JP4131015B2 - Fundus image conversion apparatus and system - Google Patents

Description

[0001]
[Field of the Invention]
The present invention relates to a fundus image conversion system, and more particularly to a fundus image conversion apparatus and system for converting a fundus image into a diagnostic image by CG.
[0002]
[Prior art]
Diabetic retinopathy, one of the three major complications of diabetes, is currently the leading cause of blindness in Japan. This is because there are almost no initial symptoms of diabetic retinopathy and it is difficult for general physicians to determine early stage signs from fundus images.
[0003]
In order to detect retinopathy with certainty, it is necessary to work with an ophthalmologist to enable patients to undergo regular ophthalmic examinations, but it will take time to generalize internal medicine / ophthalmology collaboration as a social system. It takes reality. On the other hand, as another method, if a screening judgment can be made even in general internal medicine to some extent, the possibility of solving the problem is high.
[0004]
[Problem to be Solved by the Invention]
However, in order to be able to make a screening judgment even in general internal medicine, a system that allows a clinician to judge a lesion that was previously visible only to an ophthalmologist is necessary, but such a system has not been developed at present.
[0005]
In order to solve the problem, the present invention simplifies the operation of medical staff by adding a CG process to a digitized fundus image so that a clinician can determine a lesion that was previously visible only to an ophthalmologist . A fundus image conversion apparatus and system capable of measuring an image with high accuracy are provided.
[0007]
[Means for Solving the Problems]
In order to solve such a problem, the fundus image conversion apparatus of the present invention is a fundus image conversion apparatus that converts an acquired fundus image into a diagnostic image, an acquisition unit that acquires the fundus image, and the acquired An adjusting means for adjusting the contrast and / or contrast of the fundus image, an extracting means for extracting a contour from the fundus image whose brightness and / or contrast has been adjusted, and a fundus image on which the extracted contour is displayed in an identifiable manner. Output means for outputting, and repeating means for repeating the adjustment by the adjustment means, the extraction by the extraction means, and the output by the output means until a desired fundus image is obtained in accordance with an instruction from an operator , The adjustment means stores in advance a plurality of sets of setting means for setting light and darkness and / or contrast as initial adjustment values, and transition values of light and darkness and contrast adjustment values. A set of displacement values selected by the selection means, comprising: storage means; display means for displaying the plurality of sets in a selectable manner; and selection means for selecting one of the plurality of sets displayed. according, characterized Rukoto by displacing the adjustment value of brightness and / or contrast of the fundus image.
[0008]
Here, the extraction means outputs the pixel value when the difference between adjacent pixel values exceeds a predetermined threshold value, and outputs zero as the pixel value when the difference does not exceed the predetermined threshold value. By doing so, the contour is extracted. The output means outputs the extracted contour so that it can be identified by a predetermined color.
[0009]
Moreover, the fundus image transformation system of the invention, a communication terminal, a fundus image conversion system comprising an information processing apparatus for converting the fundus image acquired via the network from the communication terminal into an image for diagnosis, the A communication terminal for acquiring a fundus image; a first transmitting unit for transmitting the acquired fundus image to the information processing apparatus via a network; and a fundus image with adjusted brightness and / or contrast and First receiving means for receiving a fundus image in which the extracted contour is displayed in an identifiable manner and a screen for inputting an instruction to select the brightness and / or contrast adjustment by the operator, and according to the operator's selection instruction Until the desired fundus image is obtained in which the extracted contours are identifiably displayed. And a instruction means for instructing repeatedly to the information processing apparatus, the information processing apparatus, via a second receiving means for receiving the fundus image through the network from the communication terminal, the network from the communication terminal In accordance with the selected instruction, the adjusting means for adjusting the contrast and / or contrast of the received fundus image, the extracting means for extracting the contour from the fundus image whose brightness and / or contrast is adjusted, and the extracted contour and a second transmission means that sends the fundus image to be distinguishably displayed on the communication terminal via a network, said adjusting means comprises a setting means for setting the brightness and / or contrast to the initial adjustment value, Storage means for storing a plurality of sets of transition values of brightness and contrast adjustment values in advance, and the communication so as to selectably display the plurality of sets. Display instruction means for instructing at the end, and selection means for selecting any one of the plurality of sets in accordance with a selection instruction from the communication terminal by an operator, and the set of displacement values selected by the selection means according, characterized Rukoto by displacing the adjustment value of brightness and / or contrast of the fundus image.
