JP2019111396A - Image processing apparatus and image processing method - Google Patents

Abstract

To provide an image processing apparatus and image processing method allowing an operator to easily correct a result obtained by automatically classifying a depolarization area into a plurality of types.SOLUTION: The image processing apparatus includes: detection means for detecting a depolarization area from a polarized tomographic image of a subject's eye; classification means for classifying the detected depolarization area into a plurality of types; display control means for causing display means to display a plurality of display modes corresponding to the plurality of classified types, with the display modes being superimposed on the polarized tomographic image; and correction means for correcting a type corresponding to a display mode designated by the operator, into another type.SELECTED DRAWING: Figure 7

Description

  The present invention relates to an image processing apparatus and an image processing method for processing a tomographic image of an eye to be examined.

  Optical coherence tomography (hereinafter referred to as "OCT") using multi-wavelength light wave interference can obtain a tomographic image of a sample (in particular, a fundus) with high resolution. In recent years, in addition to a normal OCT image for imaging the shape of fundus tissue, acquisition of a functional OCT image for imaging optical characteristics, movement and the like of fundus tissue has been attempted in an ophthalmic OCT apparatus.

  Polarized OCT, which is one of functional OCT, performs imaging using a polarization parameter (retardation and orientation), which is one of the optical properties of the fundus tissue. Polarized OCT can use polarization parameters to construct a polarized OCT image to distinguish or segment fundus tissue. Polarized OCT uses light modulated into circularly polarized light as measurement light for observing a sample, divides and detects interference light as two orthogonal linear polarized lights, and generates a polarized OCT image (see Patent Document 1).

  In addition, in Non-Patent Document 1, an RPE region and a Choroid region are identified for the depolarization region extracted from the polarization OCT image. Here, the depolarization is an index indicating the degree of depolarization in the subject. Depolarization is considered to be caused, for example, by the change in the direction and phase of polarization randomly due to the reflection of measurement light at a minute structure (for example, melanin) in a tissue. First, the depolarization area is extracted from the polarization OCT image, and the curve of the RPE area is estimated in the extracted depolarization area. Then, a curve greatly deviated to the deep or shallow part from the curve estimated in the depth direction is identified (classified) in the Choroid region. At this time, Non-Patent Document 1 discloses that different colors are displayed in the RPE region and the Choroid region in the polarization OCT image.

WO 2010/122118 A1

Baumann et. al. "Segmentation and quantification of retinal lesions in age-related macular degeneration using polarization-sensitive optical coherence tomography" Journal of Biomedical optics Feb. 2010

  Here, in the case of an eye to be examined having a disease, there is a possibility that the discrimination (classification) of the RPE and the Choroid region may fail due to a decrease in signal strength due to a cataract or the like or the presence of a lesion in the fundus.

  An object of the present invention is to provide an image processing apparatus and an image processing method capable of easily correcting the result that the depolarization area is automatically classified into a plurality of types (a plurality of areas). It is.

One of the image processing apparatuses according to the present invention is
A tomographic luminance image and a polarization tomographic image of the subject's eye obtained by processing a common OCT signal by return light from the subject's eye irradiated with the measurement light and the reference light, wherein the tomographic image of the subject's eye on the image Tomographic image acquisition means for acquiring tomographic luminance images and polarization tomographic images in which the positions of the parts correspond to each other;
Detection means for detecting a depolarization area in the polarization tomographic image;
Classification means for automatically classifying a plurality of areas in the detected depolarization area into a plurality of types including a type of a lesion of an eye to be examined;
A display control unit that causes the display unit to display the tomographic luminance image in a state in which a plurality of display forms corresponding to the plurality of classified types are superimposed on a plurality of positions corresponding to the plurality of areas; Correction means for correcting the type corresponding to the display form of the position specified by the operator, which is any position of the plurality of positions in the image, to another type specified by the operator ,
The correction means specifies a type corresponding to a display form of the designated position by the operator specifying the other type after the operator designates any one of the plurality of positions. Modify to the other type specified above,
The display control means causes the display means to display the tomographic luminance image in a state where the display form of the designated position is changed to the display form corresponding to the designated other type.

Further, one of the image processing apparatuses according to the present invention is
A tomographic luminance image and a polarization tomographic image of the subject's eye obtained by processing a common OCT signal by return light from the subject's eye irradiated with the measurement light and the reference light, wherein the tomographic image of the subject's eye on the image Tomographic image acquisition means for acquiring tomographic luminance images and polarization tomographic images in which the positions of the parts correspond to each other;
Detection means for detecting a depolarization area in the polarization tomographic image;
Classification means for automatically classifying a plurality of areas in the detected depolarization area into a plurality of types including a type of a lesion of an eye to be examined;
A display control unit that causes the display unit to display the tomographic luminance image in a state in which a plurality of display states corresponding to the plurality of classified types are superimposed on a plurality of positions corresponding to the plurality of regions; Correction means for correcting the type corresponding to the display form of the position specified by the operator, which is any position of the plurality of positions in the image, to another type specified by the operator ,
The correction means specifies a type corresponding to the display form of the designated position by designating any one of the plurality of positions by the operator after the other type is designated by the operator. Modify to the other type specified above,
The display control means causes the display means to display the tomographic luminance image in a state where the display form of the designated position is changed to the display form corresponding to the designated other type.

Further, one of the image processing methods according to the present invention is
A tomographic luminance image and a polarization tomographic image of the subject's eye obtained by processing a common OCT signal by return light from the subject's eye irradiated with the measurement light and the reference light, wherein the tomographic image of the subject's eye on the image Acquiring a tomographic luminance image and a polarization tomographic image in which the positions of the parts correspond to each other; detecting a depolarization region in the polarization tomographic image;
Automatically classifying a plurality of areas in the detected depolarization area into a plurality of types including a type of a lesion of an eye to be examined;
Displaying the tomographic luminance image on a display unit in a state in which a plurality of display forms corresponding to the plurality of classified types are superimposed on a plurality of positions corresponding to the plurality of regions; And correcting the type corresponding to the display form of the position designated by the operator to any other position designated by the operator.
In the correcting step, after the operator designates any one of the plurality of positions, the operator designates the other type, thereby the type corresponding to the display form of the designated position. To the other types specified above,
In the displaying step, the tomographic luminance image is displayed on the display unit in a state where the display form of the designated position is changed to the display form corresponding to the other designated type.

Further, one of the image processing methods according to the present invention is
A tomographic luminance image and a polarization tomographic image of the subject's eye obtained by processing a common OCT signal by return light from the subject's eye irradiated with the measurement light and the reference light, wherein the tomographic image of the subject's eye on the image Acquiring a tomographic luminance image and a polarization tomographic image in which the positions of the parts correspond to each other; detecting a depolarization region in the polarization tomographic image;
Automatically classifying a plurality of areas in the detected depolarization area into a plurality of types including a type of a lesion of an eye to be examined;
Displaying the tomographic luminance image on a display unit in a state in which a plurality of display states corresponding to the plurality of classified types are superimposed on a plurality of positions corresponding to the plurality of regions; And correcting the type corresponding to the display form of the position designated by the operator to any other position designated by the operator.
In the correcting step, after the operator designates the other type, the operator designates any one of the plurality of positions, thereby the type corresponding to the display form of the designated position. To the other types specified above,
In the displaying step, the tomographic luminance image is displayed on the display unit in a state where the display form of the designated position is changed to the display form corresponding to the other designated type.

