JP5931163B2 - Ophthalmic apparatus and method for measuring movement of eye to be examined - Google Patents

Description

  The present invention relates to an ophthalmologic apparatus and a control method, and more particularly to an ophthalmologic apparatus that obtains the movement of an eyeball and a method of measuring the movement of an eye to be examined.

  In recent years, an apparatus for measuring eye movement has attracted attention. If the eye movement can be measured, it can be applied to a visual field inspection, a fundus tomographic imaging apparatus that acquires a higher-definition image, etc., and a fundus inspection with higher accuracy can be performed.

  Various techniques such as a corneal reflection method (Purkinje image) and a search coil method are known as methods for measuring eye movement. In particular, a method for measuring eye movement from a fundus image has been studied as a simple method with less burden on the subject.

  In order to accurately measure eye movements from the fundus image, it is necessary to extract feature points from the fundus image, search for and detect feature points in the image to be processed, and then calculate the amount of movement of the feature points. Among these, the feature point extraction process is important in terms of measurement stability, accuracy, and reproducibility of eye movements. As the feature point of the fundus image, the macula, the optic disc (hereinafter referred to as the nipple), or the like is used. In addition, since there are many cases in which the macula and the nipple are incomplete in the affected eye, blood vessels may be used as feature points of the fundus image. Various methods are known as methods for extracting feature points of blood vessels. For example, in Patent Document 1, the number of blood vessels and the presence / absence of blood vessels in the center of the filter are determined from the average value of the pixel values on the outer periphery of the filter set in the fundus image, and the blood vessel crossing portion in the filter region is determined. Is disclosed.

JP 2001-70247 A

  If a feature point of the fundus is extracted using a method as in Patent Document 1 and the amount of movement of the eyeball between acquired images is calculated by comparing the positional relationship of the feature points between images, the fundus image can be accurately obtained. It is possible to detect the movement of the eyeball. However, when the extracted feature point is located at the end of the acquired image, the feature point is deviated from the image to be processed due to the movement of the eyeball, and the amount of movement of the eyeball cannot be detected. (Feature point search error). Further, when a feature point is extracted from the center of the image when attempting to detect the rotational movement of the eyeball, the position of the feature point does not change due to the rotation of the eyeball, and the rotation may not be detected.

In order to solve the above-described problem, an ophthalmologic apparatus according to the present invention includes an image acquisition unit that acquires a plurality of fundus images of an eye to be examined at different times;
Determining means for determining a partial region having a predetermined width from a peripheral portion of at least one fundus image among the plurality of acquired fundus images;
Processing means for extracting at least one feature region from another region different from the determined partial region of the at least one fundus image;
Means for measuring movement of the eye to be examined using the at least one feature region.

Further, the method for measuring the movement of the eye to be inspected according to the present invention includes a step of acquiring a plurality of fundus images of the eye to be examined at different times,
Determining a partial region having a predetermined width from a peripheral portion of at least one fundus image among the plurality of acquired fundus images;
Extracting at least one feature region from another region different from the determined partial region of the at least one fundus image;
Measuring the movement of the eye to be examined using the at least one feature region;
It is characterized by including

  According to the present invention, even when the eyeball moves, the fundus feature point does not fall outside the frame of the acquired image, and the occurrence of a feature point search error can be prevented.

It is the schematic explaining the structure of the optical system of the retinal camera in Example 1 of this invention. It is the schematic explaining the fundus image in Example 1 of the present invention. It is the schematic of the function system of the apparatus in Example 1 of this invention. It is a control flow figure in Example 1 of the present invention. It is a flowchart regarding the process A of the control flow in Example 1 of this invention. It is a flowchart regarding the process B of the control flow in Example 1 of this invention. It is the schematic explaining the structure of the optical system of the OCT apparatus and SLO apparatus in Example 2 of this invention. It is the schematic of the function system of the apparatus in Example 2 of this invention. It is a control flow figure in Example 2 of the present invention. It is a flowchart regarding the process C of the control flow in Example 2 of this invention. It is the schematic regarding the graph for calculating the moving distance of the eyeball in Example 2 of this invention. It is a flowchart regarding the process D of the control flow in Example 2 of this invention. It is the schematic explaining the SLO image in Example 2 of this invention. It is the schematic showing the example of a display in Example 2 of this invention. It is the schematic showing the fundus image in Example 3 of the present invention.

EMBODIMENT OF THE INVENTION The form for implementing this invention is demonstrated in detail using drawing in the following Examples.
In the following embodiments, an example in which the present invention is applied to a single device will be described. However, the object of the present invention is not limited to the following configuration, and a single device having the following configuration is used. It is not limited to. The present invention uses a method for realizing the following functions and supplies software (computer program) for realizing these functions to a system or apparatus via a network or various storage media. It is also realized by a process in which a computer (or CPU, MPU, etc.) reads and executes a program.

[Example 1]
Embodiment 1 of the present invention will be described below.
In this embodiment, in order to solve the above-described problem, an example in which eye movement is measured by specifying an extraction region when extracting feature points of a fundus image will be described.

(Fundus camera)
A fundus camera was used to acquire the fundus image. FIG. 1 shows a schematic diagram of a fundus camera. A digital single lens reflex camera 50 capable of imaging at a video rate as a signal acquisition unit is connected to the fundus camera main body unit 1 via a connection unit 40. On the optical path of the objective lens 11 facing the eye E, a perforated mirror 12 is provided. A relay lens 13, a black spot plate 14, a relay lens 15, a ring slit plate 16, a fluorescent exciter filter 17, and a mirror 18 are arranged on the optical path in the incident direction of the perforated mirror 12. Further, in the incident direction of the mirror 18, a condenser lens 19, a photographing light source 20 made of a xenon tube, a condenser lens 21, and an observation light source 22 made of an infrared light emitting diode are arranged. In the figure, the optical path is indicated by a dotted line.

  A focus lens 24, a fluorescent barrier filter 25, and an imaging lens 26 are arranged behind the perforated mirror 12, and a digital single lens reflex camera 50 is connected thereto. In the digital single-lens reflex camera 50, a quick return mirror 51, a focal plane shutter (not shown), and a two-dimensional sensor 53 are arranged on the same optical path as the optical path behind the objective lens 11. A pentaprism 54 and an eyepiece 55 are provided in the reflection direction of the quick return mirror 51. The signal received by the two-dimensional sensor 53 is processed by the signal processing board 52, transferred to the PC 56 having the HDD 58 via the cable, and displayed on the display 57. The fundus camera is provided with an external fixation lamp (not shown) so that fixation is stable.

(Control method)
Fig. 3 shows the functional system. The functional system includes a CPU 303 that controls the entire system, a control unit 305 that controls the fundus image acquisition unit, a camera unit 301 that acquires the fundus image, a display unit 302 that displays the system state, and an HDD that records the fundus image and imaging conditions. (Recording unit) 304. At the time of observing and photographing the fundus, the CPU 303 instructs the control unit 305 to capture image conditions, and the fundus is imaged. After the fundus is imaged, an image is sent from the camera unit 301 to the CPU 303, subjected to image processing and the like, displayed on the display unit 302, and stored in the recording unit 304 at the same time or thereafter.

  FIG. 4 shows the overall flow of measuring eye movements for a certain period of time using the functions described above. A first fundus image is acquired using the fundus camera 1 (step 402). After acquiring the first fundus image, before extracting feature points (hereinafter referred to as templates), an area having a predetermined width that is a constant width from the periphery of the first image in the center coordinate direction of the image Is set as a mask area (first area) (step 403). At this time, a mask region, which is an example of a partial region, is set so that the width of the mask region is wider than the distance that the eyeball moves due to fixation micromotion within the measurement time. A template image is extracted from the extraction region using a portion other than the mask region as an extraction region (second region) (step 404). The method for extracting a template image as a feature region is not particularly limited as long as it is a method for extracting feature points that can be searched and detected between a plurality of images. The template image, the reference position information set in the first image, and the template coordinate information are stored in the recording unit 304 (step 405). Here, the template coordinate is the value of the center coordinate of the template image when the reference position is the origin (0, 0), and means the position information (first position information) of the template image with respect to the reference position. To do. Since the fundus camera continuously captures images for a certain period of time, the next new fundus image (second fundus image) is acquired (step 406). In the process A (step 407), a template image corresponding to or similar to the extracted feature region is searched from the entire acquired second fundus image other than the fundus image from which the feature region has already been extracted. (Hereinafter also referred to as template matching). By the process B, the change in the position of the template image is calculated, and the eye movement for a certain time is calculated (step 408). As described above, the processing from step 406 to step 408 is repeated until the measurement of eye movement is completed (a new image is acquired). In addition, control for displaying the eye movement amount, the image, the measurement time, the real-time monitor image of the anterior eye portion, and the like on the display unit 302 in a specific display form may be performed using a display control unit (step 409). ).

  The process A which is a partial flow (step 407) will be described with reference to FIG. In template matching, a template image stored in the recording unit 304 is read (step 502), and template matching is performed in the newly acquired fundus image (step 503). The template matching means is not limited and can be performed by any known technique. After completion of template matching (step 504), the matching coordinates are stored in the recording unit 304. Here, the matching coordinates are the values of the center coordinates of the matching image when the point on the second image corresponding to the reference position in the first image is the origin (0, 0), and matching with the reference position is performed. It means image position information (second position information).

  Next, process B (step 408) will be described with reference to FIG. First, template coordinates and matching coordinates are read from the recording unit 304 (step 602), a coordinate difference between the template coordinates and the matching coordinates is calculated (step 603), and a movement distance is calculated from the calculated coordinate difference.

(Tracking measurement: specific example)
A specific example in which each of the above processes is performed on the acquired fundus image is shown in FIG. Using the above-described fundus camera 1, the fundus is tracked for 10 seconds under a measurement condition for acquiring a fundus image having a diameter of 10 mm at a cycle of 10 Hz.

First, FIG. 2A shows the first digital fundus image acquired. As shown in FIG. 2A, the blood vessel extends in a complicated manner from the nipple 201 toward the end. After acquiring the first fundus image, an area of Y = 1 mm from the edge of the image is set as the mask area 202 as shown in FIG. A method for setting the width Y of the mask area will be described later. A template is not extracted from the mask region 202, and a template image 203 including at least one of the branching or crossing of blood vessels as a feature portion, which is indicated by a dotted line in FIG. . Here, the template image is a square of 500 μm × 500 μm, but the shape and size of the template image can be arbitrarily determined without being limited to this. In this embodiment, assuming that the center coordinates of the fundus photograph are the origin (0, 0), the center coordinates (hereinafter referred to as template coordinates) of the template image extracted from the first fundus image shown in FIG. ₀ (0, -200). The coordinate unit is μm. Next, the template is searched from FIG. 2E which is the second fundus image to be processed. After the template matching as in FIG. 2 (f), the measured matching coordinates Z ₁ of the matched matching image with the template image. In this example, the matching coordinates _{Z 1} is (0, -400). Figure 2 determined by calculating the change of coordinates as a difference from the template coordinates _{Z 0} matching coordinates _{Z 1} as in (g), eye movement amount (in this embodiment (0μm, -200μm)) is calculated. Similarly, for the third, fourth,... Fundus image diagrams acquired at 10 Hz, the template matching in FIGS. 2E to 2G is repeated, and how much the eyeball moves from the reference position during measurement. It is possible to perform display control to measure whether or not to display in a specific display form.

  An example of the mask area setting method performed in FIG. As shown in FIG. 2B, attribute information such as the size and accuracy of the fundus image, and the distance that the eyeball moves due to microscopic fixation within a measurement time (10 seconds in this embodiment) between capturing all the fundus images. In consideration of the above, a range having a width of Ymm from the peripheral portion of the image is designated as a mask region. This time, the diameter of the image is 10 mm, and the accuracy is about 10 μm. Further, the movement amount of the human eye within the measurement time can be obtained based on a graph showing a function modeled on the relationship between the time from the start of fixation and the movement distance of the human eye. This graph can be a known graph prepared in advance for each imaging condition such as external fixation, internal fixation, affected eye or healthy subject, age, or time to image one fundus image, Any method can be selected for each measurement method and target. Fixation fine movement of a healthy person having this external fixation lamp is about 1 mm in a measurement time of 10 seconds. Therefore, under the above conditions, an area having a width of 1 mm from the peripheral portion of the image is secured as a mask area.

  This mask area is an area that may be out of the measurement range due to the movement of the human eye during measurement. Therefore, by setting the mask area 202 at the time of template extraction as described above, the template image will not deviate from the acquired image range due to the movement of the human eye during measurement. Eye movement can be measured.

[Example 2]
Embodiment 2 of the present invention will be described below.
In the second embodiment, SLO (Scanning Laser Optimalscope) is used for acquiring fundus imaging, and eye movement is measured from the SLO fundus image by the same method as in the first embodiment, and the measurement result of the eye movement is captured as a tomographic image of the fundus. An example in which a high-definition stereoscopic OCT image is acquired by feedback in real time to an optical coherence tomography (OCT) apparatus will be described.

(OCT configuration)
An OCT apparatus was used as an ophthalmic apparatus. An outline of the OCT configuration will be described with reference to FIG.
As the low-coherent light source 715, an SLD light source (Super Luminescent Diode) or an ASE light source (Amplified Spontaneous Emission) can be suitably used. As the suitable wavelength of the low-coherent light, wavelengths near 850 nm and 1050 nm are preferably used for fundus photography. In this embodiment, an SLD light source having a center wavelength of 840 nm and a wavelength half width of 45 nm is used. The low-coherent light emitted from the low-coherent light source enters the fiber coupler 710 via the fiber, and is divided into a measurement light OCT beam) and a reference light. Although an interferometer configuration using a fiber is described here, a configuration using a beam splitter in a spatial light optical system may be used.

  Measurement light (measurement light) is irradiated as parallel light from a fiber collimator 708 via a fiber 709. Further, the measurement light passes through the OCT scanner (Y) 707 and the relay lenses 706 and 705, further passes through the OCT scanner (X) 704, passes through the dichroic beam splitter 703, passes through the scan lens 702, and the eyepiece lens 701, and is covered by 700. Enter the optometry. Here, the OCT scanners (X) 707 and (Y) 704 use galvano scanners. The measurement light incident on the eye e is reflected by the retina and returns to the fiber coupler 710 through the same optical path.

The reference light is guided to the fiber collimator 716 of the fiber couplers 710 to 716 and irradiated as parallel light. The irradiated reference light passes through the dispersion correction glass 717 and is reflected by the reference mirror 719 on the optical path length variable stage 718. The reference light reflected by the reference mirror 719 follows the same optical path and returns to the fiber coupler 710.
The measurement light and the reference light returned by the fiber coupler 710 are combined and guided to the fiber collimator 711. Here, the combined light is called interference light. The fiber collimator 711, the grating 712, the lens 713, and the line sensor 714 constitute a spectrometer. The interference light is measured by the spectrometer as intensity information for each wavelength. The intensity information for each wavelength measured by the line sensor 714 is transferred to a PC (not shown) and is generated as a tomographic image of the eye e.

(SLO configuration)
Next, the optical configuration of the SLO photographing unit that acquires the fundus image will be described with reference to FIG. As the laser light source 729, a semiconductor laser or an SLD light source can be preferably used. The wavelength to be used is not limited as long as it is a light source that can be separated from the wavelength of the low-coherent light source for OCT by the wavelength separation means, but the near-infrared wavelength region of 700 nm to 1000 nm is preferably used as the image quality of the fundus observation image. . In this embodiment, a semiconductor laser having a wavelength of 760 nm is used.

  The laser emitted from the laser light source 729 is emitted as parallel light from the fiber collimator 727 via the fiber 728 and enters the cylinder lens 726. In this embodiment, a cylinder lens is used. However, any optical element capable of generating a line beam is not particularly limited, and a line beam shaper using a Powell lens or a diffractive optical element can be used. A beam (also referred to as an SLO beam) spread by the cylinder lens 726 passes through the center of the ring mirror 723 by the relay lenses 725 and 724, passes through the relay lenses 721 and 722, and is guided to the SLO scanner (Y) 720. The SLO scanner uses a galvano scanner. Further, the light is reflected by the dichroic beam splitter 703, passes through the scan lens 702 and the eyepiece lens 701, and enters the eye 700 to be examined. The dichroic beam splitter 703 is configured to transmit the OCT beam and reflect the SLO beam. The beam incident on the subject's eye is irradiated with a linear beam on the fundus of the subject's eye. This linear beam is reflected or scattered by the fundus of the subject's eye, and follows the same optical path and returns to the ring mirror 723. The position of the ring mirror 723 is conjugate with the pupil position of the eye to be examined. Of the light backscattered by the line beam applied to the fundus, the light passing through the periphery of the pupil is reflected by the ring mirror 723. The image is formed on the line sensor 731 by the lens 730. A planar image of the fundus is generated by a PC (not shown) based on the intensity information for each position of the line sensor. In the present embodiment, the SLO is described in a line scan SLO (hereinafter referred to as L-SLO) configuration using a line beam, but it may be a flying spot SLO as a matter of course.

(Internal fixation light)
In this embodiment, there is an internal fixation lamp that stabilizes the fixation fine movement by gazing at the eye e. The internal fixation lamp of this embodiment will be described with reference to FIG. The light source 752 used for the fixation lamp is a light emitting diode (LD). The lighting position of the light emitting diode is changed in accordance with the part to be imaged under the control of the PC. The light emitting diode 752 has a wavelength of 500 nm, and the beam emitted from the light source is irradiated to the eye e through the lens 751 and the dichroic mirror 750. The dichroic mirror 750 is located between the scan lens 702 and the eyepiece lens 701, and wavelength-separates light having a short wavelength (about 500 nm), OCT beam, and SLO beam (700 nm or more).

(Control method)
FIG. 8 shows a functional system used in this embodiment. The functional system includes a CPU 801 that controls the entire system, each control unit 802, 803, and 808 that controls the fixation lamp and each line sensor camera unit 804 that acquires an SLO image and an OCT image. Reference numeral 805 denotes a PC display unit 806 that displays a system state, and a PC recording unit 807 that records fundus images and imaging conditions. When photographing the fundus, the CPU 801 instructs each of the imaging conditions to the control units 802 and 803 to image the fundus. After the fundus is imaged, images are sent from the camera units 804 and 805 to the CPU 801, processed by the CPU 801, displayed on the display unit 806, and stored in the recording unit 807 simultaneously or thereafter.

  FIG. 9 shows an overall flow of measuring the eye movement while acquiring the tomographic image of the eyeball with the OCT apparatus using the above function.

  First, processing C (mask area measurement) is executed (step 901), and the mask area width is determined so that the width of the mask area becomes wider than the distance that the eyeball moves due to fixation movement within the measurement time (step 901). Step 902). Simultaneously with the above processing, the SLO device is activated to acquire a fundus image by SLO (step 903). For the SLO image, a mask area (first area) is set from the periphery of the first image to the center coordinate direction of the image (step 904), and from the extraction area (second area) other than the mask area A template is extracted (step 905). After template extraction, the template information image and the template coordinates which are the center coordinates of the template image are stored (step 906). The scanning reference position of the OCT apparatus is recorded (step 907), and measurement of the OCT apparatus is started (step 908). After acquiring a new image from the SLO device (step 909), processing A (template matching) (step 910) and processing B (eyeball movement amount calculation) are executed (step 911) and processing D, as in the first embodiment. (Feedback of eye movement amount to OCT) is performed (step 912), and steps 909 to 912 are repeated while the OCT apparatus continuously measures tomographic images (step 913). These steps are preferably repeated in this order. After the OCT imaging is completed, the eye movement measurement is terminated (step 914). Note that the processing A and the processing B are the same as those in the first embodiment, and thus description thereof is omitted.

  Processing C (mask area measurement) which is a partial flow (step 901) will be described with reference to FIG. From the OCT imaging conditions (step 1002), the OCT imaging time during imaging of all fundus images is measured (step 1003). With reference to FIG. 11, which is a graph showing a function modeled on the relationship between the time from the start of fixation and the movement distance of the human eye, the OCT imaging time is adapted to the graph 1101, and the eyeball movement distance (mask region). Is calculated (step 1004). As in Example 1, this graph is a known graph prepared in advance for each imaging condition such as external fixation, internal fixation, affected eye or healthy subject, age, or time for imaging one fundus image. It can be used and can be arbitrarily selected for each measurement method and target. The graph of FIG. 11 is eyeball movement amount information when a healthy eye is measured with a device having an internal fixation lamp.

  Process D (feedback to the OCT apparatus) (step 912) will be described with reference to FIG. The CPU reads the scan position data of the OCT apparatus (step 1202), calculates the voltage to be applied to the OCT scanner from the eyeball movement amount (step 1203), and transfers the input power from the CPU to the OCT controller (step 1204). ) Confirm the scanner movement signal (step 1205), and store the scan position change information (step 1206). Optionally, the change state, OCT image, SLO image (mask area, template position display) remaining time, etc. are displayed (step 1207).

(Tracking measurement: specific example)
FIG. 13 shows a specific example in which each of the above processes is performed on the SLO image acquired when a normal eye is measured with an apparatus having an internal fixation lamp. In the L-SLO, the size of the image at the fundus position is a line width of 10 mm and a scanning range of 10 mm, that is, 10 mm × 10 mm. SLO images can be acquired at 30 Hz. The conditions for acquiring the OCT image are the above-described SD-OCT, the camera is operated at 70 kA-scan, the B-scan image (fundus scan range 10 mm, laser spot diameter 20 μm) is composed of 1000 lines, B -Acquire 280 sheets of retinal stereoscopic images. The measurement time is 4 seconds.

FIG. 13A shows a fundus image acquired by SLO (hereinafter simply referred to as SLO image). As shown in FIG. 13A, the blood vessel extends in a complicated manner from the nipple toward the end. After acquiring the first SLO image, the imaging time of 4 seconds is calculated from the OCT imaging conditions. Referring to FIG. 11, since the eyeball moves 250 μm in 4 seconds, an area of P = 250 μm, which is an example of a predetermined width from the image end (four sides), is set as the mask area 1301 as shown in FIG. 13B. Then, the template is not extracted from the mask area 1301, and a 500 μm × 500 μm template image 1302 as shown by the dotted line in FIG. 13C is extracted. Here, the template image is a square of 500 μm × 500 μm, but the shape and size of the template image can be arbitrarily determined without being limited to this. After template extracting, set as a reference of the movement amount calculating a template coordinates X ₀ in FIG. 13 (d). In this embodiment, when the center coordinate of the SLO image is the origin (0, 0), the template coordinate of the template image is X ₀ (−50, −200). The coordinate unit is μm. Next, the template is matched from FIG. 13 (e) which is the second fundus image to be processed. After the template matching as in FIG. 13 (f), to measure the matching coordinates X ₁ matching image matching. In the second fundus image diagram, the matching coordinates are X ₁ (−50, −400). 13 obtains a change of coordinates from the template coordinates _{X 0} and matching coordinates _{X 1} as in (g), it calculates the amount of movement of the eyeball (0μm, -200μm). The calculated results are reflected on the scanners 707 and 704 of the OCT apparatus via the CPU, and the OCT scanning position is changed. The template matching shown in FIGS. 13E to 13G is repeated to match SLO images acquired at 30 Hz and feed back to the OCT apparatus. During the above processing, as shown in FIG. 14, the OCT image 1501, the eye movement measurement result 1506, the remaining measurement time 1505, the SLO image (including mask area and template image display) 1504, the imaging condition 1503, and the like are displayed on the display 1507. The user may be able to confirm the operation.

  A method for setting the mask area as shown in FIG. An examination time is calculated when an examination is performed using an ophthalmic device (an OCT apparatus according to the present embodiment) different from the fundus imaging unit. From the examination time and the imaging conditions, for example, select the graph of FIG. 11 that matches the imaging conditions such as external fixation, internal fixation, affected eye or healthy person, age, or time for imaging one fundus image. The amount of movement of the eyeball is calculated from the selected graph. The movement amount of the eyeball is set as the width of the mask area P, and the mask area is determined.

  This mask area is an area that may be out of the measurement range due to the movement of the human eye during measurement. Therefore, as described above, by setting the mask area 202 at the time of template extraction according to the OCT imaging time, the template image is not deviated from the acquired image range due to the movement of the human eye during measurement, and template detection is performed. An error can be prevented, and a three-dimensional OCT image with no image shift due to eye movement can be acquired.

(Example 3)
In the first and second embodiments, the region for setting the mask region is not limited to the peripheral portion of the image, and may be set to the central portion of the acquired first fundus image, for example.
At this time, the mask area may be set only in the center of the image, or may be set in both the peripheral and center of the image.

  In order to detect the rotation of the eyeball, a plurality of feature points may be extracted from the first fundus image as shown in FIG. At this time, as shown in the first and second embodiments, a mask region (first region) is set from the peripheral portion of the image as indicated by the shaded portion 1505. Further, a mask area is also set at the center 1506 of the image. Here, the central portion is an area defined by a circle centered on the center of the first fundus image, for example, having a radius larger than the distance that the eyeball moves during the time during which all fundus images are captured. It can be defined as being. In this embodiment, the radius of the circle is set to a value larger than the amount of movement of the eyeball. When each feature point moves beyond the center, rotation, eyeball shift, and eyeball magnification (eye axis) This is because it becomes difficult to separate the movement in the direction). Even if the radius is smaller than the moving amount of the eyeball, it can be dealt with by the conditions at the time of measurement (changing the number, size and extraction method of feature points). Similar to the first and second embodiments, the distance to move the eyeball sac can be obtained from the graph showing the function modeled on the relationship between the time from the start of fixation and the movement distance of the human eye. Similarly, a well-known graph prepared in advance for each imaging condition such as external fixation, internal fixation, affected eye or healthy subject, age, or time for imaging one fundus image can be used. The method can be arbitrarily selected for each measurement method and target.

  When feature points are extracted from this mask region, there is a possibility that the rotation cannot be measured because the coordinate difference between the template coordinates and the matching coordinates due to the rotation of the eyeball does not occur. Therefore, as described above, by setting the mask area 1506 at the time of template extraction according to the imaging time, it is possible to prevent the template from being extracted from the center of the image, and to detect the rotational movement of the eyeball more reliably. It becomes possible to do.

  Then, after determining the mask, the image is divided into four like A, B, C, and D, and feature points 1501 to 1504 are extracted from the divided areas. Thereafter, the movement distances of the feature points 1501 to 1504 are detected in the same manner as in the first and second embodiments, and the turning of the eyeball is calculated from the four points. As described above, it is possible to accurately calculate the rotation of the eyeball by extracting feature points other than the image. Although a circular image as in the first embodiment is used this time, the same can be done with a rectangular fundus image as in the second embodiment.

(Other)
In Examples 1 and 2, extraction was performed using a blood vessel template, but the same effect can be obtained by using a macula or nipple as a template. The fundus image is acquired using a fundus camera and a scanning laser ophthalmoscope (SLO). However, the fundus image may be acquired using an apparatus other than those described in the embodiments. Further, the same effect can be obtained with a visual field inspection apparatus for performing visual field inspection.
The order of the flows shown in the embodiment is not limited to this, and the same effect can be obtained by setting the mask area at the time of template extraction even if the order is different or other flows.

  In the first embodiment, since the external fixation is provided, the mask area is set to 1 mm. However, the width set as the mask area can be appropriately set according to various conditions such as the fundus image capturing condition. For example, in an apparatus equipped with an internal fixation lamp that further stabilizes the fixation, the mask area can be reduced. On the other hand, when the fixation is unstable, such as in an elderly person or an affected eye, the mask area is reduced. It is better to make it wider. When measuring the subject multiple times, a more accurate mask area can be set by using the previous measurement data.

  In the above embodiment, the eye movement is corrected in real time for the ophthalmologic apparatus, but the same effect can be obtained even if correction or post-processing is performed on the obtained image after completion of the measurement.

1: Fundus camera
50: Digital SLR camera
202: Mask area
403: Mask area setting
1301: Mask area

Claims (24)

Image acquisition means for acquiring a plurality of fundus images of the eye to be examined at different times;
Determining means for determining a partial region having a predetermined width from a peripheral portion of at least one fundus image among the plurality of acquired fundus images;
Processing means for extracting at least one feature region from another region different from the determined partial region of the at least one fundus image;
An ophthalmologic apparatus comprising: means for measuring movement of the eye to be examined using the at least one feature region. The processing means includes
Extraction means for extracting a feature region from another region different from the determined partial region;
Search means for searching an area corresponding to the extracted feature area from at least one fundus image other than the fundus image from which the feature area is extracted from the plurality of acquired fundus images;
The ophthalmic apparatus according to claim 1, comprising:   The processing means includes extraction means for extracting a feature region from at least one of the plurality of fundus images, and the processing means further includes at least one other fundus image from the plurality of fundus images. The ophthalmologic apparatus according to claim 1, further comprising a search unit that searches for an image similar to the extracted feature region. The image acquisition means acquires the plurality of fundus images based on return light by measurement light irradiated to the eye to be examined through a scanning means,
The ophthalmologic apparatus according to claim 1, wherein the image acquisition unit includes a control unit that controls the scanning unit based on a positional shift in the plurality of fundus images.   The ophthalmologic apparatus according to claim 4, wherein the processing, the measurement, the control, and the acquisition are performed in this order.   The said determination means determines the partial area | region which has a predetermined | prescribed width | variety from four sides of the at least 1 fundus image among the acquired several fundus images. The ophthalmic device described. The determining means determines a plurality of regions in another region different from the partial region;
The ophthalmologic apparatus according to claim 1, wherein the processing unit extracts a plurality of feature regions from each of the plurality of regions. Image acquisition means for acquiring a plurality of fundus images of the eye to be examined at different times;
Processing means for extracting at least one feature region from another region different from a partial region having a predetermined width from a peripheral portion of at least one fundus image among the plurality of acquired fundus images;
Means for measuring movement of the eye to be examined using the at least one feature region;
An ophthalmologic apparatus comprising:   The processing means extracts at least one feature region from another region different from a partial region having a predetermined width from four sides of at least one fundus image among the plurality of acquired fundus images. The ophthalmic apparatus according to claim 8, wherein the ophthalmic apparatus is characterized.   The ophthalmologic apparatus according to claim 1, wherein the at least one feature region includes at least one image of a branch and intersection of blood vessels. Image acquisition means for acquiring a plurality of fundus images of the eye to be examined at different times;
Determining means for determining a partial region having a predetermined width from a peripheral portion of at least one fundus image among the plurality of acquired fundus images;
Processing means for extracting a feature region from another region different from the partial region in at least two fundus images among the plurality of fundus images, and calculating a difference in coordinates between the extracted feature regions;
Means for measuring the movement of the eye to be examined based on the result of the processing;
An ophthalmologic apparatus comprising:   The ophthalmologic apparatus according to claim 11, wherein the determining unit determines a partial region having a predetermined width from four sides of at least one fundus image among the plurality of fundus images.   The ophthalmologic apparatus according to claim 1, wherein the image acquisition unit includes any one of a fundus camera, a scanning laser ophthalmoscope, an optical coherence tomography apparatus, or a visual field inspection apparatus. The plurality of fundus images includes a plurality of SLO images;
The ophthalmologic apparatus according to claim 1, wherein the ophthalmologic apparatus includes a unit that acquires a plurality of OCT images corresponding to each of the SLO images. The ophthalmologic apparatus according to claim 1, further comprising a display control unit that causes a display unit to display a display form indicating the measured movement of the eye to be examined . The ophthalmologic apparatus according to claim 1, further comprising a display control unit that causes a display unit to display a display form indicating the partial area .   The computer program for functioning a computer as an ophthalmologic apparatus of any one of Claims 1-16. Determining means for determining a partial region having a predetermined width from a peripheral portion of at least one fundus image among a plurality of fundus images of the eye to be examined acquired at different times;
Processing means for extracting at least one feature region from another region different from the determined partial region of the at least one fundus image;
Means for measuring movement of the eye to be examined using the at least one feature region;
A device for measuring the movement of the eye to be examined. Processing means for extracting at least one feature region from another region different from a partial region having a predetermined width from the peripheral portion of at least one fundus image among a plurality of fundus images of the eye to be examined acquired at different times When,
Means for measuring movement of the eye to be examined using the at least one feature region;
A device for measuring the movement of the eye to be examined. Determining means for determining a partial region having a predetermined width from a peripheral portion of at least one fundus image among a plurality of fundus images of the eye to be examined acquired at different times;
Processing means for extracting a feature region from another region different from the partial region in at least two fundus images among the plurality of fundus images, and calculating a difference in coordinates between the extracted feature regions;
Means for measuring the movement of the eye to be examined based on the result of the processing;
A device for measuring the movement of the eye to be examined. Determining a partial region having a predetermined width from the periphery of at least one fundus image among a plurality of fundus images of the eye to be examined acquired at different times ;
Extracting at least one feature region from another region different from the determined partial region of the at least one fundus image;
Measuring the movement of the eye to be examined using the at least one feature region;
A method for measuring movement of an eye to be examined, comprising: Extracting at least one feature region from another region different from a partial region having a predetermined width from a peripheral portion of at least one fundus image among a plurality of fundus images of the eye to be examined acquired at different times; and ,
Measuring the movement of the eye to be examined using the at least one feature region;
A method for measuring movement of an eye to be examined, comprising: Determining a partial region having a predetermined width from the periphery of at least one fundus image among a plurality of fundus images of the eye to be examined acquired at different times ;
Extracting a feature region from another region different from the partial region in at least two fundus images among the plurality of fundus images, and calculating a coordinate difference between the extracted feature regions. A process of processing;
Measuring the movement of the eye to be examined based on the result of the processing;
A method for measuring movement of an eye to be examined, comprising: A program characterized by executing the steps of the method according to the computer in any one of claims 21 ~ 2.

Images