Here, the information processing apparatus further includes the received fundus image, the fundus image in which the brightness and / or contrast is adjusted, and / or the eye according to a transmission instruction from the communication terminal by the operator via the network. Third funding means for transmitting a fundus image in which the contour is displayed in an identifiable manner to another communication terminal is provided.
[0010]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, a fundus image conversion apparatus and system according to an embodiment of the present invention will be described.
[0011]
<Configuration example of fundus image conversion apparatus of this embodiment>
FIG. 1 is a diagram illustrating a configuration example of a fundus image conversion apparatus according to the present embodiment. The fundus image conversion apparatus according to the present embodiment can be realized by a general-purpose personal computer or a dedicated computer. Although FIG. 1 is described as a stand-alone device, the present invention is not limited to this. Further, a configuration in which a plurality of devices are combined instead of one device may be used.
[0012]
In FIG. 1, reference numeral 10 denotes a calculation / control CPU which controls the entire apparatus. Reference numeral 20 denotes a ROM which stores fixed programs (for example, OS and BIOS) and fixed parameters. A communication processing unit 30 performs wired / wireless communication with an external device.
[0013]
Reference numeral 40 denotes a temporary storage RAM, which is used as a data storage area and a program storage area. In this example, contours are extracted from the area 41 for storing fundus photographic image data (hereinafter also referred to as an original image), the storage area 42 for image data adjusted for brightness / contrast and contrast, and the image 42 in the data storage area. It has an area 43 for storing an image, and an adjustment parameter section 44 for storing the above-mentioned adjustment parameters for brightness / contrast. The adjustment parameter unit 44 may store parameters input by the operator, or may store parameters selected by pattern selection described later. Considering the operability of the operator, it is desirable that a parameter set is prepared in advance as in the latter case. A close-up image described later may be stored. The program storage area stores a program loaded via the external storage device 50 or the communication processing unit 30 and is executed by the CPU 10.
[0014]
Reference numeral 50 denotes an external storage device that provides data and programs to the apparatus, and includes, for example, a hard disk, a floppy (registered trademark) disk, an optical disk such as a CD, and a memory card. In this example, the data storage area of the external storage device 50 includes, for example, a pattern storage unit 51 that stores a set of brightness / contrast adjustment parameters, an original image, and an image that has been subjected to brightness / contrast adjustment and contour extraction for each patient. It has a fundus image database 52 to be accumulated, and a program storage area stores an area 53 for storing a main program, an area 54 for storing a light / dark / contrast adjustment module, an area 55 for storing a contour extraction module, and a close-up module. An area 56 for storing data, an area 57 for storing a cutting module, and the like.
[0015]
Reference numeral 60 denotes an input interface for connecting an input from an input device. In this example, a keyboard 60a and a pointing device 60b are connected. Reference numeral 70 denotes an output interface for connecting an output to an output device. For example, a display unit 70a and a printer 70b are connected. The display unit 70a may have a touch panel and be used as an input device.
[0016]
(Preset pattern storage example)
As described above, the operator may input the light / dark / contrast adjustment parameters. However, the light / dark / contrast adjustment parameters appropriate for fundus image processing generally vary greatly if the fundus photo capture conditions do not change. It is a wasteful operation that the operator inputs each time. Therefore, in the present embodiment, several appropriate adjustment parameter sets are prepared in advance, and the operator devised to select them. If further fine adjustment is desired, it is possible.
[0017]
FIG. 2 shows an adjustment parameter set in an example having three patterns A, B, and C. The parameters shown in FIG. 2 are stored in advance in the pattern storage unit 51 shown in FIG. 1, and are read out based on the operator's selection of patterns A, B, and C, and are used for adjustment of light / dark / contrast.
[0018]
In FIG. 2, a light / dark parameter 51b and a contrast parameter 51 are stored corresponding to the pattern 51a, and read out corresponding to the pattern selection. In addition, the numerical value of FIG. 2 represents the following numerical values.
[0019]
Regarding the brightness value: When the brightness (including each color element) data included in the original image has 256 gradations, the brightness of each pixel is changed by a specified value. For example, a brightness parameter value of “+15” means that the brightness of all pixels included in the original image is increased by 15 gradations, and if it is “−15”, 15 gradations. Means darkening. As a result of this adjustment, if the brightness exceeds a limit value (0 or 255), the brightness of the pixel is made equal to the limit value.
[0020]
Contrast value: The gradient of the tone curve (the relationship between the gradation value and the actual gradation) when the brightness (including each color element) data included in the original image is 256 gradations is “1”. Then, the inclination of the tone curve of the image is changed by the specified value by the adjustment. For example, the contrast parameter value “+15” is an adjustment that increases the slope of the tone curve of the original image. Specifically, the gradation width is set to an adjustment value from both limit values (brightness 0 or 255). A tone curve is created by reducing the size, and then converted to 256 gradations. By adjusting in this way, the brightness of the original brightness of 15 gradations or less becomes brightness “0”, the brightness of 240 or more becomes brightness “255”, and the elements in between are changed from 0 to 255 Since it changes linearly, the slope of the tone curve becomes steep.
[0021]
Needless to say, the brightness and contrast adjustment methods are not limited to the above. About a regulation value, what is necessary is just to be able to acquire the same effect as the case where it adjusts by the said method. In the present embodiment, the case of storing a plurality of patterns has been described. However, a configuration in which one default parameter set is stored may be used. Furthermore, the configuration may be such that fundus photo shooting conditions are input and a suitable parameter set is selected or automatically calculated according to the shooting conditions.
[0022]
<Example of fundus image conversion of this embodiment>
An example of image conversion in the fundus image conversion apparatus having the above configuration will be described with reference to soft buttons and fundus images displayed on the display unit 70a in FIG. 1 based on FIGS.
[0023]
FIG. 3 is an example of a display screen when a fundus image is captured via the external storage device 50 or the communication processing unit 30.
[0024]
A pull-down menu for selecting processing contents is displayed at the top of the screen. The “file” menu 100 relates to file processing. In FIG. 3, the “image capture” menu 101 is selected. The “image conversion” menu 200 is a pull-down menu for selecting image conversion processing such as brightness / contrast adjustment and contour extraction for the acquired fundus image. The “edit” menu 300 is a pull-down menu for selecting an edit process such as a close-up.
[0025]
At the bottom of the screen are soft buttons that execute the process selected in the menu. A cursor (not shown) is moved onto the soft button and the pointing device 60b is clicked, or in the case of a touch panel, the soft button on the screen is touched. The P button 400 is a button for instructing execution of image capture, and FIG. 3 shows a display screen when this button is pressed. An A button 501, a B button 502, and a C button 503 are buttons for selecting preset patterns A, B, and C in the present embodiment. The R button 600 is a button for instructing execution of contour extraction. * The button 700 is a cut button for cutting a part of the selected area of the image. The Z button 800 is a close-up button for enlarging and displaying the area cut out by the * button 700.
[0026]
In FIG. 3, the “image capture” menu 101 is selected, the P button 400 is pressed, and the fundus photo original image 900 captured via the external storage device 50 or the communication processing unit 30 is displayed.
[0027]
In FIG. 4, the “brightness / contrast adjustment” menu 210 of the “image conversion” menu 200 is selected, and the “pattern A” menu 211, “pattern B” menu 212, “pattern C” menu 213, “free adjustment”. It is a figure which shows a screen when the menu 214 is displayed and the pattern A is selected with the A button 501, for example. An image 901 whose brightness / contrast / contrast is adjusted with the parameters of the pattern A in FIG. 2 is displayed to the right of the captured original image 900. In FIG. 4, an image 901 that is brighter and contrast-enhanced than the original image 900 is created.
[0028]
In FIG. 5, the “contour extraction” menu 220 of the “image conversion” menu 200 is selected, and the R button 600 is further pressed to perform the contour extraction processing on the right-side image 901 in FIG. FIG. A contour-extracted image 902 is displayed to the right of the captured original image 900. It should be noted that the displayed image depends on conditions such as the size of the screen, but three images of the original image 901, the adjustment image 901, and the extraction image 902 are displayed, or two of the adjustment image 901 and the extraction image 902 are displayed. It may be configured to display only one.
[0029]
In FIG. 6, the “close-up” menu 301 of the “edit” menu 300 is selected, the area 902a is cut out by the * button 700, and the Z button 800 is further pressed, so that a part of the contour extracted image 902 in FIG. Is displayed as a close-up image 903. As in FIG. 5, the displayed image may be displayed in any combination at any position depending on conditions such as the size of the screen.
[0030]
The operation instructions and display examples shown in FIGS. 3 to 6 are merely examples, and various modifications are possible within the scope of the technical idea of image conversion that allows a doctor who is not an ophthalmologist of the present invention to easily diagnose from a fundus photograph. The present invention includes them. For example, when there is no diagnostic information in the color itself, it is possible to add a color to further enhance the outline.
[0031]
<Example of operation procedure of fundus image conversion of this embodiment>
An example of an operation procedure of fundus image conversion according to the present embodiment represented by a program executed by the CPU 10 of FIG. 1 for realizing the fundus image conversion will be described below.
[0032]
FIG. 7 is a flowchart showing a main program for fundus image conversion according to this embodiment.
[0033]
First, in step S10, an original image is captured as shown in FIG. Note that an image that has already been subjected to the conversion process and stored in the database 52 may be captured. Proceeding to step S20, brightness / contrast adjustment is executed as shown in FIG. Details of this processing are shown in FIG. In step S30, it is evaluated whether the brightness / contrast adjustment is sufficient / appropriate. If not, the process returns to step S20 to repeat the adjustment. If it is enough, it will progress to step S40 and will perform an outline extraction process. An example of contour extraction is shown in detail in FIG. In step S50, it is evaluated whether the contour extracted image well represents the state of the fundus. If not enough, the process returns to step S20 to repeat the brightness / contrast adjustment. If sufficient, the original image, the contour extraction image, and the parameters are associated with each other and stored in the fundus image database 52 in units of patients in step S60. The stored parameters are used to adjust fundus photographs under the same conditions of the same patient to diagnose contrast and change. Many examples of the same symptoms may be grouped as clinical results and stored in a database.
[0034]
The evaluation in step S30 and the evaluation in step S50 may be performed by an operator, by a computer, or may be further learned by a computer.
[0035]
FIG. 8 is a flowchart showing the brightness / contrast adjustment in step S20 of FIG. 7 in more detail.
[0036]
First, in step S21, it is determined which pattern has been selected. When the button of pattern A is pressed, the process proceeds to step S22, and the parameters of pattern A in FIG. 2 are read. Even when pattern B or pattern C is selected, the parameters are read from FIG. 2 in step S23 or S24. In step S25, brightness / contrast adjustment is executed according to the read parameters. Next, in step S26, it is determined whether or not the free adjustment menu 214 has been selected. If free adjustment has been selected, the process proceeds to step S27, and measures such as numerical values and bar slides are performed according to a predetermined menu (not shown). Fine adjustment is performed with. If “free adjustment” is selected without selecting the preset button, the process passes from step S21 to step S25 and jumps to step S26.
[0037]
FIG. 9 is a flowchart showing an example of contour extraction in step S40 of FIG. 7 in more detail. In this example, the simplest threshold value and pixel replacement example are shown, but differentiation processing using a two-dimensional matrix or the like may be performed.
[0038]
First, in step S41, the current density of the target pixel is read. In step S42, the difference in density from the adjacent pixels is calculated. In step S43, it is determined whether or not the difference is larger than the threshold value. If the difference is larger, the process proceeds to step S44, and the read density is stored as it is. If not, the process proceeds to step S45 where the density is replaced with “0”. In step S46, it is determined whether all pixels have been processed. If not, the process returns to step S41 to process the next pixel of interest.
[0039]
In this example, vertical and horizontal processing is performed for each color element (for example, R, G, B). Each threshold is set to an appropriate value. Moreover, you may decide the priority in the case of different judgment.
[0040]
The outline of the processing procedure has been described above, but the cropping / zoom-up process is known in the image technology, and thus the description thereof is omitted here. In addition, actual processing is realized in cooperation with a system program (OS, BIOS, etc.), but illustration and description thereof are omitted.
[0041]
<Configuration and operation example of fundus image conversion system of this embodiment>
Up to now, a stand-alone fundus image conversion apparatus has been described. Hereinafter, a configuration and an operation example of a fundus image conversion system including a plurality of devices connected via a network will be described.
[0042]
FIG. 10 is a diagram illustrating a configuration example of the fundus image conversion system of the present embodiment.
[0043]
A network 1 performs fundus image conversion in the fundus image conversion system of the present embodiment, and stores a patient 1 (corresponding to the fundus image conversion apparatus), a communication terminal 2 operated by the attending physician 2, and A communication terminal 3 operated by the ophthalmologist 3 is connected to a communication terminal 4a and a mobile phone 4b operated by the patient.
[0044]
The server 1 and the communication terminals 2, 3, 4a may be general-purpose computers. In FIG. 2, only one server 1 is shown, but it may be divided into a plurality of servers, and a configuration in which some information is stored in the communication terminal is also conceivable.
[0045]
FIG. 11 is an example of the operation of the fundus image conversion system having the configuration of FIG.
(1) The attending physician 1 transmits the data to the server 1 after photographing the fundus image.
(2) The server 1 performs brightness / contrast adjustment and contour extraction of the present invention, and returns the result to the attending physician 2. In FIG. 11, although described as one transmission / reception, the doctor 2 may interactively select parameters.
(3) If requested by the attending physician 2, send it to the ophthalmologist 3. In this case as well, it is desirable that the ophthalmologist 3 interactively select parameters.
(4) The ophthalmologist 3 returns the diagnosis result to the server 1.
(5) The server 1 returns the diagnosis result from the ophthalmologist 3 to the doctor 1. At this time, charging is usually performed.
[0046]
The attending physician 2 may be replaced with the patients 4a and 4b. In this example, the server 1 is configured to intensively perform fundus image conversion processing. However, even if the fundus image conversion processing is distributed over the network, a program is downloaded from the server 1 to the terminal and executed. It may be a configuration.
[0047]
According to this embodiment, it is possible to obtain an image that makes it easier to discriminate blood vessels and bleeding situations that are more difficult to see, such as vascular abnormalities, vitiligo, edema, and color change. In the case of this embodiment, since the adapted filter can be quantified as a parameter amount, a conversion system dedicated to the fundus image can be constructed by programming the condition (filter effect). For example, as a business model A fundus image conversion service via the Internet can be considered. The clinician can convert the fundus image taken at his / her own hospital to a predetermined content on the Internet, and convert the image into the image he / she needs from the menu. Therefore. It is possible to construct a medical examination cooperation system such as transferring image data to an ophthalmologist to make a precise decision. From the findings obtained by the fundus conversion system, it is possible to perform analysis with higher accuracy than the findings of the conventional fundus camera, to improve medical care, and to bring about dramatic progress in medical services.
[0048]
As an image processing program in the present embodiment, for example, image software PHOTOSHOP (trademark) manufactured and sold by Adobe can be used for applying various filters to a digital image. For example, brightness and contrast can be adjusted, and only the contour of the image can be extracted.
[0049]
【The invention's effect】
According to the present invention, by processing the digitized fundus image with CG, it is possible to obtain a highly accurate image by simplifying the operation of a medical staff so that a clinician can determine a lesion that only a conventional ophthalmologist could see. It is possible to provide a fundus image conversion apparatus and system that can perform this .
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a configuration example of a fundus image conversion apparatus according to an embodiment.
FIG. 2 is a diagram illustrating an example of a parameter table stored in a pattern storage unit in FIG.
FIG. 3 is a diagram illustrating an example of a screen at the time when an original image is captured in the present embodiment.
FIG. 4 is a diagram illustrating an example of a screen at the time when brightness / contrast adjustment is performed in the present embodiment;
FIG. 5 is a diagram illustrating a screen example at the time when contour extraction is performed in the present embodiment;
FIG. 6 is a diagram showing an example of a screen at the time when cut / close-up is performed in the present embodiment.
FIG. 7 is a flowchart illustrating an example of a processing procedure of a main program according to the present embodiment.
8 is a flowchart showing an example of a processing procedure of brightness / contrast adjustment in FIG. 7;
9 is a flowchart illustrating an example of a processing procedure for contour extraction in FIG. 7;
FIG. 10 is a diagram illustrating a configuration example of a fundus image conversion system according to the present embodiment.
11 is a diagram illustrating an operation example in the fundus image conversion system of FIG. 10;

Claims (5)

A fundus image conversion device that converts an acquired fundus image into a diagnostic image,
Acquisition means for acquiring a fundus image;
Adjusting means for adjusting the contrast and / or contrast of the acquired fundus image;
Extraction means for extracting a contour from a fundus image with adjusted brightness and / or contrast;
An output means for outputting a fundus image in which the extracted contour is displayed in an identifiable manner ;
In accordance with an operator's instruction, the adjustment unit includes a repeating unit that repeats the adjustment by the extraction unit, the extraction by the extraction unit, and the output by the output unit until a desired fundus image is obtained ,
The adjusting means includes
Setting means for setting light and darkness and / or contrast as initial adjustment values;
Storage means for previously storing a plurality of sets of transition values of adjustment values of brightness and contrast;
Display means for selectively displaying the plurality of sets;
Selecting means for selecting any one of the plurality of sets of displays;
Wherein according to a set of selected said displacement value selecting means, the fundus image converter according to claim Rukoto by displacing the adjustment value of brightness and / or contrast of the fundus image.   The extraction means outputs the pixel value when the difference between adjacent pixel values exceeds a predetermined threshold value, and outputs zero as the pixel value when the difference does not exceed the predetermined threshold value. The fundus image conversion apparatus according to claim 1, wherein a contour is extracted.   The fundus image conversion apparatus according to claim 1, wherein the output unit outputs the extracted contour so as to be identifiable with a predetermined color. A communication terminal, the fundus image conversion system comprising an information processing apparatus for converting the fundus image acquired via the network from the communication terminal into an image for diagnosis,
The communication terminal is
Acquisition means for acquiring a fundus image;
First transmission means for transmitting the acquired fundus image to the information processing apparatus via a network;
Receiving a fundus image in which brightness and / or contrast and / or contrast are adjusted and / or an extracted contour are displayed in an identifiable manner, and a screen for inputting an instruction to select the brightness and / or contrast adjustment by an operator First receiving means;
Instructing means for repeatedly instructing the information processing apparatus via the network to adjust the brightness and / or contrast until a desired fundus image in which the extracted contour is displayed in an identifiable manner is obtained in accordance with an operator selection instruction; With
The information processing apparatus is
Second receiving means for receiving a fundus image from the communication terminal via a network ;
Adjusting means for adjusting brightness and / or contrast of the received fundus image according to a selection instruction from the communication terminal via the network ;
Extraction means for extracting a contour from a fundus image with adjusted brightness and / or contrast;
The fundus image extracted the contour is identifiably displayed via a network and a second transmission means that sends to the communication terminal,
The adjusting means includes
Setting means for setting light and darkness and / or contrast as initial adjustment values;
Storage means for previously storing a plurality of sets of transition values of adjustment values of brightness and contrast;
Display instruction means for instructing the communication terminal to display the plurality of sets in a selectable manner;
Selecting means for selecting any one of the plurality of sets according to a selection instruction from the communication terminal by an operator;
Fundus image conversion system according to a set of selected said displacement value at said selecting means, said fundus image brightness and / or of, wherein Rukoto by displacing the adjustment value of the contrast. The information processing apparatus further identifies the received fundus image, the fundus image whose brightness and / or contrast is adjusted, and / or the contour according to a transmission instruction from the communication terminal by the operator via the network. The fundus image conversion system according to claim 4, further comprising third transmission means for transmitting a fundus image that can be displayed to another communication terminal.

Images