JP5842477B2 - Corneal endothelial cell imaging device - Google Patents

Description

  The present invention relates to a corneal endothelial cell imaging apparatus that images a cell image of a corneal endothelium of a subject's eye.

  2. Description of the Related Art Conventionally, there has been known an apparatus that irradiates illumination light from an illumination light source obliquely toward a cornea, receives a reflected light beam from the cornea by an image sensor, and obtains a cell image of the corneal endothelium without contact. As such a device, as a method for evaluating an acquired captured image, a device that selectively extracts a high degree of focus of each captured image stored in a memory (see Patent Document 1) is known. . An apparatus that evaluates an image based on pixel information on a line or a plurality of lines (for example, five horizontal lines) extending in the horizontal direction of a captured image is known (see Patent Document 2).

Japanese Patent Laid-Open No. 10-113335 JP 2008-1173779 A

  However, in the case of the apparatus of Patent Document 1, it is not specifically described how to extract a good quality captured image from each captured image stored in the memory. The apparatus of Patent Document 2 is configured to evaluate a captured image before being stored in the memory, and image evaluation is performed from the viewpoint of how to reduce useless captured images stored in the memory.

  That is, an apparatus capable of selecting a good captured image suitable for analysis is desired.

  In view of the above problems, an object of the present invention is to provide an endothelial cell imaging device capable of selecting a high-quality captured image suitable for analysis from captured images stored in a memory.

  In order to solve the above problems, the present invention is characterized by having the following configuration.

(1)
An illumination optical system that irradiates illumination light from an illumination light source obliquely toward the subject's eye cornea, imaging optical that receives reflected light from the cornea including corneal endothelial cells by an imaging element and acquires a corneal endothelial cell image An apparatus body comprising a system,
Driving means for moving the device main body relative to the eye of the subject;
An imaging control unit that controls driving of the driving unit to move the apparatus main body in a predetermined direction, and causes the illumination light source to emit light a plurality of times during the movement of the apparatus main body to acquire a plurality of captured images by the imaging element. When,
Storage means for storing a plurality of captured images captured by the imaging element;
When calculating an evaluation value as an analysis target of each captured image stored in the storage unit, in the captured image, the edge of the luminance distribution in the other direction at each position in either the horizontal direction or the vertical direction. An evaluation value calculating means for detecting a number and calculating the sum of the number of edges at each position as the evaluation value;
It is characterized by providing.

  An embodiment of the present invention will be described with reference to the drawings. FIG. 1 is an external side configuration diagram of a corneal endothelial cell imaging apparatus according to the present embodiment.

  The device 100 is a so-called stationary device, and is a base 1, a face support unit 2 attached to the base 1, and a movable base provided so as to be movable on the base 1 by a sliding mechanism not shown. 3 and a photographing unit (device main body) 4 that is provided so as to be movable with respect to the movable table 3 and accommodates a photographing system and an optical system described later.

  The imaging unit 4 is moved in the left-right direction (X direction), the up-down direction (Y direction), and the front-rear direction (Z direction) with respect to the eye E by an XYZ driving unit 6 provided on the moving table 3. The movable table 3 is moved in the XZ direction on the base 1 by operating the joystick 5. Further, when the examiner rotates the rotary knob 5 a, the photographing unit 4 is moved in the Y direction by the Y drive of the XYZ drive unit 6. A start switch 5 b is provided on the top of the joystick 5. The display monitor 95 is provided on the examiner side of the imaging unit 4. In this embodiment, the photographing unit 4 is moved relative to the eye E by a sliding mechanism (not shown) or the XYZ driving unit 6.

  FIG. 2 is a schematic configuration diagram illustrating an example of an optical arrangement and a control system configuration when the optical system housed in the photographing unit 4 is viewed from above. FIG. 3 is a view of the first projection optical system and the second projection optical system as viewed from the subject side. The overall configuration of the optical system is such that the illumination optical system 10 that irradiates illumination light from the illumination light source 12 toward the cornea Ec obliquely and the reflected light from the cornea Ec including endothelial cells is received by the first imaging device 44. Imaging optical system 30 for acquiring endothelial cell images, front projection optical system 50 for projecting an alignment index from the front toward the center of the cornea Ec, and projection of an alignment index at infinity from the oblique toward the cornea Ec. 1 projection optical system 60a, 60b, and second projection optical systems 65a to 65d (see FIG. 3) for projecting alignment indices at a finite distance from the plurality of oblique directions toward the periphery of the cornea Ec, respectively, and the interior of the eye E The internal fixation optical system 70 for projecting a fixation target from the front, the anterior segment observation optical system 80 for observing the anterior segment image from the front, and the alignment state of the photographing unit 4 with respect to the eye E in the Z direction. Having a Z alignment detection optical system 85 for output, a. Hereinafter, individual specific configurations will be described.

  The illumination optical system 10 includes an illumination light source (eg, visible LED, flash lamp) 12 that emits visible light for endothelium imaging, a condensing lens 14, a slit plate 16, a dichroic mirror 18 that reflects and transmits infrared light, and light projection. A lens 20. The light emitted from the illumination light source 12 illuminates the slit plate 16 via the condenser lens 14. Then, the slit light that has passed through the slit plate 16 is converged by the light projecting lens 20 via the dichroic mirror 18 and irradiated onto the cornea. Here, the slit plate 16 and the cornea Ec are disposed at a substantially conjugate position with respect to the objective lens 20.

  The imaging optical system 30 is symmetrical with the illumination optical system 10 with respect to the optical axis L1, and includes an objective lens 32, a visible light reflecting / infrared transmitting dichroic mirror 34, a mask 35, a first imaging lens 36, and a total reflection mirror 38. The second imaging lens 42 and a dedicated first two-dimensional imaging device 44 (for example, a two-dimensional CCD, CMOS, etc.) 44 for acquiring an endothelial cell image. The mask 35 is disposed at a position substantially conjugate with the cornea Ec with respect to the objective lens 32. The first imaging lens 36 and the second imaging lens 42 form an imaging optical system that forms an endothelial image on the image sensor 44. The imaging element 44 is disposed at a position substantially conjugate with the cornea Ec with respect to the lens system of the imaging optical system 30.

  The cornea-reflected light from the illumination optical system 10 is directed in the direction of the optical axis L3 (oblique direction), converged by the objective lens 32, reflected by the dichroic mirror 34, temporarily imaged by the mask 35, and an endothelial cell image is obtained. Light that becomes noise during acquisition is blocked. Then, the light that has passed through the mask 35 is imaged on the two-dimensional imaging device 44 via the first imaging lens 36, the total reflection mirror 38, and the second imaging lens 42. Thereby, a high-magnification corneal endothelial cell image is acquired. The output of the image sensor 44 is connected to the control unit 90, and the acquired cell image is stored in the memory 92. The cell image is displayed on the monitor 95.

  The front projection optical system 50 includes an infrared light source 51, a light projection lens 53, and a half mirror 55, and projects infrared light for XY alignment detection onto the cornea Ec from the direction of the observation optical axis L1. Infrared light emitted from the light source 51 is converted into a parallel light beam by the light projecting lens 53, then reflected by the half mirror 55, and projected onto the center of the cornea Ec to form an index i10 (see FIG. 4). ).

  The first projection optical systems 60a and 60b are disposed so as to be inclined at a predetermined angle with respect to the optical axis L1. The first projection optical systems 60a and 60b have infrared light sources 61a and 61b and collimator lenses 63a and 63b, respectively, are arranged symmetrically with respect to the optical axis L1, and are infinite with respect to the eye E. An index is projected (see FIG. 2). The first projection optical systems 60a and 60b are disposed on substantially the same meridian as the horizontal direction passing through the optical axis L1 (see FIG. 3).

  The lights emitted from the light sources 61a and 61b are collimated by the collimator lenses 63a and 63b, respectively, and then projected onto the cornea Ec to form the indices i20 and i30 (see FIG. 4).

  The second projection optical systems 65a to 65d are arranged to be inclined with respect to the optical axis L1. The second projection optical systems 65a to 65d have infrared light sources 66a to 66d, are arranged symmetrically with respect to the optical axis L1, and project a finite index onto the eye E. The second projection optical systems 65a and 65b are disposed above the optical axis L1 and are disposed at the same height with respect to the Y direction. The second projection optical systems 65c and 65d are disposed below the optical axis L1 and are disposed at the same height with respect to the Y direction. The second projection optical systems 65a and 65b and the second projection optical systems 65c and 65d are arranged in a vertically symmetrical relationship with the optical axis L1 in between.

  Here, the light from the light sources 66a and 66b is irradiated obliquely upward toward the upper part of the cornea Ec, and indexes i40 and i50 which are virtual images of the light sources 66a and 66b are formed. In addition, light from the light sources 66c and 66d is irradiated obliquely downward toward the lower portion of the cornea Ec, and indexes i60 and i70 that are virtual images of the light sources 66c and 66d are formed (see FIG. 4).

  According to the index projection optical system as described above, the index i10 is formed at the corneal apex of the eye E (see FIG. 4). In addition, the indices i20 and i30 by the first projection optical systems 60a and 60b are formed symmetrically with respect to the index i10 at the same horizontal position as the index i10. Further, the indices i40 and i50 by the second projection optical systems 65a and 65b are formed symmetrically with respect to the index i10 above the index i10. The indices i60 and i70 by the second projection optical systems 65c and 65d are formed symmetrically with respect to the index i10 below the index i10.

  The internal fixation optical system 70 includes a visible light source (fixation lamp) 71, a light projecting lens 73, and a visible reflection / infrared transmission dichroic mirror 74, and emits light for fixing the eye E in the front direction. Project to eye E. Visible light emitted from the light source 71 is converted into a parallel light beam by the light projection lens 73, reflected by the dichroic mirror 75, and projected onto the fundus of the eye E. An external fixation optical system (not shown) is disposed in the vicinity of the first projection optical system and the second projection optical system.

  Returning to FIG. The anterior ocular segment observation optical system 80 has an objective lens 82 and a two-dimensional image sensor 84 for acquiring an anterior ocular segment front image, and has a second image sensor 84 different from the first image sensor 44. Then, the anterior segment image and the alignment index are captured by the second image sensor 84. As the two-dimensional imaging device 84, for example, a two-dimensional CCD image sensor (Charge coupled device image sensor) or a two-dimensional CMOS (Complementary Metal Oxide Semiconductor Image Sensor) is used.

  The anterior segment illuminated by an anterior segment illumination light source (not shown) is imaged by the two-dimensional imaging device 84 via the dichroic mirror 75, the half mirror 55, and the objective lens 82. Similarly, the corneal reflection images by the front projection optical system 50, the first projection optical systems 60a and 60b, and the second projection optical systems 65a to 65d are received by the two-dimensional imaging device 84.

  The output of the image sensor 84 is connected to the control unit 90, and the anterior segment image captured by the image sensor 84 is displayed on the monitor 95 as shown in FIG. Note that the reticle LT displayed electronically on the monitor 95 indicates a reference for XY alignment. Note that the observation optical system 80 also serves as a detection optical system for detecting the alignment state of the imaging unit 4 with respect to the eye E (alignment deviation direction / deviation amount).

  The Z alignment detection optical system 85 includes a light projecting optical system 85a that projects a detection light beam from an oblique direction toward the cornea Ec, and a light receiving optical system 85b that receives a corneal reflected light beam by the light projecting optical system 85a. . The optical axis L2 of the light projecting optical system 85a and the optical axis L3 of the light receiving optical system 85b are arranged at positions symmetrical with respect to the observation optical axis L1.

  The light projecting optical system 85a includes, for example, an illumination light source 86 that emits infrared light, a condensing lens 87, a pinhole plate 88, and a lens 20. Here, the pinhole plate 88 and the cornea Ec are disposed at a substantially conjugate position with respect to the lens 20. The light receiving optical system 85 b includes, for example, a lens 32 and a one-dimensional light receiving element (line sensor) 89. Here, the one-dimensional light receiving element 89 and the cornea Ec are disposed at a substantially conjugate position with respect to the lens 32.

  The infrared light emitted from the light source 86 illuminates the pinhole plate 88 through the condenser lens 87. Then, the light that has passed through the opening of the pinhole plate 88 is projected onto the cornea Ec through the lens 20. The corneal reflection light is received by the light receiving element 89 via the lens 32 and the dichroic mirror 34.

  The output of the light receiving element 89 is connected to the control unit 90 and used for Z alignment detection with respect to the eye E. Here, the light receiving position of the alignment light flux received on the light receiving element 89 is changed depending on the positional relationship between the imaging unit 4 and the eye E in the Z direction. For example, the control unit 90 detects the position of the corneal reflected light in the detection signal from the light receiving element 89 and detects the alignment state in the Z direction. The alignment detection using the light receiving element 89 is used for precise alignment with the eye E.

  The control unit 90 controls the entire apparatus. The control unit 90 is connected to the rotary knob 5a, the start switch 5b, the XYZ driving unit 6, the two-dimensional imaging devices 44 and 84, each light source, a memory 92 as a storage unit, and a monitor 95.

  For example, the control unit 90 controls the display on the monitor 95. In addition, the control unit 90 detects the alignment state of the imaging unit 4 with respect to the eye E in the XYZ directions based on the light reception result of the alignment index. And the control part 90 outputs the signal which instruct | indicates the movement of the imaging | photography part 4 based on the detection result. Further, the control unit 90 detects the alignment state of the photographing unit 4 in the Z direction with respect to the eye E based on the light reception result of the light receiving element 89.

The operation of the apparatus having the above configuration will be described. In the apparatus 100, after performing XYZ alignment, a plurality of endothelial cells are continuously photographed while the imaging unit 4 is advanced. Hereinafter, each operation will be described in detail.
<XYZ alignment>
4A and 4B are diagrams showing an example of an anterior ocular segment observation screen when imaging the inner skin of the cornea, FIG. 4A is a display example when there is misalignment, and FIG. 4B is a state where alignment is appropriate Is a display example.
In this case, the light source 71 is turned on and the fixation direction of the eye E is guided to the front. First, the examiner causes the subject to gaze at the fixation target. Further, the imaging unit 4 is aligned with the eye E while observing the anterior segment image displayed on the monitor 95.

  When rough alignment (manual alignment by the examiner until the indicators i40, i50, i60, and i70 are displayed on the monitor 95) is performed as described above, as shown in FIG. An image is detected on the light receiving surface of the image sensor 64. The control unit 90 searches for a bright spot from the upper left coordinate position of the image toward the lower right side of the screen. When the indices i40, i50, i60, i70 are detected, the control unit 90 detects the position of the detected bright spot.

  Then, the control unit 90 detects the center position of the rectangle including the indices i40, i50, i60, and i70 as a substantially corneal apex, and detects the misalignment direction / deviation amount of the imaging unit 4 with respect to the eye E in the XY directions. Then, the control unit 90 controls the driving of the driving unit 6 and moves the photographing unit 4 in the XY directions so that the alignment deviation falls within a predetermined alignment allowable range (for example, a range in which the index i10 is detected). This enables automatic alignment over a wide range.

  When the photographing unit 4 is moved and the index i10 is detected as described above, the control unit 90 ends the alignment using the indices i40 to i70 and performs the alignment using the index i10. Here, the control unit 90 determines the index i10 and the indices i40 to i70 from the positional relationship.

  Then, the control unit 90 detects the coordinate position of the index i10 as a substantially corneal apex, and detects the misalignment direction / deviation amount of the imaging unit 4 with respect to the eye E in the XY directions. Then, the control unit 90 controls the driving of the driving unit 6 and moves the imaging unit 4 in the XY direction so that the alignment deviation falls within a predetermined alignment allowable range (for example, the index i10 is located in the reticle LT). Let

  Further, when the index i10 is detected as described above, the indices i20 and i30 at infinity are similarly detected. Therefore, the control unit 90 compares the interval between the infinite indices i20 and i30 detected as described above and the distance between the finite indices i60 and i70 to determine the alignment deviation direction / deviation amount in the Z direction. Obtained (first alignment detection). Then, the control unit 90 determines that the alignment deviation of the imaging unit 4 with respect to the eye E in the Z direction is within a predetermined alignment allowable range (for example, the central luminescent spot of the infinite indices i20 and i30 and the central luminescent spot of the finite indices i60 and i70 The photographing unit 4 is moved in the Z direction so that the difference from the point is within plus or minus one pixel (dimension is about ± 0.1 mm) (first automatic alignment).

  In this case, when the measuring unit 3 is displaced in the working distance direction, the control unit 90 does not change the interval between the infinity indices i20 and i30 described above, whereas the image of the finite index images i60 and i70. Using the characteristic that the interval changes, the misalignment in the Z direction is obtained (for details, refer to Japanese Patent Laid-Open No. 6-46999). Note that index images i40 and i50 may be used instead of the index images i60 and i70. Also, the Z alignment may be detected based on the index distance (index height) from the optical axis L1.

  Then, when it is determined that the alignment state is appropriate in the first Z alignment detection, the control unit 90 stops the operation of the first automatic alignment and performs the second Z alignment detection using the detection optical system 85 and the detection thereof. A second automatic alignment based on the result is activated.

  The controller 90 turns on the light source 86 and projects an alignment light beam onto the cornea Ec (the light source 86 may be turned on in advance), and detects the corneal reflected light beam with the light receiving element 89. Then, the control unit 90 controls the driving of the driving unit 6 based on the light reception result from the light receiving element 89 and moves the photographing unit 4 in the Z direction.

  For example, the control unit 90 detects the peak P corresponding to the reflected light beam from the corneal epithelium based on the light reception signal output from the light receiving element 89, and detects the position Pz of the epithelial peak on the light receiving element 89 (FIG. 5). reference). Then, the control unit 90 drives the drive unit 6 so that the peak of the light reception signal due to the reflected light beam from the epithelium is at a predetermined position (for example, the center position) on the light receiving element 89.

  If the alignment state in the XYZ directions satisfies the alignment completion condition by the alignment operation described above, the control unit 90 determines that the alignment in the XYZ directions is matched, and issues a trigger signal.

<Photographing endothelial cells>
When the trigger signal is generated, the control unit 90 continuously turns on the illumination light source 12 and acquires a corneal endothelial cell image by visible illumination light with the two-dimensional imaging device 44. At this time, it is preferable that the control unit 90 causes the light source 12 to emit light with an amount of light such that epithelial reflected light is detected and endothelial reflected light is not detected. Thereafter, the control unit 90 turns on the light source 12 and controls driving of the driving unit 6 to advance the photographing unit 4 toward the eye E. During the movement of the photographing unit 4 in the Z direction, the control unit 90 continues the automatic alignment operation (tracking control using the image sensor 84) in the XY directions.
The control unit 90 detects an output image from the image sensor 44 and controls the light source 12 and the drive unit 6 based on the detection result. FIG. 6 is a diagram illustrating an example of determining the light reception state of the corneal image based on the output image from the image sensor 44. In FIG. 6, the central white rectangular area corresponds to the opening of the mask 35 disposed in front of the image sensor 44, and the black hatches on the left and right correspond to the light shielding part of the mask 35.

  For example, the control unit 90 sets a first detection region Lc1 and a second detection region Lc2 that extend in a direction orthogonal to the thickness direction of the cornea (the Z direction in FIG. 6) in order to detect the light reception state of the cornea image. . The first detection region Lc1 is set for detecting the light receiving state of epithelial reflected light, and the second detection region Lc2 is set for detecting the light receiving state of endothelial reflected light. The control unit 90 calculates the total luminance value SLC1 of each pixel in the first detection region Lc1. In addition, the total luminance value SLC2 of each pixel in the second detection region Lc2 is calculated.

  7A to 7C are diagrams showing changes in the light reception state of the corneal reflection light when the imaging unit 4 is advanced, and FIGS. 8A and 8B are total values SLC1 and SLC2 when the imaging unit 4 is advanced. It is a graph showing the change of chronologically. 8A corresponds to the total value SLC1, and FIG. 8B corresponds to the total value SLC2.

  FIG. 7A is a diagram when alignment in the XYZ directions is completed. At this time, the epithelial reflected light Ep is received on the first region Lc1. For this reason, as the first total value SLC1, a high value corresponding to epithelial reflection is calculated (see FIG. 8A).

  And if the imaging | photography part 4 is advanced, epithelial reflected light Ep will be moved to the right direction of the paper surface of FIG. When the epithelial reflected light Ep passes the detection region Lc1, the total value SLC1 is greatly reduced (see the slope A in FIGS. 7B and 8A). Then, when the total value SLC1 falls below the predetermined threshold value S1, the control unit 90 increases the light amount of the light source 12 to the extent that the endothelial image appears in the output image from the image sensor 44. Thereby, the endothelial reflected light En can be detected by the image sensor 44.

  After the light amount of the light source 12 is increased, the control unit 90 continues the forward operation of the imaging unit 4 and stores the images continuously output from the imaging element 44 in the memory 92 as needed. The two-dimensional imaging device 44 outputs an imaging signal to the control unit 90 as needed in accordance with the frame rate. Thereby, a plurality (for example, about 30 to 40) of captured images of the endothelium are acquired in 1 to 2 seconds. And the control part 90 memorize | stores in the memory 92 the image which satisfy | fills certain conditions (for example, an endothelial cell image is acquired appropriately) as a still image among output images. Thereby, an endothelial cell image is photographed. In this case, the controller 90 may store a predetermined number set in advance in the memory 92. Then, the control unit 90 outputs the captured image stored in the memory 92 to the monitor 95.

  When the imaging unit 4 is moved forward, the endothelial reflected light En is moved to the right in the image (see FIGS. 7A to 7C). When the endothelial reflected light En reaches the second detection region Lc2, the total value SLC2 increases (see the slope B in FIG. 8B). A high value is positioned while the endothelial reflected light En is received on the second detection region Lc2. Further, when the imaging unit 4 is moved forward and the endothelial reflected light En passes the detection region Lc2, the total value SLC2 is greatly reduced (see the slope C in FIGS. 7C and 8B). When the total value falls below the predetermined threshold value S2, the control unit 90 dims the light source 12 (including turning off the light), stops the driving of the driving unit 6, and stops the forward movement operation of the photographing unit 4.

  In addition, as a method of causing the light source 12 to emit light continuously, a method of continuously flashing the light source 12 is included in addition to a method of always lighting the light source 12. When the light source 12 is continuously blinked, for example, the control unit 90 blinks so that a plurality of endothelial images can be acquired while the imaging unit 4 is moving. Further, the light source 12 may blink continuously in synchronization with the frame rate of the two-dimensional image sensor 44. For example, when the imaging time of one sheet is 30 ms, the light source 12 is turned on for several ms from the start of image acquisition, and then turned off. Then, when the acquisition of the next image is started, the light source 12 is turned on. That is, such a blinking operation is repeated.

  The control unit 90 is not limited to these, and may be any control (including continuous light emission of course) that causes the light source 12 to emit light a plurality of times so that a plurality of endothelial images can be obtained by the imaging device 44.

<Tracking control using a two-dimensional image sensor for anterior segment observation during imaging>
When acquiring endothelial cell images continuously as described above, the eye E moves in the middle of the process, and an appropriate cell image may not be obtained. Therefore, the control unit 90 detects the alignment state of the imaging unit 4 with respect to the eye E based on the captured image output from the imaging element 84 during acquisition of the captured image.

  When the alignment detection result deviates from the alignment allowable range (for example, when the deviation is 0.25 mm or more in the X direction and 0.5 mm or more in the Y direction), the control unit 90 returns the photographing unit 4 to the proper alignment position. The driving of the driving unit 6 is controlled. That is, the control unit 90 performs automatic alignment using the imaging device 84 for anterior segment observation.

  For example, even after the light source 12 is turned on, the control unit 90 detects the index i10 in the captured image and detects an alignment shift in the XY directions. Then, the control unit 90 controls the driving of the driving unit 6 when the detection result is out of the predetermined alignment allowable range, and moves the photographing unit 4 in the XY directions so that the alignment deviation falls within the predetermined alignment allowable range. Move. Thereby, even if the eye E shifts from the proper alignment position, the photographing unit 4 and the eye E are smoothly returned to the proper positional relationship. That is, automatic tracking control during continuous acquisition of cell images is possible.

  In addition, since the index i10 is formed by a parallel light beam from the front direction, the image sensor 84 does not receive light when the alignment is greatly deviated. Therefore, the control unit 90 detects the indices i40, i50, i60, i70 in the captured image, and detects an alignment shift in the XY directions. Then, the control unit 90 controls the driving of the driving unit 6 and moves the photographing unit 4 in the XY directions so that the alignment deviation falls within a predetermined alignment allowable range. As a result, even if the eye E deviates significantly from the proper alignment position and cannot be restored when automatic alignment is performed using a position sensor as in the past, the photographing unit 4 and the eye E It returns smoothly to the proper positional relationship. That is, automatic tracking control over a wide range during continuous acquisition of cell images becomes possible.

As described above, when XY alignment is performed using the two-dimensional imaging device 84 for observing the anterior segment, automatic tracking control over a wide range is possible, but the return time is slower than when the position sensor is used. Therefore, there is a possibility that a predetermined number (for example, about 30 to 40) of endothelial cells cannot be imaged within a predetermined time (for example, about 1 to 2 seconds). Therefore, the return time to the proper alignment position was examined. Specifically, the time until the central bright spot of the indices i40, i50, i60, i70 was shifted from the appropriate position by 3.2 mm (detection limit range of the position sensor) and returned to the appropriate position was measured. The return time is 0.62 seconds on average, and even when the eye E is greatly deviated from the alignment proper position, a predetermined number (for example, about 30 to 40 sheets) within a predetermined time (for example, about 1 to 2 seconds). ) It was confirmed that the endothelial cells can be sufficiently photographed. As a result, even if the eye E is greatly out of the proper alignment position and it has been necessary to take a long time to shoot the endothelial cells in the past, it is not necessary to retake the image and the image is taken in a short time. Is completed, the burden on the subject can be reduced.
When the above-described automatic tracking control is performed, the control unit 90 temporarily stops the forward movement of the photographing unit 4 and resumes the forward movement of the photographing unit 4 after returning to the appropriate position. Cell images are continuously acquired using the image sensor 44 until the position of the endothelial image reaches a predetermined photographing end position. That is, the acquisition of the cell image output from the image sensor 44 is interrupted while the alignment return control in the XY directions is performed (until the apparatus main body is returned to the proper alignment position). Thereby, only a good endothelial cell image can be acquired.

  Further, the control unit 90 may detect the vignetting of the light beam caused by wrinkles and wrinkles using the index i40 and the index i50 formed on the upper cornea, and determine the open state based on the detection result. For example, when at least one of the index i40 and the index i50 disappears from the image sensor 84, or when a part of the index i40 or the index i50 is detected, the control unit 90 performs illumination light and reflected light for acquiring corneal endothelium. It is determined that there is a possibility that it will be blocked by 瞼 / 瞼, and the open state is determined to be insufficient. Thereby, it is determined whether the open state of the eye E is appropriate during the continuous acquisition of cell images. This determination result is associated with a cell image acquired as needed, and is used, for example, for selection of cell images.

  In the return control, in the case of the anterior ocular segment observation optical system 80, at least the pupil, iris, and sclera (preferably eyelids, eyelids) of the eye E are imaged when the corneal center of the eye E and the optical axis L1 coincide. It has a wide imaging range (for example, 11 mm long × 15 mm wide) that is included in the range. In other words, the imaging range is large compared to the imaging range (for example, 3.2 mm x 3.2 mm) of the position sensor used as a conventional alignment detection sensor.

  Therefore, by detecting the alignment using the image sensor 84 of the observation optical system 80 simultaneously with the continuous acquisition of the endothelium image using a dedicated image sensor, even if the alignment greatly deviates beyond the detection range of the position sensor. Automatic alignment is activated. For this reason, an endothelial cell image can be reliably acquired even for an eye whose fixation is not stable.

  By detecting the alignment index based on the captured image obtained by the two-dimensional image sensor 84, the alignment index is acquired as two-dimensional image data. Therefore, even when a plurality of point-like indicators are projected as shown in FIG. 3 or when a two-dimensional pattern image such as a ring-like indicator or an on-line indicator is projected onto the eye E, each indicator image is obtained by image processing. Identification and alignment detection using each index image are easily performed. For example, the index image is discriminated from the positional relationship among a plurality of indices, the shape of the index pattern, and the like. Further, the open state can be detected. In addition, it is easy to distinguish between alignment bright spots and ambient light.

<Automatic tracking control in the Z direction using the light receiving element 89 during imaging>
During the continuous acquisition of cell images, if the alignment is greatly shifted in the XY direction, the alignment in the Z direction may be greatly shifted. Therefore, the control unit 90 may perform alignment return control in the Z direction together with XY automatic alignment. In this case, acquisition of the cell image output from the image sensor 44 is interrupted while the alignment return control in the Z direction is performed (until the apparatus main body is returned to the proper alignment position). Thereby, only a good endothelial cell image can be acquired.

  For example, the control unit 90 acquires the alignment position information before the alignment state of the imaging unit 4 with respect to the eye E is out of the allowable range based on the light reception result of the light receiving element 89, and returns the imaging unit 4 to the original alignment position. Thus, the drive of the drive unit 6 is controlled.

  Here, the control unit 90 stores in the memory 92 the alignment position in the Z direction before the alignment deviation in the XY direction deviates from the allowable range, and temporarily stops the forward movement operation by the driving unit 6. Then, the control unit 90 drives the driving unit 6 based on the position information stored in the memory 92 at the same time as the XY automatic alignment or after the XY alignment is completed, and takes the imaging unit toward the alignment position before the alignment is deviated. 4 is moved in the Z direction.

  For example, the control unit 90 detects the position of the alignment index based on the imaging signal output from the imaging element 84 during cell image acquisition, and detects the position of the epithelial peak on the light receiving element 89. At this time, the epithelial peak moves on the light receiving element 89 in accordance with the forward movement of the imaging unit 4.

  Here, when the eye E moves and the XY misalignment detected by the image sensor 84 deviates from the allowable range, the control unit 90 stores the epithelial peak position before (preferably immediately before) deviating from the allowable range. 92. In this case, it may be a pixel position where a peak is detected, or a deviation amount between a predetermined position on the light receiving element 89 and the peak detection position.

  When the return operation is performed, the control unit 90 controls the drive unit 6 so that the XY alignment deviation is within the allowable range as described above to move the imaging unit 4 in the XY direction, and the epithelium on the light receiving element 89. The photographing unit 4 is moved in Z so that the peak is detected at the position (or the vicinity thereof) stored in the memory 92.

  When the return operation is completed, the control unit 90 resumes the forward operation of the imaging unit 4 and uses the imaging element 44 until the position of the epithelial peak on the light receiving element 89 reaches a predetermined imaging end position. Is acquired continuously.

  The control unit 90 may turn off the light source 12 once when the misalignment is out of the allowable range, and turn on the light source 12 again when it is detected that the misalignment has reached the allowable range. . In this way, the burden on the subject's eye due to the visible light source is reduced.

  In the above description, the timing of misalignment is detected using the image sensor 84, but the present invention is not limited to this. When the eye E moves, the change of the peak position on the light receiving element 89 is different from the advance of the photographing unit 4 described above.

  Therefore, for example, the control unit 90 may detect the amount of change in the peak position per unit time as needed, and may determine that an XY direction misalignment has occurred when the amount of change deviates from a predetermined range. Then, the control unit 90 may perform return control of the imaging unit 4 with respect to the eye E based on the peak position on the light receiving element 89 before the deviation occurs.

  For the alignment detection by the light receiving element 89, a peak corresponding to the endothelial reflection may be detected. Further, in addition to the return control using the light receiving element 89, the control unit 90 may activate automatic alignment in the Z direction using the imaging element 84 when the Z alignment greatly deviates.

  Moreover, although it was set as the structure which advances the imaging | photography part 4 when acquiring a cell image continuously, it is not limited to this. The apparatus moves the photographing unit 4 in a predetermined direction, causes the light source 12 to emit light a plurality of times while the photographing unit 4 is moving, and obtains a plurality of captured images by the imaging element 44. For example, a captured image is acquired while the imaging unit 4 is moved backward. Further, a plurality of captured images may be acquired during the movement of the photographing unit 4 in the direction orthogonal to the optical axis L1.

  In the above configuration, the alignment state is detected using the alignment index. However, the alignment may be detected based on the imaging signal output from the imaging device. For example, the control unit 90 may extract feature parts (for example, pupils and irises) in the anterior ocular segment image by image processing, and detect a positional shift using the extracted feature parts.

  In the above configuration, the same two-dimensional imaging device is used for anterior ocular segment observation and alignment detection. However, the present invention is not limited to this, and a two-dimensional imaging element for anterior ocular segment observation and alignment detection may be provided separately. For example, a light dividing member (for example, a half mirror) is disposed between the lens 82 and the image sensor 84, and a two-dimensional image sensor for detecting alignment is disposed in the reflection direction of the light dividing member.

<First Example Regarding Selection of Captured Image Used for Analysis>
In general, the control unit 90 calculates an evaluation value as an analysis target of each captured image stored in the memory 92 based on a two-dimensional luminance distribution in the captured image, and the control unit 90 calculates an evaluation value. Based on the result, an analysis candidate for the corneal endothelium of the eye E is selected from each captured image stored in the memory 92. Further, the control unit 90 sorts the plurality of captured images based on the calculation result of the evaluation value, and displays the sorted captured images on the monitor 95 in the upper order. In the first embodiment, the functions realized by the control unit 90 are divided into calculation of evaluation values, selection of captured images, and sorting of a plurality of captured images.

  For example, the control unit 90 processes a plurality of captured images stored in the memory 92 and performs selection processing of captured images used for analysis. Then, the number of endothelial cells per unit area is measured by analyzing at least one of the selected captured images. Of course, as the plurality of captured images stored in the memory 92, two or more captured images are conceivable as long as the storage capacity of the memory 92 allows. For example, 40 captured images are stored in the memory 92 and selected.

  FIG. 9 is a flowchart illustrating an example of a process for selecting a captured image of the corneal endothelium.

<Determination of Analyzable Area AA for Each Captured Image>
Based on the luminance information of the captured image, the control unit 90 determines the entire image area as an analyzable area AA and an unanalyzable area. For example, the control unit 90 obtains the luminance distribution of the pixels arranged in the horizontal direction in the captured image, and in the obtained luminance distribution, the horizontal position satisfying the allowable range is set as the analyzable area AA, and the horizontal position outside the allowable range is analyzed. It is determined as an impossible area (see FIGS. 10 and 11). These determination processes are performed for each captured image stored in the memory 92. As a result of the determination, in the case of a captured image in which the analyzable area AA is narrower than a predetermined width (for example, the analyzable area AA is less than half of the whole), the analysis is impossible and the processing is excluded from the following processing.

  When calculating the luminance distribution, it may be the luminance distribution of one horizontal scanning line arbitrarily set at the vertical position, or may be the average of the luminance distributions of the horizontal scanning lines at a plurality of vertical positions. Further, the luminance distribution may be a distribution obtained by calculating the sum of the luminance values of the pixels arranged in the vertical direction for each horizontal position i of the captured image.

  In the picked-up image, areas where the corneal reflection light such as endothelium and epithelium is not received are black. The epithelial region where epithelial reflected light is received has the highest luminance value. The endothelium region where the endothelium reflected light is received becomes an intermediate region between them.

  For the allowable range, for example, a first threshold SH1 is set in order to discriminate between the endothelial region and the black region (see FIG. 10). In addition, a second threshold SH2 for distinguishing between the endothelial region and the epithelial region is set (see FIG. 11). These threshold values can be set by experiments or the like from actually acquired captured images.

<Sort Processing for Collected Image Set Stored in Memory 92>
FIG. 12 is an example showing the luminance distribution in the vertical direction at a certain horizontal position of the captured image. FIG. 12A is an overall view, and FIG. 12B is an enlarged view. The control unit 90 performs edge detection on the luminance distribution of the pixels arranged in the vertical direction for each horizontal position i of the captured image, so that valleys composed of adjacent edges (see the circular portion in FIG. 12B). ) Is detected. Thereby, the cell gap (valley) between each endothelial cell is detected regarding the luminance distribution for each horizontal position i. The control unit 90 can remove small valleys (unevenness in cells, noise, etc.) that do not correspond to valleys between cells by performing a smoothing process on the luminance distribution.

  FIG. 13 is an example of the luminance information KI in which the total number of valleys in the luminance distribution in the vertical direction is arranged for each horizontal position i. The control unit 90 obtains the sum total of the number of valleys in the luminance distribution in the vertical direction for each horizontal position i, and further obtains a sum SU obtained by adding the sums for each horizontal position i. Such calculation of the sum SU is performed for each captured image stored in the memory 92.

  The sum SU is used as a reference for rearranging the set of captured images stored in the memory 92. The control unit 90 sorts the captured images in descending order of the total sum SU obtained for each captured image. As a result, the captured images are arranged in descending order of the number of valleys in the image. Here, the number of valleys such as the sum SU is used as the sorting criterion. First, the more the focus is on the endothelial cells, the higher the contrast of the image and the more valleys are detected. Secondly, it is because the more valleys are detected in the entire captured image, the more valleys are detected.

  By such sort processing, the captured image is in focus, the captured image with a large number of endothelium imaging areas is sorted in the upper rank, and the captured image is out of focus and / or the captured image with few endothelium imaging areas. Sorted down.

<Selection of images used for analysis>
The control unit 90 selects a captured image having a sum SU of 90% or more from the sum SU of the captured images having the largest sum SU in the sorted captured images. For example, when the maximum sum SU is 2000, picked-up images in which 1800 or more sums SU are detected are selected.

  FIG. 14 is a diagram in which graphs having different valley dense regions are arranged in the graph of FIG. The control unit 90 selects a first analysis candidate from a set of captured images having a total SU of 90% or more selected as described above. For each captured image, the control unit 90 determines an area that exceeds a predetermined threshold in the luminance information KI acquired as described above as an endothelial image area, and calculates the position of the center of gravity in the horizontal direction. Then, the control unit 90 selects an image whose center of gravity is closest to the horizontal center position of the captured image as an analysis first candidate image (first image) (see FIG. 14B). Accordingly, an image having a valley dense region at the center of the image is selected as an analysis candidate.

  The reason why such a reference is used is that, as the center of gravity of the endothelium region is closer to the horizontal center of the image, the focused endothelium image is located at the center of the image and closer to the center of the optical axis, so there is less influence of aberration. Because. In addition, there are few black areas corresponding to the background and saturated areas corresponding to the epithelium, and there are few unsuitable areas in the analysis.

  On the other hand, as the center of gravity of the endothelium region is decentered with respect to the horizontal center of the image, the focused endothelium image is located in the vicinity and the image quality is degraded due to the influence of aberration. In addition, a black region corresponding to the background and a saturated region corresponding to the epithelium are likely to occur, and there are many regions that are inappropriate for analysis.

<Display of sorted images>
The control unit 90 displays the first image selected as the first analysis candidate on the monitor 95 and displays the captured images in the order of the sum total SU except for the first image (it is not necessary to display all of them. The top few (for example, 5) may be sufficient). The control unit 90 displays each captured image on the monitor 95 in such a form that identification numbers such as numbers are attached to the captured images so that the ranking can be understood. Note that a lower-order captured image can be displayed by a predetermined switch operation.

  Here, when at least one captured image on the monitor 95 is selected by the switch operation of the examiner, the control unit 90 processes the selected captured image to calculate the number of endothelial cells per unit area (cells). Density) and the obtained result is displayed on the monitor 95. In this case, the cell density may be obtained from the entire endothelial image region, or the cell density may be obtained from a part of the endothelial image region (for example, near the center).

  According to the above configuration, the captured images are sorted in descending order of the number of valleys of the edges, so that images with high contrast and a large number of light receiving areas of the cell endothelium are sorted as analysis candidates. For this reason, the examiner can save the trouble of selecting the captured image, and can improve the possibility that the analysis is executed with the captured image suitable for the analysis.

  In addition, a captured image in which the position of the center of gravity in the horizontal direction of the endothelial image region is close to the horizontal center is calculated as the first analysis candidate, so that the image of the dense region in the focused valley is acquired with less aberration. Since the image is selected, the analysis accuracy is increased.

  In the above embodiment, the number of valleys in the luminance distribution in the vertical direction at each position in the horizontal direction is detected in the entire captured image, and the sum of the number of valleys in each horizontal position is calculated as the evaluation value. Conversely, the number of valleys of the luminance distribution in the horizontal direction at each position in the vertical direction may be detected, and the sum of the number of valleys at each vertical position may be calculated as the evaluation value.

  Moreover, in the said Example, although it was set as the structure which measures the number of troughs in luminance distribution, it is not limited to this, What is necessary is just a structure which can detect the number of edges of luminance distribution. For example, the number of gaps between cells may be measured by measuring the number of portions where the luminance is greatly increased / decreased in the luminance distribution (in this case, the number of the gaps is twice).

  In the above embodiment, the evaluation value is calculated based on the two-dimensional luminance distribution in the entire captured image. However, the present invention is not limited to this. The apparatus calculates an evaluation value based on a two-dimensional luminance distribution in the captured image. For example, the control unit 90 detects the number of valleys of the luminance distribution in the vertical direction at each position continuous in the horizontal direction in a part of the two-dimensional region of the captured image (for example, a rectangular region near the center). The sum of the number of valleys at the horizontal position may be used as the evaluation value.

<Second embodiment>
In the above configuration, the sort processing and selection processing of the captured image are performed by image processing on the captured image, but the present invention is not limited to this. In the second embodiment, the control unit 90 stores, in the memory 92, analysis candidates for the corneal endothelium of the eye E based on the alignment detection result of the imaging unit 4 with respect to the eye E detected each time each captured image is acquired. Select from each captured image. The control unit 90 sorts the plurality of captured images based on the alignment detection result, and displays the sorted captured images on the monitor 95 in order from the top. In the second embodiment, functions realized by the control unit 90 are divided into calculation of misalignment, selection of captured images, and sorting of a plurality of captured images.

<Storage of alignment deviation amount at the time of captured image acquisition>
The control unit 90 detects the amount of alignment deviation of the imaging unit 4 with respect to the eye E during continuous acquisition of captured images. The XY alignment deviation amount is detected based on the imaging signal from the imaging element 84, and the Z alignment deviation amount is detected based on the light reception signal from the light receiving element 89. The control unit 90 stores continuously acquired captured images in the memory 92, and stores the detection result of the alignment deviation amount corresponding to each captured image in the memory 92 in association with each captured image. .

  Regarding the XY alignment deviation amount, the linear distance between the corneal apex position with respect to the observation optical axis L1 is obtained as the XY alignment deviation amount based on the alignment deviation amount in the X direction and the alignment deviation amount in the Y direction. Is preferred.

  Regarding the Z alignment deviation amount, it is preferable to acquire the alignment deviation amount of the imaging unit 4 with respect to the inner skin of the eye E. For example, the control unit 90 detects the peak P2 corresponding to the reflected light beam from the corneal endothelium based on the light reception signal output from the light receiving element 89, and detects the position P2z of the endothelial peak on the light receiving element 89 (FIG. 15). reference). The controller 90 obtains a deviation amount between a predetermined position (focus position) on the light receiving element 89 and the endothelial peak position P2z. Thereby, the amount of alignment deviation with respect to endothelium is calculated | required. The in-focus position is determined by the light receiving position of the endothelial peak on the light receiving element 89 when the endothelium and the imaging optical system 30 are in focus.

<Sort Processing for Collected Image Set Stored in Memory 92>
The alignment deviation amount is used as a reference for rearranging the set of captured images stored in the memory 92. First, the control unit 90 sorts the captured images in ascending order of the absolute value of the Z alignment deviation amount. Thereby, the captured images are arranged in the order in which the alignment position in the Z direction is closer to the endothelium focus position. Here, the amount of alignment deviation in the Z direction was used as the sorting reference because the contrast with the image increases as the focus on the endothelial cells increases, and the imaging area of the endothelium in the entire captured image increases. Because.

  Furthermore, the control unit 90 sorts the captured images in the order of decreasing absolute values of the XY alignment deviation amount between the captured images having the same absolute value of the Z alignment deviation amount. Thereby, the captured images are arranged in the order in which the alignment positions in the XY directions are close to the appropriate positions. Here, the amount of alignment deviation in the XY direction is used as the sorting reference when the XY alignment deviation of the imaging unit 4 with respect to the corneal apex is large and the endothelial image received by the imaging element 44 is decentered vertically and horizontally. This is because such an image is sorted in a relatively low order because many black regions that are acquired and incapable of cell analysis are included.

<Selection of images used for analysis>
The control unit 90 selects a captured image having the smallest absolute value of the Z alignment deviation amount and the XY alignment deviation amount as an analysis first candidate image (first image) from the set of captured images sorted as described above. elect. Then, the control unit 90 displays the first image selected as the first analysis candidate on the monitor 95 and sequentially displays the captured images of the second and lower ranks on the monitor 95. The control unit 90 displays each captured image on the monitor 95 in such a form that identification numbers such as numbers are attached to the captured images so that the ranking can be understood. Note that a lower-order captured image can be displayed by a predetermined switch operation. Since the subsequent flow is the same as that of the first embodiment, a special description is omitted.

  As in the second embodiment, by selecting and sorting the captured images using the alignment detection result of the imaging unit 4 with respect to the endothelium of the eye E, it is possible to reliably select an image that is in focus with respect to the endothelium. Since the time required for analyzing the image itself can be shortened, the analysis time can be greatly shortened.

  In the second embodiment, the Z alignment deviation amount and the XY alignment deviation amount may be scored, and the captured images may be sorted in descending order of the total score. In this case, the smaller the deviation amount, the higher the score. Further, the score based on the Z alignment deviation amount may be weighted so that the Z alignment deviation amount is more important than the XY alignment deviation amount.

  In the second embodiment, when the captured images are sorted using the alignment deviation amount, the alignment range amount may be divided and sorted in a predetermined step. For example, for example, the Z direction is divided by 20 μm steps, and the XY direction is divided by 0.1 mm steps.

  In the second embodiment, of course, the signal from the imaging element 84 and the signal from the light receiving element 89, which are the basis for calculating the amount of alignment deviation, are stored in the memory 95. The calculation is also included in the storage of the alignment deviation amount at the time of image acquisition.

<Transformation example>
In the present embodiment, the sorting process based on the captured image shown in the first example can be combined with the sorting process based on the alignment deviation amount shown in the second example.

  For example, the control unit 90 selects a set of captured images having a small alignment deviation amount by a sorting process based on the alignment deviation amount. In the selected set, the control unit 90 calculates the center of gravity position in the horizontal direction of the endothelial image region formed in each captured image, and analyzes the image whose center of gravity is closest to the horizontal center position of the captured image. Selected as the first candidate image (first image).

  In addition, the control unit 90 selects the top of the captured image with the large sum SU by the sorting process based on the captured image. Then, in the set selected at the top, the control unit 90 selects a captured image having the smallest absolute value of the Z alignment deviation amount and the XY alignment deviation amount as an analysis first candidate image (first image).

  Moreover, in the said embodiment, although it was set as the structure which detects an alignment state optically, it is not limited to this, What is necessary is just a device which can detect the amount of alignment deviation with respect to the eye E. FIG. For example, an ultrasonic sensor or an X-ray sensor may be used.

  The embodiment is not limited to the example described above, and various modifications are possible within the scope of the design concept of those skilled in the art.

It is an external appearance side block diagram of the corneal-endothelial-cells imaging device which concerns on this embodiment. It is a schematic block diagram which shows an example of the optical arrangement | positioning when the optical system accommodated in the imaging | photography part is seen from upper direction, and the structure of a control system. It is a figure when the 1st projection optical system and the 2nd projection optical system are seen from the subject side. It is a figure which shows an example of the anterior ocular segment observation screen at the time of imaging | photography of the endothelium of the cornea center part, and is a display example in case there exists a misalignment. It is a figure which shows an example of the anterior ocular segment observation screen at the time of imaging | photography of the endothelium of a cornea center part, and is the example of a display in the state with an appropriate alignment. It is a figure which shows an example of precise alignment detection. It is a figure which shows an example at the time of determining the light reception state of a cornea image based on the output image from an image pick-up element. It is a figure which shows the change of the light reception state of a cornea reflected light when an imaging | photography part is advanced, and is a figure when the alignment of a XYZ direction is completed. It is a figure which shows the change of the light reception state of the cornea reflected light when an imaging | photography part is advanced, and is a figure after epithelial reflected light is moved. It is a figure which shows the change of the light reception state of the cornea reflected light when an imaging | photography part is advanced, and is a figure when endothelium reflected light passes the predetermined detection area. It is a graph which represents the change of total value SLC1 when an imaging part is advanced in time series. It is a graph which represents the change of total value SLC2 when an imaging part is advanced in time series. It is a flowchart which shows an example of the selection process of the picked-up image of a corneal endothelium. It is a figure for demonstrating the determination method of the analysis possible area | region AA. It is a figure for demonstrating the determination method of the analysis possible area | region AA. It is an example which shows the luminance distribution of the perpendicular direction in a certain horizontal position of a captured image. It is an example of the brightness | luminance information KI which arranged the sum total of the number of troughs in the brightness | luminance distribution of a perpendicular direction for every horizontal position i. It is the figure which arranged each the graph from which the dense area of a valley differs in the graph of FIG. It is a figure which shows an example of the detection method of an endothelial peak.

4 Shooting unit (device main unit)
6 driving unit 10 illumination optical system 12 illumination light source 30 imaging optical system 60a, 60b first projection optical system 65a to 65d second projection optical system 80 anterior ocular segment observation optical system 85 Z alignment detection optical system 85a light projection optical system 85b light reception Optical system 90 control unit 92 memory

Claims (2)

An illumination optical system that irradiates illumination light from an illumination light source obliquely toward the subject's eye cornea, imaging optical that receives reflected light from the cornea including corneal endothelial cells by an imaging element and acquires a corneal endothelial cell image An apparatus body comprising a system,
Driving means for moving the device main body relative to the eye of the subject;
An imaging control unit that controls driving of the driving unit to move the apparatus main body in a predetermined direction, and causes the illumination light source to emit light a plurality of times during the movement of the apparatus main body to acquire a plurality of captured images by the imaging element. When,
Storage means for storing a plurality of captured images captured by the imaging element;
When calculating an evaluation value as an analysis target of each captured image stored in the storage unit, in the captured image, the edge of the luminance distribution in the other direction at each position in either the horizontal direction or the vertical direction. An evaluation value calculating means for detecting a number and calculating the sum of the number of edges at each position as the evaluation value;
A corneal endothelial cell imaging device comprising:   The corneal endothelial cell imaging according to claim 1, further comprising image selection means for selecting a candidate for analysis of the corneal endothelium of a subject's eye from each captured image stored in the storage means based on a calculation result of the evaluation value calculation means. apparatus.