JP2013198587A - Fundus imaging system, fundus imaging apparatus, and fundus image management system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a remote diagnosis system using a fundus imaging apparatus.SOLUTION: A fundus image managing and diagnosing apparatus 6 processes a fundus image captured by a fundus camera 7, by which a patient can photograph an ocular fundus by the patient himself/herself, based on a reference image on a cloud, confirms the temporal change, and assists diagnosis. The fundus image managing and diagnosing apparatus 6 gives the patient a warning to urge the patient to take diagnosis again if the warning is necessary based on the result of diagnosis. Further, the image processed by the fundus image managing and diagnosing apparatus 6 can be accessed from a medical facility, and remote diagnosis by a doctor is possible. A fundus image capable of further improving the accuracy in diagnosis can be obtained in the photography of the ocular fundus by the fundus camera 7 by combining an image captured by emitting infrared rays and visible rays.

Description

  The present invention relates to a fundus imaging system, a fundus imaging apparatus, and a fundus image management system.

For example, Patent Document 1 discloses a fundus photographing apparatus 1 that irradiates the fundus with photographing illumination light, receives fundus reflection light of the photographing illumination light, and generates image data of a designated primary color component.
Patent Document 2 discloses an ophthalmologic photographing apparatus that obtains an in-focus state by photographing a plurality of images with different in-focus positions by a single photographing operation.
Patent Document 3 discloses an ophthalmic image processing method for easily and reliably comparing with past images.
Further, Patent Document 4 discloses an image processing apparatus that enables follow-up observation by comparison with past images and enables setting of an imaging region effective for diagnosis.
Patent Document 5 discloses a lightweight handheld camera that provides continuous images in order to obtain an image with the most desirable focus.
Further, in Patent Document 6, a lesion portion in the eye image is set as a region of interest (dotted line frame) for an eye image photographed in the past, and past eye image data is referred to at the time of photographing this time. Thus, an eye image capturing device that sets the position, size, exposure, and the like of a region of interest corresponding to an eye image captured this time is disclosed.

JP 2010-000192 A JP 2001-008900 A JP 2003-010131 A JP 2010-259629 A Special table 2010-505468 Japanese Patent Laid-Open No. 2001-2751979

  An object is to provide a remote diagnosis system using a fundus imaging apparatus.

  The fundus imaging system according to the present invention includes an imaging unit that captures an image of a specific wavelength component at a plurality of focus positions, and an area extraction that extracts a part of the image area from each of the plurality of images captured by the imaging unit. And means for synthesizing a plurality of image areas extracted by the area extracting means.

  Preferably, the photographing unit irradiates light of a non-visible light component and photographs a fundus image at a plurality of focus positions.

  Preferably, the photographing unit shoots an image of the fundus by irradiating the fundus with infrared rays.

  Preferably, after photographing by the photographing means, visible light photographing means for photographing a fundus image by irradiating light of a visible light component, an image synthesized by the synthesizing means, and the visible light photographing means And a visible component synthesizing unit that synthesizes the visible light image photographed by the above.

  Preferably, the area extracting means extracts an image area based on information indicating the sharpness of each image area.

  Preferably, the imaging unit irradiates light in a partial wavelength region of visible light, and images a fundus image at a plurality of focus positions.

  The fundus imaging apparatus according to the present invention includes a camera that images the fundus, a fixing unit that mounts the camera on a human head, and a transfer unit that transfers an image captured by the camera.

  Preferably, the fixing means includes a fixing portion for fixing the camera to a human face.

  Preferably, the fixing means includes at least three fixing parts that come into contact with a person's nose and ear, and the transfer means transfers the photographed image to the outside using wireless communication.

  The fundus image management system according to the present invention includes a first fundus imaging apparatus installed in a medical institution, a portable second fundus imaging apparatus, and an information processing apparatus that manages the fundus image. The apparatus uses the fundus image captured by the first fundus imaging apparatus as a reference, and a fundus for comparing with a reference fundus image from among a plurality of fundus images captured by the second fundus imaging apparatus. Image selection means for selecting an image is provided.

  Preferably, the information processing apparatus further includes an image correcting unit that corrects the fundus image selected by the image selecting unit with reference to the fundus image serving as a reference.

  Preferably, the information processing apparatus is configured to store the fundus image corrected by the image correction unit in association with the time of photographing, and a plurality of fundus images stored by the image storage unit. And a fundus oculi change specifying means for specifying a change with time.

  Preferably, the image correction unit compares the fundus image serving as a reference with the fundus image serving as a correction target, and at least performs parallel movement processing, rotation processing, and enlargement / reduction processing on the fundus image serving as a correction target. Apply.

  Preferably, the information processing apparatus further includes notification means for notifying that re-examination is necessary when the change with time specified by the blood vessel change specifying means matches a predetermined condition.

  According to the present invention, it is possible to provide a remote diagnosis system using a fundus imaging apparatus.

It is a figure explaining the outline | summary of the cloud type remote diagnosis system 1 using a small fundus camera. (A) is a diagram showing the fundus camera 7 and the client terminal 9, (B) is a diagram showing the structure of the fundus camera 7, and (C) is a structure of the photographing unit 72 of the fundus camera 7. FIG. It is a figure explaining the functional composition of fundus photography program 10. It is a figure explaining the functional structure of the fundus image composition program 20. It is a figure explaining the functional structure of the fundus image correction program. It is a figure explaining the functional structure of the diagnostic assistance program. It is a figure explaining a fundus remote diagnosis process (S10). It is a figure explaining a fundus imaging process (S20). It is a figure explaining a fundus oculi image composition process (S30). It is a figure explaining a fundus oculi image correction process (S40). It is a figure explaining a diagnosis assistance process (S50). It is a figure which shows the fundus image image | photographed by irradiating infrared rays.

Embodiments of the present invention will be described below with reference to the drawings.
FIG. 1 is a diagram illustrating an overview of a cloud-type remote diagnosis system 1 using a small fundus camera.
As illustrated in FIG. 1, the cloud type remote diagnosis system 1 according to the present embodiment includes a fundus imaging system 100 and a fundus image management system 200.

The fundus imaging system 100 includes fundus cameras 7A to 7C and client terminals 9A to 9C connected to the fundus camera 7.
The fundus camera 7 is an apparatus that allows a patient to photograph the fundus at home, and is an example of a second fundus imaging apparatus according to the present invention. The fundus camera 7 is connected to the client terminal 9 by a USB cable or the like. A fundus photographing program 10 for photographing the fundus is installed in a control unit 720 (described later) of the fundus camera 7 via a recording medium such as a CD-ROM.

  The client terminal 9 is a computer terminal and is connected to the cloud 5 via a network. The client terminal 9 may be a smartphone (multifunctional mobile phone) or a tablet terminal. The fundus image captured by the patient with the fundus camera 7 is transmitted from the client terminal 9 to the fundus image management diagnosis apparatus 6 (described later) existing in the cloud 5 via the Internet. In the client terminal 9, a fundus image synthesis program for synthesizing a fundus image captured by the fundus camera 7 is installed via a recording medium such as a CD-ROM.

  The fundus image management system 200 includes a fundus camera 2, a diagnosis terminal 3 </ b> A, and a fundus image management diagnosis device 6 that exists on the cloud 5. The fundus image management system 20 corrects the fundus image transmitted from the patient to an image suitable for diagnosis, supports diagnosis of the fundus based on the corrected fundus image, and transmits the diagnosis result to the medical institution or patient.

The fundus camera 2 is a fundus camera device installed in a medical institution, and is an example of a first fundus imaging device. Further, the fundus camera 2 captures a reference image of the fundus of the patient. The fundus camera 2 transmits the captured reference image to the diagnosis terminal 3A in association with the patient ID that uniquely identifies the patient. The fundus camera 2 is connected to the diagnostic terminal 3A via a network.
The diagnostic terminal 3A is a computer terminal installed in a medical institution and is connected to the cloud 5 via a network. The diagnosis terminal 3 </ b> A transmits the reference image or the diagnosis result of the fundus diagnosed by the medical staff to the fundus image management diagnosis apparatus 6. Further, the diagnostic terminal 3A extracts the fundus image associated with the patient ID from the fundus image management diagnosis apparatus 6. The fundus image management diagnostic apparatus 6 holds the reference image in the fundus image DB 600. The fundus image DB 600 stores the fundus image captured by the fundus camera 7 in association with the patient ID and time. The fundus image DB 600 stores the fundus image captured by the fundus camera 2 in association with the patient ID as a reference image. Further, the fundus image DB 600 stores the diagnosis result of the fundus image received from the diagnosis terminal 3A in association with the patient ID.

  The fundus image management diagnosis apparatus 6 is a virtual server configured on the cloud 5 and corrects the fundus image transmitted from the patient, and supports diagnosis of the fundus based on the corrected fundus image. The fundus image management diagnosis apparatus 6 has a fundus image DB 600. Further, the fundus image management diagnosis apparatus 6 includes a fundus image correction program 30 that corrects the fundus image and a diagnosis support program 40 that supports diagnosis of the fundus based on the corrected fundus image, such as a CD-ROM. It is installed via a recording medium.

2A is a diagram showing the fundus camera 7 and the client terminal 9, FIG. 2B is a diagram showing the structure of the fundus camera 7, and FIG. It is a figure showing the structure of the imaging | photography part 72 of.
As illustrated in FIGS. 2A and 2B, the fundus camera 7 includes a fixing unit 70 for mounting the fundus camera 7 on a person's head and a photographing unit 72 for photographing the fundus.

The fixing portion 70 includes at least three fixing portions that come into contact with the human nose and ear. Specifically, the fixing unit 70 includes a spectacle support 700 that abuts on the base of the nose, a spectacle frame 702 that abuts on the top of both ears or the base, and a mounting adjustment unit 704. When the fundus camera 7 is attached to a person's head like glasses, the eyeglass support 700 and the eyeglass frame 702 fix the fundus camera 7 at the attachment position.
The wearing adjustment unit 704 detects the angle of the spectacle support 700 at the time of fundus photographing and the horizontal angle of the spectacle frame 702, and holds the detected values. When the patient wears the fundus camera 7 again, the wearing adjustment unit 704 extracts the angle between the optimal eyeglass support 700 held by the wearing adjustment unit 704 and the eyeglass frame 702 and prompts an appropriate angle. Guide the patient to improve the position accuracy.

As illustrated in FIG. 2C, the photographing unit 72 includes a control unit 720, a C-MOS sensor 722, an LED 724, a light diffusion liquid 726, and an infrared light source unit 728.
The control unit 720 is a computer including a CPU and a recording medium, and the fundus imaging program 10 is installed.
The C-MOS sensor 722 is an image sensor and operates according to control by the control unit 720.
The LED 704 is a light source of visible light that is emitted to capture a fundus image as a color image. That is, the LED 704 emits visible light including R, G, and B components.
The light diffusion liquid 706 is a jelly-like liquid, and prevents light from being reflected in the fundus image by diffusing the light emitted from the LED 704.
The infrared light source 708 is a light source for irradiating near infrared rays. When photographed by irradiating with infrared rays, the fundus photograph is a monochrome photograph.

FIG. 3 is a diagram illustrating the functional configuration of the fundus imaging program 10.
As illustrated in FIG. 3, the fundus imaging program 10 includes an infrared imaging position determination unit 100, an infrared imaging unit 102, an infrared imaging information storage unit 104, a visible light imaging position determination unit 106, and a visible light imaging unit 108. And a visible light photographing information storage unit 110.

  The infrared imaging position determination unit 100 determines a plurality of focus positions of the fundus image. Specifically, the infrared imaging position determination unit 100 operates the imaging element at a constant interval so as to move away from the fundus. In the present embodiment, the infrared imaging position determination unit 100 operates the C-MOS sensor 722 that is an image sensor so that the C-MOS sensor 722 moves away from the vicinity of the fundus in units of 1/100 mm. Each time the C-MOS sensor 722 moves 1/100 mm, the infrared imaging position determination unit 100 instructs the infrared imaging unit 102 to take a fundus photograph. The infrared imaging position determining unit 100 operates the C-MOS sensor 722 to move to the maximum movement distance in 2 seconds.

  The infrared imaging unit 102 captures images of specific wavelength components at a plurality of focus positions. Specifically, the infrared imaging unit 100 irradiates light of a non-visible light component and images the fundus. More specifically, the infrared imaging unit 100 irradiates the fundus with infrared rays and captures an image of the fundus. In the present embodiment, the infrared photographing unit 102 shoots the fundus when the infrared light source unit 728 emits near infrared rays and the infrared photographing position determination unit 100 receives a photographing notification. The infrared photographing unit 100 takes about 100 fundus photographs in 2 seconds. The infrared photographing unit 102 stores the photographed fundus image in the infrared photographing information storage unit 104 in association with the position of the C-MOS sensor at the time of photographing (that is, the focus position).

  The infrared imaging information storage unit 104 holds the fundus image captured by the infrared imaging unit 102. In the present embodiment, the infrared imaging information storage unit 104 receives the imaging instruction by receiving the imaging instruction and the position of the C-MOS sensor 722 when the infrared imaging position determination unit 100 notifies the infrared imaging unit 102 of the imaging instruction. The fundus image photographed by the unit 102 is associated and held. Further, the infrared photographing information holding unit 104 notifies the visible light photographing position determining unit 106 of the positional information of the C-MOS sensor that is in focus on the portion of the photographed fundus image that is necessary for fundus diagnosis.

  The visible light imaging position determination unit 106 determines the imaging position based on the notified position information of the C-MOS sensor by the infrared imaging information storage unit 104 and notifies the visible light imaging unit 108 of imaging.

  The visible light photographing unit 108 irradiates light of a visible light component from the LED 724 to photograph a fundus image. In the present embodiment, when there is a photographing instruction from the visible light photographing position determining unit 106, the visible light photographing unit 106 photographs the fundus. The visible light imaging unit 106 stores the captured color image of the fundus in the visible light imaging information storage unit 110.

  The visible light photographing information storage unit 110 holds a fundus color image photographed by the visible light photographing unit 106.

FIG. 4 is a diagram illustrating the functional configuration of the fundus image synthesis program 20.
As illustrated in FIG. 4, the fundus image synthesis program 20 includes an image reception unit 200, a region extraction unit 202, an infrared image synthesis unit 204, a visible component synthesis unit 206, an image transfer unit 208, and a result reception. Part 210.

  The image receiving unit 200 receives an infrared fundus image, which is a fundus image taken by infrared imaging, held by the infrared imaging information storage unit 104 and a color image held by the visible light imaging information storage unit 110. Furthermore, the image reception unit 200 notifies the region extraction unit 202 of the received infrared fundus image, and notifies the visible image composition unit 206 of a color image.

  The area extraction unit 202 extracts some image areas from each of a plurality of images taken at a plurality of focus positions. Specifically, the area extraction unit 202 extracts an image area based on information indicating the sharpness of each image area. In the present embodiment, the region extraction unit 202 performs edge detection of a fundus image captured with infrared rays. The region extraction unit 202 cuts out an image region having a strong edge from the infrared image and notifies the infrared image composition unit 204 of the image region. More specifically, the region extraction unit 202 selects a plurality of infrared images having regions where the pixel values of the infrared image are sharply changed, and the selected infrared image and the edge of the infrared image are selected by the infrared image composition unit 204. Notify the strong part.

  The infrared image combining unit 204 combines a plurality of image regions extracted by the region extracting unit 202. In the present embodiment, the infrared image combining unit 204 combines only the strong edge portion of the infrared image notified by the region extracting unit 202 and combines the fundus image. The infrared image synthesis unit 204 notifies the visible component synthesis unit 206 of the synthesized infrared fundus image.

  The visible component synthesis unit 206 synthesizes the infrared fundus image synthesized by the infrared image synthesis unit 204 and the visible light fundus image. In the present embodiment, the combined image of the infrared fundus image notified by the infrared image combining unit 204 is combined with the color image. The visible component synthesis unit 206 stores the synthesized fundus image in the fundus image DB 500. The fundus image DB 500 holds the synthesized infrared fundus image and the fundus image obtained by combining the infrared fundus image and the color image. The visible component combining unit 206 notifies the image transmitting unit 208 of a fundus image obtained by combining the infrared fundus image and the color image.

  The image transfer unit 208 transfers an image captured using wireless communication to the outside. Specifically, the image transfer unit 208 adds the patient ID to the fundus image notified by the visible component synthesis unit 206 and transfers the fundus image to the fundus image management diagnosis apparatus 6.

  The result receiving unit 208 receives a diagnosis result from the fundus image management diagnosis apparatus 6.

FIG. 5 is a diagram for explaining the functional configuration of the fundus image processing program 30.
As illustrated in FIG. 5, the fundus image processing program 30 includes an image reception unit 300, a reference image selection unit 302, an image selection unit 304, an image correction unit 306, and an image storage unit 308.
The image reception unit 300 receives the fundus image transmitted from the image transfer unit 208 and notifies the image selection unit 304 of the fundus image. In addition, the image receiving unit 300 notifies the reference image selecting unit 302 of the patient ID of the received fundus image.

  The reference image selection unit 302 extracts the fundus image of the reference image associated with the patient ID notified by the image reception unit 300 from the fundus image DB 600. The reference image selection unit 302 notifies the image selection unit 304 of the extracted reference image.

  An image selection unit 304 compares a fundus image captured by the first fundus imaging apparatus with a reference fundus image from among a plurality of fundus images captured by the second fundus imaging apparatus. Select the fundus image. Specifically, a fundus image closer to the reference image is selected from the fundus images captured by the fundus camera 7 based on the reference image captured by the fundus camera 2. In this embodiment, the image selection unit 304 selects the fundus image closest to the reference image notified by the reference image selection unit 302 from the synthesized fundus images notified by the image reception unit 300. The image selection unit 304 notifies the image correction unit 306 of the selected fundus image.

  The image correction unit 306 corrects the fundus image selected by the image selection unit 304 with reference to the fundus image serving as a reference. Specifically, the image correction unit 306 compares the fundus image serving as the reference with the fundus image serving as the correction target, and at least performs parallel movement processing, rotation processing, and scaling on the fundus image serving as the correction target. Apply processing. In the present embodiment, the image correction unit 306 obtains the shift amount and size of the fundus image relative to the reference image, and corrects the fundus image to the same position and size as the reference image. In this embodiment, a SURF feature recognition algorithm program is used as the image correction unit 306. The image correction unit 306 notifies the image storage unit 308 of the corrected fundus image.

  The image storage unit 308 stores the fundus image corrected by the image correction unit 306 in association with the photographing. In this embodiment, the fundus image notified by the image correction unit 306 is stored in the fundus image DB 600 in association with time and patient ID.

FIG. 6 is a diagram for explaining the functional configuration of the diagnosis support program 40.
As illustrated in FIG. 6, the diagnosis support program 40 includes an image extraction unit 400, a difference detection unit 402, a threshold value comparison unit 404, and a warning unit 406.
The image extraction unit 400 extracts the fundus image and the reference image associated with the patient ID to be diagnosed from the fundus image DB 600 and notifies the difference detection unit 402 of the extracted image.

  The difference detection unit 402 identifies the temporal change of the fundus based on the reference image and the fundus image notified by the image extraction unit 400. Changes in the fundus over time include changes in the thickness and shape of blood vessels, bleeding, vitiligo, or nipple depression. Specifically, the difference detection unit 402 quantifies the local temporal change of the fundus image using a cross-correlation coefficient or the like. Changes in the fundus over time include changes in the thickness and shape of blood vessels, bleeding, vitiligo, or nipple depression. Further, the difference detection unit 402 statistically analyzes feature points of changes in the fundus image. The difference detection unit 402 notifies the threshold value comparison unit 404 of the quantified value and the analysis data.

  The threshold value comparison unit 404 compares the quantified value of the difference notified by the difference detection unit 402 and the analysis data with a threshold value. The threshold value in the present embodiment indicates a value obtained by quantifying the difference and a value within a normal change range of the difference analysis data. Further, the threshold value comparison unit 404 re-asserts the patient with respect to the warning unit 406 when the difference quantified value or the difference analysis data exceeds the threshold value. Instruct to encourage medical examination.

  The warning unit 406 notifies that re-examination is necessary when the change with time specified by the difference detection unit 402 matches the predetermined condition. In this embodiment, the warning unit 406 transmits a warning for reexamination to the client terminal 9 associated with the patient ID when the threshold comparison unit 404 prompts the patient to be warned.

FIG. 7 is a diagram for explaining fundus remote diagnosis processing (S10).
As illustrated in FIG. 7, in step 100 (S <b> 100), the patient uses the fundus camera 2 to photograph the fundus serving as a reference image at a medical institution.
In step 105 (S105), the diagnostic terminal 3A stores the captured reference image in the fundus image DB 600 of the fundus image management diagnostic apparatus 6 in association with the patient ID.
In step 110 (S110), the patient uses the fundus camera 7 to photograph the fundus at home.
In step 115 (S115), the client terminal 9 performs fundus image synthesis processing on the photographed fundus image, and further transmits the fundus image management diagnosis apparatus 6 in association with the patient ID.
In step 120 (S120), the fundus image management diagnosis apparatus 6 performs fundus image correction processing and diagnosis support processing on the transmitted fundus image.
In step 125 (S125), when a warning is received from the fundus image management diagnostic apparatus 6 as a result of the diagnosis support process, the fundus remote diagnosis process (S10) proceeds to S130. If no warning is received, the fundus remote diagnosis process (S10) ends.
In step 130 (S130), the patient undergoes reexamination at a medical institution.

FIG. 8 is a diagram for explaining fundus photographing processing (S20).
As illustrated in FIG. 8, in step 200 (S200), the infrared imaging position determination unit 100 moves the C-MOS sensor 722 to a position closest to the fundus. At the next imaging, the infrared imaging position determining unit 100 moves the C-MOS sensor 722 away from the fundus in units of 1/100 mm. Further, when the position of the C-MOS sensor is determined, the infrared imaging position determination unit 100 instructs the infrared imaging unit 102 to image the fundus.

  In step 205 (S205), the infrared imaging unit 102 irradiates the fundus with infrared rays from the infrared light source unit 728, and performs imaging of the fundus.

  In step 210 (S210), the infrared photographing unit 102 associates the photographed fundus image with the position of the C-MOS sensor at the time of photographing and stores it in the infrared photographing information storage unit 104.

  In step 215 (S215), when the C-MOS sensor 722 has reached the maximum movement distance, the fundus photographing process (S20) proceeds to S220. If the C-MOS sensor 722 has not reached the maximum movement distance, the fundus imaging process (S20) proceeds to S200.

  In step 220 (S220), the visible light imaging position determination unit 106 receives the position information of the C-MOS sensor that can capture the fundus image in focus on the part necessary for diagnosis of the fundus from the infrared imaging information storage unit 104. The visible light photographing position determining unit 106 moves the C-MOS sensor according to the received position information.

  In step 225 (S225), the visible light photographing unit 108 irradiates the visible light from the LED 724 and photographs the fundus with the visible light diffused by the light diffusion liquid 726.

  In step 230 (S230), the visible light imaging unit 108 stores the fundus image obtained by imaging using visible light in the visible light imaging information storage unit 110.

FIG. 9 is a diagram for explaining fundus image synthesis processing (S30).
As illustrated in FIG. 9, in step 300 (S300), the image receiving unit 200 receives fundus images from the infrared imaging information storage unit 104 and the visible light imaging information storage unit 110, and extracts the infrared fundus image as a region. The color image is notified to the visual component composition unit 206 to the unit 202.

  In step 305 (S305), the region extraction unit 202 detects an edge of the notified infrared fundus image, cuts out an image region having a strong edge from the infrared image, and notifies the infrared image composition unit 204 of the image region.

  In step 310 (S310), the infrared image combining unit 204 combines the image regions with strong edges notified by the region extracting unit 202 to form a fundus image. The infrared image synthesis unit 204 notifies the visible component synthesis unit 206 of the synthesized fundus image.

  In step 315 (S315), the visible component combining unit 206 combines a color image with the fundus image notified by the infrared image combining unit 204.

  In step 320 (S320), the visible component combining unit 206 stores the combined fundus image in the fundus image DB 500. The visible component combining unit 206 notifies the image transfer unit 208 of a fundus image obtained by combining the infrared fundus image and the color image in association with the patient ID.

  In step 325 (S325), the image transfer unit 208 transmits the fundus image synthesized by the visible component synthesis unit 206 to the fundus management diagnosis apparatus 6.

  The fundus imaging by the fundus camera 7 may be performed several times a day for several days. The fundus imaging process (S20) and the fundus image synthesis process (S30) To be implemented.

FIG. 10 is a diagram for explaining fundus image correction processing (S40).
As illustrated in FIG. 10, in step 400 (S400), the image receiving unit 300 receives an infrared fundus image associated with a patient ID transferred from the client terminal 9 and a fundus image obtained by combining visible light. The image reception unit 300 notifies the reference image selection unit 302 of the received patient ID, and notifies the image selection unit 304 of a fundus image obtained by combining visible rays.

  In step 405 (S405), the reference image selection unit 302 extracts the reference image of the patient ID associated with the fundus image from the fundus image DB 600 and notifies the image selection unit 304 of the reference image. The image selection unit 304 selects a fundus image closer to the reference image from the fundus images obtained by combining visible rays, and notifies the image correction unit 306 of the selected fundus image and the reference image.

  In step 410 (S410), the image correction unit 306 compares the reference image and the fundus image to obtain the position and shift amount of the fundus image with respect to the reference image.

  In step 415 (S415), the image correction unit 306 corrects the fundus image to the same position and size as the reference image. The image correction unit 306 notifies the image storage unit 308 of the corrected fundus image.

In step 420 (S420), the image storage unit 308 stores the notified fundus image in the fundus image DB 600 in association with time and patient ID.

FIG. 11 is a diagram illustrating the diagnosis support process (S50).
As illustrated in FIG. 11, in step 500 (S500), the fundus image extraction unit 400 obtains, from the fundus image DB 600, a plurality of fundus images and reference images associated with the patient ID and time to be diagnosed. Extract and notify the difference detection unit 402. The difference detection unit 402 quantifies the local temporal change of the fundus image from a plurality of fundus images based on the reference image using a cross-correlation coefficient or the like. Further, the difference detection unit 402 statistically analyzes feature points of changes in the fundus image. The difference detection unit 402 notifies the threshold comparison unit 404 of the quantified value and the analysis data.

  In step 505 (S505), the threshold value comparison unit 404 compares the notified quantified value and the analysis data with a threshold value.

  In step 510 (S510), when the value compared with the threshold and quantified or the analysis data is equal to or greater than the threshold, the diagnosis support processing (S50) proceeds to S515. If it is less than the threshold value, the diagnosis support process (S50) ends.

  In step 515 (S515), the warning unit 406 transmits a warning prompting reexamination to the client terminal 9 associated with the patient ID.

FIG. 12 is a diagram illustrating a fundus image captured by irradiating infrared rays.
As illustrated in FIG. 12, (A) is a fundus image that is extracted by the region extraction unit 202 and that has strong edges concentrated at the center. (B) is a fundus image extracted by the region extraction unit 202, in which strong edges are concentrated around the center. (C) is a fundus image that is extracted by the region extraction unit 202 and has portions with strong edges concentrated outside. The infrared image synthesis unit 204 synthesizes a portion having a strong edge from the fundus images of (A), (B), and (C) extracted by the region extraction unit 202, and generates a fundus image as illustrated in (D). Form.

As described above, the fundus camera 7 of the present embodiment is small and lightweight that can be mounted on the head and can shoot a fundus image even by one person. For example, other than an ophthalmic examination room, for example, internal medicine, bedside It can also be used in places where there is no fundus photography specialist at home.
Furthermore, the fundus camera 7 irradiates infrared rays and visible rays to perform two types of fundus image shooting, and can synthesize a respective fundus image to obtain a clearer fundus image. Thereby, it is possible to improve the accuracy of measurement of the thickness of the blood vessel.

  Furthermore, the image management diagnostic device 6 existing on the cloud according to the present embodiment corrects an image captured by the fundus camera 7 to a diagnosable image based on the reference image, and detects feature points to change over time. Is quantified. Further, the image management diagnostic device 6 compares the quantified value with a value within a normal range of change with time. As a result of the comparison, if not within the normal range, the image management diagnostic apparatus 6 transmits an alert display and a warning message to the patient. As a result, a fundus image can be diagnosed by non-ophthalmologists.

  In addition, the cloud-type remote system 1 according to the present embodiment captures a fundus image of a patient serving as a reference in a medical institution, and stores the reference image in association with a patient ID on the cloud via the diagnosis terminal 3. Thereafter, the fundus image captured by the patient using the fundus camera 7 is transmitted by the client terminal 9 to the fundus image management diagnosis device 6 on the cloud. The fundus image management diagnosis apparatus 6 diagnoses the fundus image based on the reference image, and transmits the diagnosis result to the doctor or the patient. Thereby, the temporal change of the fundus photograph taken at home or the like is measured, and telemedicine is performed based on the measured change, thereby enabling early detection of the disease.

Next, a modification of the above embodiment will be described.
In the above embodiment, the fundus image is stored in association with the patient ID. However, the present invention is not limited to this. The retinal pattern is recorded when the reference image is captured, and the retinal pattern and the fundus image are stored in association with each other. May be.
In the above embodiment, an image obtained by synthesizing the infrared fundus image and the color image is transmitted to the fundus image management diagnostic apparatus 6. Although a color image is necessary for the fundus diagnosis by the doctor, diagnosis support processing (S50) In the diagnosis of the fundus, the diagnosis can be made only with the infrared fundus image.
In the above-described embodiment, the image correction unit 306 obtains the shift amount and size of the position of the fundus image with respect to the reference image in correcting the fundus image, and corrects the fundus image to the same position and size as the reference image. However, the present invention is not limited to this, and color correction may be further performed. The image correction unit 306 may perform the same correction on the reference image.

DESCRIPTION OF SYMBOLS 1 ... Cloud type remote diagnosis system 2 ... Fundus camera 3 ... Diagnostic terminal 5 ... Cloud 6 ... Fundus image management diagnostic apparatus 7 ... Fundus camera 9 ... Client terminal 100 ... Fundus imaging system 200 ... Fundus image management system 500 ... Fundus image database 600 ... fundus image database

Claims (14)

Photographing means for photographing an image of a specific wavelength component at a plurality of focus positions;
A region extracting means for extracting a part of the image region from each of the plurality of images photographed by the photographing means;
A fundus photographing system comprising: a combining unit configured to combine a plurality of image regions extracted by the region extracting unit. The fundus imaging system according to claim 1, wherein the imaging unit irradiates light of a non-visible light component and captures images of the fundus at a plurality of focus positions. The fundus imaging system according to claim 2, wherein the imaging unit irradiates the fundus with infrared rays to capture an image of the fundus. Visible light photographing means for photographing a fundus image by irradiating light of a visible light component after photographing by the photographing means;
The fundus imaging system according to claim 2, further comprising: a visible component synthesis unit that synthesizes the image synthesized by the synthesis unit and the visible light image captured by the visible light imaging unit. The fundus imaging system according to claim 1, wherein the region extraction unit extracts an image region based on information indicating the sharpness of each image region. The fundus imaging system according to claim 1, wherein the imaging unit irradiates light in a partial wavelength region of visible light and images a fundus image at a plurality of focus positions. A camera for photographing the fundus,
Fixing means for mounting the camera on a human head;
A fundus photographing apparatus comprising: a transfer unit configured to transfer an image photographed by the camera. The fundus imaging apparatus according to claim 7, wherein the fixing unit includes a fixing part for fixing the camera to a human face. The fixing means includes at least three fixing parts that abut the human nose and ear,
The fundus imaging apparatus according to claim 8, wherein the transfer unit transfers a captured image to the outside using wireless communication. A first fundus photographing apparatus installed in a medical institution;
A portable second fundus imaging device;
An information processing device for managing fundus images,
The information processing apparatus includes:
A fundus image for comparison with a reference fundus image is selected from a plurality of fundus images captured by the second fundus imaging device, using the fundus image captured by the first fundus imaging device as a reference. A fundus image management system having image selection means to perform. The information processing apparatus includes:
The fundus image management system according to claim 10, further comprising: an image correction unit that corrects the fundus image selected by the image selection unit with reference to a fundus image serving as a reference. The information processing apparatus includes:
Image storage means for storing the fundus image corrected by the image correction means in association with the photographing,
The fundus image management system according to claim 11, further comprising: a fundus change specifying unit that specifies a temporal change of the fundus based on a plurality of fundus images stored by the image storage unit. The image correction unit compares the fundus image serving as a reference with the fundus image serving as a correction target, and performs at least a parallel movement process, a rotation process, and a scaling process on the fundus image serving as a correction target. The fundus image management system according to 11. The fundus image management system according to claim 12, further comprising notification means for notifying that re-examination is necessary when the temporal change specified by the blood vessel change specifying means matches a predetermined condition.

Images