Summary of the invention
Measurement method and device and image processing method this application provides a kind of eyeground without perfusion area, can be in scanning amount
In lesser situation, the efficiency or raising for improving eyeground nonperfusion region areal calculation obtain the effect of eyeground nonperfusion region
Rate.
In a first aspect, the embodiment of the present application provides a kind of OCTA rete vasculosum image processing method, comprising:
The optical fundus blood vessel net image of the tested eye is obtained, the optical fundus blood vessel net image is by optical coherence tomography blood
Pipe radiography (optical coherence tomographyangiography, OCTA) generates;
Clustering processing is carried out to the optical fundus blood vessel net image, is formed on the optical fundus blood vessel net image after clustering processing
Angiosomes and avascular area domain;
Corrosion treatment is carried out to the optical fundus blood vessel net image after the clustering processing, the eyeground after corrosion treatment
A first area and at least a second area are formed on blood vessel net image, the first area includes the angiosomes, institute
It states an at least second area to be included in the avascular area domain, the first area is the overall region of a connection, Mei Yisuo
State the overall region that second area is a connection;
The maximum second area of area in the optical fundus blood vessel net image after searching the corrosion treatment;
Expansion process is carried out to the optical fundus blood vessel net image after the corrosion treatment, obtains the eyeground of the tested eye
Nonperfusion region, the eyeground nonperfusion region include the maximum second area of the area, the expansion process and the corruption
Erosion processing uses the core of same size.
In one possible implementation, face in the optical fundus blood vessel net image after the lookup corrosion treatment
After the maximum second area of product, and to the optical fundus blood vessel net image after the corrosion treatment carry out expansion process it
Before, the method also includes:
To the area except the maximum second area of area described in the optical fundus blood vessel net image after the corrosion treatment
Domain is filled processing, to distinguish the maximum second area of the area.
In one possible implementation, the optical fundus blood vessel net image for obtaining the tested eye, comprising:
The optical fundus blood vessel net of the tested eye is obtained by the optical coherence tomography device including light source and optic probe
Image;
The optical coherence tomography device by including light source and optic probe obtains the eyeground blood of the tested eye
Pipe network image, comprising:
Multiple scanning is carried out using retina of the measurement light to the tested eye by the optic probe, wherein the survey
Amount light is provided by the light source, and the multiple scanning includes the scanning at least twice to the same position of the retina;
During scanning, an at least spectral domain interference signal relevant to the retina is obtained;
Data related with cell in the retina, tissue or Particles Moving are extracted by the spectral domain interference signal;
The optical fundus blood vessel net is calculated according to data related with cell in the retina, tissue or Particles Moving
Image.
It is in one possible implementation, described that clustering processing is carried out to the optical fundus blood vessel net image, comprising:
The optical fundus blood vessel net image is filtered, and using K Mean Method to the eye after filtering processing
Bottom blood vessel net image is classified.
It is in one possible implementation, described that the optical fundus blood vessel net image is filtered, comprising:
To the optical fundus blood vessel net image carry out interpolation processing, and to the optical fundus blood vessel net image after interpolation processing into
Row filtering processing.
Second aspect, the embodiment of the present application also provides a kind of OCTA rete vasculosum image processing methods, comprising:
The optical fundus blood vessel net image of the tested eye is obtained, the optical fundus blood vessel net image is by optical coherence tomography blood
Pipe radiography generates;
Clustering processing is carried out to the optical fundus blood vessel net image, is formed on the optical fundus blood vessel net image after clustering processing
Angiosomes and avascular area domain;
Corrosion treatment is carried out to the optical fundus blood vessel net image after the clustering processing, the eyeground after corrosion treatment
A first area and at least a second area are formed on blood vessel net image, the first area includes the angiosomes, institute
It states an at least second area to be included in the avascular area domain, the first area is the overall region of a connection, Mei Yisuo
State the overall region that second area is a connection;
Expansion process is carried out to the optical fundus blood vessel net image after the corrosion treatment, obtains an at least third region,
Each third region is the overall region of a connection, and each third region includes a second area, described swollen
Swollen processing uses the core of same size with the corrosion treatment;
The maximum third region of area in the optical fundus blood vessel net image after searching expansion process, the area are maximum
Third region is the eyeground of the tested eye without perfusion area;
The pixel number of the eyeground nonperfusion region is calculated, to calculate the area of the eyeground nonperfusion region.
In one possible implementation, it is described search expansion process after the optical fundus blood vessel net image in area most
After big third region, the method also includes:
To the area except the maximum third region of area described in the optical fundus blood vessel net image after the expansion process
Domain is filled processing, to distinguish the maximum third region of the area.
In one possible implementation, the optical fundus blood vessel net image for obtaining the tested eye, comprising:
The optical fundus blood vessel net of the tested eye is obtained by the optical coherence tomography device including light source and optic probe
Image;
The optical coherence tomography device by including light source and optic probe obtains the eyeground blood of the tested eye
Pipe network image, comprising:
Multiple scanning is carried out using retina of the measurement light to the tested eye by the optic probe, wherein the survey
Amount light is provided by the light source, and the multiple scanning includes the scanning at least twice to the same position of the retina;
During scanning, an at least spectral domain interference signal relevant to the retina is obtained;
Data related with cell in the retina, tissue or Particles Moving are extracted by the spectral domain interference signal;
The optical fundus blood vessel net is calculated according to data related with cell in the retina, tissue or Particles Moving
Image.
It is in one possible implementation, described that clustering processing is carried out to the optical fundus blood vessel net image, comprising:
The optical fundus blood vessel net image is filtered, and using K Mean Method to the eye after filtering processing
Bottom blood vessel net image is classified.
It is in one possible implementation, described that the optical fundus blood vessel net image is filtered, comprising:
To the optical fundus blood vessel net image carry out interpolation processing, and to the optical fundus blood vessel net image after interpolation processing into
Row filtering processing.
Fourth aspect, the embodiment of the present application provides a kind of measurement method of the eyeground without perfusion area, for measuring tested eye
Area of the eyeground without perfusion area including above-mentioned OCTA rete vasculosum image processing method further include calculating the eyeground without filling
The pixel number in region is infused, to calculate the area of the eyeground nonperfusion region.
Fourth aspect, the embodiment of the present application provide a kind of eyeground without perfusion area measuring device, and described device is for measuring
Without the area of perfusion area, described device includes: on the eyeground of tested eye
Acquiring unit, for obtaining the optical fundus blood vessel net image of the tested eye, the optical fundus blood vessel net image is by optics
Coherence tomography angiography generates;
Clustering processing unit, for carrying out clustering processing to the optical fundus blood vessel net image, the eye after clustering processing
Angiosomes and avascular area domain are formed on the blood vessel net image of bottom;
Corrosion treatment unit, it is rotten for carrying out corrosion treatment to the optical fundus blood vessel net image after the clustering processing
A first area and at least a second area are formed on erosion treated the optical fundus blood vessel net image, the first area is wrapped
The angiosomes are included, an at least second area is included in the avascular area domain, and the first area is a connection
Overall region, each second area be one connection overall region;
Searching unit, for searching maximum secondth area of area in the optical fundus blood vessel net image after the corrosion treatment
Domain;
Expansion process unit is obtained for carrying out expansion process to the optical fundus blood vessel net image after the corrosion treatment
To the eyeground nonperfusion region of the tested eye, the eyeground nonperfusion region includes the maximum second area of the area, institute
It states expansion process and the corrosion treatment uses the core of same size;
Computing unit, for calculating the pixel number of the eyeground nonperfusion region, to calculate the eyeground nonperfusion region
Area.
In one possible implementation, described device further include:
Processing unit is filled, for maximum to area described in the optical fundus blood vessel net image after the corrosion treatment
Region except second area is filled processing, to distinguish the maximum second area of the area.
In one possible implementation, the acquiring unit, specifically for by the light including light source and optic probe
Learn the optical fundus blood vessel net image that coherence tomography device obtains the tested eye.
In one possible implementation, the acquiring unit is specifically used for utilizing measurement light by the optic probe
Multiple scanning is carried out to the retina of the tested eye, wherein the measurement light is provided by the light source, the multiple scanning packet
Include the scanning at least twice to the same position of the retina;During scanning, acquisition is relevant to the retina at least
One spectral domain interference signal;And had by spectral domain interference signal extraction with cell, tissue or Particles Moving in the retina
The data of pass;And the eyeground blood is calculated according to data related with cell in the retina, tissue or Particles Moving
Pipe network image.
In one possible implementation, the clustering processing unit, comprising:
Subelement is filtered, for being filtered in the optical fundus blood vessel net image;
Clustering processing subelement, for being carried out using K Mean Method to the optical fundus blood vessel net image after filtering processing
Classification.
In one possible implementation, the filtering processing subelement is specifically used for the optical fundus blood vessel net figure
As carrying out interpolation processing, and the optical fundus blood vessel net image after interpolation processing is filtered.
5th aspect, the embodiment of the present application also provides a kind of eyeground without perfusion area measuring device, for measuring tested eye
Area of the eyeground without perfusion area, described device includes:
Acquiring unit, for obtaining the optical fundus blood vessel net image of the tested eye, the optical fundus blood vessel net image is by optics
Coherence tomography angiography generates;
Clustering processing unit, for carrying out clustering processing to the optical fundus blood vessel net image, the eye after clustering processing
Angiosomes and avascular area domain are formed on the blood vessel net image of bottom;
Corrosion treatment unit, it is rotten for carrying out corrosion treatment to the optical fundus blood vessel net image after the clustering processing
A first area and at least a second area are formed on erosion treated the optical fundus blood vessel net image, the first area is wrapped
The angiosomes are included, an at least second area is included in the avascular area domain, and the first area is a connection
Overall region, each second area be one connection overall region;
Expansion process unit is obtained for carrying out expansion process to the optical fundus blood vessel net image after the corrosion treatment
To an at least third region, each third region is the overall region of a connection, and each third region includes an institute
Second area is stated, the expansion process and the corrosion treatment use the core of same size;
Searching unit, for searching the maximum third region of area in the optical fundus blood vessel net image after expansion process,
The maximum third region of area is the eyeground of the tested eye without perfusion area;
Computing unit, for calculating the pixel number of the eyeground nonperfusion region, to calculate the eyeground nonperfusion region
Area.
In one possible implementation, described device further include:
Processing unit is filled, for maximum to area described in the optical fundus blood vessel net image after the expansion process
Region except second area is filled processing, to distinguish the maximum second area of the area.
In one possible implementation, the acquiring unit, specifically for by the light including light source and optic probe
Learn the optical fundus blood vessel net image that coherence tomography device obtains the tested eye.
In one possible implementation, the acquiring unit is specifically used for utilizing measurement light by the optic probe
Multiple scanning is carried out to the retina of the tested eye, wherein the measurement light is provided by the light source, the multiple scanning packet
Include the scanning at least twice to the same position of the retina;During scanning, acquisition is relevant to the retina at least
One spectral domain interference signal;And had by spectral domain interference signal extraction with cell, tissue or Particles Moving in the retina
The data of pass;And the eyeground blood is calculated according to data related with cell in the retina, tissue or Particles Moving
Pipe network image.
In one possible implementation, the clustering processing unit, comprising:
Subelement is filtered, for being filtered to the optical fundus blood vessel net image;
Clustering processing subelement, for being carried out using K Mean Method to the optical fundus blood vessel net image after filtering processing
Classification.
In one possible implementation, the filtering processing subelement is specifically used for the optical fundus blood vessel net figure
As carrying out interpolation processing, and the optical fundus blood vessel net image after interpolation processing is filtered.
6th aspect, the embodiment of the present application also provides a kind of eyeground without perfusion area measuring device, and described device includes place
Device and memory, the processor and memory coupling are managed, is stored with program instruction in the memory, described program refers to
When enabling by processor execution, described device is made to execute the method as described in first aspect or second aspect.
7th aspect, the embodiment of the present application provide a kind of computer readable storage medium, the computer-readable storage
It is stored with program instruction in medium, described program instructs when being executed by the processor of computer, and the processor is made to execute such as the
Method described in one side or second aspect.
Eighth aspect, the embodiment of the present application also provides a kind of computer program product, in the computer program product
Be stored with program instruction makes the processor execute such as first aspect when described program instructs the processor execution by computer
Or method described in second aspect.
Implement the embodiment of the present application, is scanned by retina of the OCTA to tested eye, to obtain optical fundus blood vessel net
Then image carries out clustering processing and corrosion treatment to optical fundus blood vessel net image, include a first area and at least to obtain
The optical fundus blood vessel net image of one second area finds the maximum second area of area and then by expansion process, Ke Yiyou
Effect makes up due to corrosion treatment and the region of reduction, and then can obtain eyeground nonperfusion region, can reduce through OCTA to quilt
The number of scan points that the retina of inspection eye is scanned, avoids the bigger scanning of density, can be improved to obtain eyeground nonperfusion region
Efficiency, guarantee it is available arrive complete FAZ area.
Specific embodiment
In order to keep the purposes, technical schemes and advantages of the application clearer, below in conjunction with attached drawing to the application make into
One step it is described in detail.
Term " first " and " second " in the description of the present application, claims and attached drawing etc. are for distinguishing difference
Object is not use to describe a particular order.In addition, term " includes " and " having " and their any deformations, it is intended that
It covers and non-exclusive includes.Such as it contains the process, method, system, product or equipment of a series of steps or units and does not limit
Due to listed step or unit, but optionally further comprising the step of not listing or unit, or optionally further comprising it is right
In the intrinsic other step or units of these process, methods, product or equipment.
It is a kind of process signal of measurement method of eyeground without perfusion area provided by the embodiments of the present application referring to Fig. 1, Fig. 1
Figure, the measurement method can be applied to eyeground without perfusion area measuring device, and eyeground can be arbitrarily can without perfusion area measuring device
Electronic equipment of the eyeground without perfusion region measurement is carried out, if the eyeground can also include optical coherence tomography without perfusion area measuring device
Scanning means, that is, OCT device, the OCT device may include light source and optic probe etc., the embodiment of the present application for eyeground without
The concrete form of perfusion area measuring device is not construed as limiting.As shown in Figure 1, measurement method of the eyeground without perfusion area includes:
S101, the optical fundus blood vessel net image for obtaining tested eye, optical fundus blood vessel net image is by optical coherence tomography blood vessel
Radiography generates.
In the embodiment of the present application, optical fundus blood vessel net image is image relevant to retina, i.e. optical fundus blood vessel net image
Image including needing to extract eyeground nonperfusion region.It is understood that optical fundus blood vessel net image is the figure generated by OCTA technology
Picture.Wherein, tested eye can be regarded as the eyes for being detected user.
Specifically, the scanning element using 256*256 is scanned tested eye, image is obtained in scanning, is being scanned
After the completion, optical fundus blood vessel net image can be obtained.Such as light is got on a point on the retina of tested eye, is beaten a branch of parallel
Light, the directional light converge on the retina of tested eye, form the light of convergence, to complete in the retina scanning to tested eye
Later, optical fundus blood vessel net image can be obtained.
It optionally, can be by including that the OCT device acquisition of light source and optic probe is detected the optical fundus blood vessel net image of eye.
Specifically, by including that the optical coherence tomography device acquisition of light source and optic probe is detected the optical fundus blood vessel of eye
Net image, comprising:
Multiple scanning is carried out using retina of the measurement light to tested eye by optic probe, wherein measurement light is mentioned by light source
For multiple scanning includes the scanning at least twice to the same position of retina;
During scanning, an at least spectral domain interference signal relevant to retina is obtained;
Data related with cell in retina, tissue or Particles Moving are extracted by spectral domain interference signal;
Optical fundus blood vessel net image is calculated according to data related with cell in retina, tissue or Particles Moving.
In the present embodiment, for how by including the optical coherence tomography device of light source and optic probe to obtain quilt
The optical fundus blood vessel net image for examining eye, can refer to Fig. 2, Fig. 2 is a kind of schematic diagram of Optical coherence tomography device.Its
In, the principle of OCT image and the principle of ultrasonic imaging are similar, are all the interference using wave.
As shown in Fig. 2, the light that light source 101 issues is measurement light and reference in the punishment of coupler 102 into coupler 102
Light.Measurement light eventually enters into tested eye 108 via optical fiber 105, scanning device 107, and to the retina 110 of tested eye 108 into
Row scanning, the retina 110 that measurement light is detected eye 108 scatter, and part measures light backtracking, is entered by optical fiber 105 and is coupled
Device 102.And reference light is beaten on reflecting mirror 104 via optical fiber 103, returns to coupler 102 by optical fiber 103 later.Measure light by
Coupler 102 returns again to the distance of coupler 102 to retina 110, is returned again by coupler 102 to reflecting mirror 104 with reference light
The distance for returning coupler 102, is not much different.The measurement light and reference light of return interfere at coupler 102, by detector
120 acquisition interference lights.
Further, OCT detection is structure, and OCTA detection is function.Vivider says, OCT detection scanning point
With the presence or absence of thing, and the thing near OCTA detection scanning point can or can not move.If the same scanning element, one has thing,
Do not have thing for a moment, then illustrates that the thing of the scanning element there can move.Therefore, OCTA at least needs to scan two width OCT figure.Tool
Body, OCTA can obtain 4 width figures by scanning.It is understood that should not be construed as the above is only the example to OCTA to this implementation
The restriction of example.
That is, the embodiment of the present application, it can be dynamically to retina by optical coherence tomography angiography
It is scanned, to obtain optical fundus blood vessel net image.
Optionally, the optical fundus blood vessel net image of tested eye is obtained, comprising:
Tested eye is scanned by OCTA, obtains the scanning signal of scanning element;
In the case where the scanning signal of scanning element is less than reference threshold, scanning element is pressed into noise processed, and by scanning element
Gray value return-to-zero;
In the case where retouching completion to tested eye, optical fundus blood vessel net image is obtained.
In the present embodiment, reference threshold is threshold value for measuring the scanning signal of scanning element, therefore, the present embodiment for
The specific value of reference threshold is not construed as limiting.If reference threshold can be a pre-set threshold value, if eyeground is without perfusion area
Reference threshold can be arranged by receiving the setting instruction of user's input in measuring device, alternatively, can also be by eyeground without perfusion
Region measurement device is arranged etc. automatically, and this embodiment is not limited.
In the present embodiment, when being such as scanned using retina of the OCTA method to tested eye, if the scanning of scanning element
Signal is less than reference threshold, then scanning element can be pressed noise treatment, such as by the gray value return-to-zero of scanning element;It then proceedes to
Next scanning element is scanned, if the scanning signal of next scanning element is greater than reference threshold, retains next scanning element, directly
It is completed to scanning, then available optical fundus blood vessel net image.
It is understood that the scanning signal of scanning element can be regarded as the luminous intensity of scanning element, at the gray value zero of scanning element
Reason, then can be regarded as the brightness return-to-zero of scanning element.As the gray value of scanning element can be between 0-155, and if scanning element
Scanning signal be less than reference threshold, then the gray value of scanning element then can use 0.
S102, clustering processing is carried out to optical fundus blood vessel net image, form blood on the optical fundus blood vessel net image after clustering processing
Area under control domain and avascular area domain.
Eyeground in the embodiment of the present application, after carrying out clustering processing to optical fundus blood vessel net image, after clustering processing
Angiosomes and avascular area domain can be formed on blood vessel net image.That is, in the eye for getting tested eye by OCTA
When the blood vessel net image of bottom, the differentiation in optical fundus blood vessel net image medium vessels region and avascular area domain may be unobvious, or can not
It distinguishes, but passes through the optical fundus blood vessel net image medium vessels area to optical fundus blood vessel net image clustering processing, after may make clustering processing
Domain and avascular area domain are effectively distinguished.It is understood that when carrying out clustering processing, using the characteristic of angiosomes and without blood
The characteristic in area under control domain carries out clustering.
Optionally, the embodiment of the present application provides a kind of clustering processing method, as follows:
Clustering processing is carried out to optical fundus blood vessel net image, comprising:
Optical fundus blood vessel net image is filtered, and using K Mean Method to the optical fundus blood vessel net after filtering processing
Image is classified.
In the present embodiment, by being filtered to optical fundus blood vessel net image, optical fundus blood vessel net figure can be effectively filtered out
Noise as in can also prepare to implement K Mean Method.Wherein, K Mean Method is i.e. for the eyeground blood after filtration treatment
Region in pipe network image is clustered, and angiosomes and avascular area domain are thus obtained.
It is understood that in the present embodiment, it, can also be to optical fundus blood vessel before being filtered to optical fundus blood vessel net image
Net image carries out interpolation processing, becomes apparent from blood vessel, is then filtered again to the optical fundus blood vessel net image after interpolation processing
Processing.
Wherein, the method for interpolation processing can by Fourier transformation method carry out interpolation processing, so as to effectively make up by
Scanning element is very few in the optical fundus blood vessel net image that noise reduction process obtains, so as to cause can not effectively distinguish angiosomes and without blood
The case where area under control domain.
S103, corrosion treatment is carried out to the optical fundus blood vessel net image after clustering processing, due to reducing before this to tested eye
The number of scan points that is scanned of retina, b and Fig. 4 c referring to fig. 4 is formed on the optical fundus blood vessel net image after corrosion treatment
One first area and at least a second area.C referring to fig. 4, first area are the region of black, and second area is the area of white
Domain.First area includes angiosomes, and at least a second area is included in avascular area domain, and first area is the whole of a connection
Body region, each second area are the overall region of a connection.
In the embodiment of the present application, by the optical fundus blood vessel net after the clustering processing for forming angiosomes and avascular area domain
Image carries out corrosion treatment, the area of angiosomes and avascular area domain can be effectively reduced or reduce, to form angiosomes
The optical fundus blood vessel net image separated with avascular area domain.Specifically, carrying out corrosion treatment to optical fundus blood vessel image, first can be formed
Region and at least one second area, which may include angiosomes, and second area may include avascular area domain.Its
In, first area can be regarded as the overall region of a connection, and each second area can be regarded as the overall region of a connection.
The maximum second area of area in optical fundus blood vessel net image after S104, lookup corrosion treatment.
It is understood that after carrying out corrosion treatment to optical fundus blood vessel net image, it may in obtained optical fundus blood vessel net image
There can be lesser white area in many noise points such as Fig. 4 c, therefore can be by searching for area in optical fundus blood vessel net image
Maximum second area, to obtain the eyeground nonperfusion region of tested eye.
S105, expansion process is carried out to the optical fundus blood vessel net image after corrosion treatment, obtains the eyeground of tested eye without perfusion
Region, eyeground nonperfusion region include the maximum second area of area, and expansion process and corrosion treatment use the core of same size.
In the embodiment of the present application, expansion process is carried out to the optical fundus blood vessel net image after corrosion treatment, it will be appreciated that for shape
At first area and an at least second area optical fundus blood vessel net image carry out expansion process.Specifically it can be regarded as to eyeground
The maximum second area of area and first area in blood vessel net image carry out expansion process.
In the embodiment of the present application, the i.e. eyeground nonperfusion region of the maximum second area of area is reduced by corrosion treatment
Area, to carry out expansion process again, can effectively make up corrosion reduction area so that obtaining true eyeground without filling
Infuse region area.It is understood that the extent of corrosion of corrosion treatment and the degrees of expansion of expansion process are identical in the embodiment of the present application,
That is, the nuclear phase of core and expansion process when corrosion treatment is same.
Optionally, after the maximum second area of area and right in the optical fundus blood vessel net image after searching corrosion treatment
Before optical fundus blood vessel net image after corrosion treatment carries out expansion process, method further include:
C referring to fig. 4, to the region except the maximum second area of area in the optical fundus blood vessel net image after corrosion treatment into
Row filling processing, to distinguish the maximum second area of area.
C referring to fig. 4 has lesser white area in noise point such as Fig. 4 c in the optical fundus blood vessel net image after corrosion treatment
Domain, this will affect subsequent identification or calculating.In the present embodiment, by optical fundus blood vessel net image except area maximum the
Region except two regions is filled processing, can effectively reduce the influence of noise in optical fundus blood vessel net image, consequently facilitating knowing
The maximum second area of other area calculates it.
S106, the pixel number for calculating eyeground nonperfusion region, to calculate the area of eyeground nonperfusion region.
In the embodiment of the present application, by calculating the number of pixels of eyeground nonperfusion region, the eyeground can be calculated without perfusion
The area in region.
Implement the embodiment of the present application, is scanned by retina of the OCTA to tested eye, to obtain optical fundus blood vessel net
Then image carries out clustering processing and corrosion treatment to optical fundus blood vessel net image, include a first area and at least to obtain
The optical fundus blood vessel net image of one second area finds the maximum second area of area and then by expansion process, Ke Yiyou
Effect makes up due to corrosion treatment and the region of reduction, and then can obtain eyeground nonperfusion region, can reduce through OCTA to quilt
The number of scan points that the retina of inspection eye is scanned, avoids the bigger scanning of density, can be improved to obtain eyeground nonperfusion region
Efficiency, guarantee it is available arrive complete FAZ (foveal avascular zone, central fovea avascular area domain) area.Its
In, it reduces scanning times and namely reduces number of scan points in fact.
It is at a kind of image provided by the embodiments of the present application referring to Fig. 3, Fig. 3 for vivider understanding method shown in FIG. 1
The flow diagram of reason method.The image processing method of the embodiment of the present application is specifically OCTA rete vasculosum image processing method.Such as
Shown in Fig. 3, image processing method includes:
S301, it is scanned, is swept with retina of the Optical coherence tomography angiography to tested eye
The scanning signal of described point, in the case where the scanning signal of scanning element is lower than a preset threshold value, directly according to noise
Processing, the gray value return-to-zero of scanning element.In the case where retina is completed in scanning, optical fundus blood vessel net image is obtained.
As shown in fig. 4 a, Fig. 4 a is a kind of schematic diagram of optical fundus blood vessel net image provided by the embodiments of the present application, eyeground blood
The eyeground nonperfusion region that pipe network image includes is the middle section in figure.
S302, interpolation processing is carried out to optical fundus blood vessel net image, is filtered and uses K mean value (k-means) method
Clustering processing.
It is understood that after processing by step S302 angiosomes can be formed on optical fundus blood vessel net image and without blood vessel
Region.
As shown in Figure 4 b, Fig. 4 b is showing for the optical fundus blood vessel net image after a kind of clustering processing provided by the embodiments of the present application
It is intended to.It include angiosomes and avascular area domain in optical fundus blood vessel net image after clustering processing, wherein black is blood vessel
Region, it is white for avascular area domain.Since number of scan points is very little, the blood vessel of black has very big missing in the figure, causes
Avascular area domain and angiosomes can be connected to together.It, can be as shown in subsequent step in order to mark off avascular area domain.
S303, corrosion treatment is carried out to the optical fundus blood vessel net image after clustering processing.
After corrosion treatment, a first area and an at least second area can be formed on the optical fundus blood vessel net image.
As illustrated in fig. 4 c, Fig. 4 c is showing for the optical fundus blood vessel net image after a kind of corrosion treatment provided by the embodiments of the present application
It is intended to.Due to having carried out corrosion treatment to optical fundus blood vessel net image, while corrosion can reduce the area in the region of white, thus
The region being connected is separated.Since there are the noise point of very little (such as white areas of figure small areas) for centre, so corrosion
The small cube of a black is had afterwards, but this will not influence final result.Then it looks in optical fundus blood vessel net image most
Big white area, and maximum white area is filled, as shown in S304.
(i.e. eyeground is without perfusion area for maximum white area in optical fundus blood vessel net image after S304, lookup clustering processing
Domain), and processing is filled to the region except maximum white area.
As shown in figure 4d, Fig. 4 d is a kind of showing for filling treated optical fundus blood vessel net image provided by the embodiments of the present application
It is intended to.
S305, expansion process is carried out to optical fundus blood vessel net image, obtains the eyeground nonperfusion region of tested eye.
As shown in fig 4e, Fig. 4 e is a kind of schematic diagram of eyeground nonperfusion region provided by the embodiments of the present application.From Fig. 4 e
In can be seen that through expansion process, eyeground nonperfusion region (white area) finally can be obtained.Wherein, at expansion process and corrosion
Reason selects an equal amount of core, thus can make up and corrode reduced region so that the area of obtained eyeground nonperfusion region with
The area in former region is consistent or almost the same.
S306, the pixel number for calculating eyeground nonperfusion region, obtain the area of eyeground nonperfusion region.
Implement the embodiment of the present application, can effectively avoid the bigger scanning of density and cause sweep time too long, can effectively mention
High eyeground nonperfusion region measuring device gets the efficiency of central no blood irrigating region, improves and gets the complete face FAZ
Long-pending efficiency.
It is the process signal of measurement method of another eyeground without perfusion area provided by the embodiments of the present application referring to Fig. 5, Fig. 5
Figure, the measurement method can also be applied to eyeground without perfusion area measuring device.As shown in figure 5, measurement method of the eyeground without perfusion area
Include:
S501, the optical fundus blood vessel net image for obtaining tested eye, optical fundus blood vessel net image is by optical coherence tomography blood vessel
Radiography generates.
S502, clustering processing is carried out to optical fundus blood vessel net image, form blood on the optical fundus blood vessel net image after clustering processing
Area under control domain and avascular area domain.
S503, corrosion treatment is carried out to the optical fundus blood vessel net image after clustering processing, the optical fundus blood vessel net after corrosion treatment
A first area and at least a second area are formed on image, first area includes angiosomes, at least a second area packet
It includes in avascular area domain, first area is the overall region of a connection, and each second area is the overall region of a connection.
S504, expansion process is carried out to the optical fundus blood vessel net image after corrosion treatment, obtains an at least third region, it is each
Third region is the overall region of a connection, and each third region includes a second area, and expansion process and corrosion treatment use
The core of same size.
S505, the maximum third region of area in the optical fundus blood vessel net image after expansion process, area maximum the are searched
Three regions are the eyeground of tested eye without perfusion area.
Optionally, in the optical fundus blood vessel net image after searching expansion process after the maximum third region of area, method is also
Include:
Place is filled to the region except the maximum third region of area in the optical fundus blood vessel net image after expansion process
Reason, to distinguish the maximum third region of area.
S506, the pixel number for calculating eyeground nonperfusion region, to calculate the area of eyeground nonperfusion region.
It is understood that the implementation with reference to described in Fig. 2 or Fig. 3 can be corresponded to for specific implementation shown in fig. 5, this
In be no longer described in detail one by one.
It is above-mentioned to illustrate the method for the embodiment of the present application, the device of the embodiment of the present application is provided below.
Fig. 6 is referred to, Fig. 6 is structural representation of a kind of eyeground without perfusion area measuring device provided by the embodiments of the present application
Figure, image processing apparatus can be used for executing method shown in Fig. 2 to Fig. 5.As shown in fig. 6, eyeground is without perfusion area measuring device packet
It includes:
Acquiring unit 601, for obtaining the optical fundus blood vessel net image of tested eye, optical fundus blood vessel net image is broken by optical coherence
Layer vessel scanning radiography generates;
Clustering processing unit 602, for carrying out clustering processing to optical fundus blood vessel net image, the optical fundus blood vessel after clustering processing
Angiosomes and avascular area domain are formed on net image;
Corrosion treatment unit 603, for carrying out corrosion treatment, corrosion treatment to the optical fundus blood vessel net image after clustering processing
A first area and an at least second area are formed on optical fundus blood vessel net image afterwards, first area includes angiosomes, until
A few second area is included in avascular area domain, and first area is the overall region of a connection, and each second area is a company
Logical overall region;
Searching unit 604, for searching the maximum second area of area in the optical fundus blood vessel net image after corrosion treatment;
Expansion process unit 605 is detected for carrying out expansion process to the optical fundus blood vessel net image after corrosion treatment
The eyeground nonperfusion region of eye, eyeground nonperfusion region includes the maximum second area of area, and expansion process is adopted with corrosion treatment
With the core of same size;
Computing unit 606, for calculating the pixel number of eyeground nonperfusion region, to calculate the face of eyeground nonperfusion region
Product.
Implement the embodiment of the present application, is scanned by retina of the OCTA to tested eye, to obtain optical fundus blood vessel net
Then image carries out clustering processing and corrosion treatment to optical fundus blood vessel net image, include a first area and at least to obtain
The optical fundus blood vessel net image of one second area finds the maximum second area of area and then by expansion process, Ke Yiyou
Effect makes up due to corrosion treatment and the region of reduction, and then can obtain eyeground nonperfusion region, avoids that density is bigger to sweep
It retouches, and also improves to obtain the efficiency of eyeground nonperfusion region, guarantee available to complete FAZ area.
Optionally, as shown in fig. 7, eyeground is without perfusion area measuring device further include:
Processing unit 607 is filled, for the maximum second area of area in the optical fundus blood vessel net image after corrosion treatment
Except region be filled processing, to distinguish the maximum second area of area.
Optionally, acquiring unit 601, specifically for by the optical coherence tomography device including light source and optic probe
Obtain the optical fundus blood vessel net image of tested eye.
Optionally, acquiring unit 601, specifically for carrying out weight to the retina of tested eye using measurement light by optic probe
Multiple scanning, wherein measurement light is provided by light source, and multiple scanning includes the scanning at least twice to the same position of retina;?
During scanning, an at least spectral domain interference signal relevant to retina is obtained;And it is extracted by spectral domain interference signal and retina
Middle cell, tissue or the related data of Particles Moving;And according to data related with cell in retina, tissue or Particles Moving
Optical fundus blood vessel net image is calculated.
Optionally, as shown in figure 8, clustering processing unit 602, comprising:
Subelement 6021 is filtered, for being filtered to optical fundus blood vessel net image;
Clustering processing subelement 6022, for being carried out using K Mean Method to the optical fundus blood vessel net image after filtering processing
Classification.
Optionally, subelement 6021 is filtered, is specifically used for carrying out optical fundus blood vessel net image interpolation processing, and to slotting
Treated that optical fundus blood vessel net image is filtered for value.
It should be noted that the realization of each unit can also correspond to referring to Fig.1-embodiment of the method shown in fig. 5 phase
It should describe, no longer be described in detail one by one here.
Fig. 9 is referred to, Fig. 9 is structural representation of a kind of eyeground without perfusion area measuring device provided by the embodiments of the present application
Figure, image processing apparatus can be used for executing method shown in Fig. 2 to Fig. 5.As shown in figure 9, eyeground is without perfusion area measuring device packet
It includes:
Acquiring unit 901, for obtaining the optical fundus blood vessel net image of tested eye, optical fundus blood vessel net image is broken by optical coherence
Layer vessel scanning radiography generates;
Clustering processing unit 902, for carrying out clustering processing to optical fundus blood vessel net image, the optical fundus blood vessel after clustering processing
Angiosomes and avascular area domain are formed on net image;
Corrosion treatment unit 903, for carrying out corrosion treatment, corrosion treatment to the optical fundus blood vessel net image after clustering processing
A first area and an at least second area are formed on optical fundus blood vessel net image afterwards, first area includes angiosomes, until
A few second area is included in avascular area domain, and first area is the overall region of a connection, and each second area is a company
Logical overall region;
Expansion process unit 904 obtains at least for carrying out expansion process to the optical fundus blood vessel net image after corrosion treatment
One third region, each third region are the overall region of a connection, and each third region includes a second area, expansion process
The core of same size is used with corrosion treatment;
Searching unit 905, for searching the maximum third region of area in the optical fundus blood vessel net image after expansion process, face
The maximum third region of product is the eyeground of tested eye without perfusion area;
Computing unit 906, for calculating the pixel number of eyeground nonperfusion region, to calculate the face of eyeground nonperfusion region
Product.
Optionally, as shown in Figure 10, eyeground is without perfusion area measuring device further include:
Processing unit 907 is filled, for the maximum second area of area in the optical fundus blood vessel net image after expansion process
Except region be filled processing, to distinguish the maximum second area of area.
Optionally, acquiring unit 901, specifically for by the optical coherence tomography device including light source and optic probe
Obtain the optical fundus blood vessel net image of tested eye.
Optionally, acquiring unit 901, specifically for carrying out weight to the retina of tested eye using measurement light by optic probe
Multiple scanning, wherein measurement light is provided by light source, and multiple scanning includes the scanning at least twice to the same position of retina;?
During scanning, an at least spectral domain interference signal relevant to retina is obtained;And it is extracted by spectral domain interference signal and retina
Middle cell, tissue or the related data of Particles Moving;And according to data related with cell in retina, tissue or Particles Moving
Optical fundus blood vessel net image is calculated.
Optionally, as shown in figure 11, clustering processing unit 902, comprising:
Subelement 9021 is filtered, for being filtered to optical fundus blood vessel net image;
Clustering processing subelement 9022, for being carried out using K Mean Method to the optical fundus blood vessel net image after filtering processing
Classification.
Optionally, subelement 9021 is filtered, is specifically used for carrying out optical fundus blood vessel net image interpolation processing, and to slotting
Treated that optical fundus blood vessel net image is filtered for value.
It should be noted that the realization of each unit can also correspond to referring to Fig.1-embodiment of the method shown in fig. 5 phase
It should describe, no longer be described in detail one by one here.
As described above, existing OCTA rete vasculosum scan method is that Multiple-Scan obtains relatively good image, scan
Points be 304*304 it is even more, the blood vessel that surrounds of the obtained center without vascular perfusion area is very continuous, can directly obtain
Eyeground nonperfusion region.But there are many points that such mode scans, and it is very long to scan the required time, low efficiency.Reduce scanning
Points such as reduce single sweep operation points or scanning times can reduce the scanning required time, but obtained blood-vessel image
Be it is desultory, be unable to get eyeground nonperfusion region according to existing method.In the embodiment of the present application, optical coherence is used
Tomography angiography is scanned retina, for each scanning element, if the scanning signal of the point is low
In reference threshold, directly according to noise treatment and the gray value return-to-zero of the point;To treated, image is carried out at cluster
Reason, wherein clustering processing is included interpolation processing, filtering processing and is classified using K Mean Method, then carries out corrosion treatment;
It is formed with after corrosion treatment in the optical fundus blood vessel net image of a first area and an at least second area, it is maximum to search area
Then second area is filled processing to the region except the maximum second area of area, then using same with corrosion treatment
The core of sample ruler cun carries out expansion process, and complete eyeground nonperfusion region can be obtained;Alternatively, to being formed with one after corrosion treatment
The optical fundus blood vessel net image in one region and an at least second area carries out at expansion using with the core of corrosion treatment same size
Reason is formed with the optical fundus blood vessel net image in an at least third region, and each third region includes a second area, searches area most
Big third region is filled processing to the region except the maximum third region of area, complete eyeground can be obtained without filling
Infuse region;In this way, being to obtain eyeground without perfusion in desultory image from capillary in the case where number of scan points is reduced
Region, time needed for scanning can be reduced.In the embodiment of the present application, it is scanned by retina of the OCTA to tested eye
Number of scan points is lower than 304*304, still can obtain complete eyeground nonperfusion region when number of scan points is 256*256, therefore
Number of scan points can be set to 256*256 to 304*304.The eyeground nonperfusion region of acquisition can be used for image recognition or based on
Calculate FAZ area.
It is structural representation of a kind of eyeground without perfusion area measuring device provided by the embodiments of the present application referring to Figure 12, Figure 12
Figure, eyeground includes processor 1201, memory 1202 and input/output interface 1203 without perfusion area measuring device, wherein processing
Device 1201, memory 1202 and input/output interface 1203 are connected with each other by bus.
Memory 1202 include but is not limited to be random access memory (random access memory, referred to as:
RAM), read-only memory (read-only memory, ROM), Erasable Programmable Read Only Memory EPROM (erasable
Programmable read only memory, EPROM) or portable read-only memory (compact disc read-
Only memory, CD-ROM), memory 1202 is for storing dependent instruction and data.In the embodiment of the present application, memory can
For storing reference threshold, it may also be used for store the extent index of corrosion treatment and the extent index of expansion process etc., this Shen
Please embodiment be not construed as limiting.
Input/output interface 1203 is for inputting and/or outputting data.In the embodiment of the present application, which can
It is connect with display, which can be used for showing optical fundus blood vessel net image.
Processor 1201 can be one or more central processing units (central processing unit, CPU),
In the case that processor 1201 is a CPU, CPU can be monokaryon CPU, be also possible to multi-core CPU.Optionally, processor is also
It may be other kinds of processor (such as image processor).Optionally, processor, which can also be, is made of multiple processors
Processor group, pass through one or more buses connections between multiple processors.
In one embodiment, processor can be used for executing acquiring unit, clustering processing unit, corrosion treatment unit, look into
Look for method performed by unit, expansion process unit and computing unit.Here it is no longer described in detail one by one.
In one embodiment, which can also be optical coherence tomography device, the light without perfusion area measuring device
Learn coherence tomography device may include light source and optic probe and the optical coherence scanning means may also include processor,
Memory etc., the embodiment of the present application are not construed as limiting.
It should be noted that the realization of each operation can also correspond to referring to Fig.1-embodiment of the method shown in fig. 5 phase
It should describe.
It is understood that the above is only a kind of structural representation of the eyeground without perfusion area measuring device provided by the embodiments of the present application
Figure, in the concrete realization, eyeground can have components more more or fewer than the part shown without perfusion area measuring device, can group
Two or more components are closed, or different configuration realizations that there can be different components etc..
Those of ordinary skill in the art will appreciate that realizing all or part of the process in embodiment method, process can be by
Computer program instructs relevant hardware to complete, and program can be stored in computer-readable storage medium, program is executing
When, it may include such as the process of each method embodiment.And storage medium above-mentioned includes: ROM or random access memory RAM, magnetic
The medium of the various program storage codes such as dish or CD.