  According to one aspect of the present invention, there is provided an image processing apparatus and an image processing method that allow an operator to easily correct the result that the depolarization area is automatically classified into a plurality of types (a plurality of areas). Can.

BRIEF DESCRIPTION OF THE DRAWINGS Schematic of the whole structure of the image processing apparatus in this embodiment. The example of the image produced | generated by the signal processing part 190. FIG. Processing flow in the present embodiment. The figure for demonstrating the image analysis in this embodiment. The figure for demonstrating the image analysis in this embodiment. 6 is a display example on the display screen of the display unit of the image processing apparatus according to the present embodiment. The figure for demonstrating the manual correction | amendment in this embodiment. The figure for demonstrating the manual correction | amendment in this embodiment. 6 is a display example on the display screen of the display unit of the image processing apparatus according to the present embodiment.

  The image processing apparatus according to the present embodiment has detection means for detecting a depolarized area from a polarization tomographic image of an eye to be examined. In addition, the image processing apparatus according to the present embodiment includes a classification unit that classifies the detected depolarization region into a plurality of types. The classification means can also classify the detected depolarization area into a plurality of areas. For example, the classification unit classifies (identifies) the region corresponding to the RPE estimation curve into the RPE region. Also, the classification means classifies (identifies) an area separated from the RPE area (an area not connected to the RPE area) and located on the shallower side than the RPE estimation curve as a Particle area. Further, the classification means classifies (identifies) a region separated from the RPE region (a region not connected to the RPE region) and located on the deeper side than the RPE estimation curve as the Choroid region. The Particle area and the Choroid area will be described later.

  In addition, the image processing apparatus according to the present embodiment has a display control unit that causes the display unit to display a plurality of display forms corresponding to a plurality of classified types in a state of being superimposed on the polarization tomographic image. In addition, the display control means can also be displayed on the display means so as to be superimposed on the polarization tomographic image in a distinguishable display state of the plurality of classified areas. For example, in the polarization tomographic image, the display control means causes the display means to display the RPE area in red, the Particle area in blue, and the Choroid area in yellow-green.

  Then, the image processing apparatus according to the present embodiment has a correction unit that corrects the type corresponding to the display form designated by the operator to another type (a correct type designated by the operator). The correction means can also correct the display state corresponding to the area designated by the operator (the correct area designated by the operator) to the display state corresponding to the other area. This allows the operator to easily correct the result that the depolarization area is automatically classified into a plurality of types. For example, consider a case where an algorithm is used in which calculation of DOPU is not performed in a region where the pixel value of the tomographic luminance image is lower than the threshold value. At this time, in an area which is darkened by the shadow of the blood vessel in the RPE area (false image area) or the like, it is determined as an area having a luminance lower than the threshold. Therefore, the RPE regions are not connected (separated) in the region of luminance lower than the threshold in the RPE region. As described above, there are cases where the RPE region is not accurately detected even when using the DOPU image. At this time, an area separated from the RPE area may be automatically classified into a type different from the RPE area. Further, for example, in an eye to be examined in which the RPE is curved due to a lesion, a case is considered where there is a region where a part of the RPE region is separated from the RPE region and detected because the intensity of the polarization tomographic image is low. At this time, there is a possibility that the separately detected area is automatically classified into a type different from the RPE area. As described above, even if it is determined that the area is different from the RPE area despite the fact that it is the RPE area, the operator can change the display form corresponding to this area to the display form corresponding to the correct area, that is, the RPE area It can be easily corrected to the corresponding display form. Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

[Overall configuration of device]
FIG. 1 is a schematic view of the entire configuration of an image processing apparatus according to the present embodiment. This device includes polarized OCT (Polarization Sensitive OCT; hereinafter, PS-OCT) 100, Polarization Sensitive Scanning Laser Ophthalmoscope (hereinafter, PS-SLO) 140, anterior segment imaging unit 160, internal A fixation lamp 170 and a control unit 200 are provided. With the internal fixation lamp 170 turned on and focused on the subject's eye, alignment of the apparatus is performed using the image of the anterior segment of the subject observed by the anterior segment observation unit 160. After alignment is completed, imaging of the fundus by PS-OCT 100 and PS-SLO 140 is performed.

<Configuration of PS-OCT 100>
The configuration of the PS-OCT 100 will be described. The light source 101 is an SLD light source (Super Luminescent Diode) which is a low coherent light source, and emits, for example, light having a central wavelength of 850 nm and a bandwidth of 50 nm. Although the SLD is used as the light source 101, any light source that can emit low coherent light such as an ASE light source (Amplified Spontaneous Emission) may be used. Light emitted from the light source 101 is guided to a fiber coupler 104 having a polarization maintaining function through a PM (Polarization Maintaining) fiber 102 and a polarization controller 103, and measurement light (also referred to as OCT measurement light) and reference light (also referred to as OCT measurement light). Branch to the reference light corresponding to the OCT measurement light). The polarization controller 103 adjusts the polarization state of the light emitted from the light source 101, and is adjusted to linear polarization. The branching ratio of the fiber coupler 104 is 90 (reference light): 10 (measurement light).

  The branched measurement light is emitted from the collimator 106 as parallel light through the PM fiber 105. The emitted measurement light comprises an X scanner 107 consisting of a galvano mirror that scans the measurement light in the horizontal direction on the fundus Er, lenses 108 and 109, and a galvano mirror that scans the measurement light in the vertical direction on the fundus Er The light reaches the dichroic mirror 111 via the Y scanner 110. The X scanner 107 and the Y scanner 110 are controlled by the drive control unit 180 and scan the measurement light in a desired range of the fundus Er (also referred to as acquisition area of tomographic image, acquisition position of tomographic image, irradiation position of measurement light). be able to. The dichroic mirror 111 has a characteristic of reflecting light of 800 nm to 900 nm and transmitting light other than that.

  The measurement light reflected by the dichroic mirror 111 passes through the lens 112 and passes through the λ / 4 polarizing plate 113 (an example of the polarization adjusting member) installed at an inclination of 45 °, so that the phase is shifted by 90 °, and circularly polarized Polarized to light. The inclination of the λ / 4 polarizing plate 113 is preferably an angle (an example of arrangement) corresponding to the inclination from the optical axis of the polarization splitting surface of the fiber coupler 123 incorporating the polarization beam splitter, for example. Preferably, the λ / 4 polarizing plate 113 can be inserted into and removed from the optical path. For example, a mechanical configuration may be considered in which the λ / 4 polarizing plate 113 is rotated about an axis parallel to the optical axis. As a result, it is possible to realize a compact device that can easily switch between the SLO optical system and the PS-SLO optical system. In addition, a compact device capable of easily switching between the OCT optical system and the PS-OCT optical system can be realized.

  Here, light incident on the eye to be examined is polarization-controlled to circularly polarized light by setting the λ / 4 polarizing plate at an angle of 45 °, but it is not circularly polarized at the fundus Er due to the characteristics of the eye to be examined There is. Therefore, by the control of the drive control unit 180, the inclination of the λ / 4 polarizing plate can be finely adjusted.

  The measurement light whose polarization is controlled to be circularly polarized light is focused on the retinal layer of the fundus oculi Er by the focusing lens 114 mounted on the stage 116 via the anterior eye Ea of the eye as the object. The measurement light irradiated to the fundus Er is reflected and scattered by each retinal layer, and the above optical path is returned to the fiber coupler 104.

  On the other hand, the reference light branched by the fiber coupler 104 is emitted from the collimator 118 as parallel light through the PM fiber 117. The emitted reference light is polarization-controlled by a λ / 4 polarizing plate 119 installed by being inclined from P-polarization to 22.5 ° S-polarization in the same manner as the measurement light. The reference light is reflected by the mirror 122 on the coherence gate stage 121 through the dispersion compensation glass 120 and returns to the fiber coupler 104. The reference light passes through the λ / 4 polarizing plate 119 twice so that the linearly polarized light returns to the fiber coupler 104.

  The coherence gate stage 121 is controlled by the drive control unit 180 in order to cope with the difference in the axial length of the subject and the like. The measurement light and the reference light returned to the fiber coupler 104 are multiplexed to form interference light (multiplexed light), which is incident on the fiber coupler 123 incorporating a polarization beam splitter and is light of different polarization directions (in this embodiment, P polarization) And S-polarized light) at a branching ratio of 50:50.

  The P-polarized light is split by the grating 131 through the PM fiber 124 and the collimator 130, and is received by the lens 132 and the line camera 133. Similarly, S-polarized light is dispersed by the grating 127 through the PM fiber 125 and the collimator 126 and received by the lens 128 and the line camera 129. It goes without saying that the gratings 127 and 131 and the line cameras 129 and 133 are disposed in alignment with the direction of each polarization. The light received by each of the line cameras 129 and 133 is output as an electrical signal according to the intensity of the light, and is received by the signal processing unit 190 (an example of a tomographic image generation unit). The inclination of the λ / 4 polarizing plate 113 is adjusted based on the polarization beam splitter, but the inclination may be adjusted with respect to a straight line connecting the optic papilla center of the fundus and the macula lutea center. Also, the same effect can be obtained by adjusting the polarization beam splitter and the λ / 4 polarizing plates 113 and 119 based on the vertical direction as the polarization reference.

<Configuration of PS-SLO 140>
The configuration of the PS-SLO 140 will be described. The light source 141 is a semiconductor laser, and in the present embodiment, for example, emits light having a central wavelength of 780 nm. The measurement light (also referred to as SLO measurement light) emitted from the light source 141 is controlled to be linearly polarized by the polarization controller 145 via the PM fiber 142 and emitted from the collimator 143 as parallel light. The emitted measurement light passes through the perforated portion of the perforated mirror 144 and passes through the lens 155 to scan the measurement light in the horizontal direction on the fundus Er by the X scanner 146 composed of a galvano mirror, lenses 147 and 148, and the fundus The light beam reaches the dichroic mirror 154 via the Y scanner 149 configured of a galvano mirror that scans the measurement light in the vertical direction at Er. The X scanner 146 and the Y scanner 149 are controlled by the drive control unit 180, and can scan a desired range with the measurement light on the fundus. The dichroic mirror 154 has a characteristic of reflecting 760 nm to 800 nm and transmitting light other than that.

  The linearly polarized measurement light reflected by the dichroic mirror 154 passes through the same optical path as that of the PS-OCT 100 and reaches the fundus Er. The measurement light irradiated to the fundus Er is reflected and scattered by the fundus Er, follows the above optical path, and reaches the perforated mirror 144. The light reflected by the perforated mirror 144 is divided into light of different polarization directions (in the present embodiment, light of P polarization and light of S polarization) by the polarization beam splitter 151 via the lens 150, and an avalanche photodiode The light is received by an APD 152, 153, converted into an electric signal, and received by a signal processing unit 190 (an example of a fundus image generating unit). Here, the position of the perforated mirror 144 is conjugate to the pupil position of the eye to be examined, and of the light reflected and scattered from the measurement light emitted to the fundus Er, the light passing through the pupil peripheral portion is It is reflected by the perforated mirror 144. Although PM fiber is used for PS-OCT and PS-SLO in this embodiment, the same configuration and effect can be obtained by controlling polarization using a polarization controller even in single mode fiber (SMF).

Anterior Eye Imaging Unit 160
The anterior eye imaging unit 160 will be described. The anterior segment imaging unit 160 illuminates the anterior segment Ea with an illumination light source 115 including LEDs 115-a and 115-b that emit illumination light having a wavelength of 1000 nm. The light reflected by the anterior segment Ea reaches the dichroic mirror 161 via the lens 114, the polarizing plate 113, the lens 112, and the dichroic mirrors 111 and 154. The dichroic mirror 161 has a characteristic of reflecting light of 980 nm to 1100 nm and transmitting light other than that. The light reflected by the dichroic mirror 161 is received by the anterior eye camera 165 through the lenses 162, 163, and 164. The light received by the anterior eye camera 165 is converted into an electrical signal and received by the signal processing unit 190.

<Internal fixation lamp 170>
The internal fixation lamp 170 will be described. The internal fixation lamp 170 includes an internal fixation lamp display section 171 and a lens 172. A plurality of light emitting diodes (LDs) arranged in a matrix is used as the internal fixation lamp display section 171. The lighting position of the light emitting diode is changed in accordance with the part to be imaged by the control of the drive control unit 180. The light from the internal fixation lamp display section 171 is guided to the subject's eye through the lens 172. The light emitted from the internal fixation lamp display unit 171 is 520 nm, and the control unit 180 displays a desired pattern.

<Control unit 200>
The control unit 200 for controlling the entire apparatus will be described. The control unit 200 includes a drive control unit 180, a signal processing unit 190, a display control unit 191, and a display unit 192. The drive control unit 180 controls each unit as described above. The signal processing unit 190 includes an image generation unit 193, an area detection unit 194, and a classification unit 195. The signal processing unit 190 generates an image, analyzes the generated image, and generates visualization information of an analysis result based on the signals respectively output from the line cameras 129 and 133, the APDs 152 and 153, and the anterior segment camera 165. . The correction unit 196 corrects the type corresponding to the display form designated by the operator to another type (the correct type designated by the operator). The details of generation and analysis of the image will be described later.

  The display control unit 191 displays, for example, an image obtained by acquiring an image generated by the tomographic image generation unit and the fundus image generation unit by the fundus image acquisition unit (not illustrated) and the tomographic image acquisition unit (not illustrated). It is displayed on the display screen of (for example, a display such as liquid crystal). The image data generated by the signal processing unit 190 may be transmitted to the display control unit 191 by wire or may be transmitted wirelessly. Further, although the image processing apparatus is described in this embodiment, as an ophthalmologic apparatus or an ophthalmologic system according to another embodiment of the present invention, the fundus image acquisition unit includes an SLO optical system, and the tomographic image acquisition unit is OCT You may comprise so that an optical system may be included.

  Under the control of the display control unit 191, the display unit 192 displays a display mode indicating various information as described later. The image data from the display control unit 191 may be transmitted to the display unit 192 by wire or may be transmitted wirelessly. Although the display unit 192 and the like are included in the control unit 200, the present invention is not limited to this, and may be provided separately from the control unit 200. Moreover, the apparatus (tablet) which a user can carry which the display control part 191 and the display part 192 were comprised integrally may be sufficient. In this case, it is preferable that the display unit be equipped with a touch panel function so that operations such as movement of display position of an image, enlargement / reduction, change of an image to be displayed, etc. can be performed on the touch panel.

[Image processing]
Next, image generation in the image generation unit 193 will be described.

