CN104224175B - Method of fusing two-dimensional magnetic resonance spectrum and three-dimensional magnetic resonance navigation image - Google Patents

Method of fusing two-dimensional magnetic resonance spectrum and three-dimensional magnetic resonance navigation image Download PDF

Info

Publication number
CN104224175B
CN104224175B CN201310454669.3A CN201310454669A CN104224175B CN 104224175 B CN104224175 B CN 104224175B CN 201310454669 A CN201310454669 A CN 201310454669A CN 104224175 B CN104224175 B CN 104224175B
Authority
CN
China
Prior art keywords
magnetic resonance
image
navigation
wave spectrum
dicom
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201310454669.3A
Other languages
Chinese (zh)
Other versions
CN104224175A (en
Inventor
姚成军
吴劲松
汤伟军
刘翌勋
庄冬晓
邱天明
毛颖
周良辅
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huashan Hospital of Fudan University
Original Assignee
Huashan Hospital of Fudan University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Huashan Hospital of Fudan University filed Critical Huashan Hospital of Fudan University
Priority to CN201310454669.3A priority Critical patent/CN104224175B/en
Publication of CN104224175A publication Critical patent/CN104224175A/en
Application granted granted Critical
Publication of CN104224175B publication Critical patent/CN104224175B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The invention belongs to the field of medical image postprocessing and relates to a method of fusing a two-dimensional magnetic resonance spectrum and a three-dimensional magnetic resonance navigation image. According to the method, three-dimensional magnetic resonance navigation information fused with a spectrum image is generated by means of spatial information and image registration by the aid of spatial information fields defined in header files of magnetic resonance original data and a non-rigid registration algorithm. The method has the advantages that the problem that the two-dimensional magnetic resonance spectrum image and the three-dimensional magnetic resonance navigation image cannot be fused is solved, metabolic information is increased for navigation, and the method is simple and convenient to operate, short in consumed time (20 minutes averagely), stable, reliable, free of extra introduction of hardware facilities and convenient for clinical practical application.

Description

A kind of method merging twodimensional magnetic resonance wave spectrum and three-dimensional magnetic resonance navigation image
Technical field
The invention belongs to medical image post processing field, it is related to a kind of twodimensional magnetic resonance wave spectrum and three-dimensional magnetic resonance navigation shadow Method as merging;The method is applied to all brain structures image and the post processing of metabolism image.
Background technology
The appearance of Use of Neuronavigation technology is the milestone in neurosurgery development history, and it makes neurosurgery from simple Rely on patient's subjective experience and enter into the epoch that objective image guides, greatly improve operation safety.Use of Neuronavigation is the most often adopted Guide image is nuclear magnetic resonance image, and three-dimensional nuclear magnetic resonance image can comprehensive display brain anatomy structure;But regrettably, Current three-dimensional magnetic resonance navigation image does not also include Magnetic Resonance Spectrum image.Described Magnetic Resonance Spectrum is for diagnostic imaging Mature technology, it is possible to provide the metabolic information of cerebral tissue or pathological changes, differentiating tumor property, grade malignancy, invading profit scope etc. side Face has substantial worth, if Magnetic Resonance Spectrum visual fusion can be entered navigation directions operation, will further improve operation safety.
The main cause that described twodimensional magnetic resonance wave spectrum image and three-dimensional magnetic resonance navigation image cannot merge is the two lattice Formula is different, and wave spectrum image is .rda form, and pixel format used by navigation sequence is DICOM format, there is no at present and merges the two Software.And in process of clinical application, spectral information can only be reduced in three-dimensional navigation image with subjective memory, method is coarse, Error is larger, and credibility is poor.
Therefore, how by Imaging processing method, twodimensional magnetic resonance wave spectrum accurately to be melted with three-dimensional magnetic resonance navigation image Close, be clinical at present demand urgently to be resolved hurrily.
Content of the invention
The purpose of the present invention is defect or the deficiency overcoming prior art, provides a kind of twodimensional magnetic resonance wave spectrum and Three-Dimensional Magnetic The method that resonance navigation image merges;The method adopts spatial information registration and non-rigid surface adjustment fusion evaluation, solves two Tie up the problem that Magnetic Resonance Spectrum image and three magnetic resonance navigation images can not merge, increase metabolic information, operation letter for navigation Just, time-consuming short, it is easy to Clinical practice.
The method that the twodimensional magnetic resonance wave spectrum of the present invention is merged with three-dimensional magnetic resonance navigation image is it is characterised in that adopt Spatial information field defined in the header file of magnetic resonance raw data and non-rigid surface adjustment algorithm, carry out spatial information and shadow As registration, generate the magnetic resonance three-dimensional navigation information merging wave spectrum image, concretely comprise the following steps:
(1)Obtain twodimensional magnetic resonance initial data, preliminary post processing carried out by the business software that magnetic resonance equipment carries, With Cho/NAA for value interested, generate twodimensional magnetic resonance wave spectrum image;
(2)Obtain three-dimensional magnetic resonance navigation DICOM data, with twodimensional magnetic resonance wave spectrum image as reference, with intracerebral lesion Centered on area, define region of interest;
(3)Above-mentioned data is imported work station, carries out wave spectrum image registering with the spatial information of navigation image;Two-dimensional magnetic is altogether Defined in the header file of vibration wave spectrum initial data, the field of spatial information is mainly:SliceThickness (thickness), PositionVector(Position vector),RowVector(Row vector),ColumnVector(Column vector), PixelSpacingRow(Pixel between-line spacing),PixelSpacingCol(Pixel column is spaced);Three-dimensional magnetic resonance navigation image Defined in DICOM image header file, the field of spatial information is mainly:SliceThickness (thickness), ImagePostionPatient(Patient image position),ImageOrientationPatient(Patient image positions), PixelSpacing(Pixel separation);Transformational relation between this two groups of data is(As shown in Figure 2):
SliceThickness(DICOM)=SliceThickness(rda)
ImagePostionPatient(DICOM)=PositionVector(rda)
ImageOrientationPatient(DICOM)=[RowVector,ColumnVector](rda)
PixelSpacing(DICOM)=[PixelSpacingRow,PixelSpacingCol];
(4)Further non-rigid method Image registration, specific algorithm is:
As with reference to image R (i, j, k) (i, j, k are body bundle index), its characteristic point is set to three-dimensional navigation imageAs floating image F (i, j, k) (i, j, k are body bundle index), its characteristic point is set to two-dimentional wave spectrum imageDeformation field U (x | D);Due to point set T and S concordance it needs to be determined that, the energy function of therefore design is,
J (R, F, S, T | D)=IntensityMatching (R, F (U (P | D)))
+λ×PointMatching(T,U(S|D)
For measuring the similarity of above-mentioned function, using cross-correlation(Cross-Correlation)As module,
Σ i , i , k ( R ( i , j , k ) - R ‾ ) ( F ( U ( p ijk | D ) ) - F ‾ ) Σ i , j , k ( R ( i , j , k ) - R ‾ ) 2 Σ i , j , k ( F ( U ( p ijk | D ) ) F ‾ ) 2
(5)By setting alpha value, adjust the pixel weight of two images, if Alpha is 1, show navigation picture completely Picture;If 0.5, respectively show 50%;If 0, only display wave spectrum;Can arbitrarily adjust as needed;
(6)If stage construction two dimension wave spectrogram, repeat the above steps;
(7)The image that registration is merged ultimately generates DICOM file, imports navigator, is applied to clinic.
Compared with prior art, its advantage is the present invention:Solve twodimensional magnetic resonance wave spectrum image and three magnetic resonance The problem that navigation image can not merge, easy and simple to handle, take short(Average 20 minutes), reliable and stable it is not necessary to additionally introduce hard Part facility, is easy to clinical practice.
Brief description
The present invention is further described with reference to the accompanying drawings and examples.
Fig. 1 is the overall flow chart of steps of the method for the invention.
Fig. 2 is twodimensional magnetic resonance wave spectrogram and three magnetic resonance navigation image data conversion relation schematic diagrams, wherein,
(1)Represent row pixel separation,(2)Represent row pixel separation.
Fig. 3 shows that twodimensional magnetic resonance wave spectrum selects region of interest and definition space information.
Twodimensional magnetic resonance wave spectrum interested position is merged after showing registration by Fig. 4 with three-dimensional magnetic resonance navigation data, and Realize three dimensional display, wherein, highlighted square is wave spectrum target spot interested.
Specific embodiment
Embodiment 1
The method that this twodimensional magnetic resonance wave spectrum is merged with three-dimensional magnetic resonance navigation image, concretely comprises the following steps:
(1)Obtain Magnetic Resonance Spectrum and magnetic resonance navigation picture initial data, preliminary treatment generates two dimension and 3-D view;
(2)Region of interest is set respectively on twodimensional magnetic resonance wave spectrum image and three-dimensional magnetic resonance navigation image;
(3)Carry out wave spectrum image registering with the spatial information of navigation image;
(4)Carry out non-rigid method and carry out the fusion registering with three-dimensional magnetic resonance navigation image of twodimensional magnetic resonance wave spectrum image, Generate the DICOM file merging, export to navigator and be used for clinic;
(5)Two kinds of image weights can be adjusted as needed at any time, reach optimum display effect.
Embodiment 2
The present invention adopts spatial information field and the calculation of non-rigid surface adjustment defined in the header file of magnetic resonance raw data Method, carries out spatial information and Image registration, generates the magnetic resonance three-dimensional navigation information merging wave spectrum image, concretely comprises the following steps:
(1)Obtain twodimensional magnetic resonance initial data, preliminary post processing carried out by the business software that magnetic resonance equipment carries, With Cho/NAA for value interested, generate twodimensional magnetic resonance wave spectrum image;
(2)Obtain three-dimensional magnetic resonance navigation DICOM data, with twodimensional magnetic resonance wave spectrum image as reference, with intracerebral lesion Centered on area, define region of interest;
(3)Above-mentioned data is imported work station, carries out wave spectrum image registering with the spatial information of navigation image;Two-dimensional magnetic is altogether Defined in the header file of vibration wave spectrum initial data, the field of spatial information is mainly:SliceThickness (thickness), PositionVector(Position vector),RowVector(Row vector),ColumnVector(Column vector), PixelSpacingRow(Pixel between-line spacing),PixelSpacingCol(Pixel column is spaced);Three-dimensional magnetic resonance navigation image Defined in DICOM image header file, the field of spatial information is mainly:SliceThickness (thickness), ImagePostionPatient(Patient image position),ImageOrientationPatient(Patient image positions), PixelSpacing(Pixel separation);Transformational relation between this two groups of data is(As shown in Figure 2):
SliceThickness(DICOM)=SliceThickness(rda)
ImagePostionPatient(DICOM)=PositionVector(rda)
ImageOrientationPatient(DICOM)=[RowVector,ColumnVector](rda)
PixelSpacing(DICOM)=[PixelSpacingRow,PixelSpacingCol];
(4)Further non-rigid method Image registration, specific algorithm is:
As with reference to image R (i, j, k) (i, j, k are body bundle index), its characteristic point is set to three-dimensional navigation imageAs floating image F (i, j, k) (i, j, k are body bundle index), its characteristic point is set to two-dimentional wave spectrum imageDeformation field U (x | D);Due to point set T and S concordance it needs to be determined that, the energy function of therefore design is,
J (R, F, S, T | D)=IntensityMatching (R, F (U (P | D)))
+λ×PointMatching(T,U(S|D)
For measuring the similarity of above-mentioned function, using cross-correlation(Cross-Correlation)As module,
Σ i , i , k ( R ( i , j , k ) - R ‾ ) ( F ( U ( p ijk | D ) ) - F ‾ ) Σ i , j , k ( R ( i , j , k ) - R ‾ ) 2 Σ i , j , k ( F ( U ( p ijk | D ) ) F ‾ ) 2
(5)By setting alpha value, adjust the pixel weight of two images, if Alpha is 1, show navigation picture completely Picture;If 0.5, respectively show 50%;If 0, only display wave spectrum;Can arbitrarily adjust as needed;
(6)If stage construction two dimension wave spectrogram, repeat the above steps;
(7)The image that registration is merged ultimately generates DICOM file, imports navigator, application.

Claims (2)

1. a kind of fusion twodimensional magnetic resonance wave spectrum and the method for three-dimensional magnetic resonance navigation image are it is characterised in that adopt magnetic resonance Spatial information field defined in the header file of initial data and non-rigid surface adjustment algorithm, carry out spatial information and image are joined Standard, generates the magnetic resonance three-dimensional navigation information merging wave spectrum image, it includes step:
(1) obtain twodimensional magnetic resonance initial data, preliminary post processing is carried out by the business software of magnetic resonance equipment, with Cho/ NAA is value interested, generates twodimensional magnetic resonance wave spectrum image;
(2) obtain three-dimensional magnetic resonance navigation DICOM data, with twodimensional magnetic resonance wave spectrum image as reference, with intracerebral lesion area be Center, defines region of interest;
(3) the three-dimensional magnetic resonance navigation DICOM data of the two dimension generating resonance wave spectrum image and acquisition is imported work station, carry out Wave spectrum image is registering with the spatial information of navigation image;Space defined in the header file of described twodimensional magnetic resonance wave spectrum initial data The field of information is mainly:SliceThickness,PositionVector,RowVector,ColumnVector, PixelSpacingRow,PixelSpacingCol;Defined in the DICOM image header file of described three-dimensional magnetic resonance navigation image The field of spatial information is mainly:
SliceThickness,ImagePostionPatient,ImageOrientationPatient,PixelSpacing;
(4) non-rigid method Image registration further;
(5) pass through to set alpha value, adjust the pixel weight of wave spectrum image and navigation image, if Alpha is 1, show completely Navigation image;If 0.5, then wave spectrum image each display 50% with navigation image;If 0, only display wave spectrum image;As needed Arbitrarily adjust the size of alpha value;
(6) if stage construction two dimension wave spectrogram, repeat the above steps;
(7) image merging registration ultimately generates DICOM file, imports navigator.
2. the method as described in claim 1 is it is characterised in that in described step (3), described twodimensional magnetic resonance wave spectrum original number According to the transformational relation between the DICOM image head file data of three-dimensional magnetic resonance navigation image it is:
SliceThickness (DICOM)=SliceThickness (rda)
ImagePostionPatient (DICOM)=PositionVector (rda)
ImageOrientationPatient (DICOM)=[RowVector, ColumnVector] (rda)
PixelSpacing (DICOM)=[PixelSpacingRow, PixelSpacingCol].
CN201310454669.3A 2013-09-27 2013-09-27 Method of fusing two-dimensional magnetic resonance spectrum and three-dimensional magnetic resonance navigation image Active CN104224175B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310454669.3A CN104224175B (en) 2013-09-27 2013-09-27 Method of fusing two-dimensional magnetic resonance spectrum and three-dimensional magnetic resonance navigation image

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310454669.3A CN104224175B (en) 2013-09-27 2013-09-27 Method of fusing two-dimensional magnetic resonance spectrum and three-dimensional magnetic resonance navigation image

Publications (2)

Publication Number Publication Date
CN104224175A CN104224175A (en) 2014-12-24
CN104224175B true CN104224175B (en) 2017-02-08

Family

ID=52213725

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310454669.3A Active CN104224175B (en) 2013-09-27 2013-09-27 Method of fusing two-dimensional magnetic resonance spectrum and three-dimensional magnetic resonance navigation image

Country Status (1)

Country Link
CN (1) CN104224175B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105913461A (en) * 2016-03-31 2016-08-31 中国人民解放军第四军医大学 Rat brain function magnetic resonance imaging standard space map template construction method
CN107389720B (en) * 2017-06-15 2019-11-19 上海联影医疗科技有限公司 Magnetic Resonance Spectrum exchange method and system
CN110021025B (en) * 2019-03-29 2021-07-06 上海联影智能医疗科技有限公司 Region-of-interest matching and displaying method, device, equipment and storage medium
CN111166332B (en) * 2020-03-04 2023-01-24 南京鼓楼医院 Radiotherapy target region delineation method based on magnetic resonance spectrum and magnetic resonance image

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5951475A (en) * 1997-09-25 1999-09-14 International Business Machines Corporation Methods and apparatus for registering CT-scan data to multiple fluoroscopic images
CN1926574A (en) * 2004-02-20 2007-03-07 皇家飞利浦电子股份有限公司 Device and process for multimodal registration of images
CN1973779A (en) * 2005-10-14 2007-06-06 美国西门子医疗解决公司 Method and system for cardiac imaging and catheter guidance for radio frequency (RF) ablation
CN101006933A (en) * 2006-01-23 2007-08-01 西门子公司 Method and device for displaying 3d objects

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7697972B2 (en) * 2002-11-19 2010-04-13 Medtronic Navigation, Inc. Navigation system for cardiac therapies
JP2007526788A (en) * 2003-07-10 2007-09-20 コーニンクレッカ フィリップス エレクトロニクス エヌ ヴィ Apparatus and method for operating an instrument in an anatomical structure

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5951475A (en) * 1997-09-25 1999-09-14 International Business Machines Corporation Methods and apparatus for registering CT-scan data to multiple fluoroscopic images
CN1926574A (en) * 2004-02-20 2007-03-07 皇家飞利浦电子股份有限公司 Device and process for multimodal registration of images
CN1973779A (en) * 2005-10-14 2007-06-06 美国西门子医疗解决公司 Method and system for cardiac imaging and catheter guidance for radio frequency (RF) ablation
CN101006933A (en) * 2006-01-23 2007-08-01 西门子公司 Method and device for displaying 3d objects

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
低磁场术中磁共振导航的融合弥散张量成像功能研究;姚成军等;《中国临床神经科学》;20071231;第15卷(第3期);第241-247页 *
基于VTK磁共振波谱多体素空间位置的配准;仲花等;《中国组织工程研究与临床康复》;20101224;第14卷(第52期);第9748-9751页 *
多影像融合技术在神经导航手术的临床应用;吴劲松等;《中华神经外科杂志》;20050430;第21卷(第4期);第227-231页 *

Also Published As

Publication number Publication date
CN104224175A (en) 2014-12-24

Similar Documents

Publication Publication Date Title
CN104224175B (en) Method of fusing two-dimensional magnetic resonance spectrum and three-dimensional magnetic resonance navigation image
de Senneville et al. EVolution: an edge-based variational method for non-rigid multi-modal image registration
US10949973B2 (en) Medical image analysis using mechanical deformation information
Hu et al. Obtaining PET/CT images from non-attenuation corrected PET images in a single PET system using Wasserstein generative adversarial networks
CN103932796A (en) Encephalic electrode individualization locating method based on multimode medical image data fusion
CN102651124B (en) Image fusion method based on redundant dictionary sparse representation and evaluation index
CN101987019A (en) Positron emission tomography (PET) and computed tomography (CT) cross-modality medical image fusion method based on wavelet transform
CN102903103A (en) Migratory active contour model based stomach CT (computerized tomography) sequence image segmentation method
CN103914823A (en) Method for registering stereoscopic medical image rapidly, accurately and non-linearly based on sparse representation
CN106504221A (en) Based on the Medical image fusion new method that quaternion wavelet converts context mechanism
Dai et al. Multi-organ auto-delineation in head-and-neck MRI for radiation therapy using regional convolutional neural network
Dai et al. TD-Net: Trans-Deformer network for automatic pancreas segmentation
CN104933759B (en) A kind of human brain tissue higher-dimension method for visualizing
CN104463885A (en) Partition method for multiple-sclerosis damage area
Qian et al. 3D automatic segmentation of brain tumor based on deep neural network and multimodal MRI images
Chen et al. Transformer-based multilevel region and edge aggregation network for magnetic resonance image segmentation
Zheng et al. Automated multi-modal transformer network (amtnet) for 3d medical images segmentation
CN103679796B (en) Bidimensional based on range conversion watershed algorithm transmission function body data identification method
Lei et al. Hippocampus substructure segmentation using morphological vision transformer learning
Zhou et al. Multimodality MRI synchronous construction based deep learning framework for MRI-guided radiotherapy synthetic CT generation
Gahm et al. Surface-based Tracking of U-fibers in the Superficial White Matter
CN103070698A (en) System and method for generating isocenter laser parameter in radiotherapy
CN103393408B (en) A kind of optoacoustic and fluorescent circulation mutual promotion formation method
Zheng et al. L2-norm scaled transformer for 3d head and neck primary tumors segmentation in pet-ct
CN110085301A (en) The synchronization delineation method of multisequencing image in a kind of radiotherapy planning system based on WEB

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant