JP5242716B2 - Fundus image processing device - Google Patents

Description

  The present invention relates to a fundus image processing apparatus for comparing a first fundus tomographic image and a second fundus tomographic image acquired by an optical coherence tomography with respect to the same part of the fundus at a time.

2. Description of the Related Art Conventionally, devices that non-invasively image retinal functions are known. This apparatus has illumination means for illuminating the retina and retinal stimulation illumination means for irradiating stimulation light that induces a functional response of the retina, and retinal function based on the state of the retinal image before and after irradiating the stimulation light to the retina. Is measured, imaged, and evaluated (see Patent Document 1).
An apparatus having an interference optical system for obtaining an image of the fundus of a subject's eye non-invasively using an optical interference technique is known (see Patent Document 2).
Special Table 2002-521115 JP-A-10-33484

The device for measuring the function of the retina as described above reads the change in the brightness of the retinal image after the irradiation with respect to the brightness of the retinal image before the irradiation with the stimulation light. Since the change in the activity of the nerve tissue due to the stimulation light is very small, it is important to accurately detect the change. In addition, in order to measure the retinal function in more detail, it is necessary to perform measurement over the depth direction of the retina.
In addition, although the retinal tomogram is evaluated in the interference optical system described above, there is room for improvement.
In view of the above-described problems, an object of the present invention is to provide a fundus image processing apparatus capable of performing accurate detailed retina evaluation.

In order to solve the above problems, the present invention is characterized by having the following configuration.
(1)
In a fundus image processing apparatus for comparing a first fundus tomographic image and a second fundus tomographic image acquired at a time by an optical coherence tomography with respect to the same part of the fundus,
First image acquisition means for acquiring a first fundus three-dimensional tomographic image acquired by an optical coherence tomography and a first fundus front image associated with the first fundus three-dimensional tomographic image;
A second fundus 3D tomographic image acquired by an optical coherence tomography at a time after acquisition of the first fundus 3D tomographic image and a second associated with the second fundus 3D tomographic image. Second image acquisition means for acquiring a fundus front image of
The first fundus front image and the second fundus front image are aligned, and based on the alignment result, the first fundus three-dimensional tomographic image acquired in advance and the previously acquired Association means for associating the second fundus three-dimensional tomographic image;
It is characterized by having.

  According to the present invention, detailed retina evaluation can be performed with high accuracy.

  Embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a diagram illustrating an optical system and a control system of the ophthalmologic photographing apparatus according to the present embodiment. In this embodiment, the depth direction of the eye to be examined is the Z direction (optical axis L1 direction), the horizontal component on the plane perpendicular to the depth direction (the same plane as the face of the subject) is the X direction, and the vertical direction. The component is described as the Y direction.

  In FIG. 1, the optical system includes an interference optical system (hereinafter referred to as an OCT optical system) 100 for noninvasively obtaining an image of the fundus of the eye to be examined using an optical interference technique, and an infrared light. A scanning laser off-salmoscope optical system 200 (hereinafter referred to as an SLO optical system) for acquiring a fundus SLO image for illuminating and observing the fundus of the optometry, and stimulating light for stimulating the retina toward the fundus of the eye to be examined The stimulus light irradiation optical system 300 is generally divided into two types. In the present embodiment, the OCT optical system 100 is synchronized with the scanning unit that scans the measurement light two-dimensionally on the fundus and the scanning of the measurement light by the scanning unit to obtain an interference signal in the depth direction of the eye to be examined. And an optical path length changing unit that changes the optical path length of the reference light.

  The configuration of the OCT optical system 100 will be described below. Reference numeral 10 denotes an OCT light source that emits low-coherent light used as measurement light and reference light of the OCT optical system 100. For example, an SLD light source or the like is used. For the OCT light source 10, for example, a light source having a center wavelength of 840 nm and a bandwidth of 50 nm is used. Reference numeral 11 denotes a fiber coupler that doubles as a light splitting member and a light coupling member. The light emitted from the OCT light source 10 is divided into reference light and measurement light by the fiber coupler 11 through an optical fiber 30a as a light guide. Therefore, the measurement light is directed to the eye E through the optical fiber 30b, and the reference light is directed to a reference mirror unit 40 described later via the optical fiber 30c.

  In the optical path for emitting the measurement light toward the eye E, the end 31b of the optical fiber 30b that emits the measurement light, the relay lens 12 that can move in the optical axis direction according to the refractive error of the eye to be examined, and the galvanometer mirror drive A scanning unit 20 including a pair of galvanometer mirrors capable of scanning the measurement light in the XY directions at high speed on the fundus by driving the mechanism 51, and a relay lens 21 for condensing the fundus of the eye E to a conjugate position. Is arranged. Reference numeral 13 denotes a dichroic mirror, and reference numeral 14 denotes an objective lens. The dichroic mirror 13 has a characteristic of reflecting light (infrared light) from the OCT light source 10 and transmitting stimulus light and fixation lamp light (visible light) described later. Further, the end 31b of the optical fiber 30b is arranged so as to be conjugate with the fundus of the eye to be examined. In addition, the reflection surface of the galvanometer mirror of the scanning unit 20 is arranged at a position conjugate with the eye pupil to be examined (in this embodiment, the middle position of the pair of galvanometer mirrors and the eye pupil to be examined are arranged in a conjugate relationship). Have been).

The measurement light emitted from the end 31b of the optical fiber 30b reaches the galvanometer mirror of the scanning unit 20 via the relay lens 12, and the reflection direction is changed by driving the pair of galvanometer mirrors. Then, the measurement light reflected by the galvanometer mirror passes through the optical axis L2, passes through the relay lens 21, passes through the dichroic mirror 13, passes through the optical axis L1, and collects on the eye fundus through the objective lens 14. Lighted.
The measurement light reflected from the fundus is reflected by the dichroic mirror 13 through the objective lens 14, travels toward the OCT optical system 100, passes through the relay lens 21, the galvano mirror of the scanning unit 20, and the relay lens 12, and the optical fiber 30 b. Is incident on the end 31b of the. The measurement light incident on the end 31b reaches the fiber coupler 15 through the optical fiber 30b, the fiber coupler 11, and the optical fiber 30d.

On the other hand, in the optical path for emitting the reference light from the fiber coupler 11 toward the reference mirror unit 40, the end 31c of the optical fiber 30c that emits the reference light, the collimator lens 16, the reference mirror unit 40, the condensing lens 17, and the reference An end 31e of the optical fiber 30e on which light is incident is disposed. The reference mirror unit 40 is configured to be movable in the optical axis direction by the reference mirror driving mechanism 50 in order to change the optical path length of the reference light. The reference mirror unit 40 includes a mirror 40a and a mirror 40b.
The reference light emitted from the end portion 31c of the optical fiber 30c is converted into a parallel light flux by the collimator lens 16, reflected by the mirror 40a and the mirror 40b constituting the reference mirror unit 40, and then condensed by the condenser lens 17 to be light. The light enters the end 31e of the fiber 30e. The reference light incident on the end 31e reaches the fiber coupler 15 via the optical fiber 30e.

  Here, the measurement light is reflected by each layer of the fundus oculi and becomes reflected measurement light having different time delays and different intensities, and is merged with the reference light by the fiber coupler 15. By utilizing the interference phenomenon that occurs when the optical path lengths of the two lights become equal, the intensity of the reflected measurement light is detected by the light receiving element 18, and the reference mirror unit 40 is moved (scanned) in the optical axis direction. The reflection intensity distribution in the optical axis direction (depth direction of the eye to be examined) can be obtained. In the present embodiment, a method of obtaining a reflection intensity distribution in the optical axis direction by changing the optical path length of the reference mirror unit at one point on a certain fundus is referred to as A scan. Further, by scanning the measurement light in the X direction or the Y direction on the fundus by the scanning unit 20, a tomographic image on the XZ plane or the YZ plane of the eye fundus to be examined can be acquired. In the present embodiment, the method of obtaining a retinal tomographic image by performing one-dimensional scanning with the measurement light on the fundus in this way and changing the optical path length due to movement of the reference mirror unit in the optical axis direction is referred to as B-scan. . Furthermore, it is also possible to obtain a fundus image (XY plane) two-dimensionally by scanning the measurement light two-dimensionally in the XY direction with the scanning unit 20 while the reference mirror unit 40 is fixed. In the present embodiment, the measurement light is scanned two-dimensionally with respect to the fundus, and a two-dimensional planar OCT image is obtained by an interference signal obtained when the optical path of the reference mirror matches within the coherence length. The method is C scan. Furthermore, by utilizing these, it is possible to acquire three-dimensional OCT image information based on the scanning of the measurement light by the scanning unit 20 and the interference signal obtained in synchronization with the reference mirror unit 40. In this case, three-dimensional OCT image information of the fundus can be acquired by moving the reference mirror unit 40 in the optical axis direction and scanning the measurement light two-dimensionally with respect to the fundus by the scanning unit 20.

  Note that the A-scan signal is measured from the front in the optical axis direction, and a position where a strong signal can be obtained first from a position where there is no signal is information on the fundus surface layer (retina surface). Therefore, the surface of the fundus is two-dimensionally connected by first connecting the reflection intensities of strong signals for the A scan signals at each measurement position (imaging position) on the fundus when the measurement light is scanned in two axes. OCT image (en-face image) of the fundus plane expressed in In addition, since the tomographic image acquired by the B scan is constructed by uniaxial scanning of the A scan signal, a partial image signal of the fundus surface OCT image and the fundus surface layer part of the acquired tomographic image By obtaining the position to be compared with the image signal, it is possible to accurately detect the position of the tomographic image on the fundus surface OCT image.

  Next, the SLO optical system 200 will be described. In the present embodiment, an OCT light source (light source 10) is also used as a light source of the SLO optical system, and a half mirror 60 is provided between the relay lens 12 and the end portion 31b of the optical fiber 30b, and is shared in the reflection direction of the half mirror 60. A condensing lens 61 for constituting a focusing optical system, a confocal aperture 62 conjugate to the fundus, and a light receiving element 63 for SLO are provided. In the SLO optical system 200, the optical path from the half mirror 60 to the eye E is shared with the OCT optical system 100. The galvanometer mirror (scanning unit 20) is also used as the OCT optical system 100 in order to scan light used for the SLO optical system in the XY directions on the fundus. In this case, the fundus SLO image can be acquired by two-dimensionally scanning the measurement light in the XY directions by the scanning unit 20.

Next, the stimulation light irradiation optical system will be described. The stimulation light irradiation optical system 300 includes a light source 22 such as a flash lamp, a wavelength selection filter 23, a light projection lens 24, a half mirror 25, and an objective lens 14. The wavelength selection filter 23 has a characteristic of selectively transmitting a wavelength capable of stimulating the retina. The wavelength selection filter 23 is prepared by replacing a plurality of wavelength selection filters such as a cone stimulation filter, a rod stimulation filter, and a color filter that transmits only red, green, and blue, respectively. You can also.
The light source 22 is arranged at a position conjugate with the vicinity of the pupil of the eye E through the light projection lens 24 and the objective lens 14. The light (stimulation light) emitted from the light source 22 is reflected by the half mirror 25 after passing through the wavelength selection filter and the light projection lens 24. The stimulation light reflected by the half mirror 25 passes through the dichroic mirror 13, passes through the objective lens 14, and is once condensed near the pupil of the eye E, and then uniformly illuminates the fundus of the eye to be examined.

  Reference numeral 1 denotes a light source for fixation composed of an LED or the like, 2 denotes an aperture, 3 denotes a collimator lens, and 4 denotes a relay lens. The diaphragm 2 is placed at a position conjugate with the fundus of the eye E through the collimator lens 3, the relay lens 4, and the objective lens 14. The light beam emitted from the fixation light source 1 illuminates the diaphragm 2. The light transmitted through the diaphragm 2 passes through the collimator lens 3, the relay lens 4, and the half mirror 25, and then converges once in the vicinity of the dichroic mirror 13 (fundus conjugate position), and then converges on the eye E to be examined via the objective lens 14. To do.

  A control unit 70 includes a CPU and the like for controlling the entire apparatus. The control unit 70 is provided with light sources 1, 10 and 22, a reference mirror driving mechanism 50, a galvano mirror driving mechanism 51, light receiving elements 18 and 63, an image processing unit 71, a storage unit 72, and switches for performing various operations. The control unit 73, the monitor 74, and the like are connected. The image processing unit 71 has a function of forming an image of the fundus of the eye to be examined and imaging a retinal function based on signals received by the light receiving element 18 and the light receiving element 63. In addition, the fundus image formed by the image processing unit 71 and retinal function information are displayed on the monitor 74. The control unit 73 includes a condition setting switch group 73a for setting an imaging condition in the OCT image, an imaging switch 73b for acquiring an SLO image and an OCT image, a stimulation light irradiation switch 73c, and irradiation before and after the stimulation light irradiation. An analysis switch 73d or the like for analyzing a change in the state of the retina later is prepared. The condition setting switch group 73a selects a B-scan image (retinal tomographic image), a C-scan image (fundus plane image), or a three-dimensional image (B-scan + C-scan), and the depth direction ( It is possible to set an imaging step interval in the depth direction).

  The operation of the apparatus having the above configuration will be described. Here, a method for obtaining a C-scan image as an OCT image and measuring the function of the retina will be described. First, the examiner uses the condition setting switch group 73a of the control unit 73 to set the interval between the capture of the C scan image and the imaging step in the depth direction (for example, imaging 1 mm depth in 50 μm steps). .

  Measurement light (low coherent light) is scanned two-dimensionally on the fundus of the eye E from the light source 10 via the OCT optical system 100, and reflected light from the fundus is received by the SLO optical system 200. . The image processing unit 71 displays a fundus observation image (SLO image) as shown in FIG. 2 on the screen of the monitor 74 in real time based on the obtained light reception signal. The examiner focuses the fundus on the basis of the SLO image displayed on the display monitor 75 and then turns on the fixation light source 1 to fix the eye E to be examined.

  When the alignment between the eye E and the apparatus is completed, the examiner next uses the imaging switch 73b to capture an OCT image before the stimulation light irradiation. When the photographing switch 73b is used, the control unit 70 drives the reference mirror driving mechanism 50 to light the reference mirror unit 40 in accordance with the driving of the scanning unit 20 so as to obtain an image with a predetermined number of scanning steps. Move in the axial direction. The light receiving element 18 sequentially detects interference light by combining the reference light and the reflected measurement light from the fundus, and the image processing 71 acquires a reflection intensity distribution in the XY direction of the measurement light. In addition, since the reference mirror unit 31 is moving in the optical axis direction, it is possible to acquire a reflection intensity distribution at predetermined intervals in the depth direction. Then, the image processing unit 71 obtains an OCT image (C scan image) of the fundus based on the obtained reflection intensity distribution in the XZ direction for each depth in the depth direction. In addition, the image processing unit 71 acquires an SLO image (still image) of the fundus surface at the same time, and stores the OCT image in association with the SLO image in the storage unit 72 together with the SLO image.

Next, the examiner uses the stimulation light irradiation switch 73c to irradiate the fundus of the eye to be examined with the stimulation light. When the stimulation light irradiation switch 73c is pressed, the control unit 70 emits flash light using the light source 22 to stimulate the retina of the eye E to be examined. By irradiating the fundus of the eye E with the stimulation light, the cells constituting the retina are stimulated, and the activity of nerve cells based on this is caused.
After the stimulation light irradiation, the control unit 70 uses the OCT optical system 100, the SLO optical system 200, and the image processing unit 71 to capture the fundus image after the irradiation of the stimulation light in the same manner as described above, and the SLO image and the corresponding OCT. An image is acquired and stored in the storage unit 72. Shooting after stimulating light irradiation is not only performed once, but also so that changes in the retinal function can be seen, the fundus images are taken over time at predetermined intervals (for example, 1 second after irradiation, 2 seconds after ...) after the stimulation light irradiation. It may be taken and stored. In the present embodiment, the process from the stimulation light irradiation to the acquisition of the fundus image after the irradiation is automatically performed. However, the present invention is not limited to this, and individual operations may be performed manually. .

  After storing a predetermined fundus image in the storage unit 72, the analysis switch 73d of the control unit 73 is pressed. When the analysis switch 73d is pressed, the image processing unit 71 measures the retinal function from the luminance difference based on the OCT image before irradiation with stimulation light and the OCT image after irradiation stored in the storage unit 72, The result is displayed on the monitor 74. When stimulation light that stimulates the retina is irradiated to the fundus of the eye E, and the cells that make up the retina are stimulated, changes in the activity of nerve cells occur with this stimulation, and reflection of the site where the nerve activity occurs The intensity (reflectance) of light changes. For this reason, by reading the change in the brightness of the fundus image before and after the stimulation light irradiation, an intrinsic signal change resulting from the change in the activity of the nerve cell can be obtained, whereby the retinal function can be measured. .

Next, a method for measuring this retinal function will be described below using the flowchart of FIG.
The image processing unit 71 first aligns the pre-irradiation SLO image and the post-irradiation SLO image stored in the storage unit 72 when measuring the retinal function. For alignment, feature points (for example, blood vessel shape, nipple, macula, etc.) are extracted from the pre-irradiation SLO image and post-irradiation SLO image by image processing, and the two images are relatively moved, enlarged, reduced, etc. In this way, the position where the feature points of the two images most closely match is obtained by calculation processing. Note that the alignment method is not limited to this, and a known image processing technique may be used.

  By such alignment of the SLO images before and after irradiation, a relative positional shift amount between both SLO images can be obtained. Next, based on the relative positional deviation amount (positional deviation information) between the two SLO images, the image processing unit 71 compares the relative positions of the OCT images stored corresponding to the two SLO images. Arithmetic processing is performed so that the deviation is eliminated. The image processing unit 71 obtains, for each pixel, a change in the brightness of the OCT image after irradiation with respect to the OCT image before irradiation with respect to the OCT image before irradiation in a state where the positional deviation between the OCT images before irradiation and after irradiation is offset. The change in brightness can be obtained by obtaining a difference, a ratio, or the like. The image processing unit 71 displays the obtained brightness change information on the monitor 74 in association with each pixel. The brightness change information includes graphic and numerical information such as a method of displaying as a shaded image, numerical information of differences and ratios, information obtained by processing this numerical information by a predetermined analysis program for evaluating retinal function, etc. Can be expressed as Note that OCT images that are obtained at the same depth (depth) are used as OCT images to be compared for measurement of retinal function. Therefore, in this embodiment, since OCT images are obtained at predetermined intervals in the depth direction, OCT images having the same depth before and after irradiation are sequentially compared, and change information is displayed on the monitor 74. By doing so, the change of the retinal function can be measured in detail with respect to the depth direction of the fundus. Moreover, it is also possible to select the depth position by the control unit 73 and compare the corresponding OCT images before and after irradiation without comparing all acquired OCT images.

As described above, the OCT image has a very good resolution in the depth direction and is suitable for measuring the function of the retina. On the other hand, since the resolution is not good in the horizontal direction, it is difficult to extract feature points and save time. Therefore, it is not suitable for correcting the positional deviation between the images taken with the camera. On the other hand, the SLO image is inferior in resolution in the depth direction compared to the OCT image, but has very good resolution in the horizontal direction, and feature points are extracted from the obtained image. Therefore, it is very effective for correcting the positional deviation between images taken with time.
In the above description, the measurement of the retinal function using the C-scan OCT image has been described as an example, but the present invention is not limited to this. Retinal functions can also be measured using B-scan retinal cross-sectional images. The case where retinal function is measured using a B-scan image (retinal cross-sectional image) will be described below.

  First, the examiner uses the condition setting switch group 73a of the control unit 73 to set conditions for acquiring a B-scan image, and from the SLO image on the display monitor 74 observed in real time, the retinal tomography that the examiner wants to photograph. Set the position of the image. The examiner operates a measurement position setting switch (not shown) to display a line 200 representing a measurement position (acquisition position) electronically displayed on the SLO image on the screen as shown in FIG. 4 with respect to the SLO image. The measurement position is determined using the control unit 73.

When the measurement position is determined, the image processing unit 71 causes the storage unit 72 to store the formation position information of the line 200 with respect to the SLO image (observation image) displayed on the monitor 74. The formation position information of the line 200 with respect to the SLO image may be acquired as, for example, the formation position information of the line 200 with respect to the feature point (blood vessel, nipple, etc.) of the SLO image.
Thereafter, when the examiner inputs the photographing switch 74b, the control unit 70 causes the storage unit 72 to store the SLO image displayed on the monitor 74 as an observation image at the time of obtaining the B scan image, and also displays the SLO on the screen. Based on the display position of the line 200 set on the image, the scanning unit 20 is driven to scan the measurement light so that a tomographic image of the fundus at the position of the line 200 is obtained. Since the relationship between the display position of the line 200 (coordinate position on the monitor) and the scanning position of the measurement light by the scanning unit 20 is determined in advance, the control unit 70 performs scanning corresponding to the set display position of the line 20. The pair of galvanometer mirrors of the scanning unit 20 is appropriately driven and controlled so that the measurement light is scanned with respect to the range. Further, the control unit 70 drives the reference mirror driving mechanism 50, moves the reference mirror unit 40 in the optical axis direction so as to obtain an image with a predetermined number of scanning steps, and acquires a B-scan image. The obtained B-scan image is stored in the storage unit 72 in association with the previously acquired SLO image.

  Next, the examiner uses the stimulation light irradiation switch 73c to irradiate the fundus of the eye to be examined with the stimulation light. When the stimulation light irradiation switch 73c is pressed, the control unit 70 emits flash light using the light source 22 to stimulate the retina of the eye E to be examined. After the stimulation light irradiation, the control unit 70 uses the OCT optical system 100, the SLO optical system 200, and the image processing unit 71 to take a B-scan image after the stimulation light irradiation in the same manner as described above, and to correspond to the SLO image and the corresponding image. A B-scan image is acquired and stored in the storage unit 72. The B scan image of the eye E after irradiation is acquired based on the formation position information of the line 200 with respect to the SLO image stored in the storage unit 72. The obtained B-scan image after irradiation is stored in the storage unit 72.

  After storing a predetermined fundus image in the storage unit 72, the analysis switch 73d of the control unit 73 is pressed. When the analysis switch 73d is pressed, the image processing unit 71 measures the retinal function from the difference in luminance based on the B scan image before irradiation with stimulation light and the B scan image after irradiation stored in the storage unit 72. The result is displayed on the monitor 74. With such a B-scan OCT image, it is possible to evaluate the retinal function in the depth direction of the fundus.

  It is also possible to acquire a three-dimensional fundus OCT image by acquiring the B-scan and C-scan OCT images over the entire scanning range. Thus, even when acquiring a three-dimensional OCT image at a time, the OCT image is stored in correspondence with the SLO image before and after the stimulation light irradiation, and the positional deviation correction can be performed by the SLO image. That's fine. If the OCT image is acquired in advance three-dimensionally in this way, it is possible to measure the retinal function by appropriately extracting the B scan image and the C scan image.

  In the above description, TD-OCT (time domain OCT) having a configuration for acquiring an optical interference image by moving the reference mirror unit 40 on the optical axis and changing the optical path length is used. However, the present invention is not limited to this, and an interference image may be acquired by another measurement principle. For example, interference light obtained by combining measurement light and reference light is split into frequency components via a diffraction grating and received, and an interference image is obtained by calculating the received light signal using Fourier transform. SD-OCT (spectral domain OCT) may be used.

It is the figure which showed the optical system and control system of this embodiment. It is the figure which showed an example of the observation image displayed on a monitor. It is the flowchart which showed the flow of the comparison method of the OCT image before and after stimulation light irradiation. It is the figure which showed the setting of the acquisition position of a retina cross-section image.

DESCRIPTION OF SYMBOLS 1 Light source for fixation 2 Aperture 10 OCT light source 20 Scanning unit 22 Light source 23 Wavelength selection filter 40 Reference mirror unit 70 Control part 71 Image processing part 72 Storage part 73 Control part 74 Monitor 100 OCT optical system 200 SLO optical system 300 Stimulation light irradiation Optical system

Claims (1)

In a fundus image processing apparatus for comparing a first fundus tomographic image and a second fundus tomographic image acquired at a time by an optical coherence tomography with respect to the same part of the fundus,
First image acquisition means for acquiring a first fundus three-dimensional tomographic image acquired by an optical coherence tomography and a first fundus front image associated with the first fundus three-dimensional tomographic image;
A second fundus 3D tomographic image acquired by an optical coherence tomography at a time after acquisition of the first fundus 3D tomographic image and a second associated with the second fundus 3D tomographic image. Second image acquisition means for acquiring a fundus front image of
The first fundus front image and the second fundus front image are aligned, and based on the alignment result, the first fundus three-dimensional tomographic image acquired in advance and the previously acquired Association means for associating the second fundus three-dimensional tomographic image;
And a fundus image processing apparatus.

Images