary quantitative magnetic susceptibility figure of deep nuclei contrast, the method that optimal probability is taken using many people, making it is general
Forthright collection of illustrative plates.The method can improve the efficiency of graphical analysis work to a certain extent.
Initial data of the present invention to collecting is introduced individually below to pass through registration, delineate, evaluate, and obtains optimum probability
The specific implementation process of collection of illustrative plates.Wherein, data source is in 3T MR imaging apparatus system (Siemens MAGNETOM Trio a
Tim 3T), quantitative susceptibility imaging scanning sequence is three-dimensional many echo gradient echo (Gradient echo, GRE) sequences, tool
Swept-volume parameter is as follows：Repetition time (TR)=60ms, the first echo time (TE1)=6.8ms, echo sounding (Δ TE)=
6.8ms, number of echoes=8, flip angle (FA)=15 °, the visual field (FOV)=240 × 180mm2, voxel size=0.625mm ×
0.625mm × 2mm, the number of plies=96.To reduce the sampling time, parallel sampling is used in phase-encoding direction (tested left and right directions)
Technology, accelerated factor is 2.T1 weighting high resolution structure pictures prepare fast gradient echo using three-dimensional magnetization
(Magnetization-Prepared Rapid Gradient Echo, MPRAGE) sequence, design parameter is as follows：Median sagittal
Bit scan, TR=2530ms, reversing time (TI)=1100ms, FA=7 °, TE=2.34ms, FOV=256 × 256mm2, body
Plain size=1mm × 1mm × 1mm, the number of plies=192.
The present embodiment has recruited 15 Healthy subjects (7 women, 8 males, average age 24.3 ± 1.0 years old), at random
Choose wherein 10 (each 5 of men and women) to be tested as collection of illustrative plates making object, remaining 5 tested as later stage inspection collection of illustrative plates validity
Object.
The complex data collected by gradin-echo by phase-fitting, phase unwrapping around, remove ambient field, based on shape
The steps such as the dipole inversion algorithm (Morphology Enabled Dipole Inversion, MEDI) of state reconstruct cranium
The cross-section position magnetic susceptibility figure of brain.
Fig. 1-2 is shown the flow chart that collection of illustrative plates of the present invention makes, including：Tested quantitative magnetic susceptibility image registration is arrived
Brain deep layer kankar group is sketched out on the magnetic susceptibility figure of normed space in normed space (MNI spaces), after registration by hand
Region, the probability collection of illustrative plates to delineating the different situations for obtaining carry out the measurement of similarity and coverage value, take similarity and reach peak
The collection of illustrative plates of overlap proportion during value makes final probability collection of illustrative plates.Idiographic flow is as follows：
1) be redeveloped into for tested sagittal plain high-resolution T1 structure pictures cross-section by the three-dimensional the poster processing soft carried using scanner
Bit image (T1 MPRAGE figures), the present embodiment rebuild after image resolution ratio be:1mm×1mm×1mm；Then FSL is used
5.0.9 BET softwares (Brain Extraction Tool) removal scalp in kit, cranium, extract brain tissue part.
2) to being tested the mould figure (GreRawMag figures) that three-dimensional GRE sequences are obtained, also using in FSL 5.0.9 kits
BET softwares, removal scalp, cranium carry out extracting brain tissue part.
3) using the rigid body translation algorithm in the linear registration Algorithms of FLIRT in FSL 5.0.9 kits by head clearing and
The cross-section bitmap being tested after skull be registrated to after skull by die trial map space in, obtain by the T1 in die trial map space
Figure (Mag T1 figures).
4) will be by using 12 parameter affine transform methods in the linear registration Algorithms of FLIRT in FSL 5.0.9 kits
The Mag T1 figures of examination are registrated in the normed space of ICBM T1 figures (MNI), obtain T1 figures (the MNI T1 being tested in MNI spaces
Figure) and by the transition matrix of die trial map space to MNI spaces.
5) transition matrix is applied on tested QSM figures, is transformed in normed space, obtain the quilt of normed space
Examination QSM figures (MNI QSM figures).
6) on the normed space QSM figures of the above-mentioned single-subject obtained by pretreatment, utilized by a researcher
The softwares of ITK-SNAP 3.2 sketch out six bilateral deep-brain kankar group ROI (caudate nucleus manually：CN, shell core：PU, globus pallidus：
GP, black substance：SN, rubrum：RN, dentate nucleus：DN), it is tested deep grey matter segmentation figure in MNI coordinate systems so as to obtain each
(Deep Gray Matter Parcellation Map, DGMPM).
7) all collection of illustrative plates are made into the tested DGMPM figures of object, by each class ROI according to overlap proportion from 10%-
100% incremented by successively 10% mode saves as a probability collection of illustrative plates respectively, is selected using corresponding collection of illustrative plates evaluation method afterwards
The collection of illustrative plates of optimum superposing ratio is used as final probability collection of illustrative plates.
The collection of illustrative plates evaluation method is specific as follows：
The softwares of ITK-SNAP 3.2 are utilized in 5 tested normed space QSM figures for evaluating by two researchers
Deep nuclei ROI is sketched out manually, in this, as the goldstandard for evaluating the collection of illustrative plates accuracy of separation.
The quantitatively evaluating of the segmentation result accuracy for each ROI is carried out using principle shown in Fig. 3.Assuming that researcher
The pixel that segmentation obtains target core group by hand is T, and probability collection of illustrative plates of the present invention, AAL collection of illustrative plates, Johns Hopkins (JH) collection of illustrative plates are obtained
The pixel of the target core group for arriving is R, and researcher's craft segmentation and collection of illustrative plates split the overlapping region of the target image for obtaining automatically
Number of pixels is true positives (TP), and inclusion region number of pixels is true outside manual segmentation result and collection of illustrative plates automatic segmentation result
Negative (TN), the number of pixels not comprising manual segmentation area in collection of illustrative plates automatic segmentation result is false positive (FP), by hand
The number of pixels not comprising collection of illustrative plates automatic segmentation result region is false negative (FN) in segmentation result.Using Kappa coefficients, Dice
Coefficient and coverage rate (Overlap Ratio, OR) are analyzed to segmentation result reliability：
Kappa coefficients：
Dice coefficients：
Coverage rate:
Wherein, similarity Dice coefficients refer to correct segmentation result number of pixels account for whole cut zone (comprising by hand
All regions of segmentation and collection of illustrative plates segmentation automatically) ratio, its difference to two area sizes and position is very sensitive, value model
It is [0,1] to enclose, and 1 represents completely the same.It is infinity to roll into a ball number of pixels relative to target core due to TN number of pixels, so that
Kappa coefficients are equal with Dice coefficients.Coverage rate refers to that the correct number of pixels being partitioned into accounts for manual segmentation result with collection of illustrative plates point
Cut the ratio that result jointly comprises region.
The atlas calculation of the selection optimum superposing ratio is as follows：
In the present embodiment, for 5 tested magnetic susceptibility figures for evaluating and testing, respectively using 10 kinds for making of the invention not
The probability atlas registration of negative lap rate and the method delineated by hand obtain two groups of ROI, then carry out Dice to this two groups of ROI regions
Coefficient and coverage measure, and weigh to both results, select the probability collection of illustrative plates of best overlapped ratio as final general
Rate collection of illustrative plates.
In order to ensure the reliability of optimal collection of illustrative plates selection, the method for multiple averaging is used：First, obtain each it is tested by
Dice coefficients and coverage rate between the ROI that every manual ROI for delineating of researcher and collection of illustrative plates are split automatically, by two researchers
The numerical value for obtaining is averaged, and obtains two, each tested each region average value of researcher's evaluation and test；Secondly, to all tested
Average Dice coefficients and average coverage rate be averaging again, so as to obtain each research core group final Dice coefficient values and cover
Lid rate value.The collection of illustrative plates of each class situation using the method for multiple averaging obtain each study core roll into a ball final Dice coefficients and
Coverage rate, then analyzes their distribution trend, finally draws optimal collection of illustrative plates, refering to shown in accompanying drawing 4,5.
It with 5 normal being tested is test object that a and b are respectively in Fig. 4, carries out the segmentation and two automatically of each probability collection of illustrative plates
Position researcher is split the average similarity (Dice coefficients) between the kankar group of all area-of-interests of Basal ganglia and is covered by hand
The comparative result of lid rate, each core group is expressed as in figure：Caudate nucleus (CN), shell core (PU), globus pallidus (GP), left (L), the right side
(R), it is seen that the probability collection of illustrative plates overlapping percentages Dice coefficients that all cores are rolled into a ball when taking 50% reach maximum, and now coverage rate reaches
More than 70%, the most of region of core group is contained, therefore Basal ganglia part finally takes collection of illustrative plates when overlapping percentages are 50%
Make final probability collection of illustrative plates.
The same with Fig. 4, a and b are the kankars to the area-of-interest in 5 basis craniis and cerebellum of test object in Fig. 5
Group is estimated, and each core group is expressed as in figure：Black substance (SN), rubrum (RN), dentate nucleus (DN), left (L), right (R), it is seen that
The Dice coefficients of the same core groups all at 50% of probability collection of illustrative plates overlapping percentages reach maximum, and now coverage rate reaches 60%
More than, contain the most of region of core group, therefore the collection of illustrative plates of core group in basis cranii and cerebellum when being also 50% with overlapping percentages
Make final probability collection of illustrative plates.
Protection content of the invention is not limited to above example.Under the spirit and scope without departing substantially from inventive concept, this
Art personnel it is conceivable that change and advantage be all included in the present invention, and with appending claims be protect
Shield scope.
Claims (2)
1. it is a kind of based on quantitative susceptibility imaging human brain gray matter core roll into a ball probability collection of illustrative plates construction method, it is characterised in that the side
Method is comprised the following steps：
Step 1：A collection of Healthy subjects are recruited, which part is randomly selected tested as collection of illustrative plates making object, remaining tested conduct
The object that later stage inspection collection of illustrative plates is evaluated；
Step 2：Using MR imaging apparatus gathered data and data rebuild obtain quantitative magnetic susceptibility image；
Step 3：In being MNI spaces to normed space by all tested quantitative magnetic susceptibility image registrations；
Step 4：On the magnetic susceptibility figure of the normed space after all collection of illustrative plates make the tested registration of object, brain depth is sketched out by hand
Layer kankar group region, and make different probability collection of illustrative plates according to different overlap proportion；
Step 5：In the magnetic susceptibility image of the normed space after for the tested registration of collection of illustrative plates evaluation object, by two researcher's hands
Work is delineated and obtains brain deep layer kankar group region；
Step 6：The different probability collection of illustrative plates obtained using step 4 is carried out certainly to the magnetic susceptibility image being tested for collection of illustrative plates evaluation object
Dynamic segmentation, and will split two researchers in the core group and step 5 for obtaining automatically and delineate the core for obtaining by hand and roll into a ball i.e. goldstandard, enter
The measurement of row similarity and coverage value, takes the probability graph described in the map construction of overlap proportion when similarity reaches peak value
Spectrum.
2. construction method according to claim 1, it is characterised in that quantitative magnetic susceptibility image that will be tested described in step 3
It is registrated in normed space, including step in detail below：
Step a1：Tested sagittal plain high-resolution T1 structure pictures are redeveloped into cross-section position, and remove scalp, cranium, extract brain group
Knit part；
Step a2：Mould figure to being tested is removed scalp, cranium, extracts brain tissue part；
Step a3：Tested T1 structures picture is carried out with tested mould figure using linear registration Algorithm registering, obtained by die trial figure
T1 figures in space；
Step a4：Used what step a3 was obtained in linear registration Algorithm and normed space by the T1 figures in die trial map space
ICBM T1 figures carry out registration, obtain the tested T1 structure charts in normed space and are changed from by die trial map space to normed space
A matrix；
Step a5：The transition matrix obtained using step a4 is transformed to tested magnetic susceptibility figure in normed space, obtains standard
Tested magnetic susceptibility image in space.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710020265.1A CN106897993B (en) | 2017-01-12 | 2017-01-12 | Construction method based on quantitative susceptibility imaging human brain gray matter core group probability map |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710020265.1A CN106897993B (en) | 2017-01-12 | 2017-01-12 | Construction method based on quantitative susceptibility imaging human brain gray matter core group probability map |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106897993A true CN106897993A (en) | 2017-06-27 |
| CN106897993B CN106897993B (en) | 2019-07-26 |
Family
ID=59199173
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710020265.1A Active CN106897993B (en) | 2017-01-12 | 2017-01-12 | Construction method based on quantitative susceptibility imaging human brain gray matter core group probability map |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106897993B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109034136A (en) * | 2018-09-06 | 2018-12-18 | 湖北亿咖通科技有限公司 | Image processing method, device, picture pick-up device and storage medium |
| CN109272514A (en) * | 2018-10-05 | 2019-01-25 | 数坤（北京）网络科技有限公司 | The sample evaluation method and model training method of coronary artery parted pattern |
| CN110458869A (en) * | 2019-05-10 | 2019-11-15 | 珠海慧脑云计算有限公司 | A kind of method for registering and its system of newborn's magnetic resonance configurations brain image |
| CN111127526A (en) * | 2019-11-27 | 2020-05-08 | 华东师范大学 | Method for positioning deep nuclei of brain by using quantitative susceptibility map |
| CN111445520A (en) * | 2020-05-26 | 2020-07-24 | 河南省人民医院 | Thalamus abdominal intermediate nucleus positioning method and positioning device based on magnetic resonance quantitative magnetic sensitive image |
| CN111481827A (en) * | 2020-04-17 | 2020-08-04 | 上海深透科技有限公司 | Quantitative magnetic susceptibility imaging and methods for DBS potentially stimulated target region localization |
| CN111681230A (en) * | 2020-06-10 | 2020-09-18 | 华中科技大学同济医学院附属同济医院 | System and method for scoring high-signal of white matter of brain |
| US10896508B2 (en) | 2018-02-07 | 2021-01-19 | International Business Machines Corporation | System for segmentation of anatomical structures in cardiac CTA using fully convolutional neural networks |
| CN113066113A (en) * | 2021-04-27 | 2021-07-02 | 河北医科大学第二医院 | Method and equipment for constructing spontaneous hypertensive rat brain template and map set |
| CN113592847A (en) * | 2021-08-10 | 2021-11-02 | 浙江大学 | QSM deep brain nuclear mass automatic segmentation method based on deep learning |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101887583A (en) * | 2010-06-24 | 2010-11-17 | 东软集团股份有限公司 | Method and device for extracting brain tissue image |
| CN103442765A (en) * | 2011-03-29 | 2013-12-11 | 瓦尔克公司 | Device and method for altering neurotransmitter level in brain |
| CN103826536A (en) * | 2011-09-26 | 2014-05-28 | 大日本印刷株式会社 | Medical image processing device, medical image processing method, program |
| CN104523275A (en) * | 2014-12-25 | 2015-04-22 | 西安电子科技大学 | Construction method for health people white matter fiber tract atlas |
| CN104644173A (en) * | 2015-01-14 | 2015-05-27 | 北京工业大学 | Depression risk three-grade early warning method and depression risk three-grade early warning system |
| CN104921741A (en) * | 2014-03-19 | 2015-09-23 | 柯尼卡美能达株式会社 | Image analysis device, imaging system, and image analysis method |
| US20150310639A1 (en) * | 2014-04-25 | 2015-10-29 | Berkin Bilgic | Systems and methods for fast reconstruction for quantitative susceptibility mapping using magnetic resonance imaging |
| US20160054410A1 (en) * | 2014-08-19 | 2016-02-25 | General Electric Company | System and method for locating and quantifying a biomarker for neurological disease |
| CN105816192A (en) * | 2016-03-03 | 2016-08-03 | 王雪原 | Method for three-dimensional registration and brain tissue extraction of individual human brain multimodality medical images |
-
2017
- 2017-01-12 CN CN201710020265.1A patent/CN106897993B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101887583A (en) * | 2010-06-24 | 2010-11-17 | 东软集团股份有限公司 | Method and device for extracting brain tissue image |
| CN103442765A (en) * | 2011-03-29 | 2013-12-11 | 瓦尔克公司 | Device and method for altering neurotransmitter level in brain |
| CN103826536A (en) * | 2011-09-26 | 2014-05-28 | 大日本印刷株式会社 | Medical image processing device, medical image processing method, program |
| CN104921741A (en) * | 2014-03-19 | 2015-09-23 | 柯尼卡美能达株式会社 | Image analysis device, imaging system, and image analysis method |
| US20150310639A1 (en) * | 2014-04-25 | 2015-10-29 | Berkin Bilgic | Systems and methods for fast reconstruction for quantitative susceptibility mapping using magnetic resonance imaging |
| US20160054410A1 (en) * | 2014-08-19 | 2016-02-25 | General Electric Company | System and method for locating and quantifying a biomarker for neurological disease |
| CN104523275A (en) * | 2014-12-25 | 2015-04-22 | 西安电子科技大学 | Construction method for health people white matter fiber tract atlas |
| CN104644173A (en) * | 2015-01-14 | 2015-05-27 | 北京工业大学 | Depression risk three-grade early warning method and depression risk three-grade early warning system |
| CN105816192A (en) * | 2016-03-03 | 2016-08-03 | 王雪原 | Method for three-dimensional registration and brain tissue extraction of individual human brain multimodality medical images |
Non-Patent Citations (1)
| Title |
|---|
| ISSEL ANNE L.LIM等: "Human brain atlas for automated region of interest selection in quantitative susceptibility mapping:Application to determine iron content in deep gray matter structures", 《NEUROIMAGE》 * |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10896508B2 (en) | 2018-02-07 | 2021-01-19 | International Business Machines Corporation | System for segmentation of anatomical structures in cardiac CTA using fully convolutional neural networks |
| CN109034136B (en) * | 2018-09-06 | 2021-07-20 | 湖北亿咖通科技有限公司 | Image processing method, image processing apparatus, image capturing device, and storage medium |
| CN109034136A (en) * | 2018-09-06 | 2018-12-18 | 湖北亿咖通科技有限公司 | Image processing method, device, picture pick-up device and storage medium |
| CN109272514A (en) * | 2018-10-05 | 2019-01-25 | 数坤（北京）网络科技有限公司 | The sample evaluation method and model training method of coronary artery parted pattern |
| CN109272514B (en) * | 2018-10-05 | 2021-07-13 | 数坤(北京)网络科技股份有限公司 | Sample evaluation method and model training method of coronary artery segmentation model |
| CN110458869A (en) * | 2019-05-10 | 2019-11-15 | 珠海慧脑云计算有限公司 | A kind of method for registering and its system of newborn's magnetic resonance configurations brain image |
| CN110458869B (en) * | 2019-05-10 | 2021-10-19 | 珠海慧脑云计算有限公司 | Registration method and system for brain image of magnetic resonance structure of newborn |
| CN111127526A (en) * | 2019-11-27 | 2020-05-08 | 华东师范大学 | Method for positioning deep nuclei of brain by using quantitative susceptibility map |
| CN111127526B (en) * | 2019-11-27 | 2023-03-31 | 华东师范大学 | Method for positioning deep nuclei of brain by using quantitative susceptibility map |
| CN111481827A (en) * | 2020-04-17 | 2020-08-04 | 上海深透科技有限公司 | Quantitative magnetic susceptibility imaging and methods for DBS potentially stimulated target region localization |
| CN111481827B (en) * | 2020-04-17 | 2023-10-20 | 上海深透科技有限公司 | Quantitative susceptibility imaging and method for locating target area of potential stimulation of DBS |
| CN111445520A (en) * | 2020-05-26 | 2020-07-24 | 河南省人民医院 | Thalamus abdominal intermediate nucleus positioning method and positioning device based on magnetic resonance quantitative magnetic sensitive image |
| CN111681230A (en) * | 2020-06-10 | 2020-09-18 | 华中科技大学同济医学院附属同济医院 | System and method for scoring high-signal of white matter of brain |
| CN113066113A (en) * | 2021-04-27 | 2021-07-02 | 河北医科大学第二医院 | Method and equipment for constructing spontaneous hypertensive rat brain template and map set |
| CN113066113B (en) * | 2021-04-27 | 2022-01-11 | 河北医科大学第二医院 | Method and equipment for constructing spontaneous hypertensive rat brain template and map set |
| CN113592847B (en) * | 2021-08-10 | 2023-10-10 | 浙江大学 | Deep learning-based QSM brain deep nucleolus automatic segmentation method |
| CN113592847A (en) * | 2021-08-10 | 2021-11-02 | 浙江大学 | QSM deep brain nuclear mass automatic segmentation method based on deep learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106897993B (en) | 2019-07-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106897993B (en) | Construction method based on quantitative susceptibility imaging human brain gray matter core group probability map | |
| Renvall et al. | Automatic cortical surface reconstruction of high-resolution T1 echo planar imaging data | |
| Heiervang et al. | Between session reproducibility and between subject variability of diffusion MR and tractography measures | |
| Gao et al. | A universal intensity standardization method based on a many-to-one weak-paired cycle generative adversarial network for magnetic resonance images | |
| Wei et al. | Learning-based single-step quantitative susceptibility mapping reconstruction without brain extraction | |
| Oishi et al. | Quantitative evaluation of brain development using anatomical MRI and diffusion tensor imaging | |
| Roy et al. | MR contrast synthesis for lesion segmentation | |
| Feng et al. | MoDL-QSM: Model-based deep learning for quantitative susceptibility mapping | |
| CN102999917B (en) | Cervical cancer image automatic segmentation method based on T2-MRI and DW-MRI | |
| JP6840898B2 (en) | Parkinson's disease diagnostic device and method | |
| Rathi et al. | Statistical analysis of fiber bundles using multi-tensor tractography: application to first-episode schizophrenia | |
| Peng et al. | Development of a human brain diffusion tensor template | |
| CN104523275A (en) | Construction method for health people white matter fiber tract atlas | |
| Hwang et al. | An automatic cerebellum extraction method in T1-weighted brain MR images using an active contour model with a shape prior | |
| CN106898000A (en) | A kind of automatic division method of magnetic resonance imaging cerebral gray matter core group | |
| Gulban et al. | Improving a probabilistic cytoarchitectonic atlas of auditory cortex using a novel method for inter-individual alignment | |
| Fripp et al. | Automated segmentation of the menisci from MR images | |
| CN105997074A (en) | Magnetic resonance quantitative magnetic susceptibility imaging multi-echo phase fitting method | |
| CN111751770A (en) | Method for quantitatively measuring exchange speed of water molecules inside and outside blood vessels by means of non-magnetic resonance contrast agent | |
| Studholme | Dense feature deformation morphometry: Incorporating DTI data into conventional MRI morphometry | |
| German et al. | Brain tissues have single-voxel signatures in multi-spectral MRI | |
| Loyek et al. | Detection of focal cortical dysplasia lesions in MRI using textural features | |
| Ma et al. | Column-based cortical depth analysis of the diffusion anisotropy and radiality in submillimeter whole-brain diffusion tensor imaging of the human cortical gray matter in vivo | |
| Maher et al. | Neuroinformatics and analysis of connectomic alterations due to cerebral microhemorrhages in geriatric mild neurotrauma: microhemorrhages in geriatric neurotrauma | |
| Zhang et al. | A framework of automatic brain tumor segmentation method based on information fusion of structural and functional MRI signals |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |