CN109171670A - A kind of 3D blood vessel imaging algorithm based on reverse Principal Component Analysis - Google Patents
A kind of 3D blood vessel imaging algorithm based on reverse Principal Component Analysis Download PDFInfo
- Publication number
- CN109171670A CN109171670A CN201810658187.2A CN201810658187A CN109171670A CN 109171670 A CN109171670 A CN 109171670A CN 201810658187 A CN201810658187 A CN 201810658187A CN 109171670 A CN109171670 A CN 109171670A
- Authority
- CN
- China
- Prior art keywords
- signal
- blood
- blood vessel
- algorithm
- component analysis
- 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
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
- A61B5/02007—Evaluating blood vessel condition, e.g. elasticity, compliance
-
- 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
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- 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
- A61B5/7225—Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation
Abstract
The present invention provides a kind of 3D blood vessel imaging algorithm based on reverse Principal Component Analysis, it is related to medical blood vessel technical field of imaging, first, ecg signal acquiring is carried out using ecg-gating method, gate-control signal is exported according to electrocardiosignal and is transmitted to spectrometer, generates structural image data and structural image data is registrated.Secondly, building complex signal statistical model, extracts the blood vessel red blood cell in complex signal statistical model using reverse Principal Component Analysis and reflects signal, and generate blood-stream image.Finally, obtaining the independent entry axis of blood-stream image profile point cluster using independent component analysis model, the translational movement and rotation amount of independent entry axis are calculated using difference searching algorithm, then blood-stream image is registrated.The technical solution improves the signal-to-noise ratio of three-dimensional blood vessel imaging, reduces the mixed and disorderly background information generated by biological tissue's reflection, improves image quality, alleviate the technical problem that blood vessel imaging quality of the existing technology is low, noise is serious.
Description
Technical field
The present invention relates to medical blood vessel technical field of imaging, more particularly, to a kind of 3D based on reverse Principal Component Analysis
Blood vessel imaging algorithm.
Background technique
Modern medicine believes that, the pathological condition of biological tissue and the vascular morphology of region of interest have close relationship,
In, optical coherence tomography (OCT) is a kind of novel imaging technique, has non-invasive, high-resolution, noninvasive and imaging deep
Spend the advantages that higher.However in lesion early stage, the difference of the scattering properties between normal tissue and pathological tissues is unobvious, causes
OCT structured image can not provide the results such as effective information guiding clinical treatment in time.The imaging of optical coherence tomography capilary
(OCTA) be a kind of based on realizing on the basis of OCT to the technology of miniature blood vessel imaging, both may be implemented it is quick, noninvasive, without mark
Note, high-resolution imaging, it is also possible to obtain the three-dimensional angiography of tissue blood vessel.OCTA technology can be from the micro-structure of tissue
Blood vessel is isolated, such as Phase-resolved optical Doppler tomography PRODT, this method, which is based primarily upon, compares one B- of OCTA signal
The phase difference between adjacent A- scanning in scanning, the phase difference and blood flow velocity have direct relation.
Although current PRODT is widely used, since it is lower to the sensitivity of blood flow, it is difficult to be clearly observed
Flow velocity is the capillary under 0.1~0.9mm/s or the lower morbid state of flow velocity.In order to improve its sensitivity, people is studied
Member has been presented for some improved methods, such as is scanned using forward and backward B-, and utilizes the phase between adjacent B- scanning
Variance.Since the time interval between adjacent B- scanning is relatively long (ms magnitude), enable the method to slow to possessing
The capillary of blood flow is imaged, and Phase-resolved optical Doppler tomography method needs longer acquisition time
(25min), and it is very sensitive to object of which movement artifact.
In realizing process of the present invention, at least there are the following problems in the prior art: OCTA system living body for inventor's discovery
In imaging process, due to there is the inevitably biology shake such as heartbeat, breathing, imaging signal to noise ratio is caused to reduce, image
Compromised quality；In OCTA data acquisition, correlated noise caused by the positional shift of adjacent smooth sequence scanning leads into image quality
Amount reduces；In OCTA image reconstruction procedure, the positional shift of adjacent two field pictures leads in image that there are jittering noises.It is wide at present
The general blood flow imaging algorithm sensitivity used is lower, can not extract the image of minute blood vessel；Existing blood-stream image restructing algorithm without
Method, which is efficiently removed, reflects caused mixed and disorderly background information by biological tissue.Therefore, that there are blood vessel imaging quality is low, makes an uproar for the prior art
The serious technical problem of sound.
Summary of the invention
In view of this, the purpose of the present invention is to provide a kind of, the 3D blood vessel imaging based on reverse Principal Component Analysis is calculated
Method, to alleviate the technical problem that blood vessel imaging quality existing in the prior art is low, noise is serious.
The embodiment of the invention provides a kind of 3D blood vessel imaging algorithm based on reverse Principal Component Analysis, including walk as follows
It is rapid:
Data collection steps: carrying out ecg signal acquiring using ecg-gating method, exports gate-control signal according to electrocardiosignal
And it is transmitted to spectrometer, structural image data is generated according to output gate-control signal, structural image data is registrated；
Image reconstruction step: complex signal statistical model is constructed according to structural image data, complex signal statistical model is to include
The Tissue reflectance signal of non-blood flow tissue ingredient, the red blood cell of flow components reflect the linear statistical mould of signal and white Gaussian noise
Type；Specifically, complex signal statistical model is I=Ic+Ib+ N, wherein I is tissue complex signal intensity matrix, IcFor non-blood flow tissue
Reflected signal strength matrix, IbFor blood flow tissue medium vessels red blood cell reflected signal strength matrix, N is then white Gaussian noise point
Amount；
Blood vessel red blood cell in complex signal statistical model is extracted using reverse Principal Component Analysis and reflects signal, calculates blood vessel
Red blood cell reflects the characteristic value and feature vector of signal, and the blood vessel red blood cell reflection signal of extraction is Ib=(1-H (w)) × I, and
Generate blood-stream image, wherein H (w) is the characteristic value that signal is reflected according to red blood cell and the filtering signal sound of feature vector building
Answer function；
3D rendering step of registration: the profile point cluster F of the wantonly two width blood-stream image of constructioni, profile point cluster is Fi=[xi,
yi,zi]T, i=(1,2,3 ..., n), the profile point cluster F of two width blood-stream imagesoAnd FtIt is expressed as oi=xoie1+yoie2+
zoie3And ti=xtie1+ytie2+ztie3；
Profile point cluster F is obtained using independent component analysis modeliIndependent entry axis, independent component analysis model be based on
The optimization algorithm model of objective function；
The translational movement and rotation amount that independent entry axis is calculated using difference searching algorithm, according to translational movement and rotation amount to blood flow
Image is registrated.
Further, in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis, the heart
Electric switch control method specifically: gating module initialization, while electrocardiosignal is acquired, using electrocardiosignal as input data and gate mould
Preset gate value is compared in block, judges whether electrocardiosignal is higher than threshold value, if so, output corresponds to the current input heart
The gate-control signal of electric signal.
Further, it in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis, adopts
It is specific that electrocardio-data collection is carried out with ecg-gating method further include: is added sensitive bit shift compensation model in collection process, is utilized
Frequency domain filtering method and multi-modal search method calculate the sensitive moving displacement parameter of sensitive bit shift compensation model.
Further, it in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis, adopts
The characteristic value and feature vector of the blood vessel red blood cell reflection signal in complex signal statistical model are extracted with reverse Principal Component Analysis
Later, further includes: blood flow red blood cell is reflected by signal precision using superposition mean value phase elimination.
Further, right in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis
Structural images be registrated and are specifically registrated using the method for registering based on feature to structural images, the registration side based on feature
Method are as follows: orthogonal transformation is carried out to artifact matrix Q；Specifically, Q=Ii-Ii+1, wherein Ii、Ii+1Respectively two width consecutive frame images
Picture element matrix.
Further, excellent in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis
Change algorithm is artificial bee colony algorithm, ant group algorithm, differential evolution algorithm, bat algorithm, any in cuckoo algorithm.
Further, in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis, base
The objective function in the optimization algorithm model of objective function specifically:
Establish F=AfSf, solve mixed matrix WfMeet Yf=WfF=WfAfSf, wherein SfFor point set, AfFor point set matrix, F is
Point cluster, YfIt is FiThe estimation of isolated component after separation, WfAs objective function.
Further, in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis, meter
Calculate the rotation amount for corresponding to independent entry axis and translation measurer body are as follows:
Rotation amount are as follows:Translational movement Δ C=Co-Ct, wherein lo-first,
lo-secondThe first yuan of axis and second yuan of axis of respectively the first width blood flow structure image, lt-first, lt-secondRespectively the second width blood flow
The first yuan of axis and second yuan of axis of structure width image, wherein
Respectively indicate the centroid of two width blood flow structure image outline point models.
Further, quick in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis
Sense bit shift compensation model is si(n)=d (n)+ri(n)+ci(n)+hi(n), wherein siIt (n) is sensitive moving displacement parameter, d (n)
For respiratory movement component, riIt (n) is cardiac motion components, ciIt (n) is translational motion component, hiIt (n) is noise component(s), sensitive position
Moving each component in compensation model is the function that displacement changes over time.
Further, it in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis, folds
Add mean value phase elimination expression formula are as follows:Wherein, Ib_clearTo refine blood flow information intensity
Matrix, Ib_iFor the single frames blood flow information intensity matrix of the i-th frame, Ib_meanIt is strong by i-th to M frame multi-frame mean blood flow information
Matrix is spent, M is that B- scans total degree.
The embodiment of the present invention brings following the utility model has the advantages that based on reverse principal component analysis provided by the embodiment of the present invention
The 3D blood vessel imaging algorithm of method exports gate letter according to electrocardiosignal firstly, carrying out ecg signal acquiring using ecg-gating method
Number and be transmitted to spectrometer, structural image data is generated according to output gate-control signal, structural image data is registrated.Its
It is secondary, complex signal statistical model is constructed according to structural image data, complex signal statistical model is the group for including non-blood flow tissue ingredient
Knit reflection signal, flow components red blood cell reflection signal and white Gaussian noise Linear Statistical Model.Using reverse principal component
Analytic approach extracts the blood vessel red blood cell in complex signal statistical model and reflects signal, calculates the characteristic value of blood vessel red blood cell reflection signal
And feature vector, the blood vessel red blood cell of extraction reflects signal, and generates blood-stream image.Finally, the wantonly two width blood-stream image of construction
Profile point cluster, the independent entry axis of profile point cluster is obtained using independent component analysis model, and independent component analysis model is base
In the optimization algorithm model of objective function；The translational movement and rotation amount that independent entry axis is calculated using difference searching algorithm, according to flat
Shifting amount and rotation amount are registrated blood-stream image.The technical solution carries out data acquisition, then benefit by using ecg-gating method
Generation blood-stream image is combined with complex signal statistical model with reverse Principal Component Analysis, 3D rendering finally is carried out to blood-stream image
Registration, realizes the three-dimensional angiography of tissue blood vessel, improves the signal-to-noise ratio of imaging, reduces and reflects generation by biological tissue
Mixed and disorderly background information, reduce biology shake influence, improve image quality, thus alleviate the prior art presence
Blood vessel imaging quality is low, technical problem that noise is serious, while the technical solution reduces imaging to object of which movement artifact
Susceptibility improves image sensitivity, the image zooming-out suitable for minute blood vessel.
To enable the above objects, features and advantages of the present invention to be clearer and more comprehensible, preferred embodiment is cited below particularly, and cooperate
Appended attached drawing, is described in detail below.
Detailed description of the invention
It, below will be to specific in order to illustrate more clearly of the specific embodiment of the invention or technical solution in the prior art
Embodiment or attached drawing needed to be used in the description of the prior art be briefly described, it should be apparent that, it is described below
Attached drawing is some embodiments of the present invention, for those of ordinary skill in the art, before not making the creative labor
It puts, is also possible to obtain other drawings based on these drawings.
Fig. 1 is a kind of process of the 3D blood vessel imaging algorithm based on reverse Principal Component Analysis provided in an embodiment of the present invention
Figure；
Fig. 2 is in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis, using the heart
Effect picture before electric switch control method；
Fig. 3 is in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis, using the heart
Effect picture after electric switch control method；
Fig. 4 is structure chart in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis
As registration effect schematic diagram.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with attached drawing to the present invention
Technical solution be clearly and completely described, it is clear that described embodiments are some of the embodiments of the present invention, rather than
Whole embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art are not making creative work premise
Under every other embodiment obtained, shall fall within the protection scope of the present invention.
Currently, due to there is heartbeat, breathing etc., inevitably biology is trembled during OCTA system living imaging
It is dynamic, cause imaging signal to noise ratio to reduce, picture quality is impaired；In OCTA data acquisition, the position of adjacent smooth sequence scanning is inclined
Correlated noise caused by moving causes image quality to reduce；In OCTA image reconstruction procedure, the positional shift of adjacent two field pictures is led
There are jittering noises in cause image.Now widely used blood flow imaging algorithm sensitivity is lower, can not extract minute blood vessel
Image；Existing blood-stream image restructing algorithm, which can not be removed efficiently, reflects caused mixed and disorderly background information by biological tissue, is based on this,
A kind of 3D blood vessel imaging algorithm based on reverse Principal Component Analysis provided in an embodiment of the present invention, can be improved three-dimensional blood vessel at
The signal-to-noise ratio of picture reduces the mixed and disorderly background information generated by biological tissue's reflection, improves image quality.
Referring to Fig. 1, a kind of 3D blood vessel imaging algorithm based on reverse Principal Component Analysis provided in an embodiment of the present invention
Flow chart.3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis includes the following steps:
Data collection steps S100: ecg signal acquiring is carried out using ecg-gating method, is exported and is gated according to electrocardiosignal
Signal is simultaneously transmitted to spectrometer, generates structural image data according to output gate-control signal, is registrated to structural image data.?
In data acquisition, the artifact due to caused by the bounce and respiratory movement of heart will affect image quality, data acquisition
Process is acquired electrocardiosignal while acquiring sample data by the way of ecg-gating, to reach removal heartbeat and exhale
Inhale the purpose of imaging artefacts caused by periodic motion.
Further, in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis, the heart
Electric switch control method specifically: gating module initialization, while ecg signal acquiring module acquires electrocardiosignal, electrocardiosignal compares mould
Block is compared using electrocardiosignal as input data with gate value preset in gating module, judges whether electrocardiosignal is higher than
Threshold value, if so, output corresponds to the gate-control signal of current input ecg signal.Referring to fig. 2 and Fig. 3, the embodiment of the present invention
In the 3D blood vessel imaging algorithm based on reverse Principal Component Analysis provided, using the effect contrast figure before and after ecg-gating method.
Further, it in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis, adopts
It is specific that electrocardio-data collection is carried out with ecg-gating method further include: is added sensitive bit shift compensation model in collection process, is utilized
Frequency domain filtering method and multi-modal search method calculate the sensitive moving displacement parameter of sensitive bit shift compensation model.Under special circumstances,
For example some heart of patient suffer from some diseases, lead to heartbeat and non-exhibiting is periodical or patient is young children
Cause some acyclic sensitive movements, can all seriously affect the quality of imaging, be added in data acquisition at this time quick
Nastic movement displacement model can realize the compensation to the sensitive displacement of aperiodicity in a certain range.
Further, quick in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis
Sense bit shift compensation model is si(n)=d (n)+ri(n)+ci(n)+hi(n), wherein siIt (n) is sensitive moving displacement parameter, d (n)
For respiratory movement component, riIt (n) is cardiac motion components, ciIt (n) is translational motion component, hiIt (n) is noise component(s), sensitive position
Move the function that the displacement that each component in compensation model is its corresponding parameter changes over time.
Referring to fig. 4, in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis, structure
Image registration effect diagram.Further, it is provided in an embodiment of the present invention based on the 3D blood vessel of reverse Principal Component Analysis at
It is specific using the method for registering based on feature as be registrated to structural images in algorithm, in data collection steps, to adjacent two
Positional shift between secondary B- scanning compensates, removal artifact due to caused by shake, artifact, that is, noise section.Based on spy
The method for registering of sign are as follows: orthogonal transformation is carried out to artifact matrix Q；Specifically, Q=Ii-Ii+1, wherein Ii、Ii+1Respectively two width
The picture element matrix of consecutive frame structural images.Further, after the step of being registrated to structural images, benefit can also be added
Zero technology realizes the purpose for increasing the precision of structural images registration.
3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis further includes image reconstruction
Step S200, wherein step S210: constructing the complex signal statistical model of certain point in spatial domain according to structural image data, multiple
Signal statistics model can guarantee the integrality of signal, and complex signal statistical model is the Tissue reflectance for including non-blood flow tissue ingredient
Signal, flow components red blood cell reflection signal and additional white Gaussian noise Linear Statistical Model；Specifically, complex signal is united
Meter model is I=Ic+Ib+ N, wherein I is tissue complex signal intensity matrix, IcFor non-blood flow tissue reflected signal strength matrix,
IbFor blood flow tissue medium vessels red blood cell reflected signal strength matrix, N is then white Gaussian noise component.
Step S220: the blood vessel red blood cell in complex signal statistical model is extracted using reverse Principal Component Analysis and reflects letter
Number, blood vessel red blood cell reflects the characterization degree of reflection of blood vessels in tissue, is the principal component in complex signal statistical model, meter
The characteristic value and feature vector of blood vessel red blood cell reflection signal are calculated, the blood vessel red blood cell reflection signal of extraction is Ib=(1-H (w))
× I reflects the characteristic value of signal according to blood vessel red blood cell and feature vector designs filter function, and then designs PCA backward filtering
Device filters out non-blood flow tissue ingredient, retains blood flow tissue ingredient, and generate blood-stream image, wherein H (w) is anti-according to red blood cell
Penetrate the characteristic value of signal and the filtering signal receptance function of feature vector building.
Further, it in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis, adopts
The characteristic value and feature vector of the blood vessel red blood cell reflection signal in complex signal statistical model are extracted with reverse Principal Component Analysis
Later, further includes: blood flow red blood cell is reflected by signal precision using superposition mean value phase elimination.
Further, it in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis, folds
Add mean value phase elimination expression formula are as follows:Wherein, Ib_clearTo refine blood flow information intensity
Matrix, Ib_iFor the single frames blood flow information intensity matrix of the i-th frame, Ib_meanIt is strong by i-th to M frame multi-frame mean blood flow information
Matrix is spent, M is that B- scans total degree.
3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis further includes that 3D rendering is matched
Quasi- step S300,3D rendering registration technique can effectively calculate the offset and rotation amount of different time acquisition image, for doctor
It is raw to provide reference to related fields such as the preoperative and postoperative variations and the detection of the state of an illness of lesion locations.Wherein, step S310: construction is appointed
The profile point cluster F of two width blood-stream imagesi, profile point cluster is Fi=[xi,yi,zi]T, i=(1,2,3 ..., n), n in formula
For the points set of profile point, profile point company-data characterizes the location information of profile point, the profile point of two width blood-stream images
Cluster FoAnd FtIt is expressed as oi=xoie1+yoie2+zoie3And ti=xtie1+ytie2+ztie3, wherein e1、e2、e3Respectively
Unit direction vector.
Step S320: profile point cluster F is obtained using independent component analysis (ICA) model in unsupervised learningiIt is only
Vertical member axis, independent component analysis model are the optimization algorithm model based on objective function.Further, the embodiment of the present invention provides
The 3D blood vessel imaging algorithm based on reverse Principal Component Analysis in, optimization algorithm be artificial bee colony algorithm, ant group algorithm, difference
It is evolution algorithm, bat algorithm, any in cuckoo algorithm.
Further, in the 3D blood vessel imaging algorithm provided in an embodiment of the present invention based on reverse Principal Component Analysis, base
The objective function in the optimization algorithm model of objective function specifically: establish F=AfSf, specifically, being deposited according to ICA algorithm principle
In a point set SfWith a point set matrix Af, meet F=AfSf, the purpose of ICA algorithm is to find one to solve mixed matrix Wf, meet
Yf=WfF=WfAfSf, wherein SfFor point set, AfFor point set matrix, F is point cluster, YfIt is FiIsolated component after separation is estimated
Meter, WfObjective function as in the optimization algorithm model based on objective function.
Step S330: translational movement and the rotation of the independent entry axis in maximum statistical correlation direction are calculated using difference searching algorithm
Amount, is registrated blood-stream image according to translational movement and rotation amount.Further, provided in an embodiment of the present invention based on reverse main
In the 3D blood vessel imaging algorithm of componential analysis, the rotation amount for corresponding to independent entry axis and translation measurer body are calculated are as follows:
Rotation amount are as follows:Translational movement Δ C=Co-Ct, wherein lo-first,
lo-secondThe first yuan of axis and second yuan of axis of respectively the first width blood flow structure image, lt-first, lt-secondRespectively the second width blood flow
The first yuan of axis and second yuan of axis of structure width image, wherein
Respectively indicate the centroid of two width blood flow structure image outline point models.
3D blood vessel imaging algorithm based on reverse Principal Component Analysis provided by the embodiment of the present invention, firstly, using the heart
Electric switch control method carries out ecg signal acquiring, exports gate-control signal according to electrocardiosignal and is transmitted to spectrometer, is gated according to output
Signal generates structural image data, is registrated to structural image data.Secondly, constructing complex signal system according to structural image data
Model is counted, complex signal statistical model is the Tissue reflectance signal for including non-blood flow tissue ingredient, the reflection of the red blood cell of flow components
The Linear Statistical Model of signal and white Gaussian noise.Blood vessel in complex signal statistical model is extracted using reverse Principal Component Analysis
Red blood cell reflects signal, calculates the characteristic value and feature vector of blood vessel red blood cell reflection signal, the blood vessel red blood cell reflection of extraction
Signal, and generate blood-stream image.Finally, the profile point cluster of the wantonly two width blood-stream image of construction, using independent component analysis model
The independent entry axis of profile point cluster is obtained, independent component analysis model is the optimization algorithm model based on objective function；Utilize difference
The translational movement and rotation amount for dividing searching algorithm to calculate independent entry axis, are registrated blood-stream image according to translational movement and rotation amount.
The technical solution carries out data acquisition by using ecg-gating method, recycles complex signal statistical model and reverse principal component analysis
Method combines generation blood-stream image, finally carries out 3D rendering registration to blood-stream image, and the three-dimensional blood vessel for realizing tissue blood vessel is made
Shadow improves the signal-to-noise ratio of imaging, reduces the mixed and disorderly background information generated by biological tissue's reflection, reduces biology shake
It influences, improves image quality, blood vessel imaging quality of the existing technology is low, the serious technology of noise to alleviate
Problem, while the technical solution reduces imaging to the susceptibility of object of which movement artifact, improves image sensitivity, is suitable for thin
The image zooming-out of thin vessels.
Embodiment described above, only a specific embodiment of the invention, to illustrate technical solution of the present invention, rather than
It is limited, scope of protection of the present invention is not limited thereto, although having carried out with reference to the foregoing embodiments to the present invention detailed
Illustrate, those skilled in the art should understand that: anyone skilled in the art the invention discloses
In technical scope, it can still modify to technical solution documented by previous embodiment or variation can be readily occurred in, or
Person's equivalent replacement of some of the technical features；And these modifications, variation or replacement, do not make corresponding technical solution
Essence is detached from the spirit and scope of technical solution of the embodiment of the present invention, should be covered by the protection scope of the present invention.Therefore,
Protection scope of the present invention should be based on the protection scope of the described claims.
Claims (10)
1. a kind of 3D blood vessel imaging algorithm based on reverse Principal Component Analysis, which comprises the steps of:
Data collection steps: carrying out ecg signal acquiring using ecg-gating method, exports gate-control signal according to electrocardiosignal and passes
Spectrometer is transported to, structural image data is generated according to output gate-control signal, structural image data is registrated；
Image reconstruction step: complex signal statistical model is constructed according to structural image data, the complex signal statistical model is to include
The Tissue reflectance signal of non-blood flow tissue ingredient, the red blood cell of flow components reflect the linear statistical mould of signal and white Gaussian noise
Type；Specifically, the complex signal statistical model is, whereinITo organize complex signal intensity matrix, I c For non-blood
Tissue reflectance signal strength matrix is flowed,I b For blood flow tissue medium vessels red blood cell reflected signal strength matrix,NIt is then Gauss white noise
Sound component；
Blood vessel red blood cell in complex signal statistical model reflection signal is extracted using reverse Principal Component Analysis, described in calculating
Blood vessel red blood cell reflects the characteristic value and feature vector of signal, and the blood vessel red blood cell reflection signal of extraction is, and generate blood-stream image, whereinH(w) it is characteristic value and the spy that signal is reflected according to blood vessel red blood cell
Levy the filtering signal receptance function of vector building；
3D rendering step of registration: the profile point cluster of the wantonly two width blood-stream image of constructionF i , the profile point cluster is,, the profile point cluster of two width blood-stream imagesF o WithF t It is expressed asWith；
Profile point cluster is obtained using independent component analysis modelF i Independent entry axis, the independent component analysis model be based on
The optimization algorithm model of objective function；
The translational movement and rotation amount that independent entry axis is calculated using difference searching algorithm, according to translational movement and rotation amount to blood-stream image
It is registrated.
2. algorithm according to claim 1, which is characterized in that the ecg-gating method specifically: gating module initialization,
Electrocardiosignal is acquired simultaneously, is compared using electrocardiosignal as input data with gate value preset in gating module, is judged
Whether electrocardiosignal is higher than threshold value, if so, output corresponds to the gate-control signal of current input ecg signal.
3. algorithm according to claim 1, which is characterized in that described to carry out electrocardio-data collection tool using ecg-gating method
Body further include: add sensitive bit shift compensation model in collection process, calculated using frequency domain filtering method and multi-modal search method quick
Feel the sensitive moving displacement parameter of bit shift compensation model.
4. algorithm according to claim 1, which is characterized in that described to extract the letter in reply using reverse Principal Component Analysis
After the characteristic value and feature vector of blood vessel red blood cell reflection signal in number statistical model, further includes: using superposition mean value phase
Blood flow red blood cell is reflected signal precision by elimination.
5. algorithm according to claim 1, which is characterized in that described to carry out being registrated specific use based on spy to structural images
The method for registering of sign is registrated structural images, the method for registering based on feature are as follows: to artifact matrixQCarry out positive alternation
It changes；Specifically,, whereinI i 、I i+1The respectively picture element matrix of two width consecutive frame images.
6. algorithm according to claim 1, which is characterized in that the optimization algorithm be artificial bee colony algorithm, ant group algorithm,
It is differential evolution algorithm, bat algorithm, any in cuckoo algorithm.
7. algorithm according to claim 1, which is characterized in that target in the optimization algorithm model based on objective function
Function specifically:
It establishes, solve mixed matrixW f Meet, whereinS f For point set,A f For point set matrix,FFor
Point cluster,Y f It isF i The estimation of isolated component after separation,W f As objective function.
8. algorithm according to claim 1, which is characterized in that described to calculate the rotation amount for corresponding to independent entry axis and translation
Amount specifically:
Rotation amount are as follows:, translational movement, whereinl o-first ,l o-second The first yuan of axis and second yuan of axis of respectively the first width blood flow structure image,l t-first ,l t-second Respectively the second width
The first yuan of axis and second yuan of axis of blood flow structure width image, wherein, Respectively indicate the centroid of two width blood flow structure image outline point models.
9. algorithm according to claim 3, which is characterized in that specifically, the sensitivity bit shift compensation model is, whereins i (n) it is sensitive moving displacement parameter,d(n) it is respiratory movement component,r i
(n) it is cardiac motion components,c i (n) it is translational motion component,h i (n) it is noise component(s), it is each in the sensitivity bit shift compensation model
A component is the function that displacement changes over time.
10. algorithm according to claim 4, which is characterized in that the superposition mean value phase elimination expression formula are as follows:, whereinI b_clear To refine blood flow information intensity matrix,I b_i It isiThe single frames of frame
Blood flow information intensity matrix,I b_mean For byiToMThe multi-frame mean blood flow information intensity matrix of frame,MIt is B- scanning total time
Number.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810658187.2A CN109171670B (en) | 2018-06-25 | 2018-06-25 | 3D blood vessel imaging algorithm based on reverse principal component analysis method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810658187.2A CN109171670B (en) | 2018-06-25 | 2018-06-25 | 3D blood vessel imaging algorithm based on reverse principal component analysis method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109171670A true CN109171670A (en) | 2019-01-11 |
| CN109171670B CN109171670B (en) | 2021-02-05 |
Family
ID=64948449
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810658187.2A Active CN109171670B (en) | 2018-06-25 | 2018-06-25 | 3D blood vessel imaging algorithm based on reverse principal component analysis method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109171670B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111145092A (en) * | 2019-12-16 | 2020-05-12 | 华中科技大学鄂州工业技术研究院 | Method and device for processing infrared blood vessel image on leg surface |
| CN111493853A (en) * | 2020-04-24 | 2020-08-07 | 天津恒宇医疗科技有限公司 | Blood vessel parameter evaluation method and system for angiodermic diseases |
| CN116563414A (en) * | 2023-07-11 | 2023-08-08 | 天津博霆光电技术有限公司 | OCT-based cardiovascular imaging fibrillation shadow eliminating method and equipment |
Citations (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101246543A (en) * | 2008-03-18 | 2008-08-20 | 苏州纳米技术与纳米仿生研究所 | Examiner identity appraising system based on bionic and biological characteristic recognition |
| CN102077108A (en) * | 2008-04-28 | 2011-05-25 | 康奈尔大学 | Tool for accurate quantification in molecular mri |
| US20120307116A1 (en) * | 2011-06-01 | 2012-12-06 | Lansel Steven P | Learning of Image Processing Pipeline for Digital Imaging Devices |
| CN103077550A (en) * | 2012-12-10 | 2013-05-01 | 华北电力大学（保定） | Vascular four-dimensional reconstruction method in NOT gate-controlled ICUS (intravascular ultrasound) image sequence |
| CN103271734A (en) * | 2012-12-10 | 2013-09-04 | 中国人民解放军第一五二中心医院 | Heart rate measuring method based on low-end imaging device |
| CN103679816A (en) * | 2013-12-30 | 2014-03-26 | 北京师范大学 | Criminology-oriented computer-assisted facial reconstruction method for skulls of unknown body sources |
| CN104299216A (en) * | 2014-10-22 | 2015-01-21 | 北京航空航天大学 | Multimodality medical image fusion method based on multiscale anisotropic decomposition and low rank analysis |
| CN104835151A (en) * | 2015-04-24 | 2015-08-12 | 南京邮电大学 | Improved artificial bee colony algorithm-based image registration method |
| CN105662413A (en) * | 2015-12-31 | 2016-06-15 | 深圳先进技术研究院 | Myocardium T1 quantifying method and device |
| CN106021912A (en) * | 2016-05-17 | 2016-10-12 | 武汉盛世康和健康管理有限公司 | Remote online health management system |
| CN106166058A (en) * | 2016-08-04 | 2016-11-30 | 温州医科大学 | One is applied to optical coherence tomography blood vessel imaging method and OCT system |
| US20170103525A1 (en) * | 2015-10-09 | 2017-04-13 | Mayo Foundation For Medical Education And Research | System and Method for Tissue Characterization Based on Texture Information Using Multi-Parametric MRI |
| CN206209898U (en) * | 2016-12-06 | 2017-05-31 | 中国科学院深圳先进技术研究院 | Three-dimensional cardiac image re-construction system |
| CN107295217A (en) * | 2017-06-30 | 2017-10-24 | 中原智慧城市设计研究院有限公司 | A kind of video noise estimation method based on principal component analysis |
| CN107485366A (en) * | 2017-09-07 | 2017-12-19 | 天津海仁医疗技术有限公司 | A kind of optics Micro flow contrast imaging method based on microvesicle enhancing |
| CN107578381A (en) * | 2017-08-09 | 2018-01-12 | 天津恒宇医疗科技有限公司 | Interference of light fault imaging Color Mapping Approach is peeped in one kind |
| CN107595250A (en) * | 2017-09-30 | 2018-01-19 | 浙江大学 | The blood flow imaging method and system of contrast is mixed with figure based on motion |
| CN107680052A (en) * | 2017-09-18 | 2018-02-09 | 广州慧扬健康科技有限公司 | Angiographic image blood vessel strengthening system based on PCA |
| CN108042125A (en) * | 2017-05-27 | 2018-05-18 | 天津海仁医疗技术有限公司 | A kind of high speed endoscopic optical coherent flow imaging system |
| WO2018094381A1 (en) * | 2016-11-21 | 2018-05-24 | Tecumseh Vision, Llc | System and method for automatic assessment of disease condition using oct scan data |
| CN108113647A (en) * | 2016-11-28 | 2018-06-05 | 深圳先进技术研究院 | A kind of electrocardiosignal sorter and method |
-
2018
- 2018-06-25 CN CN201810658187.2A patent/CN109171670B/en active Active
Patent Citations (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101246543A (en) * | 2008-03-18 | 2008-08-20 | 苏州纳米技术与纳米仿生研究所 | Examiner identity appraising system based on bionic and biological characteristic recognition |
| CN102077108A (en) * | 2008-04-28 | 2011-05-25 | 康奈尔大学 | Tool for accurate quantification in molecular mri |
| US20120307116A1 (en) * | 2011-06-01 | 2012-12-06 | Lansel Steven P | Learning of Image Processing Pipeline for Digital Imaging Devices |
| CN103077550A (en) * | 2012-12-10 | 2013-05-01 | 华北电力大学（保定） | Vascular four-dimensional reconstruction method in NOT gate-controlled ICUS (intravascular ultrasound) image sequence |
| CN103271734A (en) * | 2012-12-10 | 2013-09-04 | 中国人民解放军第一五二中心医院 | Heart rate measuring method based on low-end imaging device |
| CN103679816A (en) * | 2013-12-30 | 2014-03-26 | 北京师范大学 | Criminology-oriented computer-assisted facial reconstruction method for skulls of unknown body sources |
| CN104299216A (en) * | 2014-10-22 | 2015-01-21 | 北京航空航天大学 | Multimodality medical image fusion method based on multiscale anisotropic decomposition and low rank analysis |
| CN104835151A (en) * | 2015-04-24 | 2015-08-12 | 南京邮电大学 | Improved artificial bee colony algorithm-based image registration method |
| US20170103525A1 (en) * | 2015-10-09 | 2017-04-13 | Mayo Foundation For Medical Education And Research | System and Method for Tissue Characterization Based on Texture Information Using Multi-Parametric MRI |
| CN105662413A (en) * | 2015-12-31 | 2016-06-15 | 深圳先进技术研究院 | Myocardium T1 quantifying method and device |
| CN106021912A (en) * | 2016-05-17 | 2016-10-12 | 武汉盛世康和健康管理有限公司 | Remote online health management system |
| CN106166058A (en) * | 2016-08-04 | 2016-11-30 | 温州医科大学 | One is applied to optical coherence tomography blood vessel imaging method and OCT system |
| WO2018094381A1 (en) * | 2016-11-21 | 2018-05-24 | Tecumseh Vision, Llc | System and method for automatic assessment of disease condition using oct scan data |
| CN108113647A (en) * | 2016-11-28 | 2018-06-05 | 深圳先进技术研究院 | A kind of electrocardiosignal sorter and method |
| CN206209898U (en) * | 2016-12-06 | 2017-05-31 | 中国科学院深圳先进技术研究院 | Three-dimensional cardiac image re-construction system |
| CN108042125A (en) * | 2017-05-27 | 2018-05-18 | 天津海仁医疗技术有限公司 | A kind of high speed endoscopic optical coherent flow imaging system |
| CN107295217A (en) * | 2017-06-30 | 2017-10-24 | 中原智慧城市设计研究院有限公司 | A kind of video noise estimation method based on principal component analysis |
| CN107578381A (en) * | 2017-08-09 | 2018-01-12 | 天津恒宇医疗科技有限公司 | Interference of light fault imaging Color Mapping Approach is peeped in one kind |
| CN107485366A (en) * | 2017-09-07 | 2017-12-19 | 天津海仁医疗技术有限公司 | A kind of optics Micro flow contrast imaging method based on microvesicle enhancing |
| CN107680052A (en) * | 2017-09-18 | 2018-02-09 | 广州慧扬健康科技有限公司 | Angiographic image blood vessel strengthening system based on PCA |
| CN107595250A (en) * | 2017-09-30 | 2018-01-19 | 浙江大学 | The blood flow imaging method and system of contrast is mixed with figure based on motion |
Non-Patent Citations (2)
| Title |
|---|
| JIN等: ""Extracting contrast-filled vessels in X-ray angiography by graduated RPCA with motion coherency constraint"", 《PATTERN RECOGNITION》 * |
| 王希等: ""超微血管成像技术鉴别诊断乳腺肿瘤"", 《中国医学影像技术》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111145092A (en) * | 2019-12-16 | 2020-05-12 | 华中科技大学鄂州工业技术研究院 | Method and device for processing infrared blood vessel image on leg surface |
| CN111493853A (en) * | 2020-04-24 | 2020-08-07 | 天津恒宇医疗科技有限公司 | Blood vessel parameter evaluation method and system for angiodermic diseases |
| CN116563414A (en) * | 2023-07-11 | 2023-08-08 | 天津博霆光电技术有限公司 | OCT-based cardiovascular imaging fibrillation shadow eliminating method and equipment |
| CN116563414B (en) * | 2023-07-11 | 2023-09-12 | 天津博霆光电技术有限公司 | OCT-based cardiovascular imaging fibrillation shadow eliminating method and equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109171670B (en) | 2021-02-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Shandhi et al. | Performance analysis of gyroscope and accelerometer sensors for seismocardiography-based wearable pre-ejection period estimation | |
| Kovalski et al. | Three-dimensional automatic quantitative analysis of intravascular ultrasound images | |
| Kaisti et al. | Stand-alone heartbeat detection in multidimensional mechanocardiograms | |
| CN108618749B (en) | Retina blood vessel three-dimensional reconstruction method based on portable digital fundus camera | |
| CN105342568B (en) | The optical coherence angiographic method and system of joint phase and amplitude | |
| CN103606152B (en) | DSA vascular image segmentation method based on SIFT feature point clustering and Boolean different operation | |
| CN110662480A (en) | Device, system and method for measuring and processing physiological signals of a subject | |
| CN109171670A (en) | A kind of 3D blood vessel imaging algorithm based on reverse Principal Component Analysis | |
| CN107862724B (en) | Improved microvascular blood flow imaging method | |
| CN110236515A (en) | A kind of contactless heart rate detection method based on near-infrared video | |
| KR20130079694A (en) | Apparatus and method for measureing velocity vector imaging of blood vessel | |
| CN111387959A (en) | Non-contact physiological parameter detection method based on IPPG | |
| Tadi et al. | A new algorithm for segmentation of cardiac quiescent phases and cardiac time intervals using seismocardiography | |
| CN112001122A (en) | Non-contact physiological signal measuring method based on end-to-end generation countermeasure network | |
| CN114748032A (en) | Motion noise compensation method based on OCT blood vessel imaging technology | |
| CN110236514A (en) | The real-time heart rate detection method that mode based on video is extracted and median filtering combines | |
| Hennersperger et al. | Vascular 3D+ T freehand ultrasound using correlation of doppler and pulse-oximetry data | |
| Pahlavan et al. | A novel cyber physical system for 3-D imaging of the small intestine in vivo | |
| Nedoma et al. | A novel FBG-based triggering system for cardiac MR imaging at 3 Tesla: a pilot pre-clinical study | |
| Bajraktari et al. | Methods of Contactless Blood Pressure Measurement: A Systematic Review | |
| CN116012820A (en) | Non-contact type driver heart rate detection method | |
| Huang et al. | Unobtrusive inter-beat interval estimation from multichannel ballistocardiogram signal using kalman filter | |
| CN112155511A (en) | Method for compensating human eye shake in OCT (optical coherence tomography) acquisition process based on deep learning | |
| Nahiyan et al. | Contactless monitoring for healthcare applications | |
| CN111493832A (en) | Enface-OCT-based endoscopic imaging method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20211201 Address after: East area, 3rd floor, No.9 plant, xibadao, Tianjin Binhai New Area pilot free trade zone (Airport Economic Zone) Patentee after: TIANJIN HENGYU MEDICAL TECHNOLOGY Co.,Ltd. Address before: 300457 West Area C, 2nd floor, No.1 Workshop, No.9 West 8th Road, Tianjin pilot free trade zone (Airport Economic Zone), Binhai New Area, Tianjin Patentee before: TIANJIN HAIREN MEDICAL TECHNOLOGY Co.,Ltd. Patentee before: Tianjin Hengyu Medical Technology Co., Ltd |