<Tomogram generation and fundus image generation>
The image generation unit 193 performs reconstruction processing used for general SD-OCT (Spectral Domain OCT) on each interference signal output from the line cameras 129 and 133, based on each polarization component. Two tomographic images (also referred to as a tomographic image corresponding to the first polarization, a tomographic image corresponding to the second polarization). First, the image generation unit 193 performs fixed pattern noise removal from the interference signal. Fixed pattern noise removal is performed by extracting fixed pattern noise by averaging a plurality of detected A scan signals and subtracting the same from an input interference signal. Next, the image generation unit 193 converts the interference signal from wavelength to wave number and performs Fourier transform to generate a tomographic signal (also referred to as a tomographic signal indicating a polarization state). Two tomographic images are generated by performing the above processing on the interference signal of two polarization components.

  In addition, the image generation unit 193 aligns the signals output from the APDs 152 and 153 in synchronization with the driving of the X scanner 146 and the Y scanner 149 to obtain two fundus images (first The fundus image corresponding to the polarized light (also referred to as the fundus image corresponding to the second polarized light) is generated.

<Luminance image generation>
The image generation unit 193 generates a luminance image from the two tomographic signals described above. Luminance image intended tomographic images basically the same in the conventional OCT, the pixel value r is calculated by the equation 1 from the tomographic signals A _H and A _V obtained from the line sensors 129 and 133.

  Similarly, a fundus luminance image is generated from two fundus images.

  FIG. 2A shows an example of the luminance image of the optic disc.

<Retardation image generation>
The image generation unit 193 generates a retardation image from tomographic images of polarization components orthogonal to each other.

The value δ of each pixel of the retardation image is a digitization of the phase difference between the vertical polarization component and the horizontal polarization component at the position of each pixel constituting the tomographic image, and each tomographic signal A _H and A _V Is calculated by Equation 2.

  FIG. 2 (b) shows an example of the retardation image (also referred to as a tomographic image showing the phase difference of polarized light) of the optic nerve head thus generated, and for each B scan image, Can be obtained by calculating FIG. 2B shows in color the places where the phase difference occurs in the tomographic image, where the dark and light areas have small phase differences and the light and light areas have large phase differences. Therefore, by generating a retardation image, it is possible to grasp a layer having birefringence.

<Generation of retardation map>
The image generation unit 193 generates a retardation map from the retardation images obtained for a plurality of B-scan images. First, the image generation unit 193 detects retinal pigment epithelium (RPE) in each B scan image. Since RPE has the property of depolarizing light, each A scan is examined for the distribution of retardation in the range not including RPE from the inner boundary film (ILM) along the depth direction, and the maximum value in the A scan A representative value of retardation. The image generation unit 193 generates a retardation map by performing the above processing on all retardation images.

  An example of a retardation map of the optic disc is shown in FIG. In the figure, the dark and light areas represent small phase differences, and the light and light areas represent large phase differences. In the optic papilla, the layer having birefringence is the retinal nerve fiber layer (RNFL), and the retardation map represents the phase difference caused by the birefringence of RNFL and the thickness of RNFL. Therefore, the phase difference becomes large where the RNFL is thick, and the phase difference becomes small where the RNFL is thin. Therefore, the thickness of the RNFL of the entire fundus can be grasped by the retardation map, and can be used for diagnosis of glaucoma.

<Birefringence map generation>
The image generation unit 193 linearly approximates the value of retardation δ in the range from ILM to the retinal nerve fiber layer (RNFL) in each A scan image of the retardation image generated earlier, and the inclination thereof is the A scan image It is determined as birefringence at a position on the retina. By performing this process on all acquired retardation images, a map representing birefringence is generated. An example of a birefringence map of the optic disc is shown in FIG. 2 (d). The birefringence map maps the value of birefringence directly, so even if the thickness of the RNFL does not change, it can be depicted as a change in birefringence if the fiber structure changes.

<DOPU image generation>
The image generation unit 193 calculates a Stokes vector S according to Equation 3 for each pixel from the acquired tomographic signals A _{H and} A _V and the phase difference ΔΦ therebetween.

However, ΔΦ is calculated from the phases _{H H} and _{V V of} each signal obtained when calculating two tomographic images, as ΔΦ == _V −− _H.

  Next, the image generation unit 193 sets windows of about 70 μm and about 18 μm in the main scanning direction and in the depth direction of each B-scan image in the main scanning direction of the measurement light, and is calculated for each pixel by several C in each window. Each element of the Stokes vector is averaged, and the uniformity of polarization DOPU (Degree Of Polarization Uniformity) within the window is calculated by Equation 4.

However, Q _m , U _m and V _m are values obtained by averaging the elements Q, U and V of the Stokes vector in each window. By performing this process on all the windows in the B-scan image, a DOPU image (also referred to as a tomographic image showing the uniformity of polarization) of the optic papilla shown in FIG. 2E is generated.

  DOPU is a numerical value representing the uniformity of polarization, and is close to 1 at the place where the polarization is maintained, and smaller than 1 at the place where the polarization is canceled. In the structure in the retina, the RPE has the property of depolarizing the light, so the portion corresponding to the RPE in the DOPU image has a smaller value than other regions. In the figure, the light and dark areas 210 indicate the RPE, and the dark and light areas 220 indicate the retinal layer area where the change is maintained. Since the DOPU image is formed into an image of a depolarizing layer such as RPE, the RPE can be imaged more reliably than the change in luminance even when the RPE is deformed due to a disease or the like.

  In the present specification, the tomographic image, the retardation image, the DOPU image and the like corresponding to the first and second polarized light described above are also referred to as a tomographic image showing a polarization state. Further, in the present specification, the above-described retardation map, birefringence map, and the like will also be referred to as a fundus image showing a polarization state.

[Processing operation]
Next, the processing operation of the image processing apparatus will be described. This process is an operation for confirming the identification result of the depolarized area in the polarization OCT image and performing correction when an error occurs in the identification result. FIG. 3 is a flowchart showing the processing operation of the present image processing apparatus.

<Step S301: Acquisition of eye information>
In step S301, a subject eye information acquisition unit (not shown) acquires a subject identification number from the outside as information for identifying a subject eye. Then, based on the subject identification number, information on the subject's eye held by the storage unit (not shown) is acquired. Here, information on the subject's eye means personal information such as name, sex, age, medical history, etc., image data such as a fundus image or tomographic image, and analysis data such as image analysis.

<Step S302: Shooting, Image Generation>
The operator designates an imaging instruction button with a cursor displayed on the screen using an instruction device (not shown) such as a mouse, and performs an instruction by a click operation or the like to perform imaging of a tomographic image. In the mouse of this embodiment, for example, a sensor for detecting a movement signal when the mouse body is moved in a two-dimensional manner by the operator's hand and a sensor for detecting that the operator's hand has pressed the mouse. The left and right two mouse buttons, and the wheel mechanism that can be rotated back and forth and left and right are provided between the two left and right mouse buttons. In addition, the pointing device may have a touch panel function mounted on the display unit, and may issue a shooting instruction on the touch panel, or may have a joystick mounted on the device body and may use a joystick to issue a shooting instruction. In photographing, the measurement light is emitted from the light source 101 and the light source 141, and the return light from the retina Er is received by the line cameras 129 and 133 and the APDs 152 and 153, and the image generation unit 193 Generate

<Step S303: Image Analysis>
The area detection unit 194 performs various analysis on the image generated by the image generation unit 193 described above. Here, the region detection unit 194 detects a depolarization region which is a region of a depolarizing material (hereinafter referred to as DM) from the DOPU image. The depolarization region includes an RPE region, a Drusen region, a geographic atrophy (hereinafter referred to as GA) region, and the like. This detection will be described using FIG. 3 (b) and FIG.

<Step S331: Detection of Depolarized Region>
FIG. 4 is a diagram for explaining detection of a depolarized area from a DOPU image. Reference numeral 401 in FIG. 4A denotes a luminance image, and reference numeral 402 in FIG. 4B denotes a DOPU image corresponding to the luminance image in FIG. First, the depolarized region is determined from the DOPU image 402. The depolarization region in the DOPU image 402 is characterized by having a value smaller than one. Therefore, by performing threshold processing (for example, a threshold of 0.75) on the DOPU image, it is possible to divide the region in which the polarization is maintained and the region in which the polarization is cancelled. The depolarization area | region 403 calculated | required by these is shown in FIG.4 (c).

<Step S332: RPE Estimation>
Next, the RPE region is identified to the depolarized region 403. This will be described using FIG. 4 (d). FIG. 4 (d) shows the RPE estimation curve 404 (dashed line in the figure) to identify the RPE for the depolarized area 403. FIG. Here, an example in which the estimated curve 404 is obtained as a quadratic curve is shown. An initial curve is set to pass through the largest area of the depolarization region 403, and the curve is determined by estimating the coefficient parameters (a, b, c) of the curve using a robust estimation method (M estimation, etc.) I can do things. The method of estimating the curve is not limited to this, and it may be an estimation of an N-th order curve or may be a spline curve.

<Step S333: First Analysis? >
In step S333, it is determined whether the first analysis or the analysis after correction (re-analysis). In the case of the first analysis, since the depolarization area is not identified, the process proceeds to step S334. In the case of the second and subsequent analysis, the process proceeds to step S338. The second and subsequent processes will be described at the time of step S306, and here, step S334 will be described.

<Step S334: Identification (Classification)>
The classification unit 195 discriminates (classifies) the RPE region and the other depolarized regions (choroidal tissue, particle region) based on the RPE estimated curve 404 obtained in step S332. That is, the depolarization area is classified into a plurality of types. A depolarization region present in a region through which the RPE estimation curve 404 passes is taken as an RPE region. This represents that the connected depolarization area through which the RPE estimation curve 404 passes is the RPE area 405 (shaded area in the figure). That is, if the RPE estimation curve passes a part of the area where the depolarization areas are connected, it is determined that the entire area is the RPE area. Then, the RPE region is a region not connected, and the depolarization region present in a deeper portion than the RPE region is a Choroid (choroidal tissue) region, and the depolarization region present in a shallow portion than the RPE region is a Particle region Do. The RPE area | region 405 calculated | required by this to FIG.4 (e) is shown. FIG. 4F shows a case where the RPE area 405 is superimposed on the luminance image 401.

<Step S335: Mask Image Creation>
Next, a mask is created in an area other than the RPE from the identification results obtained in steps S334 and S338. That is, an image of only the RPE area is created. Note that since labeling is performed by identification in the previous process, it is not always necessary to create a mask image, and the next drusen calculation may be performed without creating a mask image.

<Step S336: Drusen Calculation>
Next, drusen calculation is performed using the RPE image created in step S335. Drusen sets the deepest position of the RPE image to the drusen position in each A scan or discrete A scan. The drusen curve is obtained by smoothly connecting the positions detected here with a spline function or the like. Drusen generally has a convex shape. Therefore, in order to determine the area and volume of the drusen, a reference RPE base curve is determined. This can calculate the convex hull (Convex hull) with respect to the drusen curve calculated | required here, and can obtain | require the minimum RPE base curve containing the drusen convex shape. The RPE estimation curve described above and the RPE base curve described here are different.

<Step S337: Map (Enface) Image Generation>
Finally, a map (also called Enface) image is generated for each area identified in step S333. In the present embodiment, it is assumed that a DM thickness map, an RPE thickness map, a drusen thickness map, and a GA map are created. Here, the map generation method will be described by taking the RPE thickness map as an example. The RPE thickness map counts the number of RPE pixels per A scan in the RPE tomographic image created in step S335. The thickness of the RPE can be determined by multiplying the total number of pixels counted as the RPE and the pixel resolution in the depth direction (μm / pixel).

  Other maps perform the same processing. In the case of the DM thickness map, the map is obtained by multiplying the total number of pixels in the depolarized area obtained in S331 and the pixel resolution. In the case of the drusen thickness map, the map is obtained by multiplying the total number of pixels between the drusen curve obtained in step S336 and the RPE base curve by the pixel resolution. In the GA map, not the thickness but the RPE thickness map is binarized at a certain threshold (for example, a threshold of 10 μm). This is because RPE is deficient when GA develops, so when RPE thickness map is binarized, RPE is present in the normal area and RPE is not present in the RPE deficient area due to GA. The area of GA can be determined using the defect area.

  In FIG. 4, identification of a general depolarization region has been described. Here, in order to explain an example of manual correction, a case where the depolarization region is misidentified will be described with reference to FIG. FIG. 5 shows an example where the depolarization region of the RPE region is separated and the RPE estimation curve does not pass through the separated depolarization region. Except for FIG. 5 (d) and (e), it is the same as FIG. In FIG. 5D, reference numeral 514 denotes a portion where the depolarization area is separated, 503 denotes a depolarization area, and 504 denotes an RPE estimation curve obtained in the same manner as the method of step S332. The shaded area 505 in FIG. 5E is the RPE, the lattice area 515 is the area determined as the particle, and 525 as Choroid. The following description will be made using the example shown in FIG. In the present embodiment, an example is shown in which the identification result is divided as hatching and a grid, but the present invention is not limited to this. RPE may be red, particle may be blue, and choroid may be yellowish green.

<Step S304: Display>
When generation and analysis of each image are completed in the image generation unit 193 and the area detection unit 194 in the signal processing unit 190, the display control unit 191 generates output information based on the result and outputs it to the display unit 192. Display.

  FIG. 6 is a display example on the display unit 192 in the present embodiment. In the figure, reference numeral 600 denotes a window displayed on the display unit 192, which has display areas 610, 620, 630, and 640. In the display area 610, a fundus plane image and an analysis map (Enface) image display 611, a color bar 613 for indicating the thickness of the map in color, a fundus plane image and a selection unit of the analysis map, a tomographic image position in the map And a tomographic image selection unit 616, a manual correction selection unit 617, a WW / WL selection unit 618, and a thickness graph 619. In the fundus planar image and analysis map selection unit 614, pseudo SLO, DM thickness map, RPE thickness map, drusen thickness map, GA map, Retardation map, birefringence map, and the like are switched and displayed. The tomographic image selection unit 616 switches between the luminance image, the DOPU image, the DM image, the RPE image, the Drusen image, the Retardation image, and the AxisOrientation image to enable display. Here, as the DM image, the RPE image, and the Drusen image, the luminance image obtained from the depolarization region (described above in step S303) is superimposed and displayed. That is, in the case of the DM image, all the depolarization areas are superimposed and displayed so as to distinguish the depolarization areas so that the discrimination result is known. Note that displaying a plurality of colors corresponding to a plurality of types in the depolarization area in a state of being superimposed on the polarization tomographic image is an example, and it may be displayed in a plurality of display forms capable of identifying a plurality of types. . In the case of the RPE image, only the RPE identified from the depolarization area is superimposed and displayed. In the case of the Drusen image, the Drusen curve and the RPE base curve are superimposed and displayed.

  In the display area 620, a tree of examination data is displayed. 621 is a shooting date, 622 is shooting information (right and left eyes, scan pattern, shooting time), and 623 is analysis time and analysis mode. The display area 630 displays patient information (identification information, name, age, gender). The display of patient information is not limited to this, and other information may be displayed. The display area 640 displays information for identifying a screen during work. In the present embodiment, a patient data management screen, an analysis screen, and a setting screen are displayed. In the present embodiment, the area 640 has not only display but also a selection function, and the function can be switched by selecting the place of the area 640.

  In this embodiment, an analysis map 611 is an RPE thickness map, a thickness graph 619 is an RPE thickness, and a tomographic image shows an example of displaying a DM image. The thickness graph 619 is linked to the analysis map 611. Although the types of images to be displayed on the analysis map and the tomographic image can be selected independently of each other, the present invention is not limited to this and may be displayed in conjunction with each other. For example, when a DM image is selected in the tomographic image, the analysis map image may display a DM thickness map. The thickness of the RPE thickness map created from the depolarization region identification result 505 is displayed in color as indicated by the color bar 613. In the present embodiment, the darker the color, the thinner the map, and the lighter the color, the thicker the map. In the RPE thickness map of FIG. 6, a dark area 612 near the center indicates a place where there is no place recognized as RPE, and a rectangular area at the lower right of the area 612 is the depolarization area in FIG. This is a place corresponding to the identification result 515, and shows an example in which the region 612 has a distorted shape because the RPE is erroneously identified as a particle.

<Step S305: Correction? (Fix?)>
Step S305 is processing to determine whether or not the identification result of the depolarized area has been changed. When the manual correction selection unit 617 of FIG. 6 is selected, the manual correction screen shown in FIG. 7 is displayed. Here, the manual correction screen 700 will be described with reference to FIG. FIG. 7A shows the state before manual correction, and FIG. 7B shows the state after manual correction. Here, with manual correction, when a display form of any of a plurality of display forms corresponding to a plurality of types of depolarization area is designated by the operator, the type corresponding to the display form designated by the correction unit 196 is selected. Correction to other types. The manual correction screen includes a tomographic image on which the identification result of the depolarization region is superimposed, an icon 701 for selecting an operation on the tomographic image, a region 702 for manual correction, an identification result instruction unit 703 for instructing correction of the identification result, and identification A result display / non-display designation unit 704, an area deletion designation unit 705, a reanalysis execution instruction unit 706, and a cancel instruction unit 707 are provided. Here, the icon 701-1 designates enlargement / reduction of a tomographic image, 701-2 designates a rectangular area, and 701-3 an icon designates a polygon area. Further, the identification result instruction unit 703 designates the identification results as RPE, Choroid, and Particle, respectively.

  When the user selects an area to be changed with respect to the identification result of the depolarized area, after selecting the icon 701-2 or 701-3, the area is designated on the tomographic image. In the present embodiment, it is assumed that the icon 701-2 is selected and a rectangular area 702 (dotted line in FIG. 7) is set. The position, size and rotation of the rectangular area 702 can be arbitrarily set by the user. In addition, when performing a manual correction, when it is desired to confirm a tomographic image of a luminance value, only the luminance image is displayed by removing the check from the identification result display / non-display designation unit 704. FIG. 7A shows the case where the rectangular area 702 is set in the depolarization area 515. FIG. 7B shows an example in which the RPE is selected as the identification result instruction unit 703 after the region 702 is set. As shown in FIG. 7B, the depolarization region 515 is corrected (corrected) as RPE 505 for all regions by manual correction. When it is desired to correct the set area 702, it is possible to delete the area 702 by selecting the area deletion designation unit 705 and to set a new area again.

  As a method of correcting the identification result of the depolarized area, after the area is selected, the identification result is instructed, but the present invention is not limited thereto. For example, the instruction 703 for the identification result may be selected first, and then the area designation may be performed. In this case, since the identification result is set by the identification result instructing unit 703, the area is automatically corrected to the one specified by the identification result instructing unit 703 each time the area is set, unless the setting is released. Ru. As the timing of applying the identification result, for example, in the case of a rectangular area, the size specification of the area is performed by dragging from the left click of the mouse, and the timing when the mouse operation is finished. The correction of the identification result can be performed at the timing when the Alternatively, the image may be saved by pressing a save button (not shown).

  In addition, when a certain place is designated in the identification result of the depolarized area instead of the designation of the area 702, the identification result of the entire connected area may be corrected. For example, if the instruction part 703 of the identification result is selected first and left-clicked one point of the depolarization area 515 in FIG. 7A, all the areas connected to the depolarization area at the clicked coordinates The result specified by the identification result instruction unit 703 may be corrected.

  Furthermore, in the case of tomographic image data on which volume scanning is performed, a plurality of slices exist. In that case, even when the slice position of the B scan is changed in the area 702 set for a certain B scan image, the area 702 is kept as it is, and the already set area 702 can be applied to another B scan. You may

  In the manual correction screen 700, when correction is performed and the reanalysis execution instruction unit 706 is selected, the process proceeds to step S306, and when the cancel instruction unit 707 is selected, the process proceeds to step S308. In addition, the shape of the area | region which performs the manual correction mentioned above, and the kind of identification result instruction | indication part of a depolarization area | region are not restricted to this. The shape may be circular or N-gonal, and the type may be such that Hard Exudate or the like can be specified.

  Furthermore, FIG. 8 shows another form of the manual correction screen. FIG. 8 shows the tomographic image 501 of luminance and the tomographic image on which the discrimination results 505, 515, and 525 of the depolarization area are superimposed and displayed side by side on the manual correction screen described in FIG. In the case where the tomographic image of luminance and the tomographic image in which the identification result of the depolarization area is superimposed and displayed are displayed side by side, it is preferable to display the same location in conjunction with each other at each slice position.

<Step S306: Image Analysis>
In step S306, reanalysis is performed based on the result corrected by the manual correction. Here, the flow shown in FIG. 3B is performed only on the corrected B scan. In the case of reanalysis, only step S 338 in FIG. 3B is different, and the other processes are the same. In the case of reanalysis, since the depolarization region is identified by manual correction, the following calculation is performed using the corrected identification result.

  In order to speed up, only the manually corrected B scan is corrected, but it is not limited to this. When identification is performed using the information of the adjacent B scan at the time of image analysis, it is desirable to re-analyze using all the information necessary for identification, so the process is repeated again for B scans other than the corrected one. It may be performed.

<Step S307: Display>
In step S307, the corrected and re-analyzed result is displayed. An example of this display is shown in FIG. FIG. 9 shows a state in which the analysis maps 911 and 912 and the identification result of the depolarization area are corrected as a result of the manual correction. Furthermore, before the analysis time of the inspection data tree, an index 923 for showing the correction presence or absence of the analysis result is displayed. Not only the presence or absence of correction for one image volume data but also the presence or absence of correction may be known in B scan units. For example, an index (icon or the like) representing correction execution may be displayed in B scan units at the end of the tomographic image in FIG.

<Step S308: End? >
In step S308, execution of confirmation of the analysis result and termination of the manual correction is selected. When the process is shifted to the end of the manual correction or the selection of the photographing mode, the manual correction result obtained by the above processing and the analysis result by the area detection unit 194 are stored in the storage unit.

  As described above, according to the present invention, the identification result of the depolarized area in the polarization OCT image is confirmed, and there is a mechanism for performing correction when an error occurs in the identification result. Therefore, according to the present invention, it is possible to identify the depolarization area and to store the result correctly. In addition, in this embodiment, although the example which analyzes a depolarization area | region was shown, it does not restrict to this. As the image analysis mode, it is possible to select a Drusen analysis mode, a GA analysis mode, a Glaucoma analysis mode, an analysis mode corresponding to a disease, a full analysis mode for executing all processes, and the like. Then, in the Glaucoma analysis mode, a mode for analyzing polarization components related to the nerve fiber layer such as Retardation, AxisOrientation, etc. may be mounted, and the user may select and execute an arbitrary analysis mode.

(Other embodiments)
The present invention is also realized by executing the following processing. That is, software (program) for realizing the functions of the above-described embodiments is supplied to a system or apparatus via a network or various storage media, and a computer (or CPU, MPU or the like) of the system or apparatus reads the program. It is a process to execute.

One of the image processing apparatuses according to the present invention is
A tomographic luminance image and a polarization tomographic image of the subject's eye obtained by processing a common OCT signal by return light from the subject's eye irradiated with the measurement light and the reference light, wherein the tomographic image of the subject's eye on the image Tomographic image acquisition means for acquiring tomographic luminance images and polarization tomographic images in which the positions of the parts correspond to each other;
Detection means for detecting a depolarization area in the polarization tomographic image;
A classifying means for automatically classifying a plurality of regions in the detected depolarization regions of the type of multiple, automatically classifying a region including a curve at the detected depolarization regions in RPE, the A classification means for automatically classifying an area different from the area including the curve in the detected depolarization area into a type other than RPE ;
A display control unit that causes the display unit to display the tomographic luminance image in a state in which a plurality of display forms corresponding to the plurality of classified types are superimposed on a plurality of positions corresponding to the plurality of areas, Display control means for causing the display means to display the tomographic luminance image in a state in which different display forms are superimposed on the position corresponding to the area including the curve and the position corresponding to the different area ;
A correction unit configured to correct a type corresponding to a display form of the position designated by the operator at any one of the plurality of positions in the tomographic luminance image, to another type designated by the operator; Have
The correction unit is a type corresponding to the display form of the designated position by the operator designating the RPE as the other type after the operator designates the position corresponding to the different area. To the RPE from a type other than the RPE ,
The correction means displays the designated position by the operator designating a type other than the RPE as the other type after the operator designates a position corresponding to the area including the curve. Modify the type corresponding to the form from the RPE to a type other than the RPE
The display control means causes the display means to display the tomographic luminance image in a state where the display form of the designated position is changed to the display form corresponding to the designated other type.

Further, one of the image processing apparatuses according to the present invention is
A tomographic luminance image and a polarization tomographic image of the subject's eye obtained by processing a common OCT signal by return light from the subject's eye irradiated with the measurement light and the reference light, wherein the tomographic image of the subject's eye on the image Tomographic image acquisition means for acquiring tomographic luminance images and polarization tomographic images in which the positions of the parts correspond to each other;
Detection means for detecting a depolarization area in the polarization tomographic image;
A classifying means for automatically classifying a plurality of regions in the detected depolarization regions of the type of multiple, automatically classifying a region including a curve at the detected depolarization regions in RPE, the A classification means for automatically classifying an area different from the area including the curve in the detected depolarization area into a type other than RPE ;
A display control unit that causes the display unit to display the tomographic luminance image in a state in which a plurality of display states corresponding to the plurality of classified types are superimposed on a plurality of positions corresponding to the plurality of areas, Display control means for causing the display means to display the tomographic luminance image in a state in which different display forms are superimposed on the position corresponding to the area including the curve and the position corresponding to the different area ;
A correction unit configured to correct a type corresponding to a display form of the position designated by the operator at any one of the plurality of positions in the tomographic luminance image, to another type designated by the operator; Have
The correction means is a type corresponding to the display form of the designated position by the operator designating the position corresponding to the different area after the RPE is designated as the other type by the operator. To the RPE from a type other than the RPE ,
The correction means displays the designated position by designating the position corresponding to the area including the curve by the operator after the operator designates the type other than the RPE as the other type. Modify the type corresponding to the form from the RPE to a type other than the RPE
The display control means causes the display means to display the tomographic luminance image in a state where the display form of the designated position is changed to the display form corresponding to the designated other type.

Further, one of the image processing methods according to the present invention is
A tomographic luminance image and a polarization tomographic image of the subject's eye obtained by processing a common OCT signal by return light from the subject's eye irradiated with the measurement light and the reference light, wherein the tomographic image of the subject's eye on the image Acquiring a tomographic luminance image and a polarization tomographic image in which the positions of the parts correspond to each other; detecting a depolarization region in the polarization tomographic image;
Wherein the plurality of detected regions in depolarizing regions comprising the steps of automatically classifying the type of multiple, automatically classifying a region including a curve at the detected depolarization regions in RPE, the detection Automatically classifying into the type other than RPE an area different from the area including the curved line in the depolarization area .
Displaying the tomographic luminance image on the display unit in a state in which a plurality of display forms corresponding to the plurality of classified types are superimposed on a plurality of positions corresponding to the plurality of regions, the curve Displaying the tomographic luminance image on the display means in a state in which different display forms are superimposed on the position corresponding to the area including and the position corresponding to the different area ;
Correcting the type corresponding to the display form of the position designated by the operator, which is any position of the plurality of positions in the tomographic luminance image, to another type designated by the operator Have
In the correcting step, after the position corresponding to the different area is designated by the operator, the RPE is designated as the other type by the operator, thereby corresponding to the display form of the designated position. Modify the type from the type other than the RPE to the RPE ,
In the correcting step, after the operator designates a position corresponding to the area including the curved line, the operator designates the type other than the RPE as the other type, thereby designating the position of the designated position. Modify the type corresponding to the display form from the RPE to a type other than the RPE,
In the displaying step, the tomographic luminance image is displayed on the display unit in a state where the display form of the designated position is changed to the display form corresponding to the other designated type.

Further, one of the image processing methods according to the present invention is
A tomographic luminance image and a polarization tomographic image of the subject's eye obtained by processing a common OCT signal by return light from the subject's eye irradiated with the measurement light and the reference light, wherein the tomographic image of the subject's eye on the image Acquiring a tomographic luminance image and a polarization tomographic image in which the positions of the parts correspond to each other; detecting a depolarization region in the polarization tomographic image;
Wherein the plurality of detected regions in depolarizing regions comprising the steps of automatically classifying the type of multiple, automatically classifying a region including a curve at the detected depolarization regions in RPE, the detection Automatically classifying into the type other than RPE an area different from the area including the curved line in the depolarization area .
Displaying the tomographic luminance image on the display unit in a state in which a plurality of display states corresponding to the plurality of classified types are superimposed on a plurality of positions corresponding to the plurality of regions, the curve Displaying the tomographic luminance image on the display means in a state in which different display forms are superimposed on the position corresponding to the area including and the position corresponding to the different area ;
Correcting the type corresponding to the display form of the position designated by the operator, which is any position of the plurality of positions in the tomographic luminance image, to another type designated by the operator Have
In the correcting step, after the RPE is designated by the operator as the other type, the operator designates the position corresponding to the different area, thereby corresponding to the display form of the designated position. Modify the type from the type other than the RPE to the RPE ,
In the correcting step, after the operator designates a type other than the RPE as the other type, the operator designates the position corresponding to the area including the curve, thereby designating the position of the designated position. Modify the type corresponding to the display form from the RPE to a type other than the RPE,
In the displaying step, the tomographic luminance image is displayed on the display unit in a state where the display form of the designated position is changed to the display form corresponding to the other designated type.

Claims (13)

A tomographic luminance image and a polarization tomographic image of the subject's eye obtained by processing a common OCT signal by return light from the subject's eye irradiated with the measurement light and the reference light, wherein the tomographic image of the subject's eye on the image Tomographic image acquisition means for acquiring tomographic luminance images and polarization tomographic images in which the positions of the parts correspond to each other;
Detection means for detecting a depolarization area in the polarization tomographic image;
Classification means for automatically classifying a plurality of areas in the detected depolarization area into a plurality of types including a type of a lesion of an eye to be examined;
A display control unit that causes the display unit to display the tomographic luminance image in a state in which a plurality of display forms corresponding to the plurality of classified types are superimposed on a plurality of positions corresponding to the plurality of areas; Correction means for correcting the type corresponding to the display form of the position specified by the operator, which is any position of the plurality of positions in the image, to another type specified by the operator ,
The correction means specifies a type corresponding to a display form of the designated position by the operator specifying the other type after the operator designates any one of the plurality of positions. Modify to the other type specified above,
The display control means causes the display means to display the tomographic luminance image in a state in which the display form of the designated position is changed to the display form corresponding to the designated other type. Image processing device.   When the operator designates any one of the plurality of positions first, the correction unit may designate the other type by the operator after any one of the plurality of positions is designated. By being designated, the type corresponding to the display form of the designated position is corrected to the designated other type, and when the other type is designated first by the operator, the other type is designated. Correcting the type corresponding to the display form of the designated position to the designated other type by designating any one of the plurality of positions by the operator after the type is designated. The image processing apparatus according to claim 1, characterized in that A tomographic luminance image and a polarization tomographic image of the subject's eye obtained by processing a common OCT signal by return light from the subject's eye irradiated with the measurement light and the reference light, wherein the tomographic image of the subject's eye on the image Tomographic image acquisition means for acquiring tomographic luminance images and polarization tomographic images in which the positions of the parts correspond to each other;
Detection means for detecting a depolarization area in the polarization tomographic image;
Classification means for automatically classifying a plurality of areas in the detected depolarization area into a plurality of types including a type of a lesion of an eye to be examined;
A display control unit that causes the display unit to display the tomographic luminance image in a state in which a plurality of display states corresponding to the plurality of classified types are superimposed on a plurality of positions corresponding to the plurality of regions; Correction means for correcting the type corresponding to the display form of the position specified by the operator, which is any position of the plurality of positions in the image, to another type specified by the operator ,
The correction means specifies a type corresponding to the display form of the designated position by designating any one of the plurality of positions by the operator after the other type is designated by the operator. Modify to the other type specified above,
The display control means causes the display means to display the tomographic luminance image in a state in which the display form of the designated position is changed to the display form corresponding to the designated other type. Image processing device.   The image processing apparatus according to any one of claims 1 to 3, wherein the plurality of display forms are different colors. The plurality of regions include a region including a curve estimated using the detected depolarization region, and a region discontinuous with a region including the curve.
The classification means automatically classifies the region including the curve into the type of the region of the subject eye and the discontinuous region into the type of the lesion.
The display control means causes the display means to display the tomographic luminance image in a state in which different display forms are superimposed on respective positions corresponding to the area including the curve and the discontinuous area, and the correction means When the position designated by the operator is a position corresponding to the discontinuous area, and the other type designated by the operator is the type of the portion, the display form of the designated position The image processing apparatus according to any one of claims 1 to 4, wherein the type corresponding to B is corrected from the type of the lesion to the type of the site. The area including the curve is a first area,
The discontinuous area includes a second area shallower than the estimated curve and a third area deeper than the estimated curve in the depth direction of the polarization tomographic image;
The classification means automatically classifies the second area and the third area into different lesion types;
The display control means displays the tomographic luminance image on the display means in a state in which different display forms are superimposed on the respective positions corresponding to the first area, the second area and the third area. Let
When the position specified by the operator is either the second area or the third area, and the other type specified by the operator is the type of the part The image processing apparatus according to claim 5, wherein the type corresponding to the display form of the designated position is corrected from the type of the different lesion to the type of the site.   The image processing apparatus according to claim 5, wherein the type of the part is RPE.   An analysis map regarding a region including the curve is generated using a region including the curve in the plurality of polarization tomographic images at different positions of the eye to be examined, and corresponds to the region including the curve after correction by the correction unit The image processing apparatus according to any one of claims 5 to 7, further comprising an image generation unit configured to generate the analysis map again using a correction result on a type corresponding to a display format of the position.   When the second tomographic luminance image at an acquisition position different from the acquisition position of the tomographic luminance image and the polarization tomographic image is displayed on the display unit, the correction is performed on the displayed second tomographic luminance image The image processing apparatus according to any one of claims 1 to 8, wherein a correction result by means is held.   The tomographic image acquiring unit may divide the combined light of the return light from the subject's eye irradiated with the measurement light and the reference light corresponding to the measurement light into a plurality of lights obtained by dividing into light of different polarizations. The image processing apparatus according to any one of claims 1 to 9, wherein a luminance image and the polarization tomographic image are acquired. A tomographic luminance image and a polarization tomographic image of the subject's eye obtained by processing a common OCT signal by return light from the subject's eye irradiated with the measurement light and the reference light, wherein the tomographic image of the subject's eye on the image Acquiring a tomographic luminance image and a polarization tomographic image in which the positions of the parts correspond to each other; detecting a depolarization region in the polarization tomographic image;
Automatically classifying a plurality of areas in the detected depolarization area into a plurality of types including a type of a lesion of an eye to be examined;
Displaying the tomographic luminance image on a display unit in a state in which a plurality of display forms corresponding to the plurality of classified types are superimposed on a plurality of positions corresponding to the plurality of regions; And correcting the type corresponding to the display form of the position designated by the operator to any other position designated by the operator.
In the correcting step, after the operator designates any one of the plurality of positions, the operator designates the other type, thereby the type corresponding to the display form of the designated position. To the other types specified above,
In the displaying step, the tomographic luminance image is displayed on the display unit in a state in which the display form of the designated position is changed to the display form corresponding to the designated other type. Image processing method. A tomographic luminance image and a polarization tomographic image of the subject's eye obtained by processing a common OCT signal by return light from the subject's eye irradiated with the measurement light and the reference light, wherein the tomographic image of the subject's eye on the image Acquiring a tomographic luminance image and a polarization tomographic image in which the positions of the parts correspond to each other; detecting a depolarization region in the polarization tomographic image;
Automatically classifying a plurality of areas in the detected depolarization area into a plurality of types including a type of a lesion of an eye to be examined;
Displaying the tomographic luminance image on a display unit in a state in which a plurality of display states corresponding to the plurality of classified types are superimposed on a plurality of positions corresponding to the plurality of regions; And correcting the type corresponding to the display form of the position designated by the operator to any other position designated by the operator.
In the correcting step, after the operator designates the other type, the operator designates any one of the plurality of positions, thereby the type corresponding to the display form of the designated position. To the other types specified above,
In the displaying step, the tomographic luminance image is displayed on the display unit in a state in which the display form of the designated position is changed to the display form corresponding to the designated other type. Image processing method.   The program for making a computer perform each process of the image processing method of Claim 11 or